Swedish picture of Russia

(1)

Swedish picture of Russia

Analysis of media coverage of Putin, Russia and Russians in Swedish media

By: Diana Bologova

Supervisor: Nina Springer, Elena Johansson

Södertörn University | International journalism

Master thesis 15 credits, spring term 2020

(2)

/Abstract/

The study examines how print media in Sweden portray Putin, Russia and Russians and what makes an impact on Swedish journalists when they write on that subject.

The first part presents a content analysis of articles published in the two main Swedish newspapers Dagens Nyheter and Svenska Dagbladet (DN and SvD). The two most significant years of coverage of Russia in Sweden (2014-2015) were chosen as a time frame for the analysis. It examines the predominant tones that were used in the articles towards three keywords

“Putin”, “Russia” and “Russians”, as well as the sources that were chosen by the journalists.

In line with previous researchers, the author of this study observed a certain tendency shared in two Swedish printed media to report negatively about the Russian president. At the same time tones towards the country in general and its citizens were more neutral than negative.

The second part of the research focuses on qualitative interviews with journalists representing different media in Sweden (radio, TV channels and newspapers). During the interviews, journalists were invited to reflect how they see Russian society, what connection they have with Russian culture and what impacts their choice of topics. According to the findings, a journalist’s work is affected by their individual background, economical situation in media organisations and expectations from the audience.

/Acknowledgements/

I would specially like to thank one of my supervisors, Nina Springer, who made a strong con- tribution in helping me work on this paper. She positively impacted the study during the most important period of its preparation providing academic advice and strong personal support.

Her willingness to carefully check the material and detailed responses to my questions created a positive working atmosphere. I would also like to thank Priyanka Shankar for detailed proofreading of this paper. Last but not least, I thank all the interview respondents who, des- pite special conditions caused by Covid-19, found time to answer my questions and enriched the data for this study.

Keywords: Market logic, Agenda setting, Framing, tones, sources, Russia, Sweden

(3)

1. Introduction and background

“The fear of the Russian” (Treijs, 2017) was the name for the series of articles published by ¹ the quality Swedish newspaper SvD. The editor questioned: “Why are so many [people]

around the Baltic Sea concerned about Russia?” (Ibid). In February 2019 Swedish Radio ² published a radio programme called “Is the Russian coming?” (Farah I P3, 2019). “Should ³ we be afraid of Russia? Should you learn Russian already and have a survival bag packed in the forest or is this too overdone?” - was presented in the summary of the programme (Ibid). ⁴ The idea that Russia is a dangerous neighbour is not a new concept for Sweden. The long history of Russian-Swedish wars created a certain picture of Russia among Swedish people.

Starting from the twelfth century there were about twelve wars between Sweden and Russia (Scobbie, 2007). These and other unpleasant historical events formed a mindset of many gen- erations in Sweden.

In 2016 Andrei P. Tsygankov wrote an article about the American media perception of Russia between 2008 and 2014. According to him "the media’s image of Russia has been overwhelmingly negative since 2008” (Tsygankov, 2016, para.1) and Russia has been pursued as a dangerous threat to American democracy. This article covers the American ways of portraying Russia and adds value to my study. At the same time, I want to find out whether the Swedish media may have the same or different approach to covering Russia.

In October 2019 Greg Simons published an article called "Putin’s international political image” which referred to Putin as a "well-known (high profile) global political figure" who stimulated an "overwhelmingly negative flow of information in mainstream mass media, especially after the Ukraine crisis” (Simons, 2019, para.1). The author refers to negative char- acterisations of Vladimir Putin in comparison with Hitler and Stalin. This article is interesting for my research since I examine the tone of articles about Putin. The difference is that I focus on Swedish media and Simons examined materials published in English-speaking outlets.

“Rädslan för ryssen”

1

“Varför känner så många runt Östersjön en oro för Ryssland?”

2

“Kommer ryssen?”

3

“Borde vi vara rädda för Ryssland? Ska man lära sig ryska redan nu och ha en överlevnadsväska

4

packad i skogen eller är det överdrivet?”

(5)

In the same year, the Nordic Council of Ministers published an extensive study by Lars Kabel and other researchers named “Russia in Nordic News Media”. The authors who examined the Swedish coverage of Russia concluded that Russia in Swedish media is presented as an “opposite antithesis to Sweden” and that Putin “seems to be the only person shown to exhibit any power which gives the impression of a president governing his country in a vacuum” (Nor- ström, 2019, pp. 74-75). This study has a good overview of how all Nordic countries cover news about Russia and partly how Russia covers news about Nordic countries. But as in many other researches, the authors didn’t draw a connection between the coverage of Russia and what impacts this coverage. Questions like “Why Russia is an opposite antithesis to Sweden?” and “What makes journalists portray Putin in a certain way?” remained un- answered.

I could see a clear research gap where scholars seemed to agree that Swedish media portray Russia and, especially, Putin in a negative way. But it was unclear what creates an influence on journalists when they report on Putin, Russia and Russians. My research filled this gap answering two main questions:

1. What prominent picture of Putin, Russia and Russians is created by Sweden’s print media?

2. What factors influence Swedish journalists when they write about Russia?

To prove or disregard the findings from previous studies I conducted a content analysis of selected articles published in two main Swedish newspapers (DN and SvD) in order to observe tendencies that appeared in articles about Putin, Russia and Russians. I looked whether journalists used diverse sources from Russia; if they mentioned Putin more often than Russi- ans; and if the tone of the articles was predominantly negative. Besides the content analysis, I conducted six interviews with journalists representing quality media in Sweden (SVT, TV4, DN, Sydsvenskan and Sveriges Radio). During the interviews I explored whether individual characteristics of the journalists, expectations of the audience and other factors have an impact on how they cover news about Russia.

(6)

2. Theoretical framework

The structure of my research will be further built after the analytical model of influences on journalism (see Figure 1). I developed this model based on the hierarchy of influences model by Pamela Shoemaker and Stepher D. Reese (1996) and on the conceptual roadmap for the comparative study of journalistic culture by the group of researchers in the conceptual framework of The World of Journalism Study (n.d.). The analytical model which I developed shows that there are three levels of influence in journalism: individual, organisational and so- cietal. I organised the theory, research questions, methods and results chapters according to these three levels. I started from the individual level since I assumed that the process of form- ing the image of Russia in Swedish media starts in the heads of the journalists who create this image. Then I continued with the organisational level discussing how media organisations can influence the work of journalists. I concluded with the societal level focusing on media effects which contribute to agenda setting and framing.

Based on Shoemaker & Reese 2013 p.9 and Worlds of Journalism Study, n.d.

Figure 1. Analytical model of influences on journalism

(7)

2.1 Individual influences

Authors of the conceptual framework in the World of Journalism Study argued that individual influences matter because journalists need to make decisions based on their perceptions.

These perceptions come from their “personal and professional background”, from “their oc- cupation and political orientations” (World of Journalism Study, n.d.). I separated this chapter dedicated to the individual influences on journalism into two subchapters where one explores the knowledge of journalists about the subject and the other introduces stereotypes which journalists might have. In the notion "Knowledge about Russia" I include: whether journalists had any personal connections with Russians, if they ever lived in Russia and if they speak Russian. Term “Stereotypes" in the second subchapter refers to both positive and negative perceptions of Putin, Russia and Russians which Swedish journalists may carry.

2.1.1 Knowledge about Russia

The more one knows about the subject the more s/he can talk about it. For example, a painter who can only see yellow and blue will make a less rich painting than the one who sees the entire palette. The ability to navigate freely in a country and culture allows journalists to see more and to talk to more people. In this way journalists can absorb and reflect different sides of the story in their materials. As Erin Meyer the author of the famous book “Culture map”

stated, ”success in our globalised world requires the skills to navigate through cultural differences” (Meyer, 2014, para. 1). Knowing how to enter the office of a local authority, who should speak first and how to shake hands can immensely impact the kind of information one receives, and what is held back. Thus general knowledge of Russian culture and Russian people could be a strong factor that may impact the way journalists cover news about the country. Another important factor is the language. The level of knowledge of foreign lan- guages in Russia is rather low (World English Proficiency Index, 2020) so communicating with local people could be rather hard without a good knowledge of Russian.

2.1.2 Stereotypes about Russia

Stereotypes are often not created just by one individual but they can be results of social and collective reflections which work to justify a certain type of behaviour towards another

(8)

group. People make judgements about others, basing their opinion on the dominant beliefs of the society they live in. Two psychologists Jost and Banaji (1994) argued that stereotypes develop because of the need to protect one’s position and the status of the social group one be- longs to. Stereotypes are believed to characterise systems or structures that separate "people into roles, classes, positions, or statuses, because such arrangements tend to be explained and perceived as justifiable by those who participate in them” (Jost and Banaji, 1994, p. 3). Al- though, commonly, stereotypes bring a negative image, some scholars argue that there can also be positive stereotypes representing a positive evaluation of a group that has an advantage of another group (Czopp, 2008). Positive stereotypes same as negative stereotypes can be based on gender, nationality or other characteristics of the group. In my research, I want to give space to any positive or negative opinion about Putin, Russia and Russians shared by the Swedish journalists.

Old historical book by Swedish author Alfred Jensen describes Russian peasants as people with a lot of negative features. Among them: "lethargy, timidity, tendency to stealing, distrust, laziness, begging, untidiness and drunkenness, which is explained by Mongol enslavement and serfdom lasted for centuries" (Jensen, 1896, ch.5). According to some researchers, a similar tradition of portraying Russians has continued in the modern Swedish press. They are presented as people from the undeveloped and dark East differing from people who belong to the civilised white West (Hurd, Olsson and Aker, 2002).

Kristina Riegert did extensive research of how Scandinavian media talk about their Eastern neighbours and concluded that for Scandinavians Russia is an unknown “eastern neighbour”

and Russians are presented as the “others” who are fundamentally different from “us”. Be- sides that, Riegert stated that Swedish media prefer to introduce the “others” as a "unanimous group of people" (Riegert, 2004), which means that almost 146 million Russian people (World population review, 2020) appear as a unanimous group of “the others”.

The individual level of influence on journalism is followed by the organisational level. Most journalists need to sell their stories to the media, unless they can earn their living by publish- ing materials on their platform. Once journalists financially depend on a media organisation that distributes their materials, they might be falling into dependency on the organisation’s market logic. That is why I use "market logic" as a term in the next subchapter.

(9)

2.2 Organisational influences

Market logic and time limits

According to previous studies, Swedish journalists work with a high degree of professional autonomy. In 2016 Monica Löfgren Nilsson questioned 675 Swedish journalists and 80 per cent of them "said they had complete or a great deal of freedom when it comes to selecting stories” (Löfgren Nilsson, 2916, p.3). Almost 85 percent "said that they had complete or a great deal of freedom in deciding over which aspects to emphasise in a news story” (Löfgren Nilsson, 2916, p.3).

Together with this high level of autonomy, Swedish journalists might experience a gap between their own goals and goals of their organisation. Journalists to a high extent are interested in professional reporting while their organisation might be more interested in financial benefits that this reporting brings. Journalists need to sell stories to their superiors, who have to prioritise and allocate time and space for these stories. Time and space in the news broad- cast company or newspaper is usually rather limited (Gans, 1980), this makes the newsroom a part of some kind of a free market. Hard internal competition among journalists gives big- ger power to chief-editors to prioritise and exercise their influence. This competition between journalists is aggravated by another important factor: time limits.

The number of news items produced during a week has been constantly increasing between 1992 and 2008 without a corresponding increase in the number of journalists (Lund, 2009).

Because of this, less time was allocated per news item than a decade before that. Journalists got in a hurry to produce as much content as they could during their working day. Working hours of journalists became organised according to the series of deadlines (Schlesinger, 1979). The quality of reporting got compromised because time limits created a condition where journalists had very little time to search for all the sources they needed. Thomas Han- itzsch found these procedural influences much stronger than organisational, economic and political (Hanitzsch, 2010). In the next subchapter, I discuss the last level of influences on journalism: societal influences.

(10)

2.3 Societal influences

Media effects on the audience: agenda setting and framing

Media is the "main channel for cultural representation and expression", which helps us to

"maintain a social identity" (Deuze and McQuail, 2020, p.7). "Being part of a cultural in- dustry it responds to the demand from society for information and entertainment". And at the same time, media "stimulates innovation" and contributes to socio-cultural change (Deuze and McQuail, 2020, p. 107). In their book on mass communication theory, Deuze and Mc- Quail give high importance to the role of media calling it both the "storage of our memories, map of identity, images and ideas” and the "mirror of the society and social changes” (Deuze and McQuail, 2020, p. 81). The danger is that media shaping reality and reflecting the situation in society, might also become a source of stereotypes. While media, according to the scholars, is the main channel for representation and expression, the stereotypes created by journalists stay in the heads of the consumers and sometimes become part of their social identity. In order to “shape reality” for the reader, viewer and listener journalists as well, as experts and institutions, may sometimes use agenda setting and framing.

Agenda setting is a process when media attention is given to certain items, issues and per- ceptions in order to influence public awareness. Where one news event can be more significant than another.

"Agenda setting is a process of (intended or unintended) media influence by which the relative importance of news events, issues or personages in the public mind is affected by the order of presentation (or relative salience) in news reports. It is assumed that the more media attention is given to a topic, the greater is the importance attrib- uted to it by the news audience" (McQuail, 2005, p. 379).

To set an agenda the media need to unite and report on similar or the same news items. The main characteristics of which usually need to be:

•The news items need to be significant for the local audience. For example the refugee crisis in Sweden, a country which receives a lot of asylum applications.

•The news items don’t necessarily need to be connected to the reality of the world, and they need to serve “inside purposes”. For example the news about the Russian submarine in

(11)

Swedish waters is not significant for the world arena, but people who have something to say about it make this news significant for Sweden. Once news becomes important on the media agenda, it gets importance on the public agenda, thus it becomes a shared perception.

Agenda setting is an effect which is convenient both for journalists and for the audience. Lo- gically, the media and journalists have limitations on how many news items they can produce during a day. “Only a fraction” of the day can be covered by media, thus only issues which are believed to be “important” make it to the news reports (Coleman et al., 2009. p.149).

When it comes to the audience, agenda setting helps the readers, viewers and listeners to navigate in the new environment. People need explanations of the reality and with that need they turn towards the media. Another media effect called “framing" has a very similar function, that is why agenda setting and framing are often used together.

Framing. Journalists have a capability of framing the information in their materials and there is a certain probability that the audience is affected by these frames.

Framing is “selecting some aspects of a perceived reality and making them more salient in a communicating context, in such a way as to promote a particular problem definition, causal interpretation, moral evolution and/or a treatment recommendation for the item described" (Entman, 1993, p. 52).

Framing has many other names in scientific studies, for example, it can be called: “frame of reference”, “context”, “theme”, “news angle”, “news value”, “common sense notion”.

Frames “define problems, diagnose causes, make moral judgements and suggest remedies” (Entman, 1993, p.52).

Scholars highlighted several tools which journalists use in order to frame:

•"Usage of certain words, phrases, making certain contextual references, choosing certain pictures, referring to certain sources” (Deuze and McQuail, 2020, p.416);

•Usage of visual, graphic tools, labels and metaphors, familiar narrative structures;

• "Repeated words and symbols that connect with the cultural associations of many citizens"(Entman, Matthes and Pellicano, 2009. p.177);

•Reference to shared patterns;

(12)

•Identical synonymous words and symbols which "function to promote an interpretation of a problematic situation or actor and support a desirable response” (Entman, Matthes and Pellicano, 2009. p.177).

•Calling to mind common memories. For example, comparison of Putin with Hitler and Stalin as observed by Simons (2019).

Frames repeat themselves over time, which gives a chance “to notice, understand, store and recall the mental association for future application” (Entman, Matthes, Pellicano, 2009, p.177). “Once a frame has appeared enough to be widely stored in the citizenry’s cognitive systems it no longer needs to be repeated in concentrated bursts, nor must it be fully elabor- ated; citizens can summon the stored associations years later in response to single vivid com- ponent” (Aday, Entman and Livingston, 2012, p. 329). Frames stay in the minds of people and it’s hard to change them. Although frames reduce objectivity in the journalistic materials, the audience needs them. Every single human uses frames to organize his/her thoughts on the world’s daily events. People need to understand what is dangerous to protect themselves and their family from it. At the same time, they need to understand what is good and whom they can trust. Kitzinger in her article about framing mentioned that the biggest danger in the usage of framing is that frames may become invisible. "The most powerful frame can be invisible or so obvious that it would be taken for granted. That is when the frame becomes unques- tioned" (Kitzinger, 2007, p. 134).

In theory chapter I explored three areas of influence on journalists: individual, organisational and societal factors. In the next chapter, I will present the research questions of this study.

(13)

3. Research questions

Overall, this study primarily focuses on two main research questions (RQ):

1. What prominent picture of Putin, Russia and Russians is created by Sweden’s print media?

o Is Vladimir Putin mentioned more often than Russians?

o What are the tones towards three keywords “Putin”, “Russia” and “Russians”?

o Which Russian sources are predominantly used by Swedish journalists?

2. What factors influence Swedish journalists when they write about Russia?

o Knowledge about Russia o Stereotypes about Russia o Market logic

o Media effects (framing and agenda setting)

In the first research question, I mentioned only “print” media because I focused only on the print press in my content analysis. I based my decision on the statistical data. Newspapers are one of the most important media in Nordic countries. The research about newspaper con- sumption published by Nordicom in 2017 revealed that the percentage of the Swedish population with a subscription to a newspaper stayed the same for ten years (between 2006 and 2016). 72 percent of the population aged between 9 and 79 years had a subscription to either the print or online version of the newspaper (Harrie, 2017). I did a quantitative study on articles published in two main Swedish quality newspapers: DN and SvD. I made this choice because at the time of my content analysis these two newspapers were top quality print media by circulation between 2000 and 2014/2016 in Sweden (Harrie, 2017). The second research question was answered through interviews with journalists from different media. I situated these two questions according to the chronological order of my methodology where I first did the content analysis and then the interviews. I found it necessary to familiarise myself with the way Putin, Russia and Russians are presented in Swedish newspapers before I started contacting journalists.

(14)

4. Methods

4.1 Content analysis

4.1.1 Design and sampling procedure

The first method of my research was a content analysis of the articles. This method was chosen to answer the first research question which concerned the picture of Putin, Russia and Russians created by the Sweden’s print media.

Content analysis is a research technique for the objective, systematic and quantitative description of the manifest content of communication (Berelson, 1952, p.18).

I chose content analysis for my study because it is a transparent method. The participants of the study (authors of the articles) were not aware of the fact that they had been studied. Thus this method is also nonreactive. Another advantage of the content analysis was that it allowed observing general patterns in a larger material. Nevertheless, there was one disadvantage of this method. The content analysis allowed some degree of interpretation by the author. To avoid subjectivity I created a coding scheme (see Appendix 2). Further I describe how I selected the time frames and the keywords for the content analysis.

The keywords which I had chosen for this study cover three main aspects of the country: its current leader, its name and its population. To understand what years of coverage I need to focus on I decided to observe how often Putin, Russia and Russians were mentioned over 15 years. Starting from January 2000, a year when the current Russian president started his first presidential term and finishing in December 2015, a year when I began conducting my content analysis. I searched for each keyword separately with the help of Retriever, an online database for the media analysis. This search also helped me to observe whether Putin was mentioned more often than Russians during 15 years. Later in the Results chapter I will present some of the received tables and findings.

All three words were mentioned most often between 2014 and 2015. Thus the time span of the content analysis narrowed to two years: January 2014 to December 2015. During these two years Putin, Russia and Russians seemed to be on the headlines rather often. The choice of these two particular years of coverage resonates with the answer that I got during my interview with one of the journalists from DN, who said: “Between 2014 and 2015 there was a

(15)

little more coverage than normal because of the materials from Ukraine, Luhansk , Maidan ⁵ ⁶ and Sochi ”. ⁷

Further I describe the method which I used to select specific articles for my analysis. Again with a help of the Retriever I checked how many articles including all three keywords at the same time were published between January 2014 and December 2015. I got the number - 322 articles in total. Each newspaper had published approximately the same number of articles (DN– 180, SvD – 142). Even though I narrowed down my search to certain media, a certain time frame and certain keywords, I still had to select a sample of articles which I could actu- ally read within the time limitations I had. I selected 46 articles out of the 322, by choosing the article that was published first in each month during the years 2014 and 2015. According to logical calculation, if I had analysed one published article per month of the two years of coverage, I was supposed to get 48 articles. But it happened that SvD had no articles published with keywords in June and July 2015. Thus, the final number of the articles reduced to 46. This probably occurred due to the specific procedures that I used in my research, following which I received a limited number of articles. Selecting the articles in Retriever I typed all three keywords in the search tab at the same time. Thus only the articles which had two or three keywords at the same time came up. Articles that had only “Putin”, only “Russia” or only “Russians” mentioned were not tested in my content analysis.

Random and constant errors. This sampling method could have contained random errors.

According to Deacon and other researchers (1999) some degree of sampling error was un- avoidable due to the random variation that occurs when a smaller number of cases represents a larger population (Deacon et al., 1999). In my case 46 articles were a rather small number out of the total amount of 322 articles. There was also a risk that the sample could contain constant errors – structural biases that systematically distort its representativeness. Constant errors occur when the sampling method systematically favours articles that follow a specific pattern which overemphasises this and marginalises any other aspect represented in the popu-

A city in eastern Ukraine where lot of military actions happened during the Russian-Ukrainian con

5 -

flict of 2014-2015

Euromaidan, a wave of demonstrations and civil unrest in Ukraine, which began on the night of 21

6

November 2013 and continued until February 2014 Sochi 2014 Olympic Winter Games

7

(16)

lation (Deacon et al., 1999). In my case, there were a few possible constant errors. One was that these three keywords could not cover all forms of possible mentions of Putin, Russia and Russians. For example, “Ryssar” could also be written as “Ryssarna”, “Ryss” or “Ryssen”

which are different forms of the same word. Thus the application of the selected three keywords might not have covered all articles about Russia, Russians and Putin written in the selected newspapers. The other possible constant error was the fact that not every article, where one of the keywords was mentioned was dedicated to Russia, Putin or Russians. For example, Russians could have been mentioned in the article which was about the situation in Ukraine or another country. In that case, the article was still examined in order to see what tone a journalist used for the keyword/s.

4.1.2 Operationalisation

Once the 46 articles were selected I created a coding scheme with 5 variables:

V1: Tone towards Putin (Values: negative, neutral, positive, not mentioned) V2: Tone towards Russia (Values: negative, neutral, positive, not mentioned) V3: Tone towards Russians (Values: negative, neutral, positive, not mentioned) V4: Subject (Values: politics, sport, economy, culture, society)

V5: Russian sources (Values: official, unofficial, both, none)

Since I had three keywords I decided to create one variable per one keyword. I took this decision to check if there was a difference in tone when an article was about Russian citizens, the Russian president or the country in general. The fourth variable tested the subject of the article. The last variable examined which sources journalists had used in their reporting.

To test the first three variables each article was examined to classify its tone towards each keyword separately. To assign the codes, I examined the words that surrounded a keyword and the sentence where the keyword was present. Generally, a tone of the article tells readers how the reporter interprets the story – whether the story is reported in a positive, neutral or negative manner (Hossain, 2015) towards a key word. Taking into consideration that most people as members of the same interpretive community share a “common sense” the types of news are generally positive, negative or neutral. For example, negative news tends to elicit

(17)

the emotions of sadness, fear and/or anger (Leung and Lee, 2016). And positive news brings hope, inspiration or calms down the reader. Among the articles I tested, the ones with a negative tone were critical towards the keyword and the ones with the positive tone were favour- able or complimentary towards the keyword. In cases where there was no clear negative or positive tone towards the keyword, I coded it as “neutral" tone. Some of the articles didn’t include all three keywords at the same time and would have only one or two of them mentioned. For example “Putin” and “Russia” were mentioned, but not “Russians”. In this case I used “not mentioned” value for the variable concerning tone towards Russians. Important to say, one article was assigned to one code per variable. For example in cases where "Russia”

was mentioned twice: once in a negative tone and once in a positive tone, I would not assign two different codes, but would code the entire article as “neutral”. This situation was very rare, most of the time the same tone was observed to one keyword throughout the entire article.

For the fourth variable concerning the subjects of the news items I inductively generated five variable values. I derived them from the empirical data of the previous search where I observed which subjects of the articles were most common. I selected 5 predominant themes, namely:

•politics

•sports

•economy

•culture

•society

The fact that I assigned one article to only one code per variable made it challenging to classify the “grey zones” where an article, for example, contained both political but also societal references. In my study, I focused on the predominant subject of an article. In order to do that I read the entire article and observed what sources journalists used and what facts they were referring to. If the article predominantly included citations of political authorities and references to political events I would code it as “politics”. If the article described the Olympic

(18)

Games in Sochi and referred only a few times to political events and facts, I would code this article as “sports”.

The last and fifth variable dedicated to the sources used in the articles included four values:

•official

•unofficial

•official/unofficial

•none

To define what are official and what are unofficial sources, I used a book by Jack D. Herring- ton. Although his book about podcasting was not directly connected to newspaper production routines, the question of the choice of sources is similar for any type of journalism. Journal- ists from TV, Radio and print media have similar guidelines on how to find, select, approach and question sources. According to Herrington, official sources are the civic, municipal or commercial institutions that “employ people for the specific purpose of talking to the media” (Herrington, 2005). Official sources include military, police or fire departments, gov- ernment entities and large corporations which benefit from the taxpayers. In my research, official sources were coded if journalists cited representatives of the Russian civic, municipal or commercial institutions. Unofficial sources are usually regular citizens who “have some personal experience related to the issue or long history following that issue” (Herrington, 2005). According to Herrington “unofficial sources are integral to good reporting”. In my research, unofficial sources were ordinary Russian citizens with no official role and who had not been related to any kind of institutions. “Official/unofficial” was coded if journalists cited both official and unofficial Russian sources. And “none” if no Russian sources were cited.

Important to say that I coded the article as “none” if the Russian sources were only mentioned but not cited.

4.1.3 Data collection

To classify selected articles according to the variables, I created an Excel sheet with six columns, where the first column listed the number of the articles and five other columns reflected each variable. In the coding scheme, each value received a number as a code. Thus,

(19)

my excel file only contains numbers which I later worked with. Then I used the Statistical Package for the Social Sciences (SPSS) to receive tables displaying the frequencies for each variable of my content analysis. Every separate table showed both the number and percentage of how frequently each value was repeated in one of the variables. For example, the first table was dedicated to the tone towards Putin which was my first variable. I created tables that showed data for both newspapers at the same time, as well as tables which showed data for each newspaper separately. In total, I received 15 tables where the first five shows the ag- gregated data and the rest of them go into detailed analysis displaying the findings of two individual outlets.

Once the entire work was done and I got the results of the content analysis, I decided to add one more parameter of classification to check which articles had an author and which didn’t.

The reason for that was that during the interview with one of the journalists I learned that some newspapers publish articles that are not written by their correspondents but translated from the articles published by news agencies like TT, Reuters and AFP (Respondent 4, Syds- venskan, 2020). According to the respondent, these articles usually don’t have an author mentioned in the begging of the text. To make sure that I was not working with the texts which had not been written by Swedish journalists I checked whether I had articles without an author in my selection. My re-coding showed that out of the 46, there were only two articles that weren’t bylined with an author’s name. Thus, I was sure that most of the articles I worked with were written by Swedish journalists and not translated from the news agencies’

databases. Out of personal interest, I checked articles written about Russia in two of the biggest tabloids in Sweden “Expressen” and “Aftonbladet” and I saw that most of the articles only had TT as a reference. I will get back to this finding in the Discussions chapter. ⁸

4.2 Interviews

4.2.1 Justification of method

I had chosenqualitative guideline interviews as the second method of my research because they helped me understand the interviewee’s practices and their reflections about these practices. In my study, the interviews became the most elaborate source of information. The con-

Swedish news agency

8

(20)

tent analysis was a good starting point showing how Putin, Russia and Russians were covered in print media, while interviews allowed me to investigate further and answer the question

“why” these subjects are covered in a certain way.

I considered three different types of interviews: face-to-face interviews, telephone/video interviews and email interviews. After an examination of the literature written on that subject (Opdenakker, 2006) I decided to opt for telephone/video interviews since they have several benefits:

- I could create a trustful conversation with the journalist without dedicating as much time and effort as would be required to travel for a face-to-face meeting;

- I overcame my limitation of not being able to organise face-to-face meetings (due to the outbreak of COVID-19 the World Health Organisation recommended to avoid travelling and meetings in Spring 2020 when the practical part of my research took place);

- Interviews through the telephone and videos enabled me to capture more social cues (e.g.

gesture, facial expressions, tone of voice) than email interviewing would;

- Telephone/video interviews differed from email interviews since they could be recorded;

At the same time, I was aware of the limitations the phone/video interview had and did my best to overcome them. I was aware of the fact that if the same people were interviewed in future studies, the researcher would receive different data. This would happen due to changes in perceptions and working situations that can happen in the lives of the interviewees. This was a foreseeable limitation of the interview as a qualitative method.

This study handles a rather sensitive topic, thus the ethical principles for social research had to be accounted for. A letter of information was sent out before any interview stating the purpose of the study, handling and storage of data and the rights of participants. The letter also had the contact information of the student/researcher, supervisor, university and the data pro- tection agency. Confidentiality of the records has also been respected. They all were stored locally and not through any third parties. The process of keeping records was restricted to a limited period following GDPR guidelines. Anonymity has not been promised which was agreed with all journalists. However, in my paper, I only provided the names of the news organisations which journalists represented, and not the actual names of the journalists inter-

(21)

viewed. Hence the potential of any personal or professional harm from the study is considered unlikely. Any sensitive materials in the recordings/transcripts were considered carefully (such as personal details). This also had intersected with the issue of privacy as any personal details of respondents not related to their professional roles were not disclosed in any way in the thesis.

4.2.2 Operationalisation

I rationalised my research questions into a list of questions for interviews. I followed sug- gestions from social science research literature (e.g., Bryman, 2012). All of the respondents got the same questions. This was necessary to keep the conversation in the right direction. At the same time, interviewees were able to talk freely and I allowed myself to sometimes im- provise with new questions to get the best answers. I considered the topic of my study as rather sensitive since I explored stereotypes and factors that impact journalists in their work.

In the interviews, I did my best to stay neutral focusing on the experiences that journalists have. Below is a list of questions for the interviews:

1. PHYSICAL PRESENCE. Have you been to Russia? If yes, how much time do you spend there per year?

2. DIFFICULTIES. What difficulties do you face when you report news about Putin, Russia and Russians (search for sources for example)?

3. CULTURAL EXPERIENCE. What experiences have you had with Russian officials and regular Russian people?

4. WHAT IS MISSING IN REPORTING. Do you feel that something is missing in the coverage of Putin, Russia and Russians in Sweden and why?

5. IMPACT TO CHOOSE A TOPIC. How do you find topics when you cover Putin, Russia and Russians?

The first and third questions were designed to check the level of knowledge of Russia that journalists have. The second question opened an opportunity to observe any possible stereotypes that journalists might be having. With the fourth question, asking if something is missing in the reporting I invited journalists to reflect whether or not they feel that there are sub-

(22)

jects which are underreported in Swedish media. The last, fifth question I added after the first interview had already taken place. I found it necessary to invite journalists to reflect on how they choose topics. This question, connected to the agenda setting effect which journalists might have experienced, usually provoked discussion on what creates an influence on journalists in general. Journalists are experts in interview techniques and I didn’t expect that they would simply follow “my lead”, thus I used two last questions to make them feel slightly provoked and eager to reflect on what they do. Five questions which I created covered four levels of influences from my theoretical framework: individual (stereotypes and knowledge about Russia), organisational (market logic) and societal (media effects).

4.2.3 Sampling and data collection

In order to broaden the study’s scope, my aim was to connect with journalists from different media: newspapers, TV and radio. To find journalists for the interviews, I used two different strategies:

- I visited the websites of media (DN and SvD) and typed three individual key-words in the search. Then I scrolled through the materials written and saw who were the authors writing most frequently. After that, I checked for their emails online and wrote to them.

- I used the circle of my friends and acquaintances and contacted those who are working or worked earlier for various media (SVT, TV4, Sveriges Radio, Expressen). They connected me with journalists who cover or covered news about Russia on a regular basis.

My purpose was to hold at least six interviews with journalists from different media where at least one journalist would be from a newspaper, one from TV and one from the radio. I did my best to create gender and age diversity in the group of interviewees. Later in the text, I will explain in more detail the specific sampling procedures.

The first step of my data collection was to find which journalistsI needed to contact. Since I did the content analysis before the interviews I started the search for journalists with check- ing who were the authors of the articles I examined. I wrote down the names of the authors of the articles and searched for their contacts. It was not difficult to find contacts of the journalists from the Swedish media. Most of the media outlets provide an explanation on how to reach out to journalistson their pages. This transparency of the Swedish media saved a lot of

(23)

time for my research. Once I found contacts of journalists I started an Excel file where I col- lected their names, contacts and made drafts of the email texts to them. This Excel file will be deleted from my data storage once the research is published. Considering that I might not get any answers I decided not to stop on this rather small list. I went on the websites of both newspapers and searched for more recent articles that contained the three keywords. Then I looked at the names which I came across most frequently and created a final list of the journalists from DN and SvD. I had eight journalists from DN and eight journalists from SvD on my list. Unfortunately, only three of them were female.

Once the list of journalists was created, I looked them up in social media. It happened that with many of them I had three or more friends in common. With one of the journalists, I had fifteen friends in common. I decided to reach out to journalists with whom I had common friends on Facebook. It’s known that journalists are very hesitant when it comes to interviewing requests. By approaching those who were part of my wider network, I could expect that they were more likely to say “yes”. It appeared to me that I was wrong in my assumptions because journalists whom I had to email (because we had no friends in common) started answering the next day, while my Facebook messages stayed unread for a week. I changed my strategy and wrote emails to all journalists from my list. I used the same email template for all journalists only changing a few words according to working specifics of the particular journalist. For example, if the reporter was working from Moscow I mentioned that in my email.

I considered challenging the fact that my last name had a clear Russian origin. I was concerned that journalists may think I am affiliated with Russian institutions researching what Swedish journalists are saying about Russia. Even though I used my private Gmail account I made it clear that I am studying at Södertörn University. I wanted journalists to know that I am affiliated with a Swedish educational institution and that I can speak Swedish.

In my email, I tried to be very clear regarding why I was contacting them, who I was and what my research was about. I underlined that I am interested in “what are the difficulties journalists from Sweden experience when they need to report about Russia”. I consciously chose not to write “I want to know what influences you when you write about Russia” or

“Why the articles you write have a negative tone towards Putin/Russia/Russians” because

(24)

this could make them feel that I wanted to criticise their work. To create some mutual understanding, I wrote that I was also a journalist. Lastly, I mentioned that I would be grateful if we could talk at the journalist’s earliest convenience and that the call won’t take more than 30 minutes. I believe that these details had a positive impact and created some degree of trust between me and the journalists.

Nevertheless, out of eight journalists from SvD, I received only four answers and all of them were negative due to different reasons. Some journalists explained their refusal with their financial situation where they needed to prioritise paid assignments. One journalist requested for the research outline and after reading it, refused to give the interview. Although I was sending several reminders to people who didn’t answer, I didn’t manage to arrange an interview with somebody from SvD.

From eight DN journalists I contacted, five answered and three of them answered negatively.

Only two journalists agreed to have an interview. But at first, both of them wrote a negative answer saying: “I am not the right person to talk to since I haven’t been reporting about Rus- sia lately”. My research included several years of coverage and it wasn’t important for me whether journalists reported recently or not. Explaining that, I managed to convince these journalists to reflect on their previous work. Before holding actual interviews I had to write four-five emails to each journalist trying to find the time and a platform for the interview which would be convenient for them.

Already after having done interviews with two journalists from the DN, I started contacting journalists from other types of media. In order to find contacts, I used the capacity of my own network: former colleagues, friends and acquaintances who had some relation to the media.

Thanks to them, I received contacts of three reporters from SVT, three reporters from Sveriges Radio and the contact of one reporter from the regional newspaper Sydsvenskan who was also an author of books about Russia. All people I contacted from this list answered my requests and most of them agreed to participate in the interview. But even after agreeing for a call, all journalists mentioned that they “might be not the right person to speak to” due to the fact that they didn’t report on Russia in recent months/weeks. After contacting 26 journalists in total I held six interviews. Two people were representing a quality newspaper (DN), one person was representing a regional quality newspaper (Sydsvenskan). One person

(25)

was from the biggest news platform and a public broadcasting outlet (SVT). And one - from a commercial broadcasting alternative (TV4). The last person interviewed was from the radio (Sveriges Radio). Unfortunately, she was the only female among all interviewees. I find it important to mention that one of the interviewees is my close connection. I decided to do the interview with this journalist because his experience reporting on Russia for Swedish TV was the first inspiration for my research. Before the interview with someone who is part of my private life, I did a short examination of the literature written on that subject. Ellis (2007) in her article about ethics in research with intimate others states that research and the "friend- ship roles should weave together, expand and deepen the other” (Ellis, 2007, p.13). My close relationships with the interviewee brought more expressiveness, emotion and empathy in the research and created rich data for the analysis. At the same time, I did my best to overcome the “downsides” of “interviewing a friend”. Same as with all other interviews I made a recording and I used the same set of questions. I also turned a high level of scrutiny on myself when I was evaluating the results and choosing which quotes from this interview I could add to the text of the research.

Whether it had to be a video or phone interview was decided in accordance with the technical abilities and time resources of the journalists. Even though I strived to hold video interviews rather than phone calls, most of the journalists preferred phone calls on their Swedish phone numbers. During the data collection, I faced several challenges but I found ways to overcome them. For example:

•I couldn’t see the entire body of the interviewees thus I couldn’t see some of their gestures.

If it was a phone interview - I couldn’t see the body language of interviewees at all. That is why in cases with the phone interviews I was paying extra attention to the voice and intona- tion of interviewees. When I transcribed the interviews, I mentioned emotions expressed by the journalists in brackets. For example, I would write “laughing” if I heard that the interviewees were laughing after something they shared.

•Because I had fewer chances to observe the social cues in the telephone/video interviews it was harder for me to notice when the interviewees finished answering my question. I often used specifying questions in order to understand if journalists wanted to develop their thoughts. For example, after receiving an answer to the question “Have you been to

(26)

Russia?” I would ask “Does the tone of your material change when you write from Russia?”

•The interviewees could have been distracted more easily during the telephone/video interviews by their colleagues or family. I managed to overcome the last limitation by sticking to 30-minute-long interviews and to a structured set of five questions. This solution I gained from my previous experience working as a journalist with loads of face-to-face interviews. I learned that when an interview lasted for more than 30 minutes the interviewees were getting distracted easily. None of the interviewees was distracted during our conversation.

I also mentioned above that a telephone/video interview could be recorded. I managed to use this advantage. In the case of a video interview, I used a popular video conferencing platform called “ZOOM” to call and record both video and audio conversations. If the interview was done over the phone I used audio recording by turning on the speaker of the smartphone and recording the sound on another smartphone at the same time.

Besides recording the interviews with the help of the “voice memos” application on my smartphone, I was also taking notes on my computer in the Excel file. I was typing both my questions and the answers of the journalists. During some interviews, I managed to type the entire text and during the others, I had to stop and only focus on listening and asking questions. During one interview I couldn’t make any notes at all because the journalist was giving very short answers to each question. He was not developing his sentences, thus I had to constantly feed-in new questions to inspire him to talk. During that same interview, I came up with more than ten new questions because that journalist had a different experience from other interviewees. Besides reporting he also worked as a news desk editor and wrote several books about Russia. Even though I had to come up with new questions I did my best to stay within the framework of my research.

4.2.4 Data analysis

I analysed data according to the five steps described by Ditte Mortensen in her article “How to do a thematic analysis of user interviews” (Mortensen, 2020). Although her method didn’t

(27)

come from media studies it fitted well in my situation. The five steps of the analysis I took were:

1. Transcribing and familiarising myself with the content

2. Assigning codes to the data in order to describe the content according to my research questions

3. Searching for patterns or themes in my codings across the different interviews 4. Reviewing themes in accordance with my research questions

5. Defining and describing themes

6. Producing the report with the results of the interviews

I transcribed all six interviews in an Excel file where I allocated the answers of journalists in columns: one journalist per column. Since I had to find connections between the answers of journalists and my research questions, I added two more columns in the beginning. In the first one, I listed four areas of impact from my research question: stereotypes, knowledge about Russia, market logic and media effects. In this document, I called them "categories". The second column was allocated to the themes and patterns that could be part of the category.

For example: living in Russia, speaking Russian, having private connections with Russians were the themes of the category “Knowledge of Russia”. This structure allowed me to easily observe the results and select the quotes from the interviews according to the four categories.

The questions asked during the interview reflected very well the research questions I had. I had to delete a few themes from the document since they were not mentioned by more than one journalist. For example, I first added the theme “corruption” which concerned the way foreign journalists experienced corruption in Russia. But this theme was only mentioned once, thus I didn’t analyse the quote about it.

Further, I will reveal the results of my content analysis and the interviews. First I will present the results of the content analysis which answered the first research question concerning the image of Putin, Russia and Russians in Swedish printed media. After that, I will present the results of the interviews answering the second research question which concerned influence on the journalists.

(28)

5. Results

5.1 Content analysis

The results of the content analysis will be presented in the same order as the bullet points to the first Research question. I am deliberately not going to present the results of the fourth variable which concerned the subject of the article because I consider the findings insignific- ant (see chapter 3 Research questions). Chapter 5.2.2 about tones of the articles is separated into three parts. Each of them explores the tone towards one keyword separately. This was necessary because in some cases I supported the results of the content analysis with the quotes from the articles. I also included some of the tables in the text to demonstrate the find- ings.

5.1.1 Mentions of Putin, Russia and Russians

The results of the content analysis showed that the average frequency of mentions of Putin in DN and SvD was 200 mentions per year per one newspaper (see Diagram 1 and 2). If we put together the numbers of mentions in both newspapers (around 400 motions per year) and compare with the number of days per year (365/366) we will see that Putin was mentioned at least once a day. A similar picture was observed with mentions of Russia in both newspapers.

Diagram 1. Frequency of mentions of “Putin” and "Russians" in “Dagens Nyheter” during 15 years (January 2000 - December 2015)

(29)

An average number of mentions per year was 100 mentions per newspaper. This means that Russia was mentioned at least every second day either by DN or by SvD.

At the same time the frequency of the mentions of Putin was much higher than the frequency of the mentions of Russians in both Swedish newspapers. There is no doubt that Putin’s face is easily recognised by a general Swedish audience. Thus, the articles mentioning the current Russian leader (especially if they are supported by a portrait of him) have a big chance to be sold. Putin was mentioned more often than ordinary Russians because his figure is probably well recognised in Sweden. I found it reasonable to test Putin’s presence in the media because with so many mentions of the country’s leader, journalists tend to forget to speak about the rest of the nation.

Both newspapers (see Diagrams 1 and 2) mentioned Putin more often than Russians between January 2000 and December 2015. In DN Putin was mentioned 4100 times and Russians 1691 times during 15 years. Differences between mentions of Putin and Russians in DN reached the highest point in 2014, where Putin was mentioned 727 times during twelve months and Russians - 216 times. Which means that during 2014 Putin was mentioned almost four times more often than Russians. Similar situations were observed in SvD where Putin was mentioned 4126 times and Russians were mentioned 1628 times during 15 years of coverage. Frequency of mentions of Putin took its peak in 2015 when it was 5 times higher than the frequency of mentions of Russians. During that year Putin was mentioned 520 times and Russians only 96 times. At the same time, it is important to repeat that I might have underes-

Diagram 2. Frequency of mentions of “Putin” and "Russians" in “Svenska Dagbladet” during 15 years (Jan- uary 2000 - December 2015)

(30)

timated the mentions of the Russian people due to my coding procedure where I only examined the cases when Russians were mentioned with the use of the word “Ryssar” (see Chapter 4.1.1 Design and sampling procedures of the content analysis).

Nevertheless, this finding had great value for this and other studies about the picture of Rus- sia in media around the world. Even though it is understandable that the name of the person who holds so much power would often be on the headlines, mentions of Putin significantly exceeded mentions of Russians in Swedish media. This supports a pattern observed in 2019 by Norström when the media pictured Putin as a leader governing the country in a “vacuum”

and provides evidence for possible future research on the personification of Putin in Swedish media.

5.1.2 Tone of articles

Earlier in my research I referred to the media effect called framing where some aspects of reality are made more salient. Testing which tones were used towards three keywords helped me to observe how journalists used the frames shared in the Swedish audience. The coverage was not the same towards all keywords. The results of the content analysis showed that the tone towards Putin was more negative than positive or neutral in both newspapers (see Table 1). The tone towards Russia in both newspapers was almost equally negative or neutral (see Table 2). And the tone towards Russians was more neutral than positive or negative (see Table 3). Further in the text, I will speak more in detail about the difference in tones towards the keywords. I will accompany my data with examples, quoting some of the articles.

Putin. Table 1 shows that in three-quarters of the articles from both newspapers the tone to- wards Putin was negative and only in one-quarter of the articles it was neutral. None of the analysed articles had a positive tone towards the Russian leader. Since I also decided to check articles from each newspaper separately I could observe the difference between DN and SvD.

Putin was mentioned at least once in all 24 analysed articles published in DN. And in every article the tone towards him was negative. While in SvD less than half of the examined articles had a negative tone towards Putin, in the rest of the articles he was either not mentioned or the tone towards him was neutral.

(31)

Table 1. Tone towards "Putin" in "Dagens Nyheter"

and “Svenska Dagbladet” in the articles published between January 2014 and December 2015

Table 2. Tone towards "Russia" in "Dagens Nyheter"

and “Svenska Dagbladet” in the articles published between January 2014 and December 2015

Table 3. Tone towards “Russians" in "Dagens Nyhe- ter" and “Svenska Dagbladet” in the articles published between January 2014 and December 2015

(32)

Negative tones towards Putin were constructed through several language tools, used by the authors of the articles from both newspapers. For example, journalists used a comparison with other leaders who have negative characteristics or memories about them:

“Putin has also encouraged and supported a pro-Russian separatist uprising in the Eastern part of the country. He claims a right to protect ethnic Russians wherever they happen to live, a form of language of power similar to Hitler's against Czechoslovakia in the 1930s” (DN, 08.08.2014a) ⁹

“Turkish judges also point to purely personal similarities between the 62-year-old

"Tsar Putin" and the 60-year-old "Sultan Erdogan": the same body language, the same posture, the same authoritarian leadership and the same nationalist-coloured revenge” (Ohlsson, 2014) ¹⁰

“Putin has built a modernised version of the corporate state and follows almost in detail Mussolini's motto: "Everything within the state, nothing outside the state, nothing against the state” (Kolesnikov, 2015) ¹¹

In the examples above, journalists referred to the historical leaders that are remembered in a certain way by Swedish people. For example, Hitler and Mussolini are remembered as two cruel leaders who killed masses of people. Mentioning Putin in the same sentence with these leaders brings the reader to understanding: “Putin behaves like Hitler and Mussolini”. Creat- ing a bad character in the story helped journalists to simplify the news and to lead the audience into absorbing the information from the article. Journalists used common points of references in order to get the audience on their side. Common references to Putin as a bad character is one of the signs of the framing of the Russian leader (see Chapter 2.3 Theoretical framework on Societal influences). This effect is often used together with personification in times of a war or conflict when the “enemy is portrayed as a supra-human like a demon or

“Putin har också uppmuntrat och stöttat ett proryskt separatistuppror i östra delen av landet. Han

9

hävdar en rätt att skydda etniska ryssar varhelst de råkar bo, en form av maktspråk som liknar Hitlers gentemot 1930-talets Tjeckoslovakien”

“Turkiska bedömare pekar också på rent personliga likheter mellan den 62-årige ”tsar Putin” och

10

den 60-årige ”sultan Erdogan”: samma kroppsspråk, samma kroppshållning, samma auktoritära ledarskap och samma nationalistiskt färgade revanschlusta”

“Putin har byggt en moderniserad version av den korporativa staten och följer nästan i detalj Mus

11 -

solinis motto: "Allt inom staten, ingenting utanför staten, ingenting mot staten”

(33)

monster - capable of being a mortal threat to “us” and “our way of life”(Kemp et al., 2013, p.145). It also involves equating current leaders with leaders of the past who are considered to be demonic. For example, during both the First Gulf War and the 2003 Iraq War, great ef- forts were made to equate Saddam Hussein with Hitler (Kemp et al., 2013).

Russia. The tone towards Russia was different from one newspaper to another. A little more than half of the articles published in DN had a negative tone towards Russia and none of the articles mentioned Russia positively (see Table 4). With a small difference from DN half of the articles published in SvD had a neutral tone and one-third of them had a negative tone towards Russia.

Negative tones towards Russia in DN was created by portraying Russia as a scary big neighbour that has no justice inside the state and constantly threatens neighbouring countries.

Journalists reporting from Russia often described their fear in the articles:

“Now we are here, in a dull Russian hotel lobby. There is no backup plan and everything feels uncertain and nervous” (Sundström, 2015) ¹²

Journalists also sometimes connected all three subjects Putin, Russia and Russians in one.

They used some kind of generalisation of the notions Putin, Russia, Russians, Moscow, etc.

“Nu är vi här, i en dämpad rysk hotellobby. Det finns inte någon reservplan och allting känns osäk

12 -

ert och nervöst”.

Table 4. Tone towards "Russia" in "Dagens Nyheter”

in the articles published between January 2014 and December 2015

Swedish picture of Russia