MA Erasmus Mundus Master
Journalism, Media and Globalization
(joint degree)
FRAMING INSIDE TERROR AND JUSTICE:
The NSU-Trial in German Print Media
– A Quantitative Content Analysis –
by
Paula Johanna Rösler
Student ID: 11300442
Master’s Thesis
Graduate School of Communication
Master’s programme Communication Science
Supervisor/Examiner: mw. dr. K. Gattermann
Abstract
The quantitative content analysis at hand deals with the coverage of the trial against the extreme-right terrorist organization National Socialist Underground (NSU-trial) in German quality newspapers. For the analysis a four years time period was chosen – from the
beginning of the trial onwards. This study adds to communication science research by applying framing-theory and the concept of news values to the distinctive domain of court reporting in the context of right-wing terrorism. Five frames were developed to categorize different aspects of reporting. The overall research question is: To what extent did framing in
the coverage of the NSU-trial in German quality newspapers change over time? Findings
show that framing only changed marginally and that reporting was reduced drastically in the fourth year of court hearings. As a result, it is concluded that the coverage was restricted both by the undynamic character of court hearings and a general decrease in media attention in the course of the NSU-trial.
Introduction
Four years have elapsed since the opening of the trial against the extreme-right terrorist organization National Socialist Underground (NSU). The NSU is accused of having committed ten murders all over Germany between 2000 and 2007. Among the victims were eight citizens of Turkish and Kurdish origin, one Greek man, and one German police officer (see, e.g., Virchow, Thomas, & Grittmann, 2015; Behrens, 2017). Moreover, the NSU is charged with two attacks in the city of Cologne – one explosion in an Iranian grocery shop and one nail-filled bomb blast in a Turkish quarter, leaving several people severely injured (Virchow et al., 2015; Behrens, 2017). Next to the premeditated crimes against citizens of foreign origin the terrorist organization is charged with fifteen bank robberies. After a failed hold-up in 2011 two members of the so-called “NSU-Trio” committed suicide (Virchow et al., 2015; Behrens, 2017). Their alleged accomplice Beate Zschäpe, who turned herself in to the police a couple of days later, is the main accused on trial. She faces life imprisonment being charged among others with founding the terrorist organization NSU, complicity in ten murders and assistance in fifteen bank robberies (Behrens, 2017). Besides her, four other defendants are accused of having supported the terrorist organization with arms and false documents (Virchow et al., 2015; Behrens, 2017). The NSU-trial is the biggest court case in Germany since reunification, involving five judges, thirteen defense lawyers, three federal prosecutors and approximately fifty lawyers representing the private accessory prosecutors (victims or family members of murdered victims who have the right to join the proceedings) (Behrens, 2017). Since the beginning of the trial in May 2013 355 days of court hearing have passed as at the end of March 2017 (Behrens, 2017).
The high socio-political relevance of the trial is not only manifested in the sheer scope of the examined racist hate crimes. What adds crucially to societal interest is the fact that during many years public authorities investigated the murders (which were knowingly a
series, as all victims were shot by one and the same gun) primarily within the victims’ communities, alleging ties to mafia-like organized crime (Virchow et al., 2015; Behrens, 2017). Many victims and family members described the interrogations as prejudging and presumptuous. They felt mistreated by state authorities and left alone in their grief and anxiety (Behrens, 2017; Merkel, 2012). At that, the media, which largely reported according to official portrayals, also stigmatized the people concerned, barely questioning the
constructed theory of organized crime (Virchow et al., 2015). The defamations were taken to the extremes by using the discriminating term “Döner-Morde” (the murders of Döner kebab) as a synonym for the series of murders (Virchow et al., 2015; Behrens, 2017). An extensive discourse analysis of the coverage of the attacks in German print media between 2000 and 2007 showed that also opinion leading quality newspapers like Süddeutsche Zeitung and
Frankfurter Allgemeine Zeitung used the term (Virchow et al., 2015). It was only after the
self-uncovering of the NSU that the coverage was recognized as biased and defaming. Thus, it is crucial to keep in mind the long-lasting false accusations against the victims’ communities in order to capture the media’s immense responsibility regarding the NSU-trial. Not only German chancellor Angela Merkel promised: “We will do our utmost to solve the crimes” (Merkel, 2012). Also journalists emphasized in interviews they considered the trial as an opportunity to give reparations to the victims in the form of comprehensive well-balanced reporting (Bier, 2015).
The aim of this thesis is to analyze the coverage of the NSU-trial in German quality newspapers over a four years time period from April 2013 to March 2017. To date, no comprehensive analysis has been conducted as to the reporting on this court case. Only one study dealt with the issue of stereotyping in the coverage of the first 150 days of hearing (Hansen, 2015). The study at hand aims to further investigate the way journalists reported on the trial: Which frames did they chose based on which news factors? How did time in general
affect reporting? The thesis adds to communication science research by applying framing-theory to the distinctive domain of court reporting in the context of domestic right-wing terrorism. So far, little research has been conducted as to the role of framing in court reporting. Analyzing how newspapers frame court hearings over several years in the
particularly relevant case of the NSU-trial will provide notable insights into this journalistic field and identify challenges as to framing analysis in this matter. This is of particular interest from a socio-political perspective. After all, the public obtains information about the courts first and foremost through the media and framing is decisive when it comes to common interpretation (see, Entman, 1993; Greenhouse, 1996). Hence, analyzing framing in the coverage of the NSU-trial will help understand which information the public bases their knowledge upon. Against this background, the following questions are guiding the analysis: Which frames were most present during the four years period of court reporting? To what extent did journalists place emphasis on issues beyond trial-proceedings? And, how did special characteristics of court reporting influence the coverage? The overarching research question is: To what extent did framing in the coverage of the NSU-trial in German quality
newspapers change over time?
Theoretical background
The theoretical framework of this study is based on the concept of framing. The analysis of frames is a broadly acknowledged and well-established research method in communication science, as such studies help understand how the media present complex issues to society (Kuypers, 2006). As Entman (1993, p. 52) put it: “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 other words, where some aspects are accentuated, other aspects are neglected or even omitted. Thus, framing-theory can serve
as a tool to comprehend the hierarchy of a text and shed light on its potential power to
influence the consciousness and thinking of the reader (Entman, 1993). Using framing-theory when content-analyzing the coverage of the NSU-trial is therefore a promising method to identify patterns in reporting over time and to draw conclusions as to how the media shaped public perception of the court case. This is of particular interest, as “many of society’s most difficult questions” are examined in courts (Greenhouse, 1996, p. 1538).
According to De Vreese (2005) frame-building is influenced by both factors internal and external to journalism. The latter is a process that “takes place in a continuous interaction between journalists and elites” (De Vreese, 2005, p. 52). In the broad domain of political journalism, for instance, politicians often engage with the news and compete with each other and with journalists over news frames (Entman, 1993). Contrariwise, high court officials rarely do so. As Greenhouse (1996, p. 1540) put it:
The Justices are visible, of course, on the bench whenever the [Supreme] Court is in session, but opportunities for causal or unscheduled contact are almost nonexistent. The journalist’s job is almost entirely paper-dependent, defined by the endless flow of conference lists, order lists, petitions for certiorari, and opinions. While politicians will cheerfully or angrily critique any story in which their name has appeared, Justices rarely respond to public comment, or even rank error.
However, it is important to keep in mind that trials in general are primarily shaped by the judges, the prosecution and the advocacy. Thus, court reporting in any event is subject to a certain pre-frame based on which journalists make their observations. Nevertheless, the court reporter’s individual assessment of hearings becomes more decisive to a news story when interaction with trial participants is restricted (Greenhouse, 1996).
All the more, frame-building influenced by factors internal to journalism seems to play an important role when it comes to court reporting. Technically speaking, high-quality
journalism by definition should avoid any stronger accentuation of some aspects over others. After all, according to professional standards, journalists are supposed to strive for maximum objective reporting (Entman, 2010). The notion of pure objectivity, however, has long been debilitated, as we simply “cannot lift ourselves out of our human context and apprehend
reality apart from it” (Reese, 2009, p. 32). This is why frames come naturally to any kind of reporting – be it consciously or unconsciously (Kuypers, 2006). Next to the individual journalist, who shapes the tonality of a news story by taking a certain angle and placing emphasis on certain aspects, there are patterns of media routines that can have an impact on framing, as media organizations generally develop workflows, which are the fountainhead for content and simultaneously constrain and enable action (Reese, 2009).
What comes into play here is the taxonomy of news values, as primarily developed by Galtung and Ruge (1965) and multiple times adapted by other scholars (see, e.g., Harcup & O’Neill, 2001). The concept signifies that journalistic selection procedure is guided by certain factors, which decide upon the newsworthiness of a story (Boukes & Vliegenthart, 2016). After all, the media have a limited carrying capacity, which is why only a fractional amount of daily news actually gets published (Hilgartner & Bosk, 1988). Thus, news values can influence or even determine how journalists and news organizations “present public issues within certain frames, often reflecting broader cultural themes and narratives, that help define the fund of ideas available to citizens as they think and talk about politics and public affairs” (Gamson, 1992 in Johnson-Cartee, 2005, p. 126; see also, De Vreese, 2005; Van Gorp, 2010). In other words, news values and framing theory are conceptually connected to each other (Boesman, Berbers, d’Haenens & Van Gorp, 2017). Among the news factors that have most been validated in previous research are: continuity, personification, influence and relevance,
controversy, success and damage, and geographic proximity (Boukes & Vliegenthart, 2016).
It is yet open to question to what extent journalists incorporated these factors in their
reporting on the NSU-trial. However, journalists in courts face particular challenges to meet journalistic needs and practices. Trial days are generally unpredictable, but tight space in newspapers does not expand and editorial deadlines will not be postponed on busy hearing days with important decisions (Greenhouse, 1996). At the same time, court reporters often
have a hard time presenting the somewhat bureaucratic daily trial routines in an appealing way (Greenhouse, 1996). Accordingly, when it comes to the NSU-trial, it seems likely that news values served as an instrument to keep the topic present in the news––not least given the long time period of four years. Against this background, five frames were predefined as content analytic variables with the above stated news factors in mind.
Frames & Hypothesis
The first frame is the court case frame. It refers to the continuity factor, which
basically implies that a topic has been in the news before (Boukes & Vliegenthart, 2016). As the case already caused quite a sensation before it was eventually heard in court on a regular basis, the public was familiar with the issue and media attention was high when the trial began (Bier, 2015). Regarding the coverage of the NSU-trial, the court case frame comprises basic information about trial proceedings, such as the actual indictment, the duration of the trial, the different actor groups in court, the court premises, and the presence of the media. Obviously, the notion of continuity has strengthened as the trial proceeded. However, given the long time frame of four years, the bureaucratic functioning of courts, and the static settings (see, Greenhouse, 1996) there is reason to assume that public interest in daily trial-routines rather decreased over time and continuity actually turned into a drawback.
Consequently, the court case frame might have been considered insufficiently newsworthy to attract public attention on the long run. The first hypothesis is based on these assumptions. H1: The further the trial proceeds the less present the court case frame becomes. The court case frame is an issue-specific frame, which to date was absent in framing-theory. It was
developed for this study to generate a variable that measures the presence of daily trial routines in a news story.
Personification is the second frame. The equally named news factor implies that a
Vliegenthart, 2016). In other words, the story provides personal information about e.g. outward appearance, character traits, relationships, sexual orientation, family matters, emotional life, health, profession, or interest motives. This frame is based on the human
interest frame, which was widely analyzed in framing research, as it has high potential to
capture audience interest (see, Semetko & Valkenburg, 2000). All the more, it is of particular interest in the context of court reporting, as it raises important questions as to the portrayal of defendant and prosecution actors. In this analysis the frame differentiates between the main accused Zschäpe, the four other defendants, the victims and their relatives, the judge, and the witnesses. Thus, the personification frame also includes components comparable to the victim frame, which is often found in framing theory as well (see, e.g., Van Gorp, 2010; Iyengar, 1990). As to the defendant side, a study conducted on the coverage of the Norwegian terrorist attacks in 2011 showed that the reporting “focused on the perpetrator as an individual rather than a terrorist, giving him questionable political exposure” (Falkheimer & Olsson, 2014, p. 82). Another analysis on terrorism and framing showed that domestic perpetrators are heavily personalized and often labeled as intelligent but mentally unstable (Powell, 2011).
Assumably, certain parallels exist regarding the alleged terrorists of the NSU, particularly Zschäpe. Based on this, the second hypothesis was developed. H2: The main accused
Zschäpe is dominant within the personification frame.
Next there is the right-wing terrorism frame. It refers to the news factors influence and
relevance, which basically imply the consequences for or impact on society (Boukes &
Vliegenthart, 2016). In the context of framing research it compares to both the conflict frame and the morality frame (see, Semetko & Valkenburg, 2000). The right-wing terrorism frame, too, is important when analyzing the coverage of the NSU-trial, as it helps investigating the extent to which the media actually dealt with the issue of right-wing terrorism beyond trial routines. The frame differentiates between three angles in the context of right-wing terrorism:
First, background information about the NSU. Second, analysis of societal consequences of the NSU terror attacks in Germany. And, third, the broader context of right-wing extremism. Given the huge public shock, which followed the self uncovering of the NSU, and the media’s defaming coverage until then, there is reason to assume that the right-wing terrorism frame all the more played a major role in the reporting on the trial. After all, journalists pronounced to attach particular importance to comprehensive well-balanced reporting in order to give reparations to the victims of the NSU (Bier, 2015). Against this background, a third hypothesis was developed. H3: The further the trial proceeds, the more present the
right-wing terrorism frame becomes.
The fourth frame is the criticism of authority frame. It refers to the news factor of
controversy and investigates the extent to which journalists criticize official acting and
portrayal in their reporting. Looking at other framing studies it compares best to both the
responsibility and the conflict frame (see, Semetko & Valkenburg, 2000). Existing research
shows that news organizations tend to adopt official framing by administrative bodies instead of using independent documentation to build interpretations, which challenge these frames (Bennett, Lawrence & Livingston, 2006). However, as stated above, journalists felt
particularly responsible to provide well-balanced reporting regarding the NSU-trial. This could animate journalists to turn towards such background stories more often. Moreover, there is reason to assume that a contextual connection exist between the right-wing terrorism frame and the criticism of authority frame, as official bodies such as the German
“Verfassungsschutz” were heavily criticized regarding their investigations in the right-wing extremist scene (Behrens, 2017). Against this background, a fourth hypothesis was developed. H4: The criticism of authority frame is likely to appear in connection with the right-wing terrorism frame.
The fifth frame is named after the news factor geographic proximity, which implies that an event is linked to the area where (in this case) the newspaper is spread (Boukes & Vliegenthart, 2016). It concerns only the regional newspapers analyzed. There is a widespread area of action as to the attacks committed by the NSU. The regional newspapers analyzed in this study all represent different local regions and individual experiences with the terror-organization, for example the bomb attacks in Cologne or the murders in Nuremberg. Consequently, it is assumed that those regional newspapers place different emphasis on different points in time of the trial due to proximity. Against this background, a fifth
hypothesis was developed. H5: The geographic proximity news factor shapes the framing of
regional newspapers.
Method
The method chosen for this study is a quantitative content analysis, that is, the
systematic and objective analysis of message characteristics (Neuendorf, 2002). The analysis focuses on German quality newspapers that reported on the NSU trial in a four years time period between April 1st, 2013 and March 31st, 2017. The trial itself was supposed to start on April 13th, 2013. However, the first court day was postponed to the May 6th, 2013 due to media accreditation issues. Starting point for setting the sample was the consideration that only newspapers which were accredited for the trial with seat reservation from the very beginning should be taken into account (see, Norddeutscher Rundfunk, 2013). Thereby, the same preconditions regarding access to the court were guaranteed for all newspapers. A second criterion was to analyze a balanced selection of national and regional newspapers in order to receive a meaningful portrayal of the German press news market. This resulted in a total of six newspapers. The three opinion-leading national daily newspapers – Süddeutsche
Zeitung, Die Welt and Die Tageszeitung – represent a balanced spectrum of political
Stadtanzeiger, Nürnberger Nachrichten and Thüringer Allgemeine – all represent different
geographic regions in Germany (West, South, East). Moreover, they have their own regional connection with the NSU either due to crimes committed, as is the case with Kölner
Stadtanzeiger and Nürnberger Nachrichten, or due to the origin of the NSU members, which
is the city of Jena in the Free State of Thuringia, where the Thüringer Allgemeine is based. Moreover, all newspapers have above-average circulation numbers (Statista, 2017; Schröder, 2017): Süddeutsche Zeitung (365.938), Die Welt (170.531), Die Tageszeitung (50.487),
Kölner Stadtanzeiger (245.173), Nürnberger Nachrichten (243.455), Thüringer Allgemeine
(243.335). The articles were collected via the database Factiva. The search-term “NSU-Prozess” had to appear either in the headline or lead paragraph of the article for the following reasons: There are many instances, state-authorities and commissions of inquiry on different levels, which are in charge of coming to terms with the NSU (Behrens, 2017). For this study, however, only those articles, which first and foremost reported on the NSU-trial were
relevant. Importantly, articles with less than 500 words were not included, as the emphasis of the study is placed on in-depth reporting. This resulted in a population of 598 articles:
Süddeutsche Zeitung (n = 155), Die Welt (n = 57), Tageszeitung (n = 53), Kölner
Stadtanzeiger (n = 31), Nürnberger Nachrichten (n = 63) and Thüringer Allgemeine (n =
239). As the article counts per newspaper differ critically, and to reduce the number of articles to a feasible amount to code, a randomization for Süddeutsche Zeitung (every second article /
n = 78) and for Thüringer Allgemeine (every third article / n = 79) was necessary.
Consequently, the sampling unit consists of (n = 361) articles. The unit of analysis is the single article.
Framing measures
Five frames were developed as content analytic variables: court case, personification,
of 19 questions was developed to measure the extent to which these frames appeared in the news stories. The coder had to answer these questions with either yes (1) or no (0). Each question was related to one of the frames, for example: “Does the story contain information about the indictment?” (court case), “Does the story provide personal information about Beate Zschäpe?” (personification), “Does the story provide background information about the NSU?” (right-wing-terrorism), “Does the story contain critique regarding the functioning of the court?” (criticism of authority), “Does the story refer to events within or close to the newspaper’s distribution area?” (geographic proximity). A minimum of three questions was considered to measure each frame (Semetko & Valkenburg, 2000). The codebook with the full list of questions can be found in the appendix.
Intercoder reliability, measured with Krippendorff’s Alpha based on a subsample of 10% of the articles (36 articles) for each of the 19 framing questions, was between Kalpha = .36 for authority 3 and Kalpha = 1.00 for proximity 2 (see all figures in Table 1, appendix). For this study, Kalpha = .80 was defined as the norm for a good reliability test, with a
minimum of .60. In most studies, the minimum is set at .67 (see, De Swert, 2012). Yet, for the analysis at hand the low minimum of .60 has been accepted, as it deals with rather fragile nominal variables, which don’t have a high range. Thus, little deviation between coders already causes considerable losses in Krippendorff’s Alpha. According to these settings, there are nine variables above .80 with good reliability, eight variables above .60 with sufficient reliability and two variables below .60 with insufficient reliability, which require particularly careful interpretation.
Factor Analysis
A principal component analysis (PCA) was conducted on the 19 framing questions to test whether underlying dimensions existed. Only the items with factor loadings higher than .30 were included in the scales. As Table 2 shows, the results of the factor analysis were not
corresponding to the five frames as developed based on theory. Only one frame, geographic
proximity, was clearly identified as a self-contained component (.88, .76, .90). Nevertheless,
certain correlations are to be emphasized. For instance, as visible in Table 2, the
sub-questions court case 1 (.61) and court case 3 (.78) form a component, as well as court case 4 (-.70) and court case 5 (-.84). Moreover, the sub-questions right-wing terrorism 1 (.61) and
right-wing terrorism 3 (.69) are correlating, as do criticism of authority 2 (.65) and criticism of authority 3 (.67). In addition to the factor analysis, Cronbach’s Alphas were used to
measure the internal consistencies for the five frames. The results correspond with the PCA. Alpha values were .30 for the court case frame (5 items); .16 for the personification frame (5 items); .29 for the right-wing terrorism frame (3 items); .32 for the criticism of authority frame (3 items); and .82 for the geographic proximity frame (3 items).
However, in contrast to Semetko and Valkenburg (2000), the questions were not meant to necessarily correlate with each other. For example: the five questions within the
personification frame asked for the presence of personal information about five different
actors in court – the more actors were personalized, the more present the personification frame became – however, these actors can perfectly be personalized all by themselves, and still the personification frame is present. The same principle applies to the other frames. Looking at the court case frame, there are five different aspects of court routines, which however can be present independently from each other – the article might include information about the indictment but not about the duration of the trial. Especially for the right-wing
terrorism frame and the criticism of authority frame the idea was to rather capture a range of
possible complementary aspects of these frames than to create measurable underlying dimensions. In doing so, the analysis combines a tailor-made approach that encompasses the special characteristics of the NSU-trial with leading standards in content analysis research.
Table 2. Principle Component Analysis for the 19 Framing Items Framing items 1 2 3 4 5 6 7 Court Case 1 .61 .36 -.35 Court Case 2 -.78 Court Case 3 .78 Court Case 4 -.70 Court Case 5 -.84 Personification 1 -.30 -.63 Personification 2 .69 Personification 3 .53 .44 Personification 4 -.77 Personification 5 -.50 Right-Wing Terrorism 1 .40 .61 Right-Wing Terrorism 2 .70 Right-Wing Terrorism 3 .69 Criticism of Authority 1 -.67 Criticism of Authority 2 .65 Criticism of Authority 3 .67 Geographic Proximity 1 .88 Geographic Proximity 2 .76 Geographic Proximity 3 .90 Framing analysis
For statistical testing the sub-questions of each frame were added up. Moreover, the time period from April 2013 to March 2017 was divided into four years, each from 1st of April to 31st of March. This division was chosen in order to make preferably even
comparisons between equal seasons. After all, the court as well is bound to an annual rhythm, including e.g. summer breaks in August. Dividing the time period into shorter slots like months or quarters would have caused more complexity than insights, not least as there are months with very little reporting. Hence, dividing this study’s time frame into four years helps
conduct statistical tests that generate solid figures as to the extent to which framing changed over time. One-way Anova’s were conducted for each frame to measure whether there were significant changes as to the frame’s presence between the years. The same tests were conducted for the group of national newspapers and the group of regional newspapers.
The results below need to be interpreted with caution, as a Levene’s Test of
Homogeneity of Variances revealed that the assumption of equal variances in the population has been violated for some frames: court case frame, Levene’s F(3, 357) = 3,31, p =
.020; criticism of authority frame, Levene’s F(3, 357) = 4,57, p = .004; geographic
proximity frame, Levene’s F(3, 357) = 11,19, p = .001. For only the group of national newspapers a Levene’s test revealed that the assumption of equal variances in the population has been violated for the court case frame, Levene’s F(3, 184) = 2,95, p = .034; and for the
criticism of authority frame, Levene’s F(3, 184) = 2,99, p = .032. For the group of regional
newspapers a Levene’s test revealed no significant differences for any of the frames. Results
Five hypotheses were to be tested, each of them referring to one particular frame. The hypotheses are based on assumptions as to possible developments in framing over time. H1:
The further the trial proceeds the less present the court case frame becomes. H2: The main accused Zschäpe is dominant within the personification frame. H3: The further the trial proceeds, the more present the right-wing terrorism frame becomes. H4: The criticism of authority frame is likely to appear in connection with the right-wing terrorism frame. H5: The geographic proximity news factor shapes the framing of regional newspapers. Below, as a
first step, some general descriptive findings will be presented. Afterwards, each hypothesis will be tested using descriptives and One-Way Anova’s results (see, Table 3 and 4, appendix).
For the general understanding and interpretation of the results it is crucial to look at the basic article count for both the different types of newspapers and the separate years. There
are striking differences between the newspapers as to the amount of articles about the NSU-trial published in the given time period. For the regional newspapers the range is between 31 articles (Kölner Stadtanzeiger) and 239 articles (Thüringer Allgemeine). For the national newspapers the range is between 53 articles (Die Tageszeitung) and 155 articles (Süddeutsche
Zeitung). As mentioned above, a sub-selection was made for Thüringer Allgemeine and Süddeutsche Zeitung (see figures for all newspapers in codebook, appendix). Thereby the total
count for national and regional newspapers was aligned. The different article counts per newspaper show that they placed different emphasis on quantity in their reporting, and these differences are even bigger for the chosen regional newspapers. It is likely that editorial decisions were made at each newspaper, which led to these counts. Equally important, and true for all six newspapers analyzed, is the fact that the number of published articles decreased considerably over time (see, Graph 1). The total number of articles was 281 for year 1, 140 for year 2, 127 for year 3, and 46 for year 4. It is important to look at the frame presence in proportion to the article count to draw proper conclusions as to changes in
framing over time. Moreover, Graph 1 shows that there were highs and lows in the amount of reporting during the first three years, whereas in year 4 the article count stayed low
throughout the whole time. Most articles (41) were published in May 2013 when the trial began. Least articles (1) were published in March 2017 at the end of this analysis.
Graph 1. Article Count per Year
0 10 20 30 40 50
Apr May Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar
Year 1 Year 2 Year 3 Year 4
Hypothesis testing
The first hypothesis is H1: The further the trial proceeds the less present the court
case frame becomes. To test this hypothesis a closer look at the court case frame is needed.
Figures indicate that the court case frame is most present of all frames. It has to be taken into account, however, that the court case frame was measured with five sub-questions. To get a more balanced comparison of frame presence, only those three court case indicators with the highest count were added up and put into proportion with the other frames (the same applies for the personification frame). As Graph 2 shows, the court case frame was dominant in the reporting throughout all years. Even in year four, where there was very little coverage at all, the court case frame was still dominant. This means that against this study’s assumptions, no other frame became more present than the court case frame during the four years. It was assumed that the court case frame would become less present as the routinely trial
proceedings might not be sufficiently newsworthy throughout the years, but it seems, instead of enhancing other frames, the coverage was simply generally reduced.
Graph 2. Distribution of Frames per Year
To test the hypothesis statistically an Anova was conducted. As mentioned above, a Levene’s test revealed that the assumption of equal variances in the population has been violated for the court case frame, which is why the results are to be interpreted with particular caution. Corresponding with the descriptive figures above the Anova revealed that
0 50 100 150 200 250
Year 1 Year 2 Year 3 Year 4
Court Case Personific. Right-Wing Authority Proximity
proportional with the article count there were no significant differences between the years regarding the presence of the court case frame, F(3,357) = 0.624, p = .600. These findings also hold true for both national newspapers, F(3,184) = 0.643, p = .588, and regional newspapers, F(3,169) = 0.110, p = .954. Consequently, H1 needs to be rejected.
The second hypothesis is: H2: The main accused Zschäpe is dominant within the
personification frame. A closer look at the personification frame shows that the main accused
Zschäpe is indeed the person that has most been personalized in the overall reporting (87 times, see, Table 5). Second most personalized group is witnesses (68 times), followed by victims (56 times), other defendants (35 times), and the judge (7 times). When looking at the individual years, however, one can see that in year 1 Zschäpe was personalized in as many articles as the victims (both 35 times) and less often than the witnesses (39 times). In year 2 witnesses again have been personalized in more articles (17 times) than the main accused Zschäpe (15 times). Yet, in the next two years Zschäpe was clearly dominantly personalized (25 times in year 3 / 12 times in year 4). Consequently, Zschäpe became more dominant within the personification frame as media attention and reporting on the NSU-trial decreased. Summed up, the findings support H2 only partly, namely for year 3 and year 4.
Table 5. Personification Frame Count per Year
Personific. 1 Personific. 2 Personific. 3 Personific. 4 Personific. 5 Total
Year 1 35 24 35 6 39 139
Year 2 15 5 13 0 17 50
Year 3 25 4 6 1 10 46
Year 4 12 2 2 0 2 18
Total 87 35 56 7 68
The third hypothesis is: H3: The further the trial proceeds, the more present the
right-wing terrorism frame becomes. This hypothesis was based on the assumption that news
factors such as influence and relevance would gain increasing importance over time, as reporters might place emphasis on background information on right-wing extremism. The
frame encompasses both information about societal consequences or the impact of the NSU terror attacks on society and information about the terrorist organization itself and the extreme right scene in Germany. Figures show that the right-wing terrorism frame was the second most present frame throughout all years (see, Graph 2). What is more, for the right-wing terrorism frame an Anova has found significant differences between the years, F(3,357) = 3.939, p = .009. A Bonferroni Post Hoc Test was conducted to further investigate, which years exactly significantly differed from each other. It revealed that means significantly increased from year 1 to year 2, Bonferroni = .356, p = .001. Given these results, there was indeed an increase of the right-wing terrorism frame in the aftermath of the first year.
Throughout the following years, however, no such increase was measured. Thus, the findings only partly support H3.
The fourth hypothesis is: H4: The criticism of authority frame is likely to appear in
connection with the right-wing terrorism frame. This hypothesis was based on the assumption
that a contextual connection exists between the right-wing terrorism frame and the criticism
of authority frame. In other words, when there is information given about either societal
consequences of the NSU terror attacks or the NSU as a right-extremist terrorist organization, it is likely that there will be also criticism towards authorities such as the police or the
German “Verfassungsschutz”. In general, the criticism of authority frame is less present than the right-wing terrorism frame (see, Graph 2). Moreover, an Anova has found no significant differences between the years, F(3,357) = 2.074, p = .103. Nevertheless, the factor analysis revealed interesting correlations between certain sub-questions of both frames. One
component is formed by the right-wing terrorism question “Does the story analyze societal consequences of the NSU terror attacks in Germany?” (right-wing 2, .71) and the criticism of
authority questions “Does the story contain critique regarding state authorities such as the
critique regarding the investigation of the NSU crimes?” (authority 3, .67). Consequently, when societal consequences – such as public stigmatizations of immigrants, integration issues, social marginalization of immigrants (see codebook in appendix) – are mentioned, criticism of state authorities is likely to be mentioned as well. However, as this is not the case for the frames as a whole, H4 needs to be partly rejected.
The fifth hypothesis is: H5: The geographic proximity news factor shapes the framing
of regional newspapers. This hypothesis only concerns the regional newspapers and was
based on the assumption that regional newspapers place different emphasis on different points in time of the trial due to proximity. The regional newspapers analyzed all represent different local regions and individual experiences with the terror-organization, as for Kölner
Stadtanzeiger the bomb attacks in Cologne or for Nürnberger Nachrichten the murders in
Nuremberg. The Thüringer Allgemeine is a regional newspaper from the Free State of
Thuringia, where most of the defendants, including the main accused Zschäpe are from. First of all, the geographic proximity frame is the only frame, which sub-questions have been identified as one self-contained component by the factor analysis (.87, .79, .89). Moreover, an Anova has found significant differences between the years for this frame, F(3,169) = 3.054, p = .030. Additionally, a Bonferroni Post Hoc Test was conducted to investigate, which years exactly differed from each other significantly. This test revealed that means significantly increased from year 1 to year 2, Bonferroni = .602, p = .018. For all other frames, however, Anovas have found no significant differences between the years in regional newspapers: court
case frame, F(3,169) = 0.110, p = .954 / personification frame F(3,169) = 1.675, p = .174 / right-wing terrorism frame, F(3,169) = 2.649, p = .051 / criticism of authority frame, F(3,169) = 2.252, p = .084. Thus, the geographic proximity frame does not seem to have an
Discussion
This study’s research question is: To what extent did framing in the coverage of the
NSU-trial in German quality newspapers change over time? For the quantitative content
analysis a four years time period was chosen – from the very beginning of the trial in 2013 onwards. Based on framing theory and the concept of news values, five frames were
developed in order to categorize different aspects of reporting and to measure their presence from year to year. Each frame is accompanied by one hypothesis.
First of all, this study revealed a sharp decrease in media attention over the years, especially from year 3 to year 4. This decrease does not come as a surprise, given the
bureaucratic characteristics of court reporting (Greenhouse, 1996). As elaborated above, court reporting in general places challenges to journalists. After all, it is difficult to cover court proceedings in such a way that readers do not lose their interest in trial details as hearings go on. Naturally, the public’s interest in big court cases is higher at the beginning and declines when trial routines become ever more repetitive (Greenhouse, 1996). The NSU-trial is no exception in this regard, as findings show.
It was the assumption of this study that framing in the coverage of the NSU-trial would change over time, namely in such a way that daily trial routines become less important, whereas background information as to right-wing terrorism and societal consequences as well as criticism towards authorities would increase in order to keep the issue present and relevant in the news. These assumptions were backed up by the concept of news values (Galtung & Ruge, 1965), which journalists tend to consider in their reporting, as researchers have argued many times before (see, e.g., Harcup & O’Neill, 2001; Boukes & Vliegenthart, 2016).
However, most findings do not support the assumption that variation in frames and news values occur as time passes by. In fact, very little changes were measured. The first hypothesis H1 assumed that the court case frame would become less present over time. This
assumption could not be supported. Instead, the court case frame was dominant throughout the whole time period and there were no significant differences between the years. The series of Anovas and Bonferroni Post Hoc Tests only revealed significant differences between year 1 and year 2 for the right-wing terrorism frame and the geographic proximity frame. All other frames did not show significant differences between the years. The results show that after the first two years, instead of adjusting framing, reporting was rather reduced generally, which led to a sharp drop of articles after the third year. Hence, it seems framing and news values as journalistic tools lost influence as the trial went on.
Findings as to Hypothesis H3 intensify this impression. The right-wing terrorism frame was assumed to become more present over time. The hypothesis was partly supported by statistical results, namely between year 1 and year 2. Thus, the assumed effect could be witnessed in that period, which shows that a shift in framing took place, yet only in the first half of the analysis’s time frame. The longer the trial lasted, the less flexibility in framing was measured for the right-wing terrorism frame. One reason for these findings might be that journalists from the very beginning tried to draw a broad picture and to guarantee well-balanced reporting (Bier, 2015). But it seems, the farther the trial got extended, the harder it became for court-reporters to actually provide new insights, find new angles, and tell the reader something they have not read yet.
Nevertheless, it is important to keep in mind that a link was measured, albeit only partly, between the right-wing terrorism frame and the criticism of authority frame, as suggested in hypothesis H4. This is particularly relevant, as it shows that journalists, to a certain extent, did report about issues beyond trial routines in the form of criticism towards authorities. There are no indicators, which suggest that court reporters would have covered the NSU-trial only according to official portrayals, as Bennett, Lawrence and Livingston (2006) investigated in their study on reporting on torture in the context of the Abu Ghraib
Scandal. Regarding the NSU-trial, however, it seems that journalists indeed took their responsibility seriously to create awareness in society as to the relevance, origin and consequences of right-wing terrorism.
A split between the first half and the second half of the analyzed time period becomes also visible when looking at the results for hypothesis H2. It was assumed that the main accused Zschäpe would be dominant within the personification frame. Regarding the total figures for the coverage over the whole time period statistical results supported this
assumption. When looking at the separate years, however, it gets clear that Zschäpe was only dominant in year 3 and year 4. In the first two years, witnesses were the most personalized actors in court. In year 1, victims were personalized in as many articles as Zschäpe. These results, again, show that journalists placed particular emphasis on well-balanced reporting at the beginning of the trial. The dominance of Zschäpe in the second half of the analyzed time period suggests, however, that these endeavors were neglected the longer the trial lasted.
The fifth hypothesis H5 assumed that geographic proximity shaped framing in regional newspapers. It shall be discussed with particular regard to the regional newspaper
Thüringer Allgemeine, which of all analyzed dailies published the highest number of articles.
At the beginning of the trial Thüringer Allgemeine declared in one article they considered themselves particularly capable of reporting thoroughly and fairly on the NSU-trial, as most of the defendants grew up in the Free State of Thuringia and East German newspapers knew best their milieus, parents, relatives and friends (Thüringer Allgemeine, 2013). However, despite the sheer amount of articles, there were no particularities in framing compared to national newspapers. Indeed, statistical tests revealed that means of the geographic proximity frame significantly increased from year 1 to year 2. Yet, for all other frames no such
Limitations
The study at hand is the attempt to shed light on framing characteristics as to the coverage of the first four years of NSU-trial in German quality newspapers. Due to time restrictions, it was not possible to enlarge the sample or the time period. Although the selection was based on thorough considerations such as distribution figures and range of newspapers, the findings do not represent the German press market as a whole. Moreover, statistical results need to be interpreted with caution. First, intercoder reliability was not sufficiently reliable for all sub-questions. Second, four out of five frames were not identified as one component by the conducted factor analysis. Third, the Levene’s Test of Homogeneity of Variances revealed that the assumption of equal variances in the population has been violated for some frames. Nevertheless, future studies on framing and court reporting should take these findings into account. The developed framework can be used as guidance and adjusted were needed. Regarding the particular case of the NSU-trial it would be interesting to again look at the coverage when the rendition of judgment approaches.
Conclusion
The NSU-trial placed unprecedented challenges to journalists and media outlets. Due to the massive failures regarding the investigation towards the series of murders, high
expectations were held at the beginning of the trial. Victims and their family members as well as the general public claimed justice and it was the court reporters’ responsibility to
extensively provide information about daily trial proceedings. Additionally, journalists were in charge of holding administrative bodies to account and shedding light on organizational structures of right-wing extremism in Germany. Throughout this process media organizations needed to cope with the particularities of court reporting and cover the bureaucratic trial routines in such a way that interest would not fall off. This study showed that the longer the trail lasted, the fewer articles were published. Indeed, there was well-balanced reporting in the
first half of analyzed time period. However, the fact that framing only changed marginally over time and that reporting was reduced drastically in the fourth year of court hearings suggests that journalists were quite restricted both by the undynamic character of court hearings and the general decreasing media attention in the course of the NSU-trial.
References
Behrens, v. d., A. (2017, March 7). Summary of Political, Social and Legal Aspects of the Case against the National Socialist Underground (NSU) – as of March 7th, 2017. Retrieved April 10, 2017, from https://www.nsu-watch.info/2017/03/the-nsu-case-in-germany/
Bennett, W. L., Lawrence, R. G., & Livingston, S. (2006). None Dare Call It Torture:
Indexing and the Limits of Press Independence in the Abu Ghraib Scandal. Journal of
Communication, 56(3), 467-485. doi:10.1111/j.1460-2466.2006.00296.x
Bier, N. (2015). Rechts in der Mitte? Journalisten und der NSU-Prozess. Aus Politik und
Zeitgeschichte, 65(40), 48-54. Retrieved February 8 2017, from
http://www.bpb.de/apuz/212369/journalisten-und-der-nsu-prozess?p=all
Boesman, J., Berbers, A., D’Haenens, L., & Gorp, B. V. (2017). The news is in the frame: A journalist-centered approach to the frame-building process of the Belgian Syria fighters. Journalism, 18(3), 298-316. doi:10.1177/1464884915610988
Boukes, M. & Vliegenthart, R. (2016). A General Pattern of Newsworthiness?:
Analyzing News Factors in Tabloid, Broadsheet, Financial, and Regional Newspapers. Not yet published
De Swert, K. (2012, February 01). Calculating inter-coder reliability in media content analysis using Krippendorff’s Alpha. Retrieved May 17, 2017, from
http://www.polcomm.org/wp-content/uploads/ICR01022012.pdf
De Vreese, C. H. (2005). News framing: Theory and typology. Information Design Journal,
13(1), 51-62. doi:10.1075/idjdd.13.1.06vre
Entman, R. M. (2010). Improving Newspapers' Economic Prospects by Augmenting Their Contributions to Democracy. The International Journal of Press/Politics, 15(1), 104-125. doi:10.1177/1940161209352371
Entman, R. M. (1993). Framing: Toward Clarification of a Fractured Paradigm. Journal of
Communication, 43(4), 51-58. doi:10.1111/j.1460-2466.1993.tb01304.x
Esser, F., & D'angelo, P. (2003). Framing the Press and the Publicity Process: A Content Analysis of Meta-Coverage in Camapign 2000 Network News. American Behavioral
Scientist, 46(5), 617-641. doi:10.1177/0002764202238489
Falkheimer, J., & Olsson, E. (2014). Depoliticizing terror: The news framing of the terrorist attacks in Norway, 22 July 2011. Media, War & Conflict, 8(1), 70-85.
doi:10.1177/1750635214531109
Fekete, L. (2016, November 01). The NSU scandal – investigations into child murders re-opened. Retrieved April 11, 2017, from https://www.nsu-watch.info/2016/11/the-nsu-scandal-investigations-into-child-murders-re-opened/
Galtung, J., & Ruge, M. H. (1965). The structure of foreign news. Journal of Peace Research,
2(1), 64-91.
Greenhouse, L. (1996). Telling the Court's Story: Justice and Journalism at the Supreme Court. The Yale Law Journal, 105(6), 1537. doi:10.2307/797296
Hansen, A. (2015). Journalistische Characterisierung der Akteure im ‘NSU’-Prozess. Eine qualitative Inhaltsanalyse der Berichterstattung über die ersten 150 Prozesstage. Retrieved February 8, 2017, from
https://www.otto-brenner- stiftung.de/fileadmin/user_data/stiftung/Aktuelles/AH79/Masterarbeit_NSU-Gerichtsberichterstattung_Hansen_UHH_2015.pdf
Harcup, T., & O'neill, D. (2001). What Is News? Galtung and Ruge revisited. Journalism
Studies, 2(2), 261-280. doi:10.1080/14616700120042114
Hilgartner, S., & Bosk, C. L. (1988). The rise and fall of social problems: A public arenas model. American Journal of Sociology, 94(1), 53-78.
Behavior, 12(1), 19-40. doi:10.1007/bf00992330
Johnson-Cartee, K. S. (2005). News narratives and news framing: constructing political
reality. Lanham: Rowman & Littlefield.
Kuypers, J. A. (2006). Bush's war: media bias and justifications for war in a terrorist age. Lanham, MD: Rowman & Littlefield.
Merkel, A. (2012, February 23). "Die Hintergründe der Taten lagen im Dunkeln - viel zu lange". Retrieved April 10, 2017, from http://www.sueddeutsche.de/politik/merkels- gedenkrede-fuer-neonazi-opfer-im-wortlaut-die-hintergruende-der-taten-lagen-im-dunkeln-viel-zu-lange-1.1291733
Molotch, H., & Lester, M. (1975). Accidental News: The Great Oil Spill as Local Occurrence and National Event. American Journal of Sociology, 81(2), 235-260.
doi:10.1086/226073
Neuendorf, K. A. (2002). The content analysis guidebook. Thousand Oaks, CA: Sage.
Norddeutscher Rundfunk. (2013). Akkreditierte Medien/Medienvertreter in dem Strafverfahren gegen Beate Z. u.a. (NSU) . Retrieved February 22, 2017, from
https://web.archive.org/web/20130522093305/http://www.ndr.de/fernsehen/sendungen/ zapp/akkreditierungsliste101.pdf
Powell, K. A. (2011). Framing Islam: An Analysis of U.S. Media Coverage of Terrorism Since 9/11. Communication Studies, 62(1), 90-112. doi:10.1080/10510974.2011.533599 Reese, S. (2007). Journalism research and the hierarchy of influence model: a global
perspective. Brazilian Journalism Research, 3(2), 29-42.
Semetko, H., & Valkenburg, P. M. (2000). Framing European politics: a content analysis of press and television news. Journal of Communication, 50(2), 93-109.
doi:10.1093/joc/50.2.93
auf einen Blick › Meedia. Retrieved May 19, 2017, from
http://meedia.de/2017/01/23/die-grosse-auflagen-analyse-der-regionalzeitungen-die-78-groessten-titel-auf-einen-blick/
Schuck, A. R., & De Vreese, C. H. (2006). Between Risk and Opportunity: News Framing and its Effects on Public Support for EU Enlargement. European Journal of
Communication, 21(1), 5-32. doi:10.1177/0267323106060987
Statista. (2017). Auflagenstärkste Zeitungen Deutschlands Q1 2017. Retrieved May 19, 2017, from
https://de.statista.com/statistik/daten/studie/73448/umfrage/auflage-der-ueberregionalen-tageszeitungen/
Thüringer Allgemeine. (2013, May 9). Offener Brief aus Thüringen gegen die hochmütige FAZ. Thüringer Allgemeine. Access via Factiva.
Van Gorp B (2010) Strategies to take subjectivity out of framing analysis. In: D’Angelo P and Kuypers JA (eds) Doing News Framing Analysis: Empirical and Theoretical
Perspectives. New York: Routledge, pp. 84–109.
Virchow, F., Thomas, T., & Grittmann, E. (2015). "Das Unwort erklärt die Untat" Die Berichterstattung über die NSU-Morde – eine Medienkritik. Retrieved January 11, 2017, from https://www.otto-brenner-stiftung.de/otto-brenner-stiftung/aktuelles/das-unwort-erklaert-die-untat.html
Appendix
Table 1. Krippendorff’s Alpha Reliability Values per Variable
N = 36, observers = 2, number of bootstrap samples = 10000
Variables Kalpha Probability that alpha < .80 Probability that alpha < .60
Court Case 1 0.90 p = .080 p = .003 Court Case 2 0.81 p = .353 p = .025 Court Case 3 0.72 p = .605 p = .335 Court Case 4 0.93 p = .017 p = .001 Court Case 5 0.91 p = .083 p = .003 Personification 1 0.83 p = .355 p = .009 Personification 2 0.78 p = .586 p = .076 Personification 3 0.84 p = .316 p = .013 Personification 4 0.61 p = .775 p = .373 Personification 5 0.56 p = .936 p = .564 Right-Wing 1 0.77 p = .573 p = .094 Right-Wing 2 0.65 p = .762 p = .397 Right-Wing 3 0.72 p = .779 p = .202 Authority 1 0.60 p = .779 p = .379 Authority 2 0.88 p = .139 p = .003 Authority 3 0.36 p = .968 p = .827 Proximity 1 0.79 p = .584 p = .076 Proximity 2 1.00 p = .363 p = .363 Proximity 3 0.91 p = .079 p = .004
Table 3. Descriptives Oneway Anova Year N Mean Std. Deviation Std. Error 95% Confidence Interval for Mean
Min. Max. Lower Bound Upper Bound Court Case 1,00 177 1,6441 ,96700 ,07268 1,5006 1,7875 ,00 4,00 2,00 79 1,6076 1,18124 ,13290 1,3430 1,8722 ,00 5,00 3,00 72 1,7500 ,94571 ,11145 1,5278 1,9722 ,00 4,00 4,00 33 1,8485 ,97215 ,16923 1,5038 2,1932 ,00 3,00 Total 361 1,6759 1,01253 ,05329 1,5711 1,7807 ,00 5,00 Personific. 1,00 177 ,7853 ,78278 ,05884 ,6692 ,9014 ,00 3,00 2,00 79 ,6329 ,64405 ,07246 ,4887 ,7772 ,00 2,00 3,00 72 ,6389 ,79267 ,09342 ,4526 ,8252 ,00 3,00 4,00 33 ,5455 ,50565 ,08802 ,3662 ,7248 ,00 1,00 Total 361 ,7008 ,73728 ,03880 ,6245 ,7771 ,00 3,00 Right-Wing 1,00 177 ,7458 ,74452 ,05596 ,6353 ,8562 ,00 3,00 2,00 79 1,1013 ,81011 ,09114 ,9198 1,2827 ,00 3,00 3,00 72 ,8889 ,74220 ,08747 ,7145 1,0633 ,00 2,00 4,00 33 ,9394 ,93339 ,16248 ,6084 1,2704 ,00 3,00 Total 361 ,8698 ,78684 ,04141 ,7884 ,9512 ,00 3,00 Authority 1,00 177 ,5537 ,70606 ,05307 ,4489 ,6584 ,00 3,00 2,00 79 ,5823 ,76137 ,08566 ,4117 ,7528 ,00 3,00 3,00 72 ,3333 ,50351 ,05934 ,2150 ,4517 ,00 2,00 4,00 33 ,4848 ,83371 ,14513 ,1892 ,7805 ,00 3,00
Total 361 ,5097 ,69963 ,03682 ,4373 ,5821 ,00 3,00 Proximity 1,00 177 ,2712 ,74232 ,05580 ,1611 ,3813 ,00 3,00 2,00 79 ,5696 ,99593 ,11205 ,3465 ,7927 ,00 3,00 3,00 72 ,4444 ,87031 ,10257 ,2399 ,6490 ,00 3,00 4,00 33 ,1515 ,50752 ,08835 -,0284 ,3315 ,00 2,00 Total 361 ,3601 ,82190 ,04326 ,2750 ,4452 ,00 3,00
Table 4. Oneway Anova
Sum of Squares df Mean Square F Sig.
Court Case Between Groups 1,926 3 ,642 ,624 .600
Within Groups 367,154 357 1,028
Total 369,080 360
Personific. Between Groups 2,701 3 ,900 1,665 .174
Within Groups 192,989 357 ,541
Total 195,690 360
Right-Wing Between Groups 7,142 3 2,381 3,939 .009
Within Groups 215,739 357 ,604
Total 222,881 360
Authority Between Groups 3,018 3 1,006 2,074 .103
Within Groups 173,198 357 ,485
Total 176,216 360
Proximity Between Groups 6,815 3 2,272 3,431 .017
Within Groups 236,370 357 ,662
Codebook
Framing of the coverage of the NSU trial in German quality newspapers
General information about the NSU-trial
Four years have elapsed since the opening of the trial against the extreme-right terrorist organization National Socialist Underground (NSU). The NSU is accused of having committed ten politically motivated murders all over Germany between 2000 and 2007. Among the victims were eight citizens of Turkish and Kurdish origin, one Greek man, and one German police officer. The main accused is Beate Zschäpe, alleged member and only survivor of the so-called NSU-trio. Besides her, four other defendants are accused of having supported the terrorist organization with arms and false documents. The NSU-trial is the biggest court case in Germany since reunification, involving five judges, thirteen defense lawyers, three federal prosecutors and approx. fifty lawyers representing the private accessory prosecutors (victims or family members of murdered victims who have the right to join the proceedings).
General information about the analysis
It is the aim of the thesis to analyze the coverage of the NSU-trial in German quality newspapers over a four years time period from April 2013 to March 2017. The quantitative content analysis is built upon the theory of framing. Five frames were developed as content analytic variables (further specification and instructions below):
1. Court Case 2. Personification 3. Right-Wing Terrorism 4. Criticism of Authority 5. Geographic proximity
The articles were collected via the database Factiva. The search-term “NSU-Prozess” had to appear either in the headline or lead paragraph of the article. Articles with less than 500 words were not included. Three national newspapers and three regional newspapers were chosen. Intercoder reliability is measured based on 10 percent of the articles (36 articles).
Süddeutsche Zeitung 155 / framing analysis on every second: 78
Die Welt 57
Tageszeitung 53
Total 265 / framing analysis: 188
Kölner Stadtanzeiger 31
Nürnberger Nachrichten 63
Thüringer Allgemeine 239 / framing analysis on every third: 79
Total 333 / framing analysis: 173
Total 598 / framing analysis: 361
General principles of coding
- Please don’t interpret, code based on the manifest content!
- Please be conservative while coding, if you have a doubt, don’t code! - Please be patient, if there are long articles, just code sentence by sentence! - Please follow the instructions for frame identification precisely!
1. Identification
Article ID: (as given in Word document) Coder ID: NN (Your initials)
Date: YYYY-MM-DD Media Outlet:
National newspapers
212. Die Welt (WE) 213. Tageszeitung (TZ)
214. Süddeutsche Zeitung (SZ)
Regional newspapers
311. Kölner Stadtanzeiger (KS) 312. Nürnberger Nachrichten (NN) 313. Thüringer Allgemeine (TA)
Journalist’s name Author:
Article’s headline Headline:
Prominence of the article
Page number: Number of words:
2. Frames
Please answer the questions with Yes or No. Code YES = 1 and NO = 0.
The Court Case Frame
Note: The court case frame encompasses basic information about the trial and trial-proceedings. Parameters counting to the court case frame are: information about the
groups in court, information about the premises in court, and information about the presence of the media. Please answer the following questions:
1. Does the story contain information about the indictment?
E.g.: “Dem "Nationalsozialistischen Untergrund" werden zehn Morde zur Last gelegt - begangen zwischen den Jahren 2000 und 2007.
2. Does the story contain information about the duration of the trial?
This is not the case when only the number of trial day is given. Rather, there should be information as to the weeks or years that have passed or as to how long the trial will possibly last.
E.g.: “Unterdessen deutet sich an, dass der NSU-Prozess länger dauern dürfte als ursprünglich geplant. Am Montag verschickte das Münchner Gericht an die Prozessbeteiligten eine Liste mit Terminen bis Ende Juni 2015.“
3. Does the story contain information about the different actor groups in court?
This is the case whenever one of the actors or actor groups is present as a participant in courtroom, meaning the defendants, the defense lawyers, the prosecution, the victims and family members, the judges, or the witnesses. 4. Does the story contain information about the scene (the courtroom)?
This is the case whenever there is information about how the courtroom looks like, e.g. the size, the seating arrangements, the furniture, the break rooms, the missing windows, the air condition.
5. Does the story contain general information about media presence? This is the case whenever there is information about the media covering the trial.
E.g: “Die politische Dimension ist verkannt worden. Sonst hätte man nie auf die Idee kommen können, für die Medienplätze ein Windhundverfahren zu beschließen. Mit dem Risiko, dass es schief geht.“
The Personification Frame
Note: The personification frame implies that a story accentuates individual actors by providing personal information, such as outward appearance, character traits, relationships, sexual orientation, family matters, emotional life, health, profession, and interest motives.
1. Does the story provide personal information about Beate Zschäpe?
E.g.: “Als Zschäpe dann eintrifft, dieses Mal im grauen Anzug, die Mähne im Pferdeschwanz gebändigt, an den Ohren wieder die riesigen Ringe, …”
2. Does the story provide personal information about the other defendants? The names of the other defendants are: André E., Holger G., Carsten S., and Ralf Wohlleben.
E.g.: “Eine wichtige Rolle spielt dabei die schwierige „sexuelle Identitätsfindung“: Carsten S. ist homosexuell, hat sich aber erst spät dazu bekannt...“
3. Does the story provide personal information about the victims or their relatives? E.g.: “Hoffmann vertritt eine Frau aus der Keupstraße, die hochschwanger war, als die Bombe explodierte. Sie war mit ihrem älteren Kind in der Wohnung. Die Frau leidet unter Depressionen und Panikstörungen.“ 4. Does the story provide personal information about the judges?
E.g.: “Tatsächlich führt Götzl seine Verfahren gern in forschem Ton und Tempo.“ / „Götzl, 59 Jahre alt und offenbar von unverwüstlicher Kondition...“
5. Does the story provide personal information about the witnesses?
E.g.: “Der Vater des mutmaßlichen Mehrfach-Mörders Uwe Mundlos soll an diesem Tag aussagen beim NSU-Prozess, ..., da zieht der pensionierte Informatikprofessor einen Pack Notizen aus seiner Aktentasche und spricht.”
The Right-Wing Terrorism Frame
Note: The right-wing terrorism frame occurs whenever the NSU is discussed in more detail or in a broader societal context broaching the issue of right-wing extremism in Germany.
1. Does the story provide background information about the NSU?
Foundation, radicalization of members, past underground operations, organizational structure, supporters, domiciles, holidays, relationships among the three members
E.g.: “Offensichtlich traf man sich im Keller auch zu Schießübungen. Sie ballerten mit Waffen auf Holzplatten, die die Beamten sicherstellten. Diese waren von den Projektilen völlig durchlöchert. Auch Dutzende Straßen- und Stadtkarten fanden die Beamten, an den Rändern versengt. Im gesamten Bundesgebiet mordete und raubte das Trio.”
2. Does the story analyze societal consequences of the NSU terror attacks in Germany? E.g. public stigmatization of immigrants, integration issues, social marginalization of immigrants)
haben, dass der Druck gegen Nichtdeutsche höher wird, dass den Bewohnern das womöglich selbst in die Schuhe geschoben wird. Nach dem Strafgesetzbuch sind sowohl die Beweggründe als auch die Folgen der Tat für die Opfer die Grundlage für die Strafzumessung. Insofern ist es zwingend, die Situation der Opfer nach der Tat vollständig mit einzubeziehen.”
3. Does the story put the NSU in the broader context of right-wing extremism? E.g. information about the extreme right scene, neo-nazi organizations, NPD, Thüringer Heimatschutz, xenophobia, hate crimes against immigrants/refugees
E.g.: “Einer Analyse der Amadeu Antonio Stiftung zufolge hätten die rassistischen Morde des
"Nationalsozialistischen Untergrunds" (NSU) sogar früher aufgedeckt werden können, wenn rechtsextreme Frauen bisher nicht grundsätzlich unterschätzt worden wären. Frauen seien in der rechten Szene längst nicht mehr nur die "Freundin von", sondern genauso an Straftaten beteiligt wie Männer.”
The Criticism of Authority Frame
Note: The criticism of authority frame occurs if a story contains criticism of administrative bodies, either the court (the judges) or state authorities such as e.g. the “Verfassungsschutz”, the police or politicians.
1. Does the story contain critique regarding the functioning of the court? E.g.: “Bedeutsamer noch: Womöglich hat Richter Manfred Götzl einen ersten schweren Fehler gemacht. Er selbst berichtet am Morgen, dass ihm bereits Ende August ein Kanzleikollege Grasels, der renommierte Münchner Strafverteidiger Hermann Borchert, eine schriftliche Erklärung Zschäpes angekündigt hatte. Seitdem wurde über eine Aussage beraten. Nur die drei Altverteidiger von Zschäpe informierte Götzl nicht.“
2. Does the story contain critique regarding state authorities such as the police, politicians or the “Verfassungsschutz”?
E.g.: “Unterdessen steht derzeit eine andere Frage im Raum: Ab wann wussten Verfassungsschutz und Ermittler von der Existenz des Nationalsozialistischen Untergrunds (NSU)? Sowohl der Untersuchungsausschuss im Bayerischen Landtag in München als auch der NSU-Prozess vor dem Münchner Oberlandesgericht (OLG) lassen starke Zweifel an der bisherigen Darstellung der Behörden aufkommen.“
3. Does the story contain critique regarding the investigation of the NSU crimes? E.g.: “Ich will nicht behaupten, dass die Polizei damals die Täter hätte festnehmen können. Aber die Polizei hätte, wenn sie von Anfang an auch die Möglichkeit einer rechtsterroristischen Aktion einbezogen hätte, nicht
