Identity construction in the media : the Turkish-Dutch community : has it always been like this? : comparing the representation of- and the identity discourses applied to- the Turkish-Dutch community in Dutch newspaper

(1)

Identity Construction in the Media:

The Turkish-Dutch Community

Has it always been like this?

Comparing the representation of- and the identity discourses applied

to- the Turkish-Dutch community in Dutch newspaper media between

2004/2005 and 2016

Master Thesis in Human Geography 2017-2018

Bo Schlömer 11789565 Word count: 39.700

(2)

(3)

(4)

3 Table of content

1. Introduction and relevance………6

2. Theoretical framework………..8

2.1. National identity………...………..8

2.2. Banal nationalism……….11

2.3. Everyday nationalism………...12

2.4. Minority groups in the media………...14

2.5. Conclusion theoretical framework and aim of the present research………15

3. Methodology………...17

3.1. Research question………17

3.2. Discourse analysis………18

3.3. Dutch migration demographics………22

3.4. Selection of sources; newspapers……….22

3.5. Selection of political events……….24

3.6. Selection of articles………..26

3.6.1. Final corpus selection………27

3.7. Coding and code scheme……….29

3.8. Methodological analysis………..31

3.8.1. Linguistic expressions and representation……….32

3.8.2. The language of difference and the language of integration……….33

3.8.3. Homeland deixis and the Turkish-Dutch community………...34

3.8.4. Turkish-Dutch voices and media discourses……….34

3.9. Research guide……….35

4. During the start of the accession negotiations (2004/2005)………36

4.1. De Telegraaf……….37

4.1.5. Conclusion and identity discourses………...47

4.2. De Volkskrant………..50

(5)

4

5. In the aftermath of the military coup in Turkey (2016)………..66

5.1. De Telegraaf……….67

5.2. De Volkskrant………..78

6. The analysis of change………92

6.1. Analysis of change in De Telegraaf……….92

6.2. Analysis of change in De Volkskrant………..98

6.3. Analysis of change in general………105

6.4. Can the change be attributed to the changed political relationship of Turkey and the Netherlands?...106

7. Conclusion………108

7.1. Response to main research question………..108

7.2. Contribution to existing literature and society………...109

7.3. Final reflection and limitations………..109

7.4. Future research………...110

8. Bibligraphy………111

(6)

(7)

6 1. Introduction and relevance

The Dutch national identity was one of the biggest themes in the run-up to the parliamentary elections on March 15, 2017. Almost all political parties expressed their opinion about what the Dutch identity should imply (Trouw, 2017). But the debate did not come out of nowhere. It already gained prominence since the 1980s in the context of increasing debate about the failure of immigrant integration (van Reekum, 2016). According to Van Reeken (2014), the prominence of the national identity debate today is a response to the debates about immigrant integration, Islam and citizenship. Today‟s debate focuses strongly on the differences and contrasts between us and those who are not like us (Van Reeken, 2014).

The discussion and debate around the Dutch national identity and the increasing tension in Dutch ethnocultural diverse society ask for a better understanding of the creation and reproduction of in- and out-groups by the media and the construction of socio-spatial identities. Researchers on banal nationalism, national identities and migrant and minority representation all stress the influence of the media in the creation of in- and out-groups and the construction of identities. This research tries to make the influence of the media on this social and political debate more transparent and better understandable. Additionally, this research can create a better understanding of the position of Dutch people with a Turkish migration background (In the following of this thesis „Turkish-Dutch people‟) in society and the developments in their representation.

An extensive amount of academic literature has been written on the role of the media in the creation and (re)production of national identities, banal nationalism, and on the framing and criminalisation of migrant/ethnic groups by the media. The present research tries to combine these fields of study by looking at the representation of the Turkish-Dutch community in Dutch newspaper media. By doing this, this research tries to fill the gap in the academic literature that is the difficult relation of people with a migration background and the (re)production of the national identity by the media. Additionally, by analysing the representation of Turkish-Dutch community during two different political events that are 11 years apart, this research tries to find out to what extent the representation of- and the identity discourses applied to- Turkish-Dutch people by the media changed in recent years, and to what extent this change can be seen as the result of the changed political relationship between the Netherlands and Turkey.

On the cover of this thesis, the front page of De Telegraaf is shown. The newspaper writes “We are

the boss here”. In combination with the picture of a Turkish-Dutch protester who is bitten by a police

dog and the pictures of the Dutch Prime Minister Mark Rutte and the Turkish President Recep Erdogan, the message is clear: “We are the boss here”. This is a strong and nothing disguising message, but how have we come to this point? And has it always been like this? This is what this thesis is about!

(8)

(9)

8 2. Theoretical framework

This chapter elaborates on the existing literature and knowledge about national identities, national identity construction, banal nationalism, and the representation and evaluation of minority groups in the media. At the end of this chapter, a connection will be made between these different but related research fields.

2.1 National identity

In this paragraph, the origin and existence of national identities will be addressed. Based on the work of Anderson (1983) and Paasi (1996) we can argue that the construction of national identities is the result of the interplay between different factors. In the following of this paragraph, these factors are described and explained. Finally, Paasi‟s (1996) analytical framework will be presented. This framework points out how various territorial identities are constructed in relation to certain social distinctions. The framework will be used to connect the literature on nationalist discourses with the literature on migrant representation and integration in the multicultural society, in the following of this research.

Benedict Anderson (1983), first described nations as „imagined communities‟, this can be seen as the most influential definition of nations since then. He argues that nationalist discourse represents and signifies communities as socio-spatial units. When these discourses are reproduced, these socio-spatial communities are taken as real, especially when they are legitimized by legal means of a state (Paasi, 1996). Hassner (1993, in Paasi, 1996) therefore argues that all political and territorial identities are, in a sense, fictional identities which are connected with imagined communities (Paasi, 1996, p12). A national identity can therefore be understood as the feeling of belonging to a certain imagined community (Hernandez, 2008). Imagined communities are in essence social communities applied to a territorial state. According to Carr (in Paasi, 1996), a social community exists wherever a narrative account exists of „we‟, which has a continuous existence through its experiences and activities. This process through which individual actors and collectivities are socialized as members of specific territorially bounded spatial entities and through which they internalize collective territorial identities and shared traditions, is called „spatial socialization‟ (Paassi, 1996, p 8). What this means is that social communities bound to a territorial state hold the same imagined national identity (Paasi, 1996). National communities are thus limited because their membership is restricted to the nation‟s territory. People who live outside the territory are therefore not part of the national community (Castello, 2017).

(10)

9 The identity construction of a nation often occurs to the background of an external „other‟. The definition of the „other‟ is usually inherent to a spatial dimension because the „other‟ typically lives somewhere else (Paasi, 1996). National identities have to be imagined as unique and people who identify with a certain nation have to feel that they are different than people who identify with another nation (Anderson, 1983). The existence of an imagined national community or national identity means that there are people who are part of it and people who are not part of it; people that are part of the „we‟ and people who are part of the „other‟ (Billig, 1995).

Knight (1984, in Paasi, 1996) argues that there are also various groups within states, which do not have an attachment to their supposed national territories. These people thus live in a national territory but do not identify with the national identity of that territory, and thus do not feel as if they are part of the corresponding national community. Examples of this are; people who only identify with their region, they may want to create a new nation-state or people who identify with another nation-state. This last group of people thus lives in the territory of a certain state but identifies with the national identity of another state (Paasi, 1996).

Paasi (1996) constructed a framework or scheme based on the statements above. This framework points out how the construction of various territorial identities takes place in relation to certain social distinctions. National identities have to be seen as socio-spatial identities. Their construction is the result of the identification of „we‟ and the „other‟, and the identification of „here‟ and „there‟. The scheme can be used for a more detailed analysis of the roles of specific discourses. The framework is shown in Figure 1, and works with two variables. The vertical axis represents the social distinctions; „we‟ and „the other‟; insiders and outsiders. The horizontal axis represents the location; „here‟ and „there‟; inside the territorial state and outside the territorial state. The framework results in four fields that represent different identity discourses. The first field (top-left), points to the internal integration within territorial states. This discourse represents „we‟ in a territorial state; a social group that is seen as part of the nation and that is living inside the territorial boundaries of a state. The second field (top-right) represents the discourse of „we‟ outside the territorial state or of „we‟ that live in different territorial states. The third field (bottom-left) represents the discourse of „distinction within a territory‟. This means that there are „others‟ inside the territorial state. In this discourse, a distinction is made between different social, ethnic or cultural groups inside a territorial state. An example of this is refugees or possibly even people with a migration background that are seen as „the other‟ within a territory or state. The fourth field (bottom-right) represents the classical perception of „the other‟. This discourse

(11)

10 points out the social and territorial distinction between different territorial states. In this discourse, people living in another state are seen as different because they are seen as culturally or socially different and because they live in another territorial unit (Paasi, 1996).

Figure 1: Framework of territorial identities in relation to social distinctions.

Source: Paasi (1996).

Language plays an important role in the construction of the social world. Language can construct and reconstruct the social and political reality whereby it can be used to maintain the identity cohesion of a group (Paasi, 1996). Furthermore, we can argue that the construction of identity is based on two languages; the language of difference and the language of integration. Both are important in the construction of a national identity because there can only be a „we‟ if there is a „them‟ (Paasi, 1996).

Most societies are divided by class, gender, age and ethnicity. Power is therefore contested between these groups. Power-holding actors, groups and classes within society define the social and spatial limits of membership and the dominant ideologies of a nations structure the limits of discourse in society. The dominant ideology is thus (re)produced by the power-holding actors. Because the media is an important vehicle for the reproduction of the dominant ideology, journalists are power-holding actors (Paasi, 1996). Nationalism and the idea of national identities can be seen as dominant ideologies in western countries. The (re)production of territorial distinctions and identities is therefore partly done by the media (Paasi, 1996).

National identities can be seen as either homogeneous or heterogeneous. A national identity is homogeneous when all nationals share the same characteristics such as language, ethnicity, values and practices. Contrarily, a national identity is heterogeneous when different groups with different characteristics co-exist in a diverse and multicultural national identity (Falomir-Pichastor & Frederic, 2013). How an identity‟s heterogeneity is perceived can be affected by the extent to which the media emphasize diversity and differences. This has an effect on the way in-group members perceive in-group and intergroup relations (Falomir-Pichastor &

(12)

11 Frederic, 2013). The following paragraph will elaborate on the concept of banal nationalism. In this paragraph, the role of the media as the (re)producer of nationalism will be further addressed.

2.2 Banal Nationalism

A concept that is closely related to national identity and identity discourses is that of banal nationalism, first introduced by Michael Billig in 1995. The concept argues that nationalism is in the everyday life taken for granted, which means that it is not seen as an ideology or something that is contested (Flint & Taylor, 2011). In his book; Banal Nationalism, he argues that nationalism is part of the everyday life in all societies. The nation and nation-state are naturalized as something that is unquestioned and obvious. With the usage of coins, flags and other symbols that are related to the nation, the nation is „flagged‟ continually (Flint & Taylor, 2011). This concept of „flagging‟, points out that nationalist symbols constantly remind us of our nationality and that we are part of a state, without people noticing. This is the most outstanding characteristic of banal nationalism; the consistent reproduction of nationhood through daily life (Köse & Yilmaz, 2012). The presence of these reminders in our daily life, create the acceptance of the nation as our nation (Taylor & Flint, 2011). Because of this, nationalism is part of our daily lives and isn‟t considered nationalism per se, but is felt as a natural state of society (Köse & Yilmaz, 2012). This also reminds us of the fact that we live in a world of nations (Antonsich, 2015).

The flagging of the nation and thereby banal nationalism, have a strong presence in the media. An example of this is the usage of words as „we‟ and „us‟, by which we are constantly reminded of the fact that we are part of a nation. By doing this, the media play an important role in the daily reproduction of nationalism and nationhood; they frame the nation as the nation where the readers belong (Köse & Yilmaz, 2012). The opposite is true when the media uses words as „they‟ and „their‟. These words frame people from another country as others or outsiders. The media therefore play an important role when it comes to the identification of „us‟ and „them‟; or insiders and outsiders (Billig, 1995). This shows how the „groupness‟ of nations is something that is socially constructed (Jones & Merriman, 2009). According to Billig (1995), this groupness of nations is something that needs to be forgotten through the continual identification of „us‟ and „them‟. But although such groupness is the product of prevailing frames, it is not necessarily a measure of groupness felt and experienced by the participants (Jones & Merriman, 2009).

(13)

12 When we connect Billig‟s notion of banal nationalism to Paasi‟s (1996) framework presented in figure 1, we can see that Billig‟s notion of banal nationalism only account for two of the four fields in the framework. Or in other words; only two types of discourses are possible in Billig‟s perception of nationalist discourse. These are the discourses of the national „we‟, and the discourse of the foreign „them‟. In figure 1, these are the top-left field of „integration within a territory‟ and the bottom-right field of „distinction between us and the other‟. In the concept of banal nationalism, the nation is flagged thru the usage of coins, flags and other symbols that represent the nation (Billig, 1995). This can be seen as a nationalist discourse of integration within a territory because these symbols remind citizens of their belonging to a state and their relation with other citizens of that state. Additionally, politicians and the media use language to identify the „we‟ and „them‟. But because Billig‟s notion of nationalism is state-centric (Antonsich, 2016), these „we‟ and „them‟ have to be found in different states. Thus when Paasi‟s framework is applied to Billig‟s banal nationalism, there can only be a „we‟ that is „here‟, and „another‟ that is „there‟.

2.3 Everyday nationalism

A growing amount of literature today is focussing on the idea of “everyday” nationalism. The idea of everyday nationalism came about as a result of criticisms on Billig‟s notion of banal nationalism. Firstly, multiple scholars have argued that the distinction between banal and hotter forms of nationalism is empirically questionable (Jones & Merriman, 2009; Benwell & Dodds, 2011; Antonsich, 2015). The questionable distinction between hot and banal nationalism has to do with two other points of critique addressed by multiple scholars on Billig‟s notion of banal nationalism. Jones & Merriman (2009) for instance argue that the wider political, cultural and economic context in which nationalist discourses are viewed, decisive is for the interpretation of those discourses. Furthermore, Billig works with the unrealistic idea of a homogeneous national audience (Antonsich, 2015). In reality, most western countries are increasingly ethno-cultural diverse (Jones & Merriman, 2009). This is why, among others, Antonsich (2015), Jones & Merriman (2009) and Culcasi (2016) argue that researcher should pay more attention to the role of human agency as the receiver of the information. Additionally, “Billig tends to treat people as being passively and unconsciously exposed to banal national „flagging‟, failing to discuss how individuals deliberately make nationhood” (Antonsich, 2015, p 33). The exact same statement can therefore be viewed and interpreted by different agents differently, as hot or banal (Jones & Merriman, 2009; Antonsich, 2015).

(14)

13 Another point of critique that should be addressed is that of Brubaker (2016), he argues that the presence of nationhood in everyday life varies over time. This is contrary to Billig‟s notion of banal nationalism; he assumes that nationhood is infinite in time and space (Antonsich, 2015). The actions of the state can also influence the presence of nationhood and the national identifications of people. Especially the role of the state in various conflicts influences people in their national identification (Paasi, 1996).

The critiques presented in this paragraph, ask for further research into the position of ethno-cultural diverse groups in nationalist media discourses. When we acknowledge the fact that most western countries have an increasingly ethno-cultural population, we can argue that there also are a „we‟ and „them‟ inside the same territory. This can be connected to the bottom-left field of Paasi‟s analytical framework, presented in figure 1. This field represents the discourse of „distinction within a territory‟ and is about the „other‟ that lives „here‟.

As described in this paragraph, the notion of everyday nationalism entails multiple points of critique on Billig‟s notion of banal nationalism. These are, to summarize, the problematic distinction between banal and hot forms of nationalism, the neglected role of human agency, and the unrealistic idea of a homogeneous national audience. Most research performed in this field of study aims at a better understanding of the role of human agency as the receiver of the banal nationalist information. Antonsich (2015) for instance, studied the role of human agency by group discussions and written essays of an ethno-cultural diverse group. Calcusi (2016) studied the connection between hot and banal nationalism by analyzing national maps and interviewing Jordanians about their national identity. Benwell & Dodds (2011) researched Argentine territorial nationalism by analysing the attention placed on the Malvinas/Falklands dispute by the Argentinian government, and by interviewing young Argentinians. These studies all have one thing in common; they take on a bottom-up perspective to banal nationalism. The studies collectively offer a rich inside in the role of human agency but do not address what their critiques imply for top-down research in banal nationalism.

I therefore argue, in the present study, that the notion of everyday nationalism and the critiques on banal nationalism on which it is based offer more than only research into the role of human agency in a bottom-up manner. In this study, I try to connect the literature on national identity formation, nationalist discourses, and banal nationalism to the literature on the representation of minorities and migrants in the media. The next paragraph addresses the existing literature in the field of migrant and minority group representation in the media.

(15)

14 2.4 Minority groups in the media

People rely on information to form their attitudes about migration and migrants, and the media is an important source of information (Boomgaarden & Vliegenthart, 2009). Boomgaarden & Vliegenthart (2009, p 518) argue that; “News media provide an „information environment‟ in which certain issues and perspectives are dominant compared to others”. The information flows produced by the media on different subjects or topics, can be biased. News stories can emphasize certain ideas or opinions about immigrants and thereby encourage people to think in the same way. Even people who are not exposed to the media will be affected because the news gets transmitted via interpersonal communication (Weimann & Brosius, 1994). The media thereby influences people in their categorization of others. By highlighting the ethnicity of news subjects, the media can produce and contribute to the identification of in-groups and out-groups. The media thus has a strong influence on the public opinion (Boomgaarden & Vliegenthart, 2009).

Boomgaarden & Vliegenthart (2009) researched whether news coverage of immigrants and immigration issues relates to anti-immigrant attitudes. They found that the framing and evaluation of immigrants in the media influence the attitude towards immigrants and immigration; a negative evaluation of immigrants in the media boost, negative attitudes of people towards immigrants. Additionally, they found that the context in which the media covers immigrants matters; in times of high immigration levels, the influence of negative immigrant evaluations is stronger (Boomgaarden & Vliegenthart, 2009). Based on these results, they argue that the media might, by covering immigrant minorities in a certain way, promote or restrain social cohesion (Boomgaarden & Vliegenthart, 2009, p 537). Vergeer, Lubbers & Scheepers (2000) researched the exposure to newspapers and the attitudes to ethnic minorities. They found that newspapers differ in their coverage about minorities. De Telegraaf reported extensively on minorities and crime, particularly when compared to De Volkskrant. Furthermore, they found that people who read De Telegraaf perceive ethnic minorities as more threatening than people who read De Volkskrant (Vergeer, Lubbers & Scheepers, 2000). Based on this, we can conclude that the media coverage on ethnic minorities and migrants influences people in how they see these groups. People, who are more often exposed to negative evaluations or representations of ethnic minorities, are more negative about minorities.

Much of everyday talk between citizens about minorities is inspired by the media. In these talks, people tend to refer to television or the newspaper as their sources. This is especially true for topics that can‟t be observed in everyday life and thus for ethnic affairs for the

(16)

15 following reasons: First of all, most white readers have few daily experiences with minorities whereby they have limited alternative sources of information about minority groups. Secondly, dominant media discourses on ethnic affairs are consensual because minority groups do not have enough power to oppose possibly biased news. And thirdly, ethnic issues provide a positive identification for white readers because the media creates a polarized image, in which a positive self-presentation and a negative other-presentation are usual (Van Dijk, 2000, 37; 2002,).

Based on these findings, we can conclude that the media has a big influence on the attitude of people towards ethnic minorities and migrants, especially on white readers of the majority group. Negative evaluations of migrants and minority groups in the media result in increasing negative attitudes towards these groups in society. The media can highlight the ethnicity of news subjects, whereby they contribute to the identification of in-groups and out-groups. Additionally, the media often presents a polarized image, in which a positive self-presentation and a negative other-presentation are conventional.

2.5 Conclusion theoretical framework and aim of the present research

News reporters see themselves as offering a window to the world beyond our direct surroundings. They are presenting an on-going narrative about what is happening in the world. This narrative is presented as being true (McNair, in Crawford, 2012). The idea of reporting as a window to the world is seen as naïve today. Analysis of western media has shown that the selection of news stories and the language used in news stories serves to maintain the dominant ideology (Crawford, 2012). Banal nationalism in the media can be seen as a form of this maintenance of the dominant ideology. In this case, nationalism, nationhood and the division of the world in sovereign territorial states is the dominant ideology. This ideology entails certain power structures in society.

Research into the influence of the media on people‟s attitude towards immigrants and minority groups has shown that the media do have a strong influence on the opinion of people about immigrants and minorities. The media decides what and how they report on certain issues, and this can be biased. By emphasizing certain opinions or ideas about immigration or minority groups, they can influence the public opinion. And by highlighting the ethnicity of news subjects, the media produces and contributes to the identification of in-groups and out-groups. The media thus influences people‟s categorization of others (Boomgaarden & Vliegenthart, 2009; Vergeer, Lubbers & Scheepers, 2000). When we connect these findings to the literature on national identity discourses, we can argue that the media also play a

(17)

16 significant role in the identification of who is part of the national community and who is not. The media thus is of significant influence on the opinion and perception of people about others.

In paragraph 2.2 and 2.3, the existing literature on banal nationalism and everyday nationalism is addressed. The nation and nation-state are naturalized as something that is unquestioned and obvious. With the usage of flags, coins and other symbols the nation is flagged continually. The flagging of the nation reminds people of their nationality. The media play a significant role in the flagging of the nation. Additionally, they use words as „we‟ and „us‟ by which they remind the reader that he or she is part of the nation. On the contrary, the media uses words as „they‟ and „them‟ by which people from other nations are framed as outsiders. In paragraph 2.3, we addressed everyday nationalism, which can be seen as a subconcept of banal nationalism and is based on multiple points of critique on the original notion of banal nationalism. Here it is argued that the context of increasing ethno-cultural diverse societies means that there are also „us‟ and „them‟ inside the same country. Additionally, Brubaker (2016) argues that nationhood varies over time. The actions of the state influence the presence of nationhood and the national identification of people. Especially the role of the state in various conflicts influences people in their national identification (Paasi, 1996).

In paragraph 2.1, Paasi‟s analytical framework was described. This framework points out how the construction of various territorial identities takes place in relation to certain social distinctions. National identities have to be seen as socio-spatial identities, and their construction is the result of the identification of „we‟ and the „other‟, and the identification of „here‟ and „there‟. This framework results in four fields which represent four different discourses of national identity. Based on the literature on banal nationalism, everyday nationalism and the representation of immigrants and minority groups in the media, we can argue that these discourses are also present in the media and thereby influence the public discourse on national identity.

(18)

17 3. Methodology

3.1 Research question

In light of the Dutch national debate about the national identity and the tension in Dutch society between those who see themselves as part of the nation and those who are seen by others as not part of the nation, it is interesting to investigate how the media evaluates Dutch people with a migration background. Because Turkish-Dutch people form the biggest minority group in the Netherlands (CBS, 2016), they are an interesting case to investigate minority representation and nationalist identity discourse on minority groups in the media. Two political events that influenced the relationship between the Netherlands and Turkey will be used as the case for this investigation. These events are; the start of the accession negotiations between Turkey and the European-Union in 2005, and the military coup in Turkey in 2016. As Paasi (1996) argues, the role of the state in various conflicts influences people in their national identification, in which the media plays a significant role. Because of their migration background, Turkish-Dutch citizens can be identified by the media as Dutch or Turkish and can be connected to the location „here‟ and „there‟. By using Paasi‟s framework of the construction of territorial identities in relation to social distinctions, this research will try to identify which national identity discourses were used by the media during both political events. In this research, multiple sub-questions will be used to answer the following overarching research question:

To what extent has the representation of- and the identity discourses applied to- the Turkish-Dutch community in Turkish-Dutch newspapers changed between 2004/2005 and 2016, and to what extent can this change be attributed to the developments in the political relationship of the Netherlands and Turkey?

(19)

18 To answer this research question, we first have to investigate the representation of- and the identity discourse applied to- the Turkish-Dutch community in 2004/2005. As well as the representation of- and the identity discourses applied to- the Turkish-Dutch community in 2016. When these questions are answered, a comparison can be made between the media discourses of 2004/2005 and 2016. The following three sub-questions therefore have to be answered:

 How was the Turkish-Dutch community presented in newspapers articles, and which identity discourses can be identified in this representation, in 2004/2005?

 How was the Turkish-Dutch community presented in newspaper articles, and which identity discourses can be identified in this representation, in 2016?

 To what extent has the representation of- and the identity discourse applied to- the Turkish-Dutch community changed between 2004/2005 and 2016?

When these questions are answered, the last sub-question can be answered:

 To what extent can the possible changes in the representation of- and the identity discourses applied to- the Turkish-Dutch community be attributed to the changed political relationship of the Netherlands and Turkey?

3.2 Discourse analysis

To answer the research question, stated above, a discourse analysis will be carried out. In this paragraph, the methodology of discourse analysis will be explained. Based on this methodology, this paragraph will describe how the discourse analysis will be operationalized in the present research. In the following paragraph 3.3, the Dutch migration demographics will be described. Paragraph 3.4 will address the selection of political events. Paragraph 3.5 will describe which newspapers have been selected for analysis, and why they are selected. In paragraph 3.6, the corpus selection will be described. This paragraph will explain how the news articles for analysis were selected and why certain decisions regarding the selection were made. In paragraph 3.7 the coding procedure will be described and the codebook used for the discourse analysis will be presented. What follows is a description of the methodology of discourse analysis and a description of the operationalization of this method in the present research.

“Underlying the word „discourse‟ is the idea that language is structured according to different patterns that people‟s utterances follow when they take part in different domains of social life,

(20)

19 an example of this is political discourse” (Phillips & Jorgensen, 2002, 1). A discourse can be seen as a particular way of talking about and understanding the world or an aspect of the world. And a discourse analysis is the analysis of these patterns in social life. (Phillips & Jorgensen, 2002). According to Phillips and Jorgensen (2002), “our ways of talking do not neutrally reflect our world, identities and social relations but rather play an active role in creating and changing them” (p.1). This means that by changing discourses, the world itself is transformed.

Discourse can be produced and reflected in all forms of language in use; in writing, talking, and even in cartoons. Phillips and Hardy (2002) argue that text contributes to the constitution of social reality by making meaning (p.4). Thus, to study the social reality, we have to study individual texts for clues to the nature of discourse because it is not possible to find discourse in their entirety (p.5). Discourse analytical approaches share an interest in the constructive effect of language on the social reality. Discourse analysis therefore has to be seen as a methodology instead of simply a method. Furthermore, discourse analysis does not comprise a straightforward set of techniques for conducting research, but it involves a set of assumptions about the constructive effects of language (Phillips & Hardy, 2002, 5). This fundamental viewpoint on discourse reflects a commitment to social constructivism. Social constructivism is a term to describe a wide range of theoretical perspectives regarding culture and society in which discourse analysis is an operationalization method (Phillips & Jorgensen, 2002, 4). Social constructivism and most discourse analytical approaches are based on the following assumptions, according to Phillips and Jorgensen (2002, 5):

 Our knowledge of the world should not be seen as an objective truth. Our knowledge and representations of the world are not reflections of the reality, but rather are products of our ways of categorizing the world and thus products of discourse.

 Our knowledge of the world is a product of history and culture. The ways in which we understand and represent the world are historically and culturally specific. This means that our worldviews and our identities could have been different, and they can change over time.

 Our ways of understanding the world are created and maintained by social processes. Knowledge is created through social interaction in which we construct common truths and compete about what is true and false.

 Our worldview is decisive for which actions become natural and which are unthinkable. Different worldviews lead to different social actions.

(21)

20 Most qualitative approaches take the social world for granted and then seek to understand the meaning of the social world for various participants. Contrarily, discourse analysis tries to explore how socially produced ideas and objects were created in the first place and how they are maintained and held in place over time (Phillips & Hardy, 2002, 6). The starting point in discourse analysis is therefore that reality never can be reached outside discourse; it is therefore discourse itself that is the object of analysis (Phillips & Jorgensen, 2002, 21). The present research tries to explore how the socially produced ideas about the national identity of Turkish-Dutch citizens are created and reproduced by language in news articles. This is contrarily to earlier research performed by, among others, Benwell and Dodds (2011), Antonsich (2015), and Calcusi (2016), who also studied national identity discourses in banal and everyday nationalist discourses, but focused more on the role of human agency as the receiver of the information.

Discourse analytical approaches can have various forms depending on two dimensions; the degree to which the emphasis is on individual texts or the surrounding context, and the degree to which the researcher focuses on the role of power and ideology or the actual process of social construction that constitute social reality (Phillips & Hardy, 2002, 18-19). These two dimensions have to be seen as continuums, which means that every discourse analytical approach can be placed somewhere on both continuums. Because the present study tries to explore the development of the identity discourses applied to Turkish-Dutch citizens by the media between 2004/2005 and 2016, in light of a changing political context, this research has to be placed, more or less, in the middle of the continuum of text and context. The political context will be addressed in paragraph 3.4. This part, however, will only describe the development of the political relationship of the Netherlands and Turkey and will not analyze this extensively. It is therefore fair to argue that the emphasis in the present study is more on the analysis of text than context, in which the surrounding political context will be treated as somewhat given.

The focus of the present study is on the constructive role of language used by the media in constituting the social reality, and not explicitly on the dynamics of power, knowledge and ideology because this research tries to develop a better understanding of the constructive process, in which the media plays an important role. But discourses are not only forms of interaction or social practices, they also express meanings, and may thus influence the beliefs about immigrants and minorities. Discourses often work with positive self-presentation and negative other-presentation (Van Dijk, 2002). According to Paasi (1996), the other typically

(22)

21 lives somewhere else. But as Van Dijk (2002) argues, the other can also be an ethnic group living inside the same country. The negative other-presentation is thus deployed on this ethnic group whereby this group is presented as deviating from the dominant group, which is in power (158). Discursive practices thus contribute to the creation and production of unequal power relations (Phillips & Jorgensen, 2002, 63). Power is in this research thus important in the sense that it is the dominant group in power that, to a great extent, influences discourses. However, the present research focuses mainly on the discourses in media texts and not explicitly on the power relations underlying them.

Discourse analysis and the various approaches to discourse analysis do not comprise a clear set of tools. It is therefore possible to use different techniques to conduct a research. In this case, two perspectives; interpretive structuralism and social linguistic analysis, offer useful styles for empirical research. In social linguistic analysis, the goal is to undertake close readings of texts to create an understanding of how texts work to organize and construct other phenomena, in this case national identities. In interpretive structuralism, the focus is on analyzing the social context and the discourse that supports it (Phillips & Hardy, 2002, 23). But because discourse analysis does not work with standardized methods, researchers need to develop their own approach that makes sense in light of their particular study and establish a set of arguments to justify the approach adopted (Phillips & Hardy, 2002, 75).

In discourse analysis, it is common to use different discourse analytical approaches and supplement them with non-discourse analytical theories about specific social phenomenon under study (Phillips & Jorgensen, 2002, 153). In the present study, fragments of various discourse analytical approaches are supplemented with the non-discourse analytical theories of banal nationalism and territorial identity construction. Paasi‟s (1996) framework of territorial identity formation in relation to social distinctions argues that the territorial identity construction of people is determined on the basis of two variables; location and social distinction. In Paasi‟s framework (figure 1), four identity discourses are distinguished. These four discourses are under investigation in the discourse analysis of media texts, which means that the non-discourse analytical theory of national identity formation is supplementing the discourse analysis in the present research. The notion of banal nationalism as seen by Billig (1995), argues that deixis is an important way of identifying „we‟ and the „them‟. These words signal the position of something or someone as part of a group or not-part of a group, in this case a nation. Because the identification of „we‟ and „them‟ coincides with the variable of social distinction in Paasi‟s framework of national identity formation, the deixis of „we‟, „them‟ or similar words imply which identity is ascribed to someone. In the discourse analysis

(23)

22 approach that will be outlined in the following paragraphs, these words are therefore important components this research will look at.

In the following paragraphs, the site selection, case selection, and the article selection will be addressed. Furthermore, this chapter will describe the research approach that is adopted in this study and will argue why this is the best way of conducting the present research.

3.3 Dutch migration demographics

The most recent figures about the Dutch population divided to migration background are from the first of January 2016. At this date, 12.3% of the Dutch population had a non-Western migration background, and 9.8% had a Western migration background. Almost half of the Dutch people with a non-Western migration background are born in the Netherlands. This group can be called the „second generation‟. The group of Dutch people with a non-Western migration background consists mainly of 4 different migrant groups. These are people from two former Dutch colonies; Suriname (349.000) and the Antilles (151.000), and people from Turkey (397.000) and Morocco (386.000). The two latter groups mainly came to the Netherlands as guest-workers in the 60s and 70s. The four migrant groups are still growing in the Netherlands but only due to newborn children, and thus second or third generation migrants. First generation Turkish, and to a lesser extent Moroccan, migrants are increasingly moving back to Turkey and Morocco (CBS, 2016). As the Turkish-Dutch community is the largest migrant community in the Netherlands, this research will use this community as case and will focus on their position in Dutch society. The Turkish-Dutch community can be divided into first generation migrants and second generation migrants. The first generation Turkish migrants are born in Turkey and migrated to the Netherlands. Second generation migrants are born in the Netherlands but at least one of their parents is born in Turkey. The Turkish-Dutch community consisted, on the first of January 2016, of 190.600 first generation Turkish migrants, and 206.900 second generation migrants (CBS, 2016).

3.4. Selection of sources; newspapers

The focus of this research will be on newspaper media. In the Netherlands, there are ten national newspapers which are issued throughout the country. In order of daily printed circulation, these newspapers are: De Telegraaf, Metro, Algemeen Dagblad, De Volkskrant, NRC Handelsblad, Trouw, Het Financieele Dagblad, Reformatorisch Dagblad, NRC.Next, and Het Nederlands Dagblad. Between the second half of 2016 and the first half of 2017, a total of 1.750.426 printed newspapers were delivered each day in the Netherlands (NOM,

(24)

23 2017). Because of this rather large amount of newspapers, a selection has to be made. Because the present research attempts to analyze the media discourse on Turkish-Dutch people in relation to two political events, and because the wide acknowledgement in literature of the fact that the media influences the public opinion (Boomgaarden & Vliegenthart, 2009; Vergeer et al., 2000), it is important to select newspapers which are read by a large number of people and a wide audience. In light of this, the following two newspapers are selected; De Volkskrant and De Telegraaf. These two newspapers represent two divergent ideological and political perspectives. De Volkskrant can be described as a rather progressive newspaper and De Telegraaf can be described as a conservative newspaper (Vergeer, Lubbers & Scheepers, 2000). Collectively, these newspapers thus reach out to a wide audience and are offering a wide spread of political opinions. De Volkskrant has 260.000 daily readers of their physical newspaper (NOM, 2017). In addition, they state on their website, to reach 1 million people via their website and apps daily. The newspaper describes itself as a newspaper for highly-educated and well-informed readers (Persgroep, 2018).

De Telegraaf is the most read newspaper of the Netherlands, with a total of 462.000 daily readers of their physical newspaper (NOM, 2017). Additionally, they state on their website to reach 50% of the Dutch population via their phone/tablet apps and their website (TMG. 2018). Neither, De Volkskrant nor De Telegraaf, states any sources or research backing their claims about the total number of people they reach via apps and internet. These claims thus can‟t be checked. The only available information about the reach of newspapers thru websites and apps comes from the „Media Standaard Survey (MSS)‟. In this survey, Dutch people of 13 years and older are asked if they have read a newspaper on a computer, smartphone or tablet, at least once in the last twelve months. This study shows that; 27.4% of respondents used a computer at least once to read newspapers, 20.9% used a tablet at least once, and 31% used their smartphone at least once to read newspapers in the last 12 months (Stichting Kijkonderzoek, 2017). This information still doesn‟t say anything about the daily reach of newspapers via their website and apps because it only shows figures about respondents visiting all newspaper sites and apps in the last year. By using websites that measure the internet traffic to certain websites it is possible to make a rough estimation of the daily reach of both newspaper websites. According to Siteworthtraffic.com, Telegraaf.nl has 59.543 unique users each day, and Volkskrant.nl has 36.272 unique users each day (Siteworthtraffic, 2018). These estimations are “roughly”, as Siteworthtraffic states on their own website, they therefore have to be seen as an indication.

(25)

24 It is also important to note that multiple national and regional newspapers in the Netherlands are owned by two big publishers; Telegraaf Media Groep and De Persgroep. The same individual articles can be published in multiple newspapers, by which the reach of an article strongly increases. Even newspapers that are not part of these two large publishers can pay to copy certain articles. For example, an interview with the Dutch politician Asscher was published by ten different national and regional newspapers on the same day (Sedee, 2016). It has to be noted that between 2005 and 2016 the total number of people reading newspapers in the Netherlands decreased. In 2005 the total annual readership of newspapers was 1.426.000.000, in 2015, this decreased to 936.000.000 (MediaMonitor, 2015). This means that the annual readership of printed newspapers in the Netherlands decreased with 34.4% between 2005 and 2015. On the other hand, however, the number of people using apps and the internet increased during this period.

De Volkskrant and De Telegraaf collectively thus represent two sides of the political spectrum and thereby reach out to a wide audience. In total, they deliver 722.000 printed newspapers each day and reach out to roughly 95.815 people via their websites and apps daily. Additionally, newspaper articles in the Netherlands are often published in multiple newspapers, which increase the reach of articles. And as argued by Boomgaarden & Vliegenthart (2009), even people who are not exposed to the news will be affected because news gets transmitted via interpersonal communication.

3.5. Selection of political events

The present study tries to research the development of the framing of Turkish-Dutch people in Dutch media. In order to find out how the media discourse on Turkish-Dutch people developed and which national identity discourses are used to represent the group in the media, two political events are selected. These events are; the start of the accession negotiations between Turkey and the European-Union in 2004/2005, and the military coup in Turkey in 2016. These two political events influenced the relationship between the Netherlands and Turkey. As Paasi (1996) argues, the role of the state in various conflicts influences people in their national identification, in which the media plays a significant role. Because of their migration background, Turkish-Dutch people can be identified as either Dutch or Turkish and can be connected to the location „here‟ and „there‟. The following of this paragraph elaborates on the start of the accession negotiations with Turkey in 2004 and 2005, and the military coup in Turkey in 2016. Additionally, the Dutch social and political context in which these two political events happened will be described.

(26)

25 On the 24th of November 2004, after a debate in the second chamber of the Netherlands, the Dutch government decided that they are in favour of the accession of Turkey to the European-Union if Turkey meets all asked criteria (Kamerstukken II, 23987/29800V, nr. 47, 2005). This decision was preceded by a strong discussion within the Dutch political party; VVD. This discussion led VVD member Geert Wilders to leave the VVD in September 2004 because he could not accept the position of his party about the accession negotiations of Turkey and the European-Union (NRC, 2004). Wilders did not want to vote as his party asked him and thus left the political party and continued as an independent member of the second chamber. He then announced to start a new political party on the right-side of the political spectrum (Trouw, 2004). This political party, the Partij voor de Vrijheid (PVV), was registered on 22 February 2006 and participated in the Dutch second chamber elections of 15 March 2006. The PVV presents itself as an anti-immigration and anti-islamization party. In the first elections the PVV participated, the party got 5.9% of votes and 9 out of 150 seats in the Dutch second chamber. In the following parliament elections of 2010, 2012 and 2017, the party got respectively 24, 15 and 20 seats in the Dutch parliament (Parlement & Politiek, z.d.).

The decision of the Dutch government to be in favour of the EU plan to officially start the accession negotiations with Turkey thus represents a time in which the political relations between Turkey and the European-Union/Netherlands were good. But at the same time, this decision represents the time in which a new strong anti-immigrant party in the Netherlands came into being. The accession negotiations between Turkey and the European-Union officially started on the third of October 2005 (Europa-nu, 2018).

The second political event that is selected for this research is the military coup in Turkey on 15 July 2016 (The Guardian, 2016, 18 July). On the evening of Friday 15 July, officials of the Turkish army started a military coup to overthrow the government of Turkish president Erdogan. The coup leaders stated that their goal was to overthrow the government of Erdogan in order to restore real democracy in Turkey. At 4am, Saturday morning, the coup ended and president Erdogan succeeded in retaining his position (The Guardian, 2016, 18 July). In the aftermath of the military coup, the Turkish government used strong force to capture and detain everyone suspected of any involvement in the coup. Western media reported on detainees being tortured and raped (CNN, 2016, 27 July). In the first months after the coup, the European-Union‟s reaction was tactful. The Union expressed their support for the democratically elected government of Turkey, while diplomatically suggesting that Turkey should respect human rights (Manu, 2017). After the Turkish population voted in favour of a new constitution that greatly expanded Erdogan‟s political power, and reintroduced the death

(27)

26 penalty, the European Parliament decided to stop the accession negotiations with Turkey on 6 July 2017, because the new Turkish constitution meant a move away from the European democratic values (Europa-nu, 2018).

3.6. Selection of articles

The data for this research will be collected using LexisNexis Academic. LexisNexis offers a database of newspaper articles of almost all Dutch newspapers. Both De Telegraaf and De Volkskrant are part of this database, which means that every article published from 1980 onwards can be found in the LexisNexis database. Additionally, it is possible to use advanced search criteria. In the present research, the newspaper coverage around two political events in 2004/2005 and 2016 will be analyzed. The selected political events of 2004/2005 and 2016 different subjects, the selection criteria and search queries for both periods are therefore different. The criteria for both events will be outlined separately.

As stated in the previous paragraph, the accession negotiations between Turkey and the European-Union officially started on the third of October 2005. But in the run-up to the official nomination for full membership, there was already a wide political and public debate about the issue. In the Netherlands, the political discussion led VVD member Geert Wilders to leave his party, which eventually led him to start his own anti-immigrant political party. It is therefore decided to use the period from the 1st of September 2004 to the 1st of November 2005 for the selection of articles. This period stretches from the time Geert Wilders left the VVD to the moment the European-Union officially started the accession negotiations with Turkey. The time of the public and political discussion in the Netherlands is thereby part of the analyzed period. For both newspapers, the same search criteria are used. The criteria are as followed:

1. The article must be published between 1-9-2004 and 1-11-2005.

2. Articles must hold at least five times one of the following terms: „Turk‟ or „Turkse‟ or „Turken‟ or „gemeenschap‟. These terms are Dutch for „Turk‟, „Turkish‟, „Turks‟, and „community‟.

3. Articles can‟t hold the term; „toerisme‟, which is Dutch for „Tourism‟.

These search criteria resulted in the identification of 109 articles published in de Volkskrant, and 47 articles published in de Telegraaf during the period from the 1st of September 2004 to the 1st of November 2005. For both newspapers, all identified articles will be read to evaluate the usability and relevance of each article for the present research.

(28)

27 On the 15th of July 2016, a coup was attempted by the Turkish military against the government of Turkish President Erdogan. As stated in the previous paragraph, this coup was unsuccessful and President Erdogan stayed in power. The aftermath of the coup however, strongly influenced Erdogan‟s politics and thereby the political relationship between the European-Union/the Netherlands and Turkey. The military coup in Turkey was not expected, so that news stories were only published after the event. The period under investigation therefore starts on the 15th of July 2016, with was also the day of the coup. Due to the quantity of news articles published in the aftermath of the event, the selection period will only stretch two months. This means that the selection period for the news articles stretches from 15 July to 15 September 2016. For both newspapers, the same search criteria are used. The criteria are as followed:

1. Articles must hold the term „Coup‟.

2. Articles must hold the term „Turkije‟. „Turkije‟ is Dutch for „Turkey‟.

3. Articles must hold at least one of the following terms: „Turk‟ or „Turks‟ or „Turken‟. These terms are Dutch for „Turk‟, „Turkish‟ and „Turks‟.

4. Articles can‟t hold the term; „toerisme‟, which is Dutch for „Tourism‟.

These search criteria resulted in the identification of 46 articles published in de Volkskrant, and 38 articles published in de Telegraaf during the period from the 15th of July 2016 to the 15th of September 2016. For both newspapers, all identified articles will be read to evaluate the usability and relevance of each article for the present research.

3.6.1 Final corpus selection

Based on the selection procedure and the search queries, outlined above, a total of 232 articles were identified; 87 articles published by De Telegraaf and 155 articles published by De Volkskrant. Because of this large amount of articles, a more advanced selection had to be made. Phillips and Hardy, (2002, 72) stress the problematic nature of choosing between texts for analysis; “the challenge is not to find texts but deciding which texts to choose” (72). The difficulty is how to “identify a manageable, relatively limited corpus of texts that is helpful in exploring the construction of the object of analysis” (72). The subject of analysis in the present research is the Dutch community. News articles that address the Turkish-Dutch community or individual Turkish-Turkish-Dutch people are therefore prioritized above articles that address geopolitical developments or other related subjects because these texts are most helpful in exploring the construction of the object of analysis; the Turkish-Dutch community.

(29)

28 Because this research tries to investigate the change between two separate time periods, I have tried to select an equal amount of articles for both time periods.

The first period under investigation runs from the first of September 2004 to the first of November 2005. The second period runs from the fifteenth of July 2016 to the fifteenth of September of that same year. The two time periods under investigation, differ greatly in length. This difference is motivated by the strongly different publication density on the subject in both periods which is the consequence of the different nature of the political events underlying the periods. The start of the accession negotiations between Turkey and the European-Union had a long run-up period and political progress was slow. Contrarily, the second period starts with the military coup in Turkey, which is a rapidly evolving political event. Due to the rapid developments in Turkey, a lot of news articles are published in this period matching the search queries.

The search queries for the first period resulted in the identification of 156 articles published by De Telegraaf and De Volkskrant. The vast quantity of these articles offered a lot of information about the development of the political discussion regarding the possible start of the negotiations between Turkey and the European-Union and the problematic political issues that played an important role in these negotiations. Because the present research focuses mainly on the Turkish-Dutch community, a lot of these articles are not interesting for this research because they only report on the development of the political and diplomatic relations of the Netherlands, Turkey, and the European-Union, and do not hold information about the Turkish-Dutch community in the Netherlands. Based on this, a lot of articles of both newspapers had to be excluded for analysis. An example of an article that matched the selection and search queries but is not selected for analysis, based on this problem, is an article published by De Volkskrant on December 8th, titled; “Turkije moet Cyprus erkennen; Nederland eist eind aan martelingen”. This article reports on the issue of Cyprus, as a problem that needs to be solved before Turkey can start the accession negotiations with the European-Union. But because this article only reports on international politics and not on the Turkish-Dutch community, the article is not interesting for the present research, and is excluded for the analysis. Another problem with a substantial amount of articles is that they report on internal issues of Turkey or on Turks living in Turkey. These articles are also excluded for analysis because they do not tell anything about the Turkish-Dutch community living in the Netherlands. After the exclusion of these uninteresting articles, and articles that were too short, a total of 19 articles were identified as interesting enough for analysis. De Telegraaf published 7 interesting articles, and De Volkskrant published 12 interesting articles

(30)

29 that will be analyzed. The difference in the number of suitable articles is caused by the fact that De Telegraaf published fewer articles that matched the search criteria and because De Telegraaf published more very short articles of less than 250 words. The final list of articles that are part of the analysis can be found in Appendix 1.

The search queries used for the second period resulted in the identification of 86 articles published by De Volkskrant and De Telegraaf. Based on the same selection criteria as stated above, 26 articles were identified as interesting for the present research; 10 articles of De Telegraaf and 16 articles of De Volkskrant.

3.7.Coding and code scheme

To analyze the selected newspaper articles, a coding system will be used. The codes are based on the theoretical considerations addressed in the theoretical framework of chapter 2, and the practical or operational considerations addressed in paragraph 3.2. The coding of the articles will be done with the help of the AtlasTI program. This program enables the researcher to code the content of articles in a transparent way. The coding of the articles will be done in three procedural steps that will be outlined in the remainder of this paragraph.

The first part of coding will focus on the representation of all actors involved in the newspaper articles under investigation. These actors can be social actors as well as institutional actors. For the purpose of this research, a distinction will be made between actors of different ethnic and national backgrounds. Three types of actors will be distinguished, based on their migration, ethnic and national background; Turkish-Dutch actors, native Dutch actors and Turkish actors. The representation of actors can be studied by looking at the linguistic expressions that refer to these actors, and by looking at the qualities, characteristics and actions assigned to these actors. The linguistic expressions used to refer to certain actors, and the characteristics, actions and qualities assigned to them, thereby form the representation of actors. The result of this first part of coding will be a list of linguistic expressions used for the distinguished groups

Additionally to the coding of the linguistic expressions and actions, qualities and characteristics, the description of the similarities and differences between actors with different migration, ethnic and national backgrounds will be coded. National identities are often formulated against the background of an external other. The media tends to create a polarized image, in which a positive self-representation and a negative other-representation are usual (Van Dijk, 2000, 37).