The Figure of the Immigrant in Swedish Newspaper Coverage:

(1)

The Figure of the Immigrant in Swedish

Newspaper Coverage:

a critical discourse analysis of immigrant representations in Svenska

Dagbladet and Dagens Nyheter

MASTER THESIS

Student: Kalina Tekelieva Student nr: s2986825

Degree programme: MA Journalism 2015/2016

University of Groningen

Supervisor/first reader: Ansgard Heinrich Second reader: Elisabetta Costa

(2)

1

Abstract

Sweden has a relatively recent history as an immigration country, but the past decades have witnessed a big leap in the number of people with foreign origin that have come to live in Sweden and it goes without saying that immigration has turned into a hot social topic and has booked places in the front row of Swedish media coverage. The growing significance of the immigration topic therefore became a key incentive for me to research Swedish newspaper discourse on immigration and to find out how immigrants are represented. From a theoretical perspective, the media don’t only act as agenda-setters being able to influence which topics the public considers important, but are also responsible for the framing of information, that is, the promotion of certain attributes which trigger certain interpretations. As a consequence of the process of framing, the media offer specific representations of reality, not unbiased reflections of it. In turn, media representations serve as a basis for the formation of social representations, or the collective cognitions shared within a society. In order to analyse how immigrants are represented by the two papers Svenska Dagbladet and Dagens Nyheter, I did a critical discourse analysis, using Fairclough’s three-dimensional model. I examined the discourse as a text, as a discursive practice and as a social practice and found that immigrants are rarely depicted as active individuals, but rather as a passive multitude on the go. Sweden is often idealized as a land of dreams, willing and prepared to offer immigrants a kind of life they never had in their home countries. Yet, my analysis reveals that immigrants are portrayed as weak and in need of help and support, or in other contexts, as a political tool or an economic factor. However, I didn’t observe outright discrimination or racism in the discourse of the two newspapers. This research, although having a relatively small scope, is important in a context of rising immigration in Europe and Sweden, in particular, because the way the media depict immigrants has an influence on how the public thinks of them. The findings within this study could also provide background for larger-scope research on immigrant representations and such research is necessary, especially in the context of currently rising military and political tensions between countries all over the globe.

Keywords

(3)

2

List of Tables

Table 1. Number of headlines in SvD and DN in which immigrants are mentioned………...36

Table 2. Number of leads in SvD and DN in which immigrants are mentioned……….37

Table 3. Compound words designating immigrants..………..48

Table 4. Types of sources used by SvD and DN………..……….56

(6)

5

Introduction

It is not before World War II that we can trace the beginning of immigration into Sweden, when the country received a large number of refugees, as it was one of the few not occupied states. After the war, Swedish export didn’t have much competition from the other European countries, most of which had to recover from the devastations of the war, and the Swedish economy was flourishing. In the 1960s-1970s, the Swedish industry and public sector were booming and the large demand for labour was met by an inflow of labour migrants from, most of all, other Nordic countries. The steady economic growth resulted in very low unemployment rates and social care available to all. In the 1980s-1990s, wars and dictatorships in different parts of the world triggered huge refugee waves and many people from Iran, Iraq, Lebanon, Chile, former Yugoslavia etc. sought asylum in Sweden. In 2001, Sweden joined Schengen and opened its borders to migrants from other EU states. The past couple of years since the start of the civil war in Syria in 2011 have seen thousands of Syrians come to Sweden as well.

While in 1960 only 4% of the people living in Sweden were born outside of the country, the share had jumped to 17% in 2015 (Statistics Sweden, 20161). As a reaction to this development, the Swedish government decided in November 2015 to revise its migration policy. On 4 January 2016, a regulation was made effective, according to which all passengers crossing the border from Denmark to Sweden (via the Öresund Bridge) have their identity documents checked and refugees who don’t have identification are not allowed to enter Sweden and seek asylum (Government of Sweden, 20152).

The 2015 asylum figures in Sweden represent the highest number of refugees per capita in Europe (The Land Guide, 20163). Moreover, the high immigration rates are not seen to subside in the coming years, mainly due to instability in some regions of the world (Statistics Sweden, 20164). The recent changes in

(7)

6

Swedish immigration policy also clearly reflect this trend and position immigration as an important factor in the Swedish political, economic, social and cultural climate in the years to come.

Apparently the topic of immigration is growing in significance on many different levels and in that respect, it can’t presumably be absent from Swedish media coverage either. And here we arrive at the core of my study – the way immigrants are depicted by the Swedish media. A central task for the media is to communicate information but we shouldn’t picture media reports as merely relaying information. Instead, they have to handle the immense information flow, so they pick only certain aspects, package them for the audience and it is these ready-made packages, or representations, that the consumer lacking first-hand experience gets.

Previous research on immigrant depictions has shown that only rarely do media reports provide a fair representation of immigrants (Benett et al., 2013), giving way instead to an unbalanced and discriminatory discourse on immigration. At the same time, articles published by some online news outlets (ex. Nyheter Idag, Fria Tider, Avpixlat, Exponerat) and bloggers (ex. Angry Foreigner) tend to accuse the mainstream Swedish media of adhering to a tolerance and equality narrative, while turning a blind eye to immigration-related problems in the Swedish society and disregarding the rights of the ‘native’ Swedes.

In the context of this opposition, my thesis set out to explore to what extent immigrants are marginalized and discriminated or are rather depicted as full-blooded individuals with equal rights to those of the local people. Using the method of critical discourse analysis and Fairclough’s three-dimensional model (examining discourse as a text, a discursive practice and a social practice), I studied two of the biggest daily Swedish papers and sought to answer the following central question:

How do the Swedish newspapers Svenska Dagbladet and Dagens Nyheter represent immigrants?

To help me scale down the relatively broad research topic, I developed the following sub-questions: SQ1: How is the opposition between immigrants and non-immigrants created?

SQ2: What relations of power and dominance lie under the portrayal of immigrants? SQ3: Who is assigned a passive and, respectively, active role?

(8)

7

a specific political and social event, including events that deal with ethnic relations’. This means that the way immigrants are represented in the news has solid implications for the public perception of immigrants in societies with a diverse ethnical, cultural and religious background (Benett et al., 2013). Accordingly, the way Swedish media portray immigrants is essential for how the topic of immigration is perceived by the Swedish public and the findings of my study could provide important insights as to how immigration is made sense of by the Swedish society. In addition, there is not much research done on portrayals of immigrants in the Swedish media and I hope to fill a part of that gap.

The thesis begins with an overview of key theories and concepts for my study. The first part deals with the agenda-setting role of the media, followed by a section on processes of framing which are responsible for the construction of representations. Media representations are then argued to contribute to the formation of social representations. The end of the theoretical chapter discusses findings of previous research on immigrant portrayal in the media. The following methodology chapter outlines the research design which encompasses a qualitative strategy for collecting evidence from the sample articles, the method of critical discourse analysis (CDA) used to examine the gathered data, the exact operationalization of the analysis based on Fairclough’s three-dimensional method for doing CDA, the validity of my study and its limitations. The findings and analysis chapter, where I present what I have discovered through the analysis, is organized according to Fairclough’s three levels of analysis – discourse as a text, as a discursive practice and as a social practice. In the last chapter, I briefly summarize my work and then reflect on the findings of my analysis. Key issues I deal with are the passive role assigned to immigrants, the collectivization of the immigrant group leading to detachment and blurring of personal lines, the segregation between natives and foreign-born and the disempowering of immigrants. At the end, I give suggestions for future research on immigrant representations in the media.

Theoretical framework

(9)

8

coverage. This is a so-called first level of agenda-setting, whereas agenda-setting has also a second level which deals with the exact way the covered topics are presented. There, I introduce the concept of framing which is engaged with how information in media coverage is packaged, so that certain attributes are made prominent. Media representations are respectively argued to be no mere reflections of reality, but a partial depiction of it. The discussion then moves forward to the crucial role of media depictions in constructing social representations, that is, the role of media representations in the formation of collective social perceptions of a phenomenon. The last section reviews what previous research has found regarding the portrayal of immigrants in the media.

1. Media as agenda setters

Technological advances have made it possible for information to travel across the globe in a matter of instances. Nowadays people can watch, hear and read about distant happenings while sitting at home. So much is out there, just a click away, and yet learning about an event doesn’t equal experiencing it in person. In that sense, people don’t encounter personally most of the happenings and have to receive information through a medium which is external to them. And this is where media come as a core source. It is namely from news stories that most individuals get their knowledge and beliefs about the world surrounding them (Gemi et al., 2013). For example, in a study on the role of media in shaping citizens’ perceptions of climate change, Olausson (2011) finds indeed that a large part of the interviewees acquire their knowledge about climate issues from the media. At the same time, the informational function of the media does not only apply to average citizens – it is also the elites that, struggling to deal with the informational overload, turn to the media to learn about the important facts of the day (Sevenans, 2017).

(10)

9

This being said, however, the media should not only be thought of as an information channel that merely signals the existence of certain events. By means of selecting what to include in the news, media actors direct the attention of the audience to these very events, and thus impose importance on them (Ospina Estupinan and Geraghty, 2017; Weimann-Saks et al., 2016; Moeller et al., 2016). This ability of the media to bring certain topics and events to the front of the public agenda is referred to as the

agenda-setting role of the media (see McCombs 2004). As Coleman et al. explain, large parts of the

public come to perceive an issue as significant, as it gets to be frequently covered in the news, or in other words, ‘the more coverage an issue receives, the more important it is to people’ (Coleman et al., 2009: 147).

Many scholars (see Weimann-Saks et al., 2016; Coleman et al., 2009) trace the roots of the agenda-setting theory back to Walter Lippmann’s book Public Opinion (1922), where he talks of how the media help shape the way people perceive the world around them. However, it was not until the 1970s that the agenda-setting phenomenon was given its name currently in use. In 1968, McCombs and Shaw carried out a study among citizens in Chapel Hill, North Carolina, who had not yet made up their mind about the coming presidential elections. The study, published in 1972, demonstrates a strong correlation between the media agenda of issues and the public agenda of issues, meaning that ‘those aspects of public affairs that are prominent in the news become prominent among the public’ (McCombs, 2005: 543). The Chapel Hill study (McCombs and Shaw, 1972) consequently gave impetus to numerous other studies in various fields and settings that concentrated on the transfer of importance from the news media to the public, or the so-called first level of agenda-setting (see Coleman et al., 2009). Some scholars identify also a further second level, which is discussed later on in this chapter.

(11)

10

evening news to indicate its significance. He adds that importance is also signalled by where in the broadcast the story is placed, how much time is spent on it etc.

The agenda-setting theory presented above is useful in this paper, as it argues that topics that have found their way into the news agenda do play an important role for the public, which has implications for the public relevance of the articles analysed here. However, the so-called first level of agenda setting is just one layer of media coverage. If we assume the ability of the media to make particular events prominent in the news coverage (and hence in the public agenda), then how about the stories themselves – are they neutral reflections of real life transferred to the news? McCombs (2004: 1) maintains that ‘[f]or nearly all of the concerns on the public agenda, citizens deal with a second-hand reality, a reality that is structured by journalists’ reports about these events and situations’. Similarly, Dahlgren (2013: 156) writes that what the media do is shape ‘the way we see the world beyond our own face-to-face experience’. This is where we move on to discussing that the media are actually not capable of providing completely full and unbiased accounts of events. Instead, they construct specific representations of reality, which result from a process of framing of information. These two concepts,

representations and framing, serve as a theoretical backbone for my study and are closely examined in

the following section.

2. Representations and framing

(12)

11

In line with this, let’s draw on an example by Matthes (2012). The issue of abortion could be presented in different ways – for example as an act of killing of an unborn human being, or as an act of free choice. The representation of abortion depends respectively on the way it is ‘framed’ (Matthes, 2012: 249). So, what is framing and how does it affect media coverage?

One of the most often quoted definitions of framing belongs to Entman (1993: 52): ‘To frame is to select some aspects of a perceived reality and make them more salient in a communicating text, in such a way as to promote a particular problem definition, causal interpretation, moral evaluation, and/or treatment recommendation for the item described’. In a similar vein, Reese (2001: 7) refers to framing as ‘the way events and issues are organized and made sense of, especially by media, media professionals, and their audiences’. Tuchman (1976) is often quoted arguing that framing is a tool that journalists use to handle the immense flow of information. There are also some more general approaches to what framing means – Entman et al. (2009) quote Gamson and Modigliani (1987: 143) who put it as ‘the central organizing idea or story line that provides meaning to an unfolding strip of events’. Overall, the essence of the arguments above is that the process of framing presupposes a biased presentation of events and phenomena and in a relation to this, framing in the context of media coverage results in news stories which prioritize certain features over others and organize information in such a way as to promote a particular interpretation of the covered event or phenomenon.

Sometimes framing is also referred to as a second level of agenda-setting (Coleman et al., 2009). Whereas the first level of agenda-setting deals with which topics are central in the media agenda and hence in the public agenda (what people think about), a second level (framing) refers to which specific perspectives and attributes are conveyed, affecting what people think (see Coleman et al., 2009, McCombs, 2004).

In fact, there is hardly any consensus on a single definition of framing (Entman et al., 2009) and nor is there a single scholar or research tradition connected with the concept (Van Gorp, 2005), so I won’t engage much further with defining the term but will move on to discuss how framing actually works, because in a study of immigrant representations in the news, it is important to know what mechanism are driving the framing process leading to the formation of particular depictions.

(13)

12

often used when discussing framing, needs some elaboration. Salience refers to ‘making a piece of information more noticeable, meaningful, or memorable to audiences’ (Entman, 1993: 3). So, by selecting certain aspects and making them more salient, frames have the capacity to perform four functions - define problems (specify what an agent is doing and what the agent loses/gains); diagnose the causes of the problem; give moral judgments of the agents and their effects; and recommend remedies for the problems (Entman, 1993). Respectively, if we go back to Matthes’ (2012) example on the issue of abortion, the different framings of abortion affect how it is explained as an act, how the different reasons to do it are accounted for, whether it is justified or not from a moral perspective and what guidance is given for dealing with the issue.

As mentioned above, salience is a key feature of the framing process, so it is important to explain how salience is practically accomplished. On the one hand, the actual mechanism of framing is connected with repeatedly calling upon the same objects and traits, Entman et al. (2009) argue. They explain that this happens by means of using synonymous expressions in a communicative sequence over a period of time. Consequently, a particular interpretation of a topic or agent is put forward, together with a preferable response to them (Entman et al., 2009). What is more, the three researchers maintain that these reoccurring frames, found in different texts, have a political implication, as they promote a certain interpretation to a ‘politically significant proportion of the citizenry’ who will store these associations for further use (Entman et al., 2009: 177). Once the interpretations are widely stored in the citizens’ minds, there is no need to invoke them again and again and even a single component is enough to trigger a response (Entman et al. give ‘9/11’ and ‘Berlin Wall’ as examples). At the same time, emphasizing some characteristics of reality is only one tool. Choosing what not to mention is another way of framing a message and the omission of potential problems, explanations etc., may have an equally significant effect for the audience (Lück et al., 2017; Brimeyer et al., 2016).

Ospina Estupinan and Geraghty (2017) write that Entman’s interpretation of framing favours a perception of framing as a conscious act, but add that there can be a different perspective as well. They cite Goffman (1974) who explains framing as a process where experiences are given meaning, interpreted, organized and classified, but not consciously but rather as a mechanical and inevitable activity.

(14)

13

‘the framing and reasoning devices in a text with a frame as a cultural phenomenon’ (Van Gorp, 2005: 487). Scheufele (1999: 74) maintains similarly that frames are structures for ‘both presenting and comprehending news’. This means that when making a news report, first, journalists consciously start with a particular frame, then apply frames from ‘the repertoire of frames’ that are part of the culture and finally, the readers link these elements by interacting with their own perceptions (Van Gorp, 2005: 487). These prior structures of thought enabling the understanding of new information are often referred to as schemas/schemata (see Fiske and Taylor, 1991).

A reference can be made to Entman (1993: 52), who argues that frames have ‘at least four locations in the communication process: the communicator, the text, the receiver, and the culture’. He explains that communicators, consciously or unconsciously, make decisions about what to say and how to frame it and so they create a text which bears certain frames. The receiver has their own frames and perceptions, which may or may not correspond to the frames promoted by the communicator. Finally, the culture disposes of a stock of common frames manifested in the thinking of a social group.

When selecting issues, processing information and producing text journalists are steered by professional frames (Entman et al., 2009) but also by cultural, economic, ideological and political factors (Herman and Chomsky, 2002 as cited in Ospina Estupinan and Geraghty, 2017). At the same time, reporters and news editors are striving to follow the core principle of objectivity and normally they employ framing without consciously intending to promote a certain interpretation or a political position, according to Entman et al. (2009). However, with or without intentions, framing consistently lends itself to exerting power through the text, because media framing is not only a process of sense making, but also of sense giving (Brimeyer et al., 2016).

(15)

14

prominent in the media are prominent in the public agenda. Respectively, the attributes that the media employ in their coverage might easily influence the not so well-informed audience’s perceptions of important matters.

At the same time, this argument is equally valid in the context of immigration. Blinder and Allen (2016: 4) comment that media portrayals of immigrants, asylum seekers and refugees ‘may form the raw material from which members of the public build their understandings of these complex, heterogeneous categories’, or in another formulation, ‘mass media’s discursive construction of “immigration” shapes public perceptions or mental images of immigrants’ (Blinder and Allen, 2016: 5). As my study looks into how two Swedish newspapers represent immigrants, I accept the notion that the attributes chosen by the papers have a considerable effect on the way the Swedish public thinks of migrants. In addition, uncovering how immigrants are framed points as well to the dominant attitudes towards them (see Entman, 1993, 2010). In itself the issue of dominance is closely related to the issue of power, whereas power has to do with what Blinder and Allen (2016: 5) express – ‘discursive construction’.

The concept of discourse is discussed in the methodology chapter but the framing-dominance-discursive construction axis requires a brief introduction already here. Discourse is often referred to as language use (see Van Dijk, 2008; Fairclough, 1992 etc.) and the emphasis is put on the use of language in a social context. In terms of functions, discourse is ‘connected to the “construction” of knowledge about social object, identities, processes, etc.’ (Chilcote, 2005 in Mengibar, 2015: 41). Similarly, Van Dijk (2008) and Bryman (2004) write that discourses have a significant role in reproducing social reality in general, which meanwhile leads naturally to the issue of power. According to Wodak (1999), if we consider language as an action in a social context, then language use (discourse) always entails power relations and ideologies. We find this argument also in Mengibar (2015) – although power doesn’t directly originate from language, language is a tool that can change perceptions of reality and thus, influence the social world (Wodak, 2013). With reference to media discourse, Van Dijk (2000) maintains that it is the most influential form of public discourse and, especially when coupled with a lack of alternative sources, is capable of disempowering (marginalizing, excluding) minority groups, which brings us back to Blinder and Allen’s (2016) argument in the previous paragraph.

(16)

15

argues and insists that citizens’ ability to object to media messages should by no means be neglected. Van Dijk (1993a) also writes quoting, among others, Downing et al. (1990) that media recipients are active and to some extent independent information users and their beliefs are structured by different social and cognitive factors and other sources of influence. Moreover, research on framing effects demonstrates that ‘many if not most individuals resist communications incongruent with their prior thinking’ (Entman, 2010: 392). My stance in this research is that news coverage is indeed vital for the way a phenomenon is perceived but shouldn’t be considered an almighty influence on meaning-making.

To sum up, representations are tools with which the media communicate certain information and viewpoints. They are no neutral reflections, but rather, framed interpretations. Media representations have the power to influence people’s understandings and are accordingly not confined only within the media space. As Morgan (2009) writes citing Potter and Wetherell (1998), media portrayals serve as a basis for the construction of social representations, particularly when the audience is confronted by uncomfortable and unfamiliar issues.

3. Social representations

Moving towards a seemingly more abstract level, namely that media depictions contribute to the formation of social representations, it is necessary first to give more concrete dimensions to the concept. The term social representation was introduced by the French social psychologist Moscovici in 1961. Social representations concern ‘the contents of everyday thinking and the stock of ideas that gives coherence to our religious beliefs, political ideas and the connections we create as spontaneously as we breathe’ (Moscovici, 1988: 214). As specified by Hoijer (2011:4), social representations are about ‘different types of collective cognitions, common sense or thought systems of societies or groups of people. According to Corbetta et al. (2009: 625), social representations act as a ‘map of the semantic field relative to an object’ which makes it easier for people to ‘move at ease and in a coordinated way with other individuals’. For example, these semantic maps help orient individuals and groups what the

left and right in politics are, Corbetta et al. (2009) claim in their study of the relevance and meaning

given by Italians to the political left and right.

(17)

16

comfort. It can also embody evilness or corruption and so on. In other words, the way a phenomenon is represented has implications for how it is perceived. Moscovici emphasizes that thinking of drug use, for instance (or social reality as a whole), differs considerably ‘depending on whether it is viewed and represented as a genetic defect, a sign of family breakdown, a cultural tradition or a substance required for a group ritual’ (Moscovici, 1988: 214). If we refer to the topic of immigration, we can logically argue that the way immigrants are represented (be it as a threat to the national security of the host country, an economic boost, a financial burden, a necessary step towards multiculturalism, etc.) can certainly have an influence on how individuals make sense of the issue of immigration and what attitudes they adopt.

Before moving forward to discussing how the media can contribute to the formation of particular social representations, let me introduce Walter Lippmann, considered by some scholars (see Blinder and Allen, 2016) a predecessor of the theory of social representations. Lippmann (1922: 7) writes about the ‘pictures in our heads’, encompassing ‘the pictures of [ourselves], of others, of their needs, purposes, and relationship’. The American writer, reporter and newspaper commentator explains that people need to create such pictures because ‘the real environment is altogether too big, too complex, and too fleeting for direct acquaintance’ (Lippmann, 1922: 3).

Drawing on Lippmann’s statement that people’s actions are dependent on how the world is imagined, it would be relevant to argue that the way immigrants are imagined determines how people will act upon them. But where do the media stand in this context? Lippmann is not positive about the role of the press in creating the pictures in our heads. He claims that while the press is seen to ‘make up for all that was not foreseen in the theory of democracy, and that the readers expect this miracle to be performed at no cost or trouble to themselves’, research shows that newspapers reflect and intensify ‘the defective organization of public opinion’ (Lippmann, 1922: 7).

(18)

17

that are the first to provide information, Morgan writes (2009). At the same time, the media are not the only factor affecting the construction of social representations. Morgan maintains that according to Moscovici’s social representations theory (see 1961, 1988), the individual and social views of a topic are formed under the collective influence of:

‘mass media framing, individuals’ cognitions about the topic (largely a function of information provided by the mass media, though heavily influenced by the commonsense of an individual’s community regarding similar topics), and interpersonal, everyday communication regarding the topic’ (Morgan, 2009: 33).

The theory of social representations (see Moscovici, 1961, 1988) is also used by Hoijer (2011: 3) as a link between, among others, media and public and more specifically, as a valuable tool in studying ‘how the media and citizens construct societal and political issues’. The media have a role in naturalizing social thinking and constructing collective cognitions, Hoijer (2011) adds. Another formulation that supports the aforementioned arguments is given by Olausson (2011: 283): ‘[s]ocial representations circulate freely in every segment of society’ and can be found in different discursive spaces, among which the media ‘hold a key position’ (Moscovici, 2000 in Olausson, 2011). Furthermore, Rotaru (2016: 103) claims that ‘the mass-media represent the main vector by which the immigrants’ social representations and their cultural identity are structured and formed’. She researches the role of the media in the construction of public discourse related to the integration of immigrants and finds that German media depict Romanian immigrants in a predominantly negative manner, which supposedly triggers a predominantly negative social representation of Romanian immigrants. In this context, if my study finds that Swedish newspapers tend to portray migrants unfavourably, I could argue that immigrants are socially perceived in a negative light.

(19)

18

4. Immigrants in the media

As maintained by Benett et al. (2013), media coverage of immigrants cannot be unified under one common feature, as its tone and balance vary considerably among the different EU states. Furthermore, the way different migrant groups are represented in the media in terms of their country of origin, migration category, religion, age, sex etc. is not identical (Benett et al., 2013). Blinder and Allen (2016) quote different scholars (see ex. Crawley, 2005 and Facchini et al., 2013) who argue that negative media representations of immigrants can trigger anti-immigration sentiments but although they agree that the way media discursively construct immigration does influence how public perceptions are shaped, they maintain that ‘demonstrating causal impact is difficult’ because a model suggesting that inaccurate media portrayal of migrants fuels anti-immigration attitudes is too simplistic and implies that media coverage and public opinion can only be either positive or negative (Blinder and Allen, 2016: 4). This notion is also backed by KhosraviNik (2009: 22), who claims that different newspapers represent issues of immigration in a different manner and categorizing these accounts merely as the most positive or the most negative is an ‘unhelpful simplification’ of their discursive and linguistic characteristics.

Against these more cautious arguments about the existence of bipolar features of immigrant representations in the media, however, there is as well research whose findings have a more either-or character. Some studies observe an increase in investigative journalism, counter-argumentation and the use of immigrants as sources in the reports, Benett et al. (2013) write. KhosraviNik (2009) finds that, as opposed to conservative newspapers, British liberal papers work on picturing refugees, asylum seekers and immigrants in different activities and topics and on accounting for the differences among them. Spoonley and Butcher (2009) point out that mainstream media in New Zealand have been engaged with a more sympathetic and nuanced reporting through the late 1990s and after 2000. They explain this development with a growing engagement of journalists in a situation of a boom in cultural diversity in New Zealand after the end of the 1980s.

(20)

19

immigration as it actually occurs’. Similarly, the sources for news reports dealing with immigration tend to be other than immigrants themselves and the representations of immigrants in these reports are largely based on negative metaphors and framing (Kinefuchi and Cruz, 2015). In a study of visuals picturing immigrant minors in a Spanish paper, Martinez Lirola (2014) concludes that news stories on migrant minors report only on negative facts surrounding them and that visuals in particular do not depict them positively either. As Spoonley and Butcher (2009) emphasize, being a broker between host societies and immigrant minorities, the mass media typically endorse, if not initiate, racist imagery.

While the studies referenced above demonstrate that the media tend to depict immigrant communities in a not very loving manner, it is necessary to give a more concrete dimension to these unfair/unbalanced/prejudiced/negative representations. How do media reports actually promote an incomplete and biased image of non-natives?

One tool is to assign minorities a passive role - ‘things are being decided or done, for or against them’ (Van Dijk, 2000: 40). As an exception, the agency of people from the minorities is only highlighted when they are responsible for criminal activities and other negative acts (Van Dijk, 2000). In the cases of news stories reporting on a positive development connected with immigrants, it is portrayed as an exception rather than a norm, according to Gemi et al. (2013). In addition, the media usually don’t put efforts in investigating the social context of an immigration-related news story, but instead, focus only on controversial aspects of migration (Van Dijk, 1988 in Gemi et al., 2013).

Benett et al. (2013) carried out interviews with journalists from six European countries and confirmed findings from previous research (see the MEDIVA project5), namely that media content provides rather inaccurate group designations of third-country nationals; that migrants are labeled negatively or are presented as victims; that migrants are underrepresented as sources in news reports; that the wider European perspective of immigration is disregarded as opposed to the national and local context. In a similar direction, KhosraviNik (2009) maintains that refugees, asylum seekers and immigrants are rarely recognized with their specific names or qualities unless they can be placed within a negative topos, like a riot or violence. He gives an example with three Kurdish asylum seekers who were identified as such and individualized in British press only after sewing their lips as a sign of protest.

5

(21)

20

Blinder and Allen (2016) analyse a corpus of media reports in British press covering a three-year period and find that asylum seekers, illegal immigrants and economic migrants are largely emphasized in media coverage at the expense of family migrants, international students and workers. Blinder and Allen (2016) find that the focus on illegal immigration matches public perceptions of immigration in Britain but is disproportionate with regard to actual statistical figures. During the period of the study, 2010-2012, there were more foreign students in Britain than asylum seekers, labour migrants, or family migrants, but reports in the media didn’t mirror that pattern (Blinder and Allen, 2016). In other words, British media influence attitudes towards immigrants ‘not through pervasive arguments for or against immigration, but through selective depictions of immigrants and asylum seekers’ (Blinder and Allen, 2016: 35).

The use of metaphors also plays an important role in the portrayal of immigrants in the media. Previous research identifies ‘discursive connections between the rhetoric of immigration and the representations of other human problems such as crime or war’, Cisneros (2008: 569) argues and outlines the following metaphors used to refer to immigrants: disease, infection, criminal, infestation, invader, burden, flood. Cisneros (2008) identifies yet another metaphor in popular media – that of the immigrant as pollutant. He looks into television discourse and compares images of immigration to images of pollution resulting from toxic waste spills. According to Cisneros (2008: 579), the portrayal of undocumented immigrants by networks like Fox News and CNN creates the impression that ‘immigrants [are] collecting like piles of potentially dangerous waste or [are] approaching the viewer as mobile pollutants’. Moreover, immigrants, just like an environmental disaster, are portrayed as posing a danger to the harmony and peace of American society, Cisneros (2008) argues and adds a further dimension to the pollutant metaphor - the spread of immigration, like a toxic waste spill, can be contained, more effectively as a result of individual contributions than as a government response. The metaphorical articulation of immigrants as pollutants constructs racial and xenophobic stereotypes and promotes ‘simplistic understandings of immigration that suggest equally simplistic solutions’, according to Cisneros (2008: 591).

(22)

21

racism in media discourse. Discussing this more extreme position is important in the context of my study, because instances of racism in media discourse are indicative of harsh power misbalance, and disempowerment of minority groups. The following section, therefore, discusses arguments supporting the notion that news media are a crucial contributor to the reproduction of racial and ethnic discrimination.

4.1. Racism in the media

A major point of reference in my review of research on the relation between media and racism is Van Dijk’s (1993a) book on elite discourse and racism. He outlines five elites – political, corporate, educational, academic and media, which contribute to the construction of racial and ethnic inequality by ‘persuasively preformulating the dominant ethnic consensus on ethnic affairs’ (Van Dijk, 1993a: 8). The Dutch researcher explains that contemporary elites often exercise their power and influence through discourse, and more precisely – because they have preferential access to public discourse and can control it, which has consequences for the manufacture of consensus. This consensus, on its part, is vital for the elites to be able to legitimate and maintain their power in general and the dominance of the white group in particular (Van Dijk, 1993a). One important tool to manage the consensus on ethnic matters among the public is to polarize Us and Them - to present Ourselves, the natives, in a positive light and to portray the Others, the non-natives, in a negative context, while at the same time creating a ‘self-image as moral leaders in society’, Van Dijk writes (1993a: 283). Indeed, what can be more useful for retaining power than discrimination of the Other, of the one that could possibly pose a threat to your position as a leader? To link back to the concept of framing, public discourse provides an arena for the elites, and the media among them, to legitimate their dominance by framing the Others as inferior. Even without explicit noticing, the elites are automatically framed as superior - the opposite of the inferior Other.

(23)

22

indirect forms of structural violence, like marginalization, neglect or deprivation of juridical rights. Racism is a discursive practice which helps exercise power by creating prejudice towards immigrants and it no longer builds on biological exclusion, but rather on cultural discrimination, Löwander (1998) argues.

In historical terms, the concept of racism has evolved over time to adapt to changes in society, Morrissey and Sims (2015) write. Referring to the history of the USA, they point out that racism started as legal slavery, then went through the era of segregation (the so-called ‘Jim Crow era’) after the official abolition of slavery, and has now transitioned to a ‘colour-blind’ racism, in Bonilla-Silva’s (2003) formulation. The colour blindness indicates a blurring of the role of race and supposedly puts an end to racial discrimination by de-linking individual encounters with racism from the historical and institutional context of racism (Morrissey and Sims, 2015). However, although many US citizens do believe that equality is achieved, racial discrimination persists, but has only been redressed (Morrissey and Sims, 2015). Leonard and King (2011: 6) similarly argue that despite the recession of blatant racism, racial inequalities today are ‘as perverse and pervasive as they were at the height of the freedom struggles’ - they simply have new forms.

Van Dijk (2000) is also interested in this idea and draws on Barker’s (1981) term ‘new racism’. He writes that while the old racism in the sense of explicit derogation of minorities now can only be found in the extreme right, the new racism plays democratic and denies being racism at all. In Van Dijk’s (2000) words, the new racism moves away from segregation and open violence to more subtle, discursive forms. It tends to be:

‘expressed, enacted and confirmed by text and talk, such as everyday conversations, board meetings, job interviews, policies, laws, parliamentary debates, political propaganda, textbooks, scholarly articles, movies, TV programmes and news reports in the press’ etc. (Van Dijk, 2000: 34).

(24)

23

However, whether we are talking of more blatant or, on the contrary, more subtle forms of racism, it is in any case the elites – politicians, editors, professors, judges etc., that play a crucial role in the reproduction of racial inequality, Van Dijk (1993a) claims. Speaking specifically about the media, while they can function as a channel for other power elites (especially politicians and corporate elites) to mediate their discourse, the media are at the same time autonomous and are no passive mouthpieces for the other powerful institutions (Van Dijk, 1993a). The media have their ‘own power domain’ and actively take part in the construction of favourable self-representation and negative other-depiction (Van Dijk, 1993a: 279).

(25)

24

these discursive processes may be far from discursive, as Van Dijk (2000) points out. Bearing in mind the ‘real-life’ significance of discursive processes, I set out to explore the discourse of two Swedish newspapers and examine how immigrants are pictured. Elaborations of how I went about the analysis, what I found and what conclusions I could draw are found in the next chapters.

Methodology

With the theoretical framework as both a foundation and a point of reference, I could move forward to carrying out the analysis of the immigration discourse of the papers Svenska Dagbladet and Dagens Nyheter. The methodology chapter that follows deals with the nature of my research, the framework for execution of the analysis and the exact method of data collection. In the first part, I explain why a qualitative approach is suitable for the study, I specify the sample I analysed and the criteria I used to arrive at this exact research sample. Then, I justify my choice to use a critical discourse analysis as a technique to collect evidence for the study and I include a brief theoretical overview of the concept of

discourse, which was already introduced in the theory chapter. Next, I outline how I practically carried

out the analysis (that is, its operationalization) and I explain Fairclough’s three-dimensional method for doing CDA which I chose to use. Fairclough’s method studies discourse at three levels – as a text (looks at linguistic features), as a discursive practice (studies the production and consumption of the discourse) and as a social practice (analyses the wider social context), and it allowed me to examine the immigration discourse from the perspective of three different levels, thus yielding richer evidence. The following part justifies the reliability of the current research. However, I use the terminology suggested by Lincoln and Guba (1985) as more suitable for qualitative research and I speak of trustworthiness of my study, instead of reliability. At the end, I necessarily outline the limitations of such an analysis of media representations.

1. Research design

(26)

25

passive and who is portrayed as active; I need to include the social context which helps establish the link between the news coverage and social processes. Respectively, a qualitative strategy, which ‘emphasizes words rather than quantification’ (Bryman, 2004: 266), is guiding the collection and analysis of data in this study. Qualitative research emphasizes on the understanding of the social world (Bryman, 2004) and as I want to study the social context of the discourse on immigration, this research strategy proves useful and appropriate.

Bryman writes that, although not too relevant, an opposition tends to be created between qualitative and quantitative research and following this distinction, the qualitative approach lends itself to studying processes, rather than static conditions, to making deep and rich analyses, rather than supplying hard and reliable data, to examining meaning, rather than behaviour (Bryman, 2004: 287). Research methods typically related to qualitative research include language-based approaches, such as discourse analysis (Bryan, 2004), and the current study employs discourse analysis (in this case, critical discourse analysis) as a method. More on critical discourse analysis and discourse follows later on.

1.1. Research question and sub-questions

It is common for qualitative research to examine a particular case, which means that a single case is used to provide the context needed to answer the research question (see Bryman, 2004). However, my research doesn’t employ a case study design, as I am not interested in a particular event or case but rather in SvD and DN’s discourse on immigration in many different contexts. Instead, I analysed SvD and DN’s coverage through a period of four months and examined all instances of immigrant representations in that period. What I gain with this approach is that the analysis is not restricted to just one single case but stays open for representations of immigrants in relation to various topics.

Where the central research question of my study is:

RQ: How do the Swedish newspapers Svenska Dagbladet and Dagens Nyheter represent immigrants?, the following sub-questions were helping me handle the multi-faceted sample:

(27)

26

1.2. Newspapers

For the analysis of immigrant representations I chose to study the coverage of two Swedish dailies – Svenska Dagbladet (SvD) and Dagens Nyheter (DN). In terms of circulation, they are among the largest Swedish daily papers – SvD’s circulation in 2011 was 185,600, and DN’s – 285,700 (2011 was the last year that the two papers’ circulation was measured by the same company, TS6). In their respective editorial sections, both newspapers declare being politically and economically independent, whereas SvD has a moderate political stance7 and DN’s political alignment is liberal8. I decided to examine the coverage of these two papers, as they have a national reach and a large readership (DN ranked 2nd and SvD ranked 4th in 2011 in terms of readership6). Considering that the way immigrants are represented by the media influences the shaping of public perceptions of immigrants (see the theoretical chapter), analysing the discourse of two of the largest and most influential Swedish papers (Encyclopedia Britannica9) has implications for how a large part of the Swedish public thinks of immigrants.

1.3. Sample and sampling criteria

In order to analyse the way immigrants are represented in SvD and DN’s discourse, I needed access to a database of articles, where I could search with keywords, choose a specific time frame and be able to select in which part of the article the keywords should be searched. The widely used LexisNexis doesn’t include articles in Swedish, so I used instead an online database called Mediearkivet which allows researchers to retrieve Swedish newspaper articles.

I started by selecting the sources - in this case, the print publications of SvD and DN. Then, I had to define the keywords. The current thesis aims to look into the way immigrants are portrayed, but in order to choose the most appropriate keywords for my search, I had to make clear what exact meaning the term immigrant carries and how I should adjust my search accordingly. Encyclopedia Britannica10 defines immigration as a ‘process through which individuals become permanent residents or citizens of a new country’ and respectively, an immigrant is ‘a person who has come to a different country in order

(28)

27

to live there permanently’, according to the Cambridge Dictionary11. At the same time, the context of immigration also encompasses refugees, who are ‘persons fleeing armed conflict or persecution’ (UNHCR12), asylum seekers ‘whose request for sanctuary has yet to be processed’ (UNHCR13) and

migrants who are not fleeing persecution of death threat but choose to move mainly to improve their

lives (UNHCR12). In order to handle this complexity, in my thesis I have decided to work with immigrants as an umbrella term which includes as well refugees, asylum seekers and migrants.

Having the multi-faceted nature of the term immigrants in mind, I had to choose the keywords for my search very carefully if I were to gather relevant and at the same time rich data. At the beginning, I searched with ‘immigrant’, ‘migrant’, ‘refugee’ or ‘asylum’ as keywords, as I wanted to include instances where any of these groups is mentioned. However, there were a disproportionately huge number of articles related to the current refugee crisis in Europe, while my focus is not specifically on refugees. Moreover, the sample consisted of a number of articles which was way beyond the scope of this analysis and my capabilities as a single researcher (around 2,000 articles per newspaper). So, I had to change strategy. I decided to search only with ‘immigrant’ as a keyword (‘invandr*’, which is the base of the Swedish word for immigrant). Swedish as a North Germanic language tends to use compound words and searching with ‘invandr*’ helped me include all possible compounds with this base. Of course, this approach has its disadvantages as well and they are mentioned in the limitations section. I decided to look for the keyword not only in the headline and the lead but in the entire text, because my analysis doesn’t require that immigrants necessarily be the main topic of an article. How the two papers depict immigrants outside the central topic is equally important for my research, because these instances are also part of SvD and DN’s immigrant discourse. In other words, when immigrants are not talked about in the headline or the lead but are discussed in the body of the article, immigrants are nevertheless represented in a certain manner and including these portrayals as well is vital for a study of discourse on immigration.

(29)

28

the north (via the Öresund bridge) are required to legitimize themselves. This move intends to prevent refugees with no ID from entering Sweden and applying for asylum. The new regulation represents a shift in Sweden’s traditionally open policy towards immigration in the country and I consider it important to analyse newspaper coverage of immigration-related topics in the period before and after the government’s revision of its immigration policy. The timeframe I chose is November 4, 2015 – March 4, 2016, that is, two months before the new regulation was made effective and two months after that. Of course, a study of media coverage over just four months can hardly serve to make general statements about representations of immigrants in the Swedish press, but such is not my goal either (see the Limitations section).

With ‘invandr*’ as a keyword and the given time period, I got 311 articles by SvD and 432 by DN. To refine the sample, I manually excluded all articles under 100 words as they are usually front-page summaries of articles that are published inside the newspaper. Furthermore, I removed all duplicates as well as irrelevant articles, again manually. Hence, the sample was reduced to 200 articles by SvD and 311 by DN and the articles were exported from Mediearkivet in two Word documents, one for each newspaper.

Finally, this sample had to be further downsized for the purposes of a critical discourse analysis and so as to be manageable within the format of this thesis. To do this, I selected every 5th SvD article and every 9th DN article in the Word documents. Consequently, the sample was scaled down to 40 + 34 (n=74 articles). The slightly smaller number of DN articles seeks to compensate for the greater length of DN articles as compared to SvD. The articles to be analysed were copied to separate files and each article was named SvD1 through SvD40 and respectively, DN1 through DN34. Hereafter, when citing a concrete article, I use, for example, SvD8, DN28 to refer to the respective newspaper and number of article.

A complete list of the sample articles can be found in the appendix.

2. Method

(30)

29

make contradictions transparent. Teo (2000) argues that critical analysis requires from the researcher to expose the underlying ideologies which have become so natural that they have turned into common sense. At the same time, uncovering these naturalizations is so significant, because once the way of thinking about certain relations starts to feel ‘normal’, it may legitimate specific attitudes, behavior, and policies which often lead to discrimination, dominance and even violence (Verschueren, 2012: 2). Blommaert and Bulcaen (2000: 449) maintain that the role of CDA is not only to display the social dimensions of language use, but also to advocate intervention, that is, ‘empowering the powerless, giving voice to the voiceless, exposing power abuse’.

In other words, at its core CDA provides a platform for studying socially relevant matters, where it doesn’t simply fulfill a descriptive function (Teo, 2000), but rather ‘attempts to intervene into social processes by proposing verbally and in writing possible changes that could be implemented by practitioners (Wodak, 1999: 187). Critical discourse analysis of media coverage, in this sense, does not only serve to indicate a certain state of social affairs, but it also engages with prescriptions which journalists could consider. When immigrant discourse is concerned, CDA helps uncover the wider social context in which immigrants are placed and unmask relations of inequality, and it could potentially intervene in correcting instances of injustice.

When talking of social phenomena and forces, questions of power naturally find their way in the discussion, as ‘[n]o interaction exists in which power relations do not prevail’ (Wodak, 1999: 186). CDA respectively seeks to analyse relationships of power, dominance, discrimination, injustice (see Van Dijk, 1993b; Blommaert and Bulcaen, 2000). Teo (2000) writes as well that CDA helps expose how structures of power are reproduced through discourse. In the case of the current study, I would ideally also hope that my critical approach would serve the purpose to unmask any instances of underlying discrimination which concern the Swedish society, and would help to remedy the injustice.

(31)

30

respective social actors, which requires from my analysis to constantly draw on and refer to the social world.

It is important to have in mind that discourse does not neutrally impart meaning – when talking or writing, people seek to achieve a certain effect (Bryman, 2004). Therefore, discourse should be considered as a form of action, it is ‘much more than language as such: it is constitutive of the social world that is a focus of interest or concern’, as expressed by Bryman (2004: 370). Discourse is in that sense an outcome of specific goals and decisions. It doesn’t appear as a detached consequence, but rather fulfills a certain aim, and it is namely the uncovering of the intention behind that is central for discourse analysis.

More on the interrelation between discourse and the social world can be found in Fairclough (see 1992), one of the key figures in the CDA field. In Fairclough’s formulation, discourse is not only constituted by social processes, but it also constitutes them, or in other words, contributes to their reproduction (see Jørgensen and Phillips, 2002; Teo, 2000). To paraphrase, certain social processes and phenomena find their way into the discourse that talks about them, meaning that the discourse carries and exhibits the social world. At the same time, though, discourse is powerful enough to construct social processes and phenomena. Simply put, discourse and the social world exist in a reciprocal relation. Thus, I want to argue that studying the Swedish newspaper discourse on topics related to immigration is fruitful and important in two directions – it has the potential to uncover the nature of relations within the Swedish society (which shape the immigration discourse), and it also points to how discourse may reproduce or challenge existing structures (see Jørgensen and Phillips, 2002). The analysis can therefore both reveal relations of power and equality (or disempowerment and inequality) that are already characteristic of the Swedish society, and find out what social processes immigrant discourse itself constructs.

3. Operationalization

(32)

31

model is a text-oriented form of CDA used to examine discourse in a social context and in this analytical framework discourse is seen as a communicative event which works on the following three levels:

1) Discourse as a text: the focus is on the linguistic characteristics and organization of the discourse 2) Discourse as a discursive practice: this dimension is concerned with the production, interpretation and consumption of the discourse.

3) Discourse as a social practice: it is important to see how the discourse is related to wider social structures.

Following this framework, I analysed first the linguistic devices that the sample articles utilize. I examined the headlines and leads (where the most information in the article is contained), the use of transitivity (which has implications for expressions of agency), modal verbs (expressing opinions and attitudes), syntactic features, semantic fields (through which a certain meaning is promoted), group designations (that is, how immigrants are depicted as group), rhetorical devices, and pronouns which potentially construct an ‘us vs. them’ opposition. Next, I studied the discursive practice and I looked into what sources the two newspapers have chosen to use in their discourse (that is, which voices are heard in the reports), and I also focused on what other discourses the articles were drawing on (that is, interdiscursivity). As for the third dimension of the analysis, I looked into the social context at large and sought to find larger topics related to immigrants which recurrently appear in the two papers’ discourse. These topics were not predetermined, but emerged as I was analysing the first two dimensions.

Before I explain how I exactly carried out the analysis, I want to note that although Fairclough’s model distinguishes between three separate levels, they are not isolated from one another, and a trait which is found in one dimension can be observed in another dimension as well. This only means that discourses work as a complex unity of the intertwined dimensions in them.

(33)

32

the 1st and 2nd dimension and a sheet for the 3rd dimension with the larger social topics concerning immigrants. As I went through the sample for a second time, I added all colour-coded parts to the respective sheet in the coding trail, which in the end allowed me to have a detailed and at the same time compact look at all the findings. Thus I could systematize the findings in writing and analyse them.

4. Trustworthiness

In order to assess the quality of their studies, researchers use certain criteria, such as reliability, validity and replication. However, while these criteria are typically used in quantitative research, they are not so suitable for qualitative studies, because they are to a large extent connected with measures and their adequacy (Bryman, 2004). Therefore, some scholars have adapted reliability and validity for the purposes of qualitative research, whereas others, like Lincoln and Guba (1985), have suggested alternative criteria.

For the evaluation of my research, I decided to use one of the criteria proposed by Lincoln and Guba (1985, as quoted in Bryman, 2004) – trustworthiness. According to the two scholars, we can hardly talk of a single absolute account of the social world which researchers are to reveal (Bryman, 2004). In this respect, a qualitative study is trustworthy not because it presents an absolute truth, but because it fulfills the following sub-criteria: credibility, transferability, dependability and confirmability (as defined by Lincoln and Guba). In order to ensure that my study is credible (which means the findings can be accepted by other scholars), I was sticking as much as possible to ‘the cannons of good practice’ (as expressed by Bryman, 2004: 275). I sought to achieve transferability of my findings to other contexts (that is, the possibility to be used as a basis for other studies) by providing information on the Swedish society and immigration history in the background chapter and in the discussion. Dependability, on its part, requires that a peer audits the detailed records which the researcher keeps of all phases of the study (Bryman, 2004). In my case, however, this was not done due to time constraints. Finally, I ensured the study’s confirmability (the researcher’s good faith and lack of bias), as I was as objective as possible, striving to isolate my personal beliefs from the study.

5. Limitations

(34)

33

two newspapers over a period of 4 months and the results cannot be representative of the portrayal of immigrants in Swedish media as a whole, even less in a wider European, Western etc. perspective. Sample-wise, the analysis is also restricted only to the print versions of SvD and DN and excludes the online articles, so it is not even relevant to make general statements about SvD and DN’s coverage as a whole. Moreover, as the qualitative approach rejects the norms and practices of positivism and doesn’t rely on numerical data (see Bryman, 2004), it is often criticized for being too subjective. This is indeed a limitation of my study as well, because I can’t possibly carry out a critical discourse analysis without being engaged in the interpretation of the findings. Of course I try to stay as impartial and rational as possible, but I know that I cannot be thoroughly disconnected from my own beliefs and values. Another characteristic of CDA is that it lacks clear guidelines on how it should be performed (Mengibar, 2015). This means that the way I did the analysis here is just one possible way of studying SvD and DN’s discourse on immigrants and I kept that in mind when discussing the findings.

On a personal note, the interest in how exactly Swedish papers represent the figure of the immigrant stems from my Bachelor’s degree in Scandinavian Studies and the extensive background I have in Swedish language, culture and history. However, just as my background helps me understand the context, it is also a hindrance, because I step into this research with a lot of pre-formed perceptions and even stereotypes that I developed during my BA studies. I especially have in mind an idea of Sweden as a welfare paradise standing with its arms wide open for whoever wants to come in and enjoy equal rights with the native Swedes. I realize this ‘burden’ of mine and I admit that my analysis might have been affected by my pre-conceptions to some extent. A counterweight in this situation is the fact that I come from a country which, with the exception of the past couple of years, has always been a country of

emigration and the Bulgarian media have (again with the exception of the past couple of years) rarely

reported on immigrants in the country. So, I am not much charged with previous experience with immigration media discourse.

The Figure of the Immigrant in Swedish Newspaper Coverage: