Self-identifications Explored. ‘Am I Dutch? Yes. Am I Moroccan? Yes’

Both the survey data and the interviews indicate that the Moroccan-Dutch and Turkish-Dutch climbers strongly identify as ‘Moroccan’ or ‘Turkish’. Refuting the idea that ethnic and national identifications are zero-sum in character, the climbers not only identify with their ethnic identity but also with the national identity. They explain how they experience these dimensions as complementary, and deconstruct the identity concept but not the concept of culture. Analysis of the statistical data reveals that, especially among the Moroccan Dutch, identification with a label does not necessarily reflect a specific sociocultural content. The in-depth interviews, which expose large variations in how individuals see their ethnic and national identities, explain these findings. These results warn against assuming responses to identity questions to be all-encompassing reflections of some consistent, coherent sociocultural orientation, as is done in authoritative reports in the Netherlands. Although the interviewees spoke about their identifications in static terms, their stories also contained references to the contextual, relational, and temporal nature of identification, as Chaps. 6 and 7 show.

and of the relation to Dutch society (verhouding tot de Nederlandse samenleving). Although the researchers note that they see ethnic and national identification as independent (p. 84), they connect them in a way that suggests a one-dimensional relation. Their introductory sentence reads: 'Members of the immigrant groups differ in the extent to which they feel Dutch or, in contrast, see themselves as members of the own ethnic group' (Leden van migrantengroepen verschillen nogal in de mate waarin zij zich Nederlands voelen of zich juist tot de eigen etnische herkomstgroep rekenen) (emphasis MS, p. 84). The categories used to report on the respondents' identification also radiate this one-dimensional idea; these are: 'mainly feels as a member of the ethnic group', 'equally member of ethnic group and Dutch', 'mainly feels Dutch' (voelt zich vooral lid herkomstgroep; evenveel lid herkomstgroep als Nederlander; voelt zich vooral Nederlander) (pp. 85-87). Clearly, this chapter is based on three assumptions: (1) Assumption of substantiveness. The articulation of identification by citizens with certain ethnic-minority backgrounds is seen as something societally relevant, and the answer to a single identification question is interpreted in multiple ways; the answer is assumed to reflect many divergent dimensions of identification.
(2) Assumption of difference. Without any explanation, identification with the ethnic label is interpreted in a different way than identification with the label Dutch. The first is interpreted in terms of group membership, the second in reference to a certain image, 'Dutch'. (3) Assumption of zero-sum relation. The relevance of these identity articulations is sought in a comparison. Apparently, the researchers seek the relevance of 'identification' in the fact that that one identity articulation is stronger than the other identity articulation. This contributes to a one-dimensional image of ethnic and national identifications.
The results of this current phenomenological study help nuance these assumptions about ethnic identifications. This chapter contains an introductory quantitative analysis followed by qualitative illustrations. Survey data show that the widespread groupist assumptions are not in line with how second-generation Moroccan-Dutch and Turkish-Dutch respondents answer to survey questions about identification and cultural practices. These results call for an open in-depth exploration in order to understand what identifications mean for individuals and why they identify as they do. They also warn us to be careful with the interpretation of survey answers about identification.
In this chapter I first investigate the strength of their identifications to see if second-generation Moroccan-Dutch and Turkish-Dutch climbers identify with the ethnic labels at all. I analyze if men and women, and higher educated and lower educated, differ in their answers. And I test the assumption that ethnic identification threatens their national identification (Sect. 5.1) I then study what it means when they identify as 'Moroccan' or 'Turkish', or as 'Dutch'. First I check whether it is plausible that identification with a certain label is a consequence of a broader, coherent sociocultural orientation. I analyze the association between identification and sociocultural content in the survey data (Sect. 5.2). Second, I turn to the interview data to see how the participants speak about these identifications (Sect. 5.3). The chapter concludes with a discussion of the findings, reflecting on the adequacy of objectivist views, as they dominate in the Dutch discourse and occur in scholastic literature, to capture phenomena such as ethnic and national identifications (Sect. 5.4).

Identification with the Ethnic and National Labels
Although the participants of the in-depth interviews are university educated, the selection for the statistical analyses also contains TIES respondents with higher vocational education (HBO) ( Table 5.1) to ensure a large enough selection. This is also why the selection of higher-educated ('HE') respondents includes both respondents who have completed their degrees at these levels of education and respondents who are currently enrolled in higher education. Considering the composition of the TIES data, the TIES respondents are generally younger than the participants of the indepth interviews, who are all over 30 years old. The statistical analyses only include respondents whose parents are both born in Morocco or Turkey, to avoid discussions on the effect of having a mixed ethnic background. It turns out that having a mixed ethnic background significantly influences one's ethnic identification (see Appendix C, Tables C.1 and C.2). This is not surprising because for people with mixed ethnic backgrounds, their Moroccan or Turkish origins are only half of their ethnic stories. The effect of a mixed ethnic background is not a theme of this study.

Levels of Ethnic and National Identification
The TIES questionnaire contained several questions about one's affiliation with certain labels. The answers are solely seen as expressions of affiliations with a certain label. The first observation is that the higher-educated Moroccan Dutch and Turkish Dutch indicate that they more strongly identify with their ethnic labels than with the Dutch label. Of both groups, around 80% claim to have a strong affiliation with the ethnic label, whereas around 40% feel strongly Dutch. The answers do not differ between the Moroccan-Dutch and Turkish-Dutch respondents (γ −0.041; p 0.713). 1 The strength of ethnic identification as indicated by the second-generation respondents is nearly equal to the control group's identification as Dutch. As for the latter, the label Dutch does not only connect with their country of residence but also with their ethnic background. We can thus say that ethnic identifications are more or less equally strong for the Moroccan-Dutch, Turkish-Dutch and ethnic-Dutch respondents.
The second observation is that the higher-educated second-generation respondents state a relatively weak identification with the label Dutch. Not only is their affiliation with the Dutch label weaker than with their ethnic label, but their affiliation with the label Dutch is also much weaker than the affiliation indicated by the ethnic-Dutch respondents. This applies to both the Moroccan-Dutch respondents (γ 0.634, p < 0.005) and the Turkish-Dutch respondents (γ 0.688, p < 0.005). Moroccan Dutch identify slightly stronger as Dutch than Turkish Dutch do, but this difference is not significant (γ 0.105, p 0.300). This does not mean that their identifications as Dutch overall are weak, as some 40% of the Moroccan and Turkish participants indicated that they feel Dutch to a strong extent and roughly three-quarters feel Dutch in a neutral or strong way.
In addition, the data show that the responses of those with higher education levels do not significantly differ from those with lower education levels. This means that the difference in sociocultural orientation between lower-and higher-educated individuals as described in Chap. 4 is not reflected in the identifications with the ethnic and national labels. Although the identification with the ethnic labels of the highereducated respondents (HBO+) is slightly weaker than that of the lower-educated respondents, these differences are only small and not significant (Table 5.3). A large majority of both the lower-and higher-educated Moroccan Dutch indicate that they have a strong ethnic identification (both 82%). For the Turkish Dutch lower-and higher-educated, these percentages are 81 and 78%. In their identifications with the Dutch label, the differences are even smaller. Furthermore, a strong identification as Moroccan or Turkish does not preclude identification as Dutch (Table 5.4). Roughly 75% of the higher-educated second generation combine a neutral to strong ethnic and a neutral to strong national identification. Around one-third of the second-generation higher-educated respondents even combined a strong ethnic identification with a strong identification as Dutch. There is no significant correlation between ethnic and national identifications, either among Moroccan-Dutch (r −0.067, p 0.497) or Turkish-Dutch higher-educated respondents (r 0.153, p 0.113).

Gender and Education
Regarding the identification with the ethnic label, the large majority of the highereducated second generation is in unison. Over two-thirds of both ethnic categories claim to identify (very) strongly with the ethnic label. However, this still means that one-third respond that they identify with the ethnic label less strongly. Around 5% do not identify with the ethnic label at all or only weakly. Regarding identification as 'Dutch', both groups show even greater variation. In both groups, around 40% identify very strongly as Dutch and around one-third take a neutral position; 19% of the higher-educated Moroccan Dutch and 26% of the higher-educated Turkish Dutch feel weakly or not at all Dutch. Do gender and education level explain these variations within the two ethnic categories of higher-educated respondents?
Gender and the difference between HBO and university do not explain these variations. Again, focusing on the higher educated with mono-ethnic backgrounds, male participants show similar responses with female participants across all ethnic categories (Appendix C, Table C.3). Differences between men and women in their levels of identification with the ethnic and national labels are small and not significant. Furthermore, no significant differences exist between the responses of the HBOeducated and the university-educated respondents (Appendix C, Table C.4).
The results of this section raise some questions. Apparently, that the highereducated on average have a weaker sociocultural coethnic orientation, as shown in Chap. 4, does not mean that their identification with the ethnic label is also weak. This applies to both the Moroccan Dutch and Turkish Dutch. Apparently, the fact that the Turkish Dutch have a stronger coethnic orientation and the Moroccan Dutch are more strongly oriented towards the broader Dutch society is not reflected in a stronger ethnic identification for Turkish Dutch, nor for a stronger identification as Dutch for the Moroccan Dutch-at least not among the higher educated. Elsewhere, I have shown that this also applies to a selection of TIES respondents that includes the lower educated (Slootman 2016). It seems as if the answers to questions about ethnic and national identification are not simply reflections of sociocultural orientations, as is often assumed. This is further explored in the next section.

Label and Content Among the TIES Respondents
Does identification as 'Moroccan' or 'Turkish' reflect a broader sociocultural orientation, an embedding in an internally homogeneous, externally bounded culture, what Barth calls 'cultural stuff' (1969)? In light of these questions, it is interesting to compare the higher-educated Moroccan Dutch and Turkish Dutch. Given the stronger coethnic sociocultural orientation of the higher-educated Turkish Dutch, based on the idea that identification reflects sociocultural content it would stand to reason that higher-educated members of the Turkish-Dutch second generation identify more strongly with their ethnic label than the Moroccan Dutch. However, as we saw above, the TIES data reveal no difference between higher-educated secondgeneration Moroccan Dutch and Turkish Dutch in how they respond to the survey question about ethnic identification. Can this difference teach us more about the meaning(s) of ethnic identification? Please note that strong associations between label and sociocultural practices do not mean that individuals themselves interpret their identifications in terms of these practices. They just show that individuals with stronger sociocultural orientations also more often stronger identify with the ethnic label. In other words, when trends are revealed based on quantitative data, these findings are still inconclusive about the meanings and interpretations of the individuals themselves. Nevertheless, when revealed patterns are in line with certain models (hypotheses), such as 'identifications reflect cultural orientations', this forms support for the validity of these models. When such patterns are absent, this implies that the original models are invalid, and that alternative models and stories need to be developed.
Before I analyze the relationship between the identification with ethnic labels and sociocultural 'stuff' among the higher-educated Moroccan Dutch and Turkish Dutch, I first describe the variables that are used as indicators of these sociocultural orientations. I selected variables from the TIES database that can be seen as indications of a coethnic orientation: an orientation towards coethnics, towards practices that are associated with the 'Moroccan' or 'Turkish' culture and towards Morocco or Turkey. These variables resemble most of the indicators that Phinney identified as the most widely used indicators of ethnic identity, which are language, friendship, social organizations, religion, cultural traditions, and politics; all express some sort of 'involvement in the social life and cultural practices of one's ethnic group ' (1990, p. 505). Based on the in-depth interviews, I added three variables on morality to this selection. As we will see in Sect. 5.3, some participants described their identifications in terms of mentality. The additional variables are an attempt to include the component of mentality in the quantitative analysis. These three variables reflect three aspects of a 'progressive' attitude. In line with the definition of Dutch identity in terms of progressive standards (described in Chap. 4), many of the participants see more progressive norms as central to 'the' Dutch culture and as antipodal to 'the' Moroccan/Turkish culture. In total, 17 variables were selected for this analysis, organized into four themes (Table 5.5).  Analysis of these selected variables shows that, in support of the data presented in Chap. 4, for most of these variables, the higher-educated Turkish-Dutch TIES respondents on average have a stronger sociocultural coethnic orientation than the higher-educated Moroccan-Dutch respondents, but this is reversed for religious variables (see Appendix C, Tables C.5a and C.5b). Also, the higher educated have a less strong coethnic orientation than the lower-educated second-generation respondents in both ethnic categories (Appendix C, Tables C.6a, C.6b, C.7a and C.7b). Among the higher educated, gender does not significantly influence the coethnic orientations. In both ethnic categories, the differences between men and women are small and for most variables not significant (Appendix C, Tables C.8a, C.8b, C.9a and C.9b). Again, the respondents with a mixed ethnic background were excluded from these analyses, as this dimension affects one's coethnic social and cultural orientation but falls outside the scope of this book (see Appendix C, Tables C.10a and C.10b).

Ethnic Identification Reflecting 'Cultural Stuff'?
The following section unravels the associations between identification-with-ethniclabels and sociocultural practices. The findings are discussed per theme: (a) general coethnic practices, (b) language and social network, (c) religiosity, and (d) progressive norms. For each theme, I first assess how the various sociocultural practices correlate with each other and form coherent wholes and then successively examine the correlations between these variables and the identification with the ethnic labels.

General Coethnic Practices
Analyzing the coherence between the four variables included in this theme reveals that three of the six correlations are significant for the higher-educated Turkish Dutch (Table 5.6). For example, those who watch Turkish television channels more often also attend parties frequented by second-generation youths slightly more often and take part in activities organized by Turkish-oriented organizations more frequently. Note that even though these associations are significant, the correlations are only weak, as the coefficients are all below 0.30. 2 This means that those who watch Turkish television very frequently do not always also visit Turkey very frequently. At most, there is a slight tendency for those who watch Turkish channels more often to also visit Turkey slightly more frequently. For the higher-educated Moroccan Dutch, these four practices show no significant intercorrelations.  An examination of the association between coethnic practices and ethnic identification reveals that among the higher-educated Turkish Dutch, two of the four practices are significantly correlated with ethnic identification. For the Moroccan Dutch, this correlation is significant for only one pair of the practices. Again, these correlations are not strong, with all coefficients below 0.30.

Language and Social Network
Looking at the intercorrelations between the variables on language and social network, we see that the variables show more coherence among higher-educated Turkish Dutch than among higher-educated Moroccan Dutch (Table 5.7). Also, more variables correlate with ethnic identification for the Turkish Dutch. For example, those who have more best friends with Turkish backgrounds are more likely to speak Turkish more often, have slightly better Turkish language skills and slightly worse Dutch skills, and feel slightly more 'Turkish'. These correlations are weak to moderate. Among the Moroccan Dutch, ethnic identification is not significantly associated with these variables. Feeling Moroccan is only significantly correlated to Dutch language skills, surprisingly in a positive way-albeit only weakly.

Religiosity
Among both the higher-educated Moroccan Dutch and Turkish Dutch, religiosity variables show strong coherence which each other, having correlation coefficients exceeding 0.50 (Table 5.8). For the Turkish Dutch, religiosity in all respects-except for wearing a headscarf-significantly correlates with feeling Turkish. Among the Moroccan Dutch, the correlation between religious aspects and ethnic identification is slightly weaker; ethnic identification is also positively correlated with stronger religiosity, but this relates more to emotional than behavioral aspects.

Progressive Norms
The analysis of the three progressive norms reveals a similar picture (Table 5.9). Again, for the higher-educated Turkish Dutch, the three variables form a moderately coherent whole, whereas among the higher-educated Moroccan Dutch, this coherence is largely absent. For the Turkish Dutch, ethnic identification is negatively correlated with a permissive attitude regarding premarital sex for women as well as abortion, but for the higher-educated Moroccan Dutch, ethnic identification is not associated with these norms.

Synthesis
Among the higher-educated second-generation Moroccan-Dutch and Turkish-Dutch TIES respondents, no strong correlation exists between identification with the labels 'Turkish' and 'Moroccan' and sociocultural 'stuff'. When someone identifies more strongly with the ethnic label than someone else, this does not automatically mean she or he also has a stronger coethnic orientation towards specific practices and attitudes. This applies particularly to the Moroccan-Dutch respondents. A stronger identification with the Moroccan label hardly correlates with the variables included in the analysis. Religious identification is the only variable that (at least moderately) correlates with identification as Moroccan. The observation that identification with an ethnic label is not always associated with sociocultural content parallels the findings of studies on other groups in other contexts, such as ethnic-minority groups in Britain (Modood et al. 1997) and Chinese Dutch in the Netherlands (Verkuyten and Kwa 1996). The lack of strong associations between most sociocultural variables suggests that there is no such thing as an entirely shared and homogeneous culture. Of the four subthemes, only religious 'stuff' can be said to form a relatively strongly coherent whole. Language and social network correlate moderately at most, while there is little coherence between the other coethnic practices and the progressive norms. This means that there is much more sociocultural diversity among the higher-educated second generation than is generally assumed in the integration debate. The ideas-prominent in the Dutch integration discourse and implicit in some scholastic literature-that sociocultural practices form coherent sets, that there is 'a Moroccan culture' and 'a Turkish culture' and that people are either totally oriented towards their ethnic culture or 'Dutch' culture, thus do not reflect reality; these ideas appeared particularly inaccurate for the Moroccan-Dutch respondents.
For the higher-educated Turkish Dutch, the picture is somewhat different than for the higher-educated Moroccan Dutch. The Turkish Dutch have a stronger coethnic orientation, and feeling Turkish is associated with a set of (moderately) cohesive sociocultural practices. Although we do not know how these individuals themselves would describe their identifications, for the Turkish-Dutch respondents a stronger identification with the ethnic label tends to be associated with slightly stronger coethnic and religious orientations and slightly less progressive norms. I have shown elsewhere that these conclusions also apply to a selection of TIES respondents that includes lower-educated respondents (Slootman 2016).
The findings show that a groupist perspective is inaccurate for describing people's identifications and their broader sociocultural orientations. Identification with an ethnic label does not necessarily mirror a broader sociocultural orientation, let alone a coherent, bounded culture. Differences between the ethnic categories exist, but do not convey the full story. Large variations exist within certain categories, both between subsections (such as education level and having a mixed ethnic background), and between individuals. These analyses exemplify a more explorative use of statistical methods.

Label and Content Among the Interview Participants
The survey questions on identification are based on the idea that there is such thing as an elemental affiliation or self-identification. Although the following chapters show that how the interview participants feel and label themselves differs between contexts and varies over time, in the interviews the participants also talk as if they indeed have some sort of constant self-image (particularly when the question is posed in this way), although this self-identification leaves room for ambiguities and contextualities. Let All participants expressed, either spontaneously or in response to explicit questions, that they feel Moroccan or Turkish. Also, they said they feel Dutch. All identify in dual terms. Some described these identifications in hierarchical terms, while others did not.
We have seen that ethnic identification does not necessarily reflect a broader sociocultural orientation, at least with regard to the chosen indicators in the TIES database. The question remains: what does it mean when individuals identify in ethnic terms? Let us now turn to the in-depth interviews. How did the higher-educated participants describe what it means to them to feel Moroccan or Turkish? What elements did they mention in their descriptions? These qualitative data offer stories that can help interpret the quantitative findings.
Whereas in the case of the quantitative, structured data, identification with the ethnic labels is easy to separate from identification with the label Dutch, these two dimensions are difficult to disentangle in the in-depth interviews. Accounts of feeling Moroccan or Turkish are interwoven with narratives of feeling Dutch. Descriptions of feeling Dutch are important for understanding what it means for someone to feel (more or less) Moroccan or Turkish-and vice versa. Omitting these reflections on feeling Dutch would distort the descriptions of feeling Turkish or Moroccan. In this section, I explore what participants mean when they say they feel Moroccan, Turkish, or Dutch. Imane listed her 'Dutch' and 'Moroccan' attributes. Like Karim, she referred to mentality, but she also discussed more tangible practices and the lack of a practical and emotional connection with Morocco.
But I have elements of both. My Dutch elements are for example: I can be pretty blunt; I am down to earth. In general, I feel I understand the Dutch quite well. My Moroccan elements are: I am a Muslim, although I have shaped this my own, personal way. And I love Moroccan food.
(…) Look, I was born here, and I haven't been to Morocco very often, and I don't even have really good memories about it. Although… I haven't been there for three years now, and I have started to miss things a bit. Although 'missing' might be too strong a word. Like the colors and smells, and a specific feeling… But I could never live and work there. Furthermore, well… obviously I speak Dutch; and Berber; and Moroccan Arabic. (Imane) In describing his double affiliation, Berkant also referred to the emotional relationship with the countries. Furthermore, he distinguished particular domains in which he feels more Turkish and in which he feels more Dutch.
The thing is… I've also lived in Turkey…. I find -Every time when I arrive in Turkey, I think: 'Great!' The first days are always great. And every time I come back, here in the Netherlands, that feels great as well.
(…) There are separate 'domains'. For example music; Turkish music REALLY moves me; it makes me feel really good. My emotional domain is very Turkish, just as the more personal domain. I have been raised like that. I am not a distant person: when someone is at the door at six o clock, I don't say: 'I am watching the news or I am having dinner, can you return later?' We are inclusionary, I am very Turkish in this way, and I feel good about it. Regarding the business element, I am very Dutch. I am very formal, I can easily separate work and private life. I am the boss here. Look, the Turkish are really -the emotional side -it is hard for them to separate.
(…) Obviously, in some respects, I'm really more Turkish. That is, with emotions, sensitivity, passion. It is like that with -uh -soccer teams… I love wearing orange to a Dutch soccer game as much as I enjoy watching Turkish matches. But the funny thing is, when Turkey wins, this affects me more. Maybe because the emotions are deeper; the Dutch side is always somewhat more formal. The emotions are just slightly different. But that's also -maybe I stretch it too far now… It also has to do with your family, with your roots… How can I say this… -The older you get, the more important your family becomes. It is just this feeling, because your parents -because when I visit my parents, this is my Turkish family; with Turkish traditions. (Berkant) In emphasizing his Dutchness, Adem referred primarily to his practical involvement in Dutch society.
Marieke: And you for yourself? Do you feel Dutch?
Adem: I feel, I do MORE than enough for THIS country, more than the average Dutch person. And I would defend this country MORE than enough. And I DO. So, when THIS is the condition for being Dutch, I am Dutch for one thousand percent. When you refer to the situation of the Netherlands, or the neighborhood where you live, or the Dutch economy… -then I find it really important that the Netherlands is doing well. Because THAT's where I live. THAT's where my children will live. (…) I find it much more IMPORTANT that the Netherlands flourishes than Turkey. My own surroundings are most important. Clearly… Dutch in the sense of interests… community… um… atmosphere, and quality of life… in THAT sense I am Dutch. But when you talk about Dutch culture, then I'm not. These accounts show that self-descriptions vary somewhat between participants, who referred to various attributes to describe what 'feeling Moroccan', 'feeling Turkish', and 'feeling Dutch' means for them. However, a limited number of themes emerge from the participants' self-descriptions. Some themes pop up frequently, whereas others are mentioned less often.
One of the themes mentioned most often is that of mentality. In describing their Dutch side, Karim and Imane both referred to ways of thinking, to a deep level of understanding. They mentioned their down-to-earth mentality and directness, even the 'phony tolerance' (or indifference), which they identify as truly 'Dutch' inclinations. In many interviews, individuality and independence were mentioned as attributes that participants really valued and which for them marked their Dutchness. Many mention their having liberal values and being accustomed to the relative absence of bureaucracy. These characteristics make them realize how Dutch they feel, something which they became particularly aware of when they were in Morocco or Turkey. Several participants mentioned their appreciation of social cohesion, emotions, warmth, and hospitality as typical expressions of their 'Moroccan and Turkish sides'.
The theme of mentality emerged frequently in the interviews among all categories of participants (Moroccan-Dutch and Turkish-Dutch, male and female). It was most often mentioned in descriptions of feeling Dutch and feeling Turkish; only once did a participant mention it when describing feeling Moroccan. It is also frequently used to describe why one less strongly identifies as Moroccan or Turkish. Berkant's account shows that aspects of mentality can be used simultaneously to explain feeling more and less Dutch and more and less Turkish. He used aspects of mentality to describe how he feels more Dutch (his formal business attitude) and more Turkish (his hospitality and emotionality), and also how he does not feel fully Turkish (he is not 'emotional' in the professional sphere). The emergence of mentality as a central component of identification-content led to my inclusion of the 'progressive norms' variables in the quantitative analyses.
Language was also repeatedly mentioned in the in-depth interviews. Apparently, not only one's fluency accounts for its importance, but also the instrumental role of language. In the interviews with the Moroccan-Dutch respondents, language was mostly mentioned as an illustration of Dutchness or as an example that one does not feel fully Moroccan. Like most of the other participants, Ahmed indicates he dreams and thinks in Dutch. His limited knowledge of the language of his parents means that he cannot express his deepest feelings in the Moroccan language, and this constrained his access to information about his Moroccan background. Karim not only explained that he thinks in the Dutch language but also suggested that thinking-in-Dutch for him is related to Dutch-ways-of-thinking. Furthermore, he feels closely connected to the Dutch heritage because he has always read Dutch books. This shows how language can strongly relate to mentality. When Turkish-Dutch participants mentioned language, it always referred to Turkish and was used to describe Turkish affiliations. The difference between the Moroccan-Dutch and Turkish-Dutch participants is that the Moroccan Dutch were more familiar with the Dutch language. In line with the results of the statistical analyses, the Moroccan-Dutch interview participants generally spoke Dutch with their siblings and their coethnic peers, while this was not the case for the Turkish-Dutch participants. The broad usage of Dutch by the Moroccan Dutch might explain the distinct role of language in the accounts of the Moroccan Dutch and Turkish Dutch. Additionally, it could clarify why in the TIES data, feeling Moroccan only correlates with the parental language, whereas feeling Turkish correlates (moderately) with both the parental language and the Dutch language.
When participants described their 'Dutch' and 'ethnic' sides, they occasionally mentioned the bond with the countries, both in emotional and practical respects; this was the case for both Moroccan Dutch and Turkish Dutch, and for men and women. Imane, Karim, and Berkant show in their quoted remarks that they reflected on their relations to Morocco and Turkey. In reflecting on her ethnic side, Imane pondered about not visiting Morocco frequently and considered how she could never live and work there. Karim stated that Morocco 'is not my country'. In describing his Dutchness, Hicham reflected on the emotional bond he feels with the Netherlands.
Look at me: I am very loyal to the Netherlands. It is even that I somewhat feel like a sissy -I don't go on a transfer for a year or do a project abroad, because of the risk that I'll miss the Netherlands. Not only family, but that I'll just miss the Netherlands. It's also loyalty to small things, things you value in the Netherlands - (Hicham) References to Morocco were made in a negative sense to describe that one does not feel fully Moroccan. References to Turkey were generally more positive. Berkant explained that he feels at home both in Turkey and in the Netherlands, affirming his double identification. Adem's quote illustrated that the attachment with the Netherlands can also be expressed in rather practical terms. The Netherlands is important to him because it is the country where he lives, the society he contributes to, and the place where his children's future lies. The lack of a strong correlation between ethnic identification and the frequency of visits to Morocco or Turkey in the quantitative data might indicate that emotional bonds are not necessarily related to visiting the country in practice.
Like the quantitative analyses, the interviews reveal a strong association between feeling Muslim and feeling Moroccan and Turkish. Religion was never mentioned in relation to Dutchness, in either a positive or negative way. While some participants explicitly separate the religious and ethnic dimension and emphasize the prominence of their Muslim identification over their identification as Moroccan or Turkish, most participants describe religion as an aspect of their ethnicity and mention ethnicity and religiosity in one breath. Even those who do not feel strongly religious identify as Muslim because of their Moroccan or Turkish backgrounds. They explain they would never feel (or say) they are not Muslim. The entwinement of religion with their parental culture makes them participate in some religious traditions, as Mustapha explains: Later, I came to see religion as part of your culture again, like -it's just part of Moroccan culture. Some aspects are simply inescapable. You can't really say: I'm not a Muslim, I don't do Islam; because then you actually loose part of your identity. Because some things, like for example the Ramadan, or certain holidays -these are Islamic, but closely bound to culture.
Specific cultural practices were only sporadically mentioned in descriptions of feeling more or less Dutch, Turkish, or Moroccan. This is surprising, considering the emphasis on 'ethnic involvement' in much of the research that Phinney evaluated (1990). This explains the quantitative findings, which show that the 'general coethnic practices' are not, or only weakly, correlated with ethnic identification. When such practices are mentioned in their self-descriptions, participants did not stress participation as much as emotional attachment. When participants describe feeling Turkish or Moroccan, they mention a love of Moroccan food, feeling deeply touched by Turkish music, or becoming (extra) fanatic when a Turkish football team plays. Many of the participants do not drink alcohol. This makes some feel 'less Dutch', whereas for others this not a relevant issue.
Occasionally, the theme of birth and descent popped up. The fact that one is born in the Netherlands is mentioned once or twice to describe that one feels Dutch. In describing her Dutchness, Imane referred to the fact that she was born here. Karim hates being addressed as Moroccan given the fact that he was not born and raised in Morocco. Conversely, the fact that his parents are from Turkey makes Berkant say he feels Turkish.
In the literature, knowledge is presented as another component of ethnic identification (Verkuyten 2005, pp. 198-199). This theme pops up occasionally in the interviews. Ahmed explains that his prior lack of knowledge about Morocco had contributed to his relatively weak identification as Moroccan. For Esra, knowledge about the Turkish and Kurdish political situation heightened her orientation towards Turkey and the Kurdish people. When knowledge is mentioned, it is mentioned as cause for increasing ethnic identification rather than as a component of identification.
What did not pop up in participants' descriptions of their self-identifications is the social network (besides the family). According to Phinney, 'friendship' is regarded as a component of identification in many studies, which is why it is included in the quantitative analyses. However, in the in-depth interviews, friends are not mentioned in the descriptions of ethnic or Dutch identifications. The social environment is not absent from the interviews, but it is brought up as a reason why someone identifies in a certain way rather than as a component of identification. For example, Ahmed mentions that his rather strong 'white' identification is the result of the primarily 'white' social environment of his childhood, youth, and student years.

Synthesis
This section has shown how descriptions of feeling Moroccan, Turkish, and Dutch vary between participants. They describe their self-identifications with the ethnic and national labels in different ways. Nevertheless, from the descriptions various patterns can also be distilled. The identifications were described in terms of mentality, language, ties with the countries, religiosity, certain practices, birth, and descent. The first three themes are most central in the participants' descriptions, as they were most frequently mentioned and emphasized and discussed more emotionally in the greatest detail. Religiosity was not always explicitly mentioned, but for many it is an inherent component of being Moroccan or Turkish. I will also briefly reflect on the relationship between ethnicity and religiosity in Chap. 6. Knowledge and social network were mentioned as causes of certain identifications rather than aspects of identification. The descriptions vary in profoundness, in personal 'depth'. Some describe their identifications in more profound terms, in terms of mentality and emotions, while others describe their identifications in more superficial, instrumental and factual terms, such as residence, descent, or holiday visits. 3 The descriptions clarify why the combination of ethnic identification and identification as Dutch does not pose any problems for the participants; why these dimensions of identification are not essentially zero-sum for them. For example, it is possible to describe one's Dutch side in terms of mentality (for example one's down to earth character and directness) as well as one's ethnic side (for example the level of interpersonal warmth and emotions). While the participants label most individual behaviors and attitudes as cultural traits that are either inherently 'Dutch' or inherently 'Moroccan' or 'Turkish', they do not apply this singular labeling to themselves as persons. As individuals, they are not one or the other; they combine traits that they associate with both sides. They do so in two ways. First, they combine nonconflicting traits ('Dutch' directness, 'Turkish' hospitality, or a love for 'Moroccan' food). Second, they combine traits in different domains: in the professional domain, one can feel really Dutch and value a certain personal distance whereas in the emotional domain or in raising one's children, one can feel really Turkish and value interpersonal involvement. The fact that cultural traits are defined in oppositional ways explains why descriptions of feeling Moroccan or Turkish and feeling Dutch cannot be easily disentangled; remarks about 'Dutch' traits feature in descriptions of feeling more or less 'Moroccan' or 'Turkish' and vice versa.
Despite the differences between individuals, these findings seem to support the idea that Moroccan Dutch and Turkish Dutch identify with their ethnicity in distinctive ways. Even though Moroccan-Dutch and Turkish-Dutch participants describe feeling Dutch in similar ways-in terms of mentality and positive emotions relating to living in the Netherlands-the descriptions of their ethnic identifications differ. Turkish-Dutch participants describe feeling Turkish in more profound terms of mentality and emotions, whereas Moroccan Dutch hardly mention these components when they describe feeling Moroccan. For Turkish-Dutch participants, Turkey and the Turkish language play a larger and more positive role than Morocco and Moroccan languages do for the Moroccan-Dutch participants. 4 This suggests that ethnic identification is more substantive for the Turkish-Dutch participants than for the 3 As one may have noticed, the self-descriptions in this section were phrased both in terms of 'being' and 'feeling'. Do these expressions not refer to essentially different components of identification? Verkuyten (2005) distinguishes 'being' components (referring to ontological aspects, to 'objective' characteristics related to the applicability of the categorization) from 'feeling' components (referring to other kinds of affiliations, such as emotional attachments). However, no such distinctions seem to be made in how these terms are used in the interviews, as they are used interchangeably, both by me and by participants. In the context of the interview, the theme of 'objective' characteristics or ontological arguments appears to be largely irrelevant, as it hardly pops up. It only surfaces occasionally, when referring to the ridiculously exclusivist character of the integration discourse but barely in narrations on self-definitions. 4 Without opening up a new concept and an additional domain of literature on transnationality, here I remark that ethnicity among Turkish Dutch seems to contain more transnational elements than Moroccan Dutch. Although the large variation and the small sample make these findings tentative, the resonance with the quantitative findings, and with other literature, as described in Chap. 4, strengthens this picture.
Considering the gendered ideas on being a 'typical' or 'good' Moroccan or Turk, it is surprising that no clear differences appear in how men and women describe their identifications. For example, both men and women give attitudinal and emotional descriptions, and within both categories, varying significance is attached to their parental country. This echoes the quantitative findings, which also reveal hardly any differences between men and women.

Summary and Reflection
Neither the TIES data nor the in-depth interviews support the idea that ethnic identification is weak for the higher-educated second generation. Higher-educated secondgeneration Moroccan Dutch and Turkish Dutch do not differ from the lower-educated second-generation categories in how strongly they identify with the labels 'Moroccan' and 'Turkish'. Their ethnic identifications as indicated in the TIES survey are relatively strong, surpassing the level of their identification with the label 'Dutch'. This does not mean that their identification as Dutch is weak. Only a small minority of the selected TIES respondents say they identify as Dutch weakly or not at all. A very large majority of the TIES respondents and the interview participants identify both with the ethnic and the Dutch label. In addition, for many interview participants their feeling-Dutch is relatively 'deep'. They describe their Dutch identification in terms of mentality and emotions.
Yet, it remains ambivalent what a strong ethnic identification means for these higher-educated individuals. The TIES data show that ethnic identification is not necessarily associated with coherent sociocultural content. Whereas ethnic identification is associated with a moderately cohesive set of sociocultural orientations for the Turkish-Dutch respondents and is described in relatively profound terms by the Turkish-Dutch participants, this is not the case for the Moroccan Dutch. However, this does not mean that Moroccan Dutch consider their ethnic identity to be less relevant. The ethnic identification of Moroccan-Dutch TIES respondents is equally strong to that the Turkish-Dutch respondents. These findings undermine common assumptions about the substantive content of identifications. The idea that ethnic and national identifications are, in essence, zero-sum in character is proven wrong. Furthermore, the idea that a strong affiliation with an ethnic label necessarily reflects coherent content, a predestined coherent set of sociocultural practices, is greatly nuanced. Large variations exist, both on the level of ethnic categories as well as ethnicity among Moroccan Dutch. I therefore highly contest the inflation of ethnic identification with transnationality, as ethnicity is likely to refer more to having-a-certain-background-in-a-specificcountry than to practices that are related to two countries. The first can contain the latter, but not necessarily so. on the level of educational subsections, and on the level of the individuals. In particular, the case of the Moroccan Dutch shows that identification with the ethnic label does not necessarily reflect sociocultural content. In addition, a strong 'ethnic' identification does not necessarily imply a strong orientation towards the parents' birth country nor does a strong 'national' identification always imply a strong bond with the nation of residence. What identifications mean for individuals cannot be assumed but should be studied. We have seen that self-identifications as 'Moroccan', 'Turkish', and 'Dutch' encompass many different aspects, which vary between persons.

Analytical and Methodological Reflection
This chapter demonstrates the relevance of a consistent distinction between selfidentification-with-a-label and identification-content. It shows that identification with a certain label (for example calling oneself a Moroccan, feeling Turkish, or saying one is Dutch) is not always associated with a specific 'content' (which may be watching Turkish television, praying, or speaking Dutch language with one's friends). A systematic distinction between label and content enables us to problematize and analyze affiliation with a mere label in relation to possible content and reasons for identifications.
This chapter illustrates how quantitative and qualitative methods can complement each other. The quantitative analyses helped us assess the breadth of a phenomenon and compare categories and subsets. While they exposed the existence but particularly the absence of broader societal patterns, the descriptions from the in-depth interviews helped us interpret the quantitative findings. The unstructured descriptions of the identification content help us understand why the statistical findings hardly (in the case of the Moroccan Dutch) or only moderately (in the case of the Turkish Dutch) explain what ethnic identifications mean to the respondents. Part of the reason is that many of the aspects that were brought forward by the interview participants, particularly emotional and evaluative aspects, are not included in the statistical analyses. The personal descriptions focused more on how one values certain habits, whereas the selected variables of the TIES survey focused on the occurrence of practices and attitudes.
The chapter's findings warn us to not take expressions of ethnic or national identification as straightforward indications of broader sociocultural orientations, whether in more-structured or less-structured approaches. The findings also warn against framing identifications, such as in questionnaires but also in reporting, in a way that implies a zero-sum character; for example, when answering options to the question 'Do you feel more Moroccan or Dutch inside?' range from 'completely Dutch' to 'completely Moroccan', without providing an option for indicating that one feels both completely Dutch and Moroccan. 5

The Static and Contextual Character of Identification
How identification is discussed in this chapter suggests that individuals have stable ethnic and national orientations. Questions such as 'To what extent do you feel…?' appear to reflect the notion that identifications are stable and constant. This makes results based on structured surveys often seem to imply that people's identifications are autonomous and static. At the same time, in many of the in-depth interviews (despite my own reluctance, as I explained in Chap. 3), I asked the respondents similar questions. From these interviews, it also appeared that when people are asked in lessstructured ways how they feel in ethnic and national terms, they respond as if they have a stable identification that applies to them in general. Most participants answered the questions using straightforward terms to describe their feeling 'Moroccan', 'Turkish', or 'Dutch', and did not challenge the question.
This suggests that they experience their ethnic and national identifications as static and unproblematic givens-after all, if identifications are experienced as variable and contextual, we would expect the participants to be unable or unwilling to talk about their identifications in static terms. In the following chapters, I show this is only partly the case. In their reflections on their affiliations with the ethnic and national labels, participants often mentioned the influence of the context and developments over time. In Chaps. 6 and 7, I will explore the contextual and temporal aspects of identification, and the relationship between more stable and more contextual views of ethnic identification. Verkuyten, M. (2005). The social psychology of ethnic identity. Hove: Psychology Press. Verkuyten, M., & Kwa, G. (1996). Ethnic self-identification, ethnic involvement, and group differentiation among Chinese youth in the Netherlands. Journal of Social Psychology,136,[35][36][37][38][39][40][41][42][43][44][45][46][47][48] Open Access This chapter is licensed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made. The images or other third party material in this chapter are included in the chapter's Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the chapter's Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.