Living for the neighbourhood: marginalization and belonging for the second-generation in Berlin and Paris

In this paper, based on qualitative research on the North African second-generation in Paris and the Turkish second-generation in Berlin, we discuss ethnic minorities’ attachment to place and how living in highly diverse cities shape their perceptions and experiences of marginalization and belonging. Even though France and Germany have different state-level approaches to citizenship and belonging, the experiences of marginalization and exclusion of the second generation in the city are rather similar. In both societies, ethnic and religious minorities such as the North African or Turkish second-generation are excluded from mainstream society. This exclusion is experienced on the local level. Thereby the geography of Berlin and Paris impacts ethnic second-generation populations’ feeling of belonging to the communities in which they live, as well as how they understand their experiences of racism and exclusion. This research has implications for understanding the multivariate experiences of middle-class second-generation ethnic populations across Europe.


Sense of belonging & attachment to place in super-diverse cities
Cities all over the world are increasingly super-diverse (Crul, 2016;Vertovec, 2007). Due to various waves of migration, urban populations are heterogeneous along various dimensions, such as nationality, ethnic background, religion, citizenship status, lifestyle, sexual orientation, and socioeconomic status. Wessendorf (2013, p. 407) refers to "commonplace diversity," illustrating how "ethnic, religious and linguistic diversity [is] experienced as a normal part of social life and not as something particularly special". In such neighborhoods, diversity is generally viewed positively. Despite positive accounts of lived diversity, there are still many examples of exclusionary boundary drawing based on various social categories. People develop feelings of belonging to their place of residence, which shape their self-identification (Benson & Jackson, 2013;Blokland, 2003;Savage, Bagnall, & Longhurst, 2005;Watt, 2009). By attachment to place, we are referring to the multivaried ways that individuals form relationships with and assign meaning to place (Low & Altman, 1992). Previous research on place attachment shows that it stems from "accumulated biographical experiences" (Gieryn, 2000, p. 481). To assert a strong attachment to place is to assert the critical role that place plays in how one constructs one's identity. Attachment to place can also involve challenging existing place meanings and conferring social status. Particularly in socially and ethnically mixed neighborhoods, however, place attachment is enabled through drawing boundaries against groups that are perceived as different from oneself. In Boston (Tissot, 2007) or major European cities such as Paris, Lyon, Madrid and Milan (Andreotti, Le Galès, & Fuentes, 2015), middle-class populations feel belonging to their neighborhood because they know how to manage diversity or exit from it when needed. Such studies usually focus on processes of boundary drawing of white middle classes. Here, in contrast, we focus on visible minorities.
Boundaries sufficient to develop attachment to place are often drawn based on ethnic and/or religious factors. Throughout Europe, there is evidence that ethnic and religious minorities continue to be stigmatized and marginalized (Beaman, 2017;Bleich, 2009;El-Tayeb, 2011;Hajjat & Mohammed, 2013;Joppke, 2015;Voas & Fleischmann, 2012). Thereby, who belongs and who does not belong to the city is mediated by state-level structures of and discourses on the nation, citizenship, and belonging (Crul & Schneider, 2010). Even in super-diverse cities and neighborhoods, ethnic and religious minorities often experience stigmatization and discrimination. 1 Ehrkamp (2006) illustrates how integration discourses at the national level are reflected on the local level. In Marxloh, a multiethnic neighborhood in the German city of Duisburg, "assimilation discourses are integral to the ways that native Germans construct Turkish immigrants and their cultural practices as oriental and 'other'" (p. 1688). To illustrate, she recounts how a second-generation Turk conceptualizes the oftentimes aggressive debate about dual citizenshipa hotly debated topic at the time of her fieldworkled to tensions between the migrant and non-migrant population in Marxloh.
Thus far, the processes of place attachment and belonging have primarily focused on white middle class populations. Less is known about the dynamics of place attachment and belonging for ethnic minoritiesthus those who are usually excluded by boundaries drawn by whites. Questions about belonging for ethnic minorities primarily focus on the national level, as in, for example, the relationship between different citizen and integration regimes and their impact on different facets of integration (Ersanilli & Koopmans, 2011;Koopmans, 2010).
Instead, in this paper, we focus on two super diverse cities, Paris, France and Berlin, Germany, to show how second generation immigrants who are ethnic minorities negotiate a sense of belonging to the city and the neighborhood in which they live, despite persistent exclusion from mainstream society. Focusing on the local level also allows us to analyze migrants and their descendants' agency in shaping the city and neighborhoods, something still lacking in much urban scholarship (Glick Schiller & Schmidt, 2016;Glick Schiller & Çağlar, 2009). Specifically, we ask how do middle-class ethnic second-generation populations, children of North African immigrants in France and children of Turkish immigrants in Germany, negotiate their relationships to the urban environments in which they live amid racism and exclusion from the state? We thus address the relationship between national discourses around belonging and the lived experience on both the city and neighborhood scales. France and Germany have different approaches towards citizenship and thus to the question who belongs to the nation and who does not. Certain groups remain excluded from both societies, which impacts on their experiences of belonging and how they are negotiated on the local level.
In what follows, we discuss previous research on the second generation in the city, paying particular attention the processes of belonging on different spatial scales. Our intervention in this literature is in our focus on the middle-class segment of the second-generation. We then discuss the context of both France and Germany and the utility of a cross-national comparison. We then present our methodology and findings as to how ethnic minorities in France and Germany navigate belonging to place between the national and the local levels.

The second generation in super diverse cities
The past decade has seen more research on the second generation in Europe. The large-scale TIES (The Integration of the European Second Generation) study analyzes the integration of second-generation Turks, Moroccans and Yugoslavians in various cities of various European countries (Crul, Schneider, & Lelie, 2012). Similar to the experiences of certain ethnic and racial groups in the U.S. (Alba, 2005;Kasinitz, Waters, Mollenkopf, & Holdaway, 2009;Portes & Zhou, 1993;Vallejo, 2012;Waters & Kasinitz, 2010), the second generation throughout Europe also has complex pathways of incorporation and assimilation into mainstream society. The TIES study, for example, illustrates the multi-faceted identification of the second generation, vis-à-vis their country of residence, but also their parents' home country. Thereby, ethnic minorities living in countries with more assimilationist policies have a more transnational orientation than their counterparts in countries with more accommodating policies (Fokkema, 2011).
With regard to their experiences in the city, "second-generation groups are more 'native' to their city of residence than their peers of non-migrant parentage" (Crul et al., 2012, p. 312). They are indeed one of the most "established" populations in cities, due to their length of time living in the cities (Crul, 2016). This might be one reason why for the second generation, a strong sense of belonging to the city is less complicated than a strong sense of belonging to the nation itself. For the native comparison group, this difference is less pronounced, or even reversed. Even more important for identification than the city is the neighborhood. Many of the second generation respondents in the TIES study had lived in the same neighborhood for many years and felt attached to it, which they also expressed by a high level of involvement in local affairs, an illustration of their place-making practices. In other words, the second-generation is more likely to have local, place-based connections than national, state-level ones. In general, the TIES study thus shows the importance of the local integration context, which can reinforce or weaken national discourses around who belongs to the nation. Existing research also demonstrates the importance of the local context for identity formation (Vathi, 2013).
Together, these findings suggest that for the second generation, particularly those that are subject to discrimination and exclusion due to their ethnic or religious background, the local level is more important for processes of identification and belonging than is the national state.
Our research contributes to existing research on the second generation in cities in two ways. First, we compare the second generation in two European, but clearly distinct cities and nations, which allows us to focus on the dialectic between belonging on the national and local level. Within Europe, France is often used as the prime example of a country whose citizenship regime is based on jus soli, while Germany's functions primarily according to jus sanguine principles. Moreover, while Paris and Berlin are both highly diverse cities, with a long history of immigration, the geography of the city is clearly distinct. In Berlin, the traditional immigrant neighborhoods are in the city center, while in Paris, immigrants have primarily lived in the outer ring of the city, the suburbs or the banlieues. We therefore examine not only how national discourses on immigration and integration affect ethnic minorities in their daily lives, but also how the geography of the city affects ethnic minorities' relation to their urban environments.
Second, we explicitly focus on the upwardly mobile segment of the second generation, which has only recently become a focus of empirical research. This focus on the middle-class segment allows us to unpack the racial and ethnic nature of the marginalization these children of immigrants face, as opposed to using socioeconomic status, or other factors as explanatory variables. By focusing on the middle-class segment of the second generation, we are emphasizing how upward mobility is not a panacea for exclusion from the nation-state. The descendants of immigrants from Turkey in Germany, and from North Africa in France are the largest immigrant-origin groups. Their similar experiences make them apt for comparison. Alba (2005) argued that in the United States, boundaries based on discrimination and exclusion are primarily race-based, while in Europe they are primarily based on religion. Islam is commonly portrayed to be incompatible with Christianity (Joppke, 2015), and Western society more generally, and Muslims are therefore portrayed as 'the other.' Due to this 'bright' boundary, assimilation "is most available to secularized Maghrébins and Turks and presumably to those who have attained substantially more than the modest educational attainment and occupational status that characterize the majority of these groups" (Alba, 2005, p. 40). However, we will show that even these 'successful' second generation immigrants are still excluded despite their educational and occupational successes and how they experience the city is influenced by this constant exclusion. The focus on upwardly mobile second generation also allows us to examine in more detail the intersection between ethnicity and social class.

National context: Ethnicity and citizenship in France and Germany
Understanding the plight of ethnic minorities in France requires understanding France's Republican ideology, which does not recognize race and ethnicity as legitimate. Under Republicanism, membership in the nation supersedes any other identification or distinguishing characteristic. French identity is civic, rather than ethnic, an emphasis dating to before the French Revolution (Bell, 2003). The state interacts with individuals independent of any group categorization or special interest group (Chapman & Frader, 2004). France, in contrast with Germany, has a civic conception of nationhood based on jus solis (Brubaker, 1992). The French model of integration is based on assimilation to common French culture and values (Tribalat, 2004). Despite France's long history of immigration, it has continually been framed as a social problem, as evidenced by decades of immigration-related legislation. In terms of everyday practice, the definition of what it means to be "French" often excludes particular populations within French society, including those who were born in France to parents who are immigrants from former French colonies in North Africa. Despite living in cosmopolitan urban environments, the North African second-generation is continually reminded of their difference from "whites" 2 and their inability to be seen as French as anyone else (Beaman, 2016(Beaman, , 2017. Despite their legal belonging to France, they are often perceived as foreigners and have their "Frenchness" contested by others (Currently, children born in France to North African immigrants are considered "virtual citizens at birth" (Simon, 2012)).
In Germany, belonging to the nation is still primarily based on common ancestry. After 2000, the rules for nationality acquisition have been eased. For example, children of Turkish immigrants born in Germany get the citizenship if one parent has lived in Germany for at least eight years and has an unlimited residence permit. Until 2014, double nationality was not allowed and children had until their 24th birthday to decide whether they wanted to keep the German or their parents' nationality. This rule has been eased and double nationality is now allowed for children of immigrants born in Germany after 1990, if they have lived in Germany for at least eight years. However, many children of Turkish immigrants were born before 1990 and kept their Turkish citizenship, coupled with a secure residence status in Germany. Thus, "while multiple belongings in the context of identity are recognized as postmodern normalcy, there is at least in Germany still the criterion of a unilateral decision regarding national, ethnic and cultural belonging, which reflects the idea of assimilation as vision of successful integration" (Foroutan, 2010, p. 11). Turkish immigrants and their descendants are among the most stigmatized ethnic groups in Germany, not only due to their ethnic background but also to their assumed religious background (Barwick, 2016;Ehrkamp, 2005;Korteweg & Yurdakul, 2014;Ramm, 2010). Just as in France, Islam is often thought as being incompatible with 'European values' , such as the separation between church and state or equality between men and women (Adida, Laitin, & Valfort, 2016;Joppke, 2015).

Local context: The north African and Turkish second generation in Paris and Berlin
In Germany, just as in other European countries, the share of ethnic minorities has been steadily rising and we now find majority-minority neighborhoods in many metropolitan cities (Crul, 2016). Turkish immigrants and their children are the largest ethnic minority group in Germany, including in Berlin. According to the 2011 census, about 17.7% of all people with a migration background had a Turkish origin. In Berlinthe only city in the Eastern part of Germany with a significant number of Turksthe share of people of Turkish origin is above national average. The number of ethnic minorities and thus also Turkish minorities differs substantially between the districts. The highest percentages of ethnic minorities can be found in the central districts of Wedding, Moabit and Mitte (which since 2001 merged into the single district of Mitte), Friedrichshain and Kreuzberg (now a single district) and Neukölln, ranging from 41 to 51%. These neighborhoods are often also socioeconomically disadvantaged, with above-average percentages of welfare receivers and unemployed residents. In contrast, the districts in higher status neighborhoods in the Western part of Berlin have a more homogeneous population with fewer migrants, hovering between 25 and 35%.
In Berlin, the districts with the highest share of immigrants and their descendants, are thus in the inner citya clear difference compared to Paris. In both cases, the concentration of ethnic minorities in the respective districts has historical roots. When the guest worker immigration from Turkey to Berlin started in the late 1960s, most Turks settled in Kreuzberg and Wedding (Özüekren & Ergoz-Karahan, 2010). Berlin was still separated by the Wall and the neighborhoods just adjacent to the Wall were deteriorated and not very attractive. Due to cheap housing costs and the city's policy to use immigrants as temporary residents for housing already slated for demolition, the number of Turkish immigrants in these neighborhoods quickly increased. After the fall of the Wall, these neighborhoods suddenly formed the new center of Berlin. The location thus became very attractive, which is why parts of these districts are now being gentrified and seeing rising rents. Nevertheless, they continue to be the districts with the highest proportions of Berlin's Turkish population.
In France, especially in Paris, the greatest concentration of immigrant-origin individuals tends to be in the quartiers populaires, working-class or disadvantaged neighborhoods on the outer edges of the city, particularly the 18th, 19th, and 20th arrondissements and the banlieues, or the suburbs surrounding Paris. These residential patterns reflect historical patterns of migration from former French colonies, notably from the Maghreb and Sub-Saharan Africa. These immigrants, who were expected to only be temporary residents, often settled in the outlying banlieues of major cities because of the presence of cheaper housing and factory employment. The number of immigrants from these former French colonies increased after World War II, the end of France's Fourth Republic in 1958, and the Algerian War of Independence in 1962. More and more immigrants were living in subsidized housing complexes [or habitations à loyer modéré or HLMs] in the banlieues. As whites started to move from HLMs to private housing due to low-interest government loans in the early 1970s, these banlieues became even more associated with immigrants and their descendants.
It is difficult to have precise figures on the ethnic minority concentration of different residential communities because French Republicanism disallows racial and ethnic-based statistics. Yet, there remains a persistent concentration of African-origin (both North African and Sub-Saharan African) individuals in the banlieues and a stigma attached to banlieue residents (Shon, 2010). Several studies have demonstrated the stigma and differential treatment towards banlieue residents versus other residents (Dikeç, 2007;Silverstein, 2008;Tissot, 2007). While banlieues are not homogeneous, as French sociologist Sylvie Tissot has explained, they have become "the symbol of a bleak urban environment, deviant youth, and segregated minorities." This was perhaps made more recently clear to those outside of France with the 2005 uprisings throughout several banlieue communities throughout France (which began following the deaths of Zyed Benna, a 17-year old of Tunisian origin, and Bouna Traore, a 15-year old of Malian origin, who fled police in the Parisian banlieue of Clichy-sous-Bois) (Schneider, 2008).

Methodology
The results of the paper are derived from two distinct case studies. The Berlin study is based on semi-guided interviews with 41 upwardly mobile, second-generation Turks in Berlin. In-depth interviews, lasting mostly around 1.5 h, were conducted by Barwick in 2012-13, inquiring about the respondents' social and residential mobility, their identifications along ethnicity and class, place attachment, as well as personal networks. The respondents were born in Germany or immigrated as children, they were employed and had higher economic and/or cultural capital than their parents.
To reach potential respondents Barwick used convenience sampling. She relied on several points of entry to the field, contacting ethnic-based organizations, calling businesses that were owned by people of Turkish origin, and inquiring in primary and nursery schools. Lastly, she made use of personal contacts.
Respondents are all descendants of immigrantsthey either came to Germany as children, as part of family reunification, or they were born in Germany. They work as lawyers, tax consultants, in the social service sector, or are self-employed in the food sector. The second-generation Turks were between 29 and 63 years old at the time of the interviews, but most clustered between the ages 35 and 50. The majority was married and had children; the ones without children were all in a relationship.
Respondents are upwardly mobile, and have higher economic and cultural capital than their parents, who wereas typical for guest workersmostly employed as manual laborers, and only had basic schooling (in Turkey). All respondents have lived for the most part of their childhood and young adulthood in one of the inner-city districts with a high share of ethnic Turks and other ethnic minorities. When the respondents were younger, these districts were still much more socioeconomically disadvantaged compared to today. Some of the respondents (23 out of 41) have moved out of these inner-city districts into more middle-class areas, predominantly in the Western parts of Berlin. The one major reason for such a move was not dissatisfaction with the previous neighborhood, but a concern about local educational facilities (Barwick 2016).
There are two reasons for the focus on upwardly mobile ethnic Turks. First, Barwick was interested in neighborhood choice. As lack of economic capital, as well as a lack of language skills, inhibits residential choice, Barwick only interviewed those middle-class individuals who were proficient in German and had an income that would allow them to consider living in middle-class Berlin neighborhoods. Second, Barwick was highly interested in identification. To learn more about how people identify and how that changes over the life course, it is particularly fruitful to analyze those who are in a 'blurry' social position, such as upwardly mobile persons. Bourdieu (1987, p. 12) explained that people who find themselves in the "intermediate or middle positions of the social space" have the most room to fill the fuzzy space between practices and social positions. How they fill this space depends not only on social class, but also ethnicity, nationality, or locality.
For the Paris study, Beaman (2017) conducted semi-structured interviews with 45 adult middle-class children of North African immigrants living in the Parisian metropolitan region, while living in Paris primarily from 2008 to 2009. The original aim of this study was to unpack how an ethnic minority population understands their identity and marginalization from mainstream society. This respondent sample consists of 24 men and 21 women. Respondents range in age from 24 to 49 years old; the average age being 32 years old. In terms of North African origin, 55% of are of Algerian origin; 26% are of Moroccan origin; and 17% are of Tunisian origin. The majority of these respondents' parents emigrated from the Maghreb between 1950 and 1970, primarily for economic reasons. Many of them have low levels of educational attainment, often not past middle school, speak little French and communicate mostly in Arabic. Usually the fathers worked in low-skilled jobs, such as construction and factory employment, while mothers were homemakers or did domestic labor. All respondents live in the Paris metropolitan region, or Ile-de-France. 35% of respondents live in Paris and 65% lives in the banlieues, mostly the inner-ring départements of Seine-Saint-Denis, Val-de-Marne, and Hauts-de-Seine.
Beaman recruited respondents through contacting various organizations in the Paris metropolitan region and employed snowball sampling (Small, 2009) to form her respondent sample. This sample focuses on middle-class individuals, those who have achieved upward mobility vis-à-vis their immigrant parents. In contrast to Barwick, Beaman's focus on the middle-class for her sample was a result of her snowball sampling method. Beaman defined having a middle-class status by respondents' educational attainment levels and professional statuses. In terms of education, she focused on those who passed the Baccalauréat (BAC) exam and attended college (whether or not they actually graduated). In terms of employment, she focused on those in the French socio-professional category of cadre, or professional types of employment. Beaman also used a grounded theory approach to analyzing her ethnographic data.

Moving up but not out
For second-generation Turks in Berlin, their feelings and experiences of belonging or exclusion depend on the geography of the city and the social and ethnic makeup of the different neighborhoods. There is a clear difference between second-generation Turks living in more central, socioeconomically and ethnically mixed neighborhoods, and those living in middle-class areas with a high share of native 3 German residents. What is common is the significance in the attachment with the city and neighborhood, in contrast to identification with the nation, since "where you are born is like a lottery" (interview with Dalim 4 ).
For those in diverse neighborhoods such as Wedding, Neukölln and Kreuzberg (cf. Fig. 1), we observe a high practical and symbolic use of the neighborhood. Respondents use the neighborhood for their daily needs and in their free time, but also symbolically, for identification. These inner-city neighborhoods contain many third places (Oldenburg, 1997) where locals can regularly meet and interact. These casual, spontaneous encounters are among the most appreciated characteristics of a neighborhood, according to respondents. They also lead to public familiarity, which "arises when interdependent anonymous people keep encountering each other, and Vergemeinschaftung 5 occurs" (Blokland, 2003, p. 93). To illustrate, Nursel, a law student in her early thirties particularly enjoys that "something is always happening in the streets. And you know everybody and everybody is always ready for some small talk." Some respondents also acknowledge certain problems in their neighborhood, for example dirty streets, drug trade, or prostitution. However, they rarely draw strong boundaries against groups that are socioeconomically or ethnically different, but instead view them as part of the neighborhood. For example, Selbi who lives in Wedding, is well aware of problems connected to drug dealing in her neighborhood, but she does not view this as a major problem, even though the drug dealers are ethnically different (mostly from Sub-Saharan Africa) and mostly have a temporary right to remain): Well, from the outside, [Wedding] is always presented as bad and terrible and dangerous. That's not the case at all. Of course, there are dealers and there is drug trade. But it never affects the neighborhood as such. The people who do that, they do it among themselves, and the residents are aware of that, because they see that people deal, but they are never directly approached. As I said, I've been living here since 1985, and they never asked me whether I wanted to buy something.
Atalay is another case in point. A long-term resident of the poorer part of Kreuzberg, he values the changes that have taken place in the past years and that led to an increased heterogeneity of the population: A few years ago, I didn't want to get involved in anything here, I just didn't want to, but through the diversity, this interest, you take your time. And suddenly you get to know this culture, these different people. And that brings you further, I believe. And that's what I like in this area, in this environment.
For Atalay and other respondents, the diversity of the local population facilitates place attachment. The neighborhood is also used for processes of self-identification. Thereby, diversity also includes groups that are more noticeably different, such as drug traders.
This contrasts with what we would expect based on studies of boundary drawing in mixed neighborhoods, as previously discussed. Drawing boundaries inherent in creating place attachment as observed for white middle classes do not operate in the same way for the Turkish second-generation. A possible explanation is their status as ethnic minorities. While respondents are privileged in socioeconomic terms, they are not so in ethnic terms. They are established in the neighborhood where they do not face discrimination and exclusion, but they do so in other contexts, for example when interfacing with state agencies. Several respondents spoke of their experiences in school and in higher education where they felt they had to work much harder to receive the same results than their fellow native German students. Among those who were active in a local political party, the major complaint arose that they were always counseled for issues relating to migration, even though they were specialists in different topics. Thus, these experiences of exclusion can be a reason for why place attachment is not based on drawing boundaries against other groups.

Moving up and moving out
Those second-generation Turks, who live in middle-class areas that are dominated by native German residents, have very different experiences with discrimination and exclusion in their neighborhood. Enginalp is a lawyer and lives with his wife and son in a bourgeois neighborhood. His case is an extreme one, but it illustrates well the difficulties second-generation Turks face when being among the very few families that are visibly not 'native German'. Although he is very fond of his neighborhood and appreciates it as a place where his son can grow up in peace and tranquility, he knows that his neighbors have a problem with living next to a Turkish family. The few friends he has in the neighborhoodmostly through his son's schoolknow that the neighbors gossip about him and his family, wondering how he got the money to live in such a fancy neighborhood. For Enginalp, the neighborhood is not a safe space. Skeggs (1999, p. 216) argued that "The city is (.) simultaneously raced and sexed. It is one of the spaces where (usually white) heterosexual masculinization remains spatially intact." For ethnic minorities, a fear of victimizationwhich does not need to be physical, but also verbal or even nonverbal insultscan lead to the avoidance of 'white' places (Schuster, 2012). Enginalp and other respondents' practical and symbolic neighborhood use is directed to other places in the city, mostly ethnically diverse neighborhoods. Ferda has moved from the ethnically diverse inner-city area of Kreuzberg to a residential area in Berlin's South-West, to find a better school for her children (Barwick, 2016). She never really warmed up to the new neighborhood and neighbors, and spends much of her free-time in her old neighborhood, where she wants to move back to as soon as her youngest child leaves primary school. Ferda highly appreciates her old neighborhood, particularly that life takes place outside in the streets and you always meet people you know and can chat with. This is a public familiarity she does not find in her own neighborhood of residence. Due to experiences of exclusion and marginalization in predominantly white neighborhoods, these second-generation Turks thus develop a feeling of belonging to neighborhoods other than their own.
The example of middle-class second generation Turks in Berlin shows that living in or frequenting ethnically diverse neighborhoods, many of which are of a lower socioeconomic status, is a choice and reflects their agency. It thus goes against the assumption "that migrants' social lives are confined within ethnically defined neighbourhoods and that a diversity of backgrounds constrains urban social life and development" (Glick Schiller & Schmidt, 2016, p. 5).
Living in the banlieues versus living in Paris: Wanting to move versus wanting to stay In the case of the Paris metropolitan region, how respondents understand their difference and marginalization within Franceespecially as middle-class individualsrelates to how they understand where they live. In contrast to Berlin, the marginalized residential locations are outside of the city center of Paris. It is these places that are primarily associated with North African and Sub-Saharan African origin individuals. For the majority of respondents, living in the banlieues, particularly those that are predominately immigrant-origin, facilitates solidarity and ethnic belonging amid a larger societal context where they do not belong. In other words, how the middle-class North African second-generation relates to the places where they live helps mediate their experiences of exclusion. Previous research has shown how individual ethnic identity is most influenced by individual's interactions outside of ethnic enclaves, where their exposure to discrimination and prejudice is heightened (Eid, 2008). As such, one's relationship to place has deeper implications for marginalized populations, who may experience more racism and discrimination in predominately white communities as compared to in predominately ethnic communities. In contrast to Berlin, a major distinction in terms of Paris is in the degree of attachment respondents have to where they live and the degree of control they feel they have over their residential options.
One example of a deep attachment to place is Kamel, a 29-year old of Moroccan origin who identifies as both French and Maghrébin. He has spent most of his life in a cité in Poissy, a western banlieue about 15 miles from the center of Paris. His family settled there because his father was a factory worker for Peugeot, a French car company which is headquartered there. He describes his neighborhood as a bit "special," as it is part of an urban renewal program. Kamel is very proud of how he and other residents fought against project demolition and provided insight into his attachment to this cité: We had a mayor who wanted to destroy the neighborhood, because in this neighborhood you have a lot of people of foreign, Maghrébin, African origin, and he wanted to eradicate the population . . . that's what he [the mayor] said, I no longer want any of you in this neighborhood or in this town. You see, this is a racist person … He kept saying such violent things. … They revolted, we revolted, and we formed a collective. And that was 4 years ago, we have fought against him for four years … We hired a lawyer and we fought to delay the demolition up to the mayoral election . . . Now there is a new mayor. He launched a new plan, the residents are involved in it, we are staying in the neighborhood. 6 Here, Kamel emphasizes both his attachment to this neighborhood and its characteristic social cohesion. "It's like a family here... there is a solidarity. Everyone knows each other here. Sure the living conditions are a bit difficult, but now they are improving, it's getting better." He sees living in his Poissy cité as constitutive of his identity, something which has implications for each facet of his life. Being educated in French schools and growing up in an immigrant-rich cité have led Kamel to define himself as being both Maghrébin and French.
This combination allows Kamel to feel deeply connected to his neighborhood, and value being a French citizen, despite the marginalization he sees connected to where he lives: It's difficult for people in neighborhoods like this . . . because society put them in difficult situations . . . and as soon as we make an effort, we find a wall in front of us ... As for me, I had a lot of trouble finding my first job after I finished school. It took 1½ years for me to find my first job, but my classmates who were named François, Edouard, Frédéric, it took them 6 months or 4 months to find a job, but for me it took 1½ years. This is also an example of how banlieue residential location is a signal of ethnic and racial status. Because the address on his first CV was in a beleaguered banlieue, this was an immediate signal that he is non-white or a racial and ethnic minority. Despite the problems he sees in his neighborhood-similar to those identified by secondgeneration Turkish residents in Berlin-he has no plans to move because of his feelings of attachment to the area. Kamel also wants to be a role model for others in his neighborhood. 7 Despite his socioeconomic status which would permit him to leave to this banlieue neighborhood, he is committed to staying because of its meaning to him.
While also a banlieue resident, Hinda, a 33-year-old youth worker of Tunisian origin and divorced mother of a seven-year-old daughter, relates to where she lives very differently from Kamel. Hinda wants to move but she cannot; she does not have an attachment to her neighborhood. She has lived in the northeastern banlieue of Drancy for the past seven years. She dislikes living in Drancy, and describes her neighborhood as "sad." Hinda explains that she "did not choose this neighborhood" and would rather live in Paris but it is too expensive. "It is not particularly pretty. .. the buildings are sad, there isn't much life here, we are a little too far outside [of Paris] here," she laments. She feels unconnected to where she lives, and hopes to be able to move soon, either to Paris or another banlieue closer to Paris. Here, how Hinda relates to living in Drancy illustrates how she sees herself as a French citizen. That she is surrounded by those who look like her in Drancy is irrelevant, she dislikes living in Drancy for the same reasons she imagines any other French person would.
The connection between how one relates to the place where one lives and how one sees him or herself as an ethnic minority within French society also extends to those who do not live in banlieues and instead live within Paris. For example, Djamila is a 49-year-old divorcee who has always lived in the 20th arrondissement (cf. Fig. 2) of Paris. To her, she is French because she was born and raised in France. She characterizes her neighborhood as a quartier populaire, and acknowledges that people who live in quartiers populaires are often stigmatized in larger French society. She thinks that more neighborhoods should be as diverse or mixed as hers. However, living in a quartier populaire is not relevant for Djamila in how she self-identifies. She does not feel any strong attachment to her neighborhood, even though it is the only neighborhood she has ever lived in. Though she feels indifferent to where she lives, Djamila will most likely stay in her neighborhood but only because she feels it is the only place she knows, not because of any particular connection.
After growing up in Avignon, a small town in southern France, and moving to Paris for work about six years ago, Mohamed, a 30-year-old of Algerian origin, currently lives in a quartier populaire in the 13th arrondissement, near the Porte d'Italie. His neighborhood is comprised of people who look like him. "There are many blacks and Arabs. And I like that. .. because I feel like I am in a familiar element" he explains. It is important for Mohamed to live in a neighborhood like this, rather than a more bourgeois or predominately white neighborhood. As he is one of the few racial and ethnic minorities at the insurance office where he works, Mohamed appreciates residing in a place where that is not the case. Living among other racial and ethnic minorities buttresses against the marginalization he experiences in a predominately white work environment. Mohamed sees both being French and having Algerian origins as part of who he is. He sees having a "double culture" as a richness, as he has learned two different codes of behavior or ways of being -Maghrébin and French. Mohamed operates within a French code while at work, while also operating within a Maghrébin code while at home. How he relates to where he lives reflects this understanding. Reda, a 32-year old Human Resources consultant of Algerian origin who lives in Paris' ninth arrondissement, similarly sees himself as embodying both French and Maghrébin cultures. While he currently lives in a "nice, bourgeois neighborhood," he grew up in a quartier populaire in Meaux, a banlieue in the Seine-et-Marne département near Disneyland Paris. Due to this, Reda did not recall feeling different from others growing up; it was only when he moved to Paris a few years ago that he felt "othered" because he grew up in the banlieues (or ghettos, as he refers to them). He still sees himself as being "socially marked" because he once lived there. Reda grew up in predominately Maghrébin environment in Meaux and currently lives in a predominately white environment in his bourgeois Parisian neighborhood. Because of these different residential experiences, Reda sees himself as having lived in two different worlds -French and Maghrébin. Yet his marginalization is more marked for him living in a predominately white environment. Such examples of Reda, Mohamed, and Djamila complicate ideas about how ethnic minorities relate to place by illustrating the heterogeneity of such attachments to place among the same population.

Discussion & conclusion
We have analyzed place attachment and experiences of marginalization and exclusion of the Turkish second generation in Berlin and the North African second generation in Paris, paying particular attention to the discourses around belonging at the national level and experiences at the city and neighborhood levels. Even though the national discourses on belonging and models of immigrant integration differ between France and Germany, the experiences of belonging to place and exclusion on the local level are quite similar between the two cities and second-generation populations. National discourses around belonging can thus be seen as a referential framework, but the city and neighborhood play a more important role drawing boundaries of inclusion and exclusion.
In Paris, predominately ethnic neighborhoods and communities tend to facilitate solidarity for ethnic minorities. The banlieues and peripheral communities (i.e. the 18th, 19th, and 20th arrondissements) are foils to other areas in the city proper, as some middle-class individuals prefer to live in predominately ethnic communities to counteract the effects of working in predominately white spaces. Under the French Republican model, such ethnic communities are deemphasized as each citizen is supposed to interact with the state as an individual, rather than as part of a group identity. Yet, in response to the marginalization and racism ethnic minorities experience, such identity-based communities are actually being created. Where these middle-class respondents live and how they relate to where they live reflect the failures of the implementation of the French Republican model. Even though race and ethnicity are not supposed to be salient, they are. Living in banlieues and immigrant-concentrated neighborhoods in Paris is a proxy for racial and ethnic background; it marks one as non-white or as a visible minority within France. In Germany, belonging to the nation still rests primarily on ancestry, but second-generation Turks' experiences with exclusion in white spaces are similar to those in France. Ethnicity and religious background matter more than class, particularly for those respondents who moved to middle-class neighborhoods in more peripheral parts of Berlin with a larger proportion of whites. In contrast to Paris, where the neighborhood functions as a proxy for visible difference, there is no exclusion based on one's address in Berlin. It is only the visible otherness that matters. The consequence of these experiences of exclusion is a withdrawal to the inner-city neighborhoods with a larger proportion of ethnic minorities.
Our findings suggest that a difference between the experiences of the North-African and Turkish second generation could be in how 'white spaces' are constructed in both cities (Anderson, 2015). In Paris, these are often workplaces so that a multi-ethnic neighborhood becomes more of a safe space. In Berlin, it is actually the residential middle-class neighborhood that is predominantly white and other, more diverse, neighborhoods are used to socialize with family and friends during leisure time.
By focusing on the middle-class, we have demonstrated how race and ethnicity are more salient bases for exclusion and discrimination than is class or socioeconomic status. The differences regarding place attachment as well as experiences of exclusion and belonging between the North-African second generation in Paris and the Turkish second generation in Berlin can partly be attributed to the geography of the city and the social and ethnic makeup of neighborhoods, rather than to the national discourses of immigrant assimilation. These discourses differ between the two countries, yet they have similar lived effects on ethnic and religious minorities who still feel excluded from belonging to the nation. Regarding the geography of the two cities, many communities on the periphery of Paris are very denigrated, in contrast to the Berlin, where traditional immigrant neighborhoods in the city-center are partly being gentrified. Moreover, Berlin's typical middle class neighborhoods are outside the central city. In other words, predominately immigrant neighborhoods are less visible to mainstream society in Paris than they are in Berlin.
We thus have shown how different discourses around belonging on the national level still exclude similarly marginalized populations. Ethnic and religious minorities feel the consequences of this discourse on the city and the neighborhood levels. On the one hand, in predominantly white spaces, they are excluded as boundaries related to ethnic and religious background are more salient than boundaries related to class background. On the other hand, the makeup of neighborhoods can also mitigate against experiences of exclusion, particularly if they are ethnically diverse, as ethnic minorities develop feelings of belonging and their practices shape the neighborhood. This supports the findings of Crul et al. (2012) regarding the importance of the local integration context. We have demonstrated how this also applies to middle-class populations. As we focused on such dynamics and experiences within two global cities that are also capital cities, further research should examine how the dialectic between discourses around belonging on the national level and the experiences of belonging on the local level operates for middle-class ethnic second-generation populations in less populous or cosmopolitan environments.
Endnotes 1 Terrorist attacks in recent years throughout Western Europe have only heightened Islamophobia and anti-Muslim sentiment as well as justified xenophobic sentiment. 2 We use the term "whites" here following the lead of respondents, even though it is not an officially used legalterm in France.