Violent computer games in the German press

Abstract

The most heated public debates on the subject of violent computer games in Germany take place following incidents of school shootings. Such reactions are often conceptualized as moral panics and signs of underlying social conflict. Focus is rarely on the violent computer games themselves. Actor-network theory allows for an analysis of how phenomena are sequentially drawn together, contingent upon the material available for the press at specific times, to which violent computer games can be related. Six months of press coverage following the 2006 school shooting in the German town of Emsdetten were not a continuous narrative of violent computer games, but divided into six distinct phases. In these, violent computer games achieved several different identities. Both the way the material was brought together in the press and the contingent events beyond the context of publishing houses were decisive for which identities were generated in the press, and which of these became the most enduring.

Keywords

Actor-network theory Germany mass media moral panic press school shootings violent computer games

Introduction

On 20 November 2006, an 18-year-old armed with two antique guns, one small-bore rifle and eight pipe bombs entered his former school in the German town of Emsdetten. Sebastian B. shot and injured 37 people before killing himself. The next morning, reports on the shooting at the Geschwister-Scholl-Schule dominated the front pages of German newspapers, and with that the debate about violent computer games was launched in the columns of the daily press and persisted during the following months.

The most dominant question in scholarly papers on computer games and violence is whether there is a causal link between violent computer games and violent and aggressive acts such as school shootings. This paper does not seek to answer this question but rather it investigates the way that violent computer games, the object of discussion, have come to be represented in the press. An analysis of six months’ press coverage of violent computer games in the four largest national newspapers in Germany following the Emsdetten shooting is presented.

As will be further discussed, the literature on violent computer games in the press coverage of school shootings analyzes this almost exclusively in terms of moral panics (Burns and Crawford, 1999; Kline, 1999; Lawrence and Birkland, 2004; Lawrence and Mueller, 2003; Muschert, 2007a, 2007b; Springhall, 1999; Squire, 2002). This literature treats press coverage in terms of perceptions (Cohen, 2002), problem definitions (Lawrence and Birkland, 2004), types of explanations (Springhall, 1999) and negotiations of meaning (McRobbie, 1994; McRobbie and Thornton, 1995), or concentrates in other ways on peoples’ perspectives on the object reported upon. Moral panic literature has a fine sense of how press interpretations fuel the self-perception of the social groups causing the panic. However, it provides less attention to the object under discussion − violent computer games. Instead of presenting an analysis of peoples’ perspectives on violent computer games this paper focuses on the moral identities that violent computer games themselves achieved by being mediated through the press. Such identities are not necessarily from any particular perspective or perception but result from the association of a variety of events, archive materials and processes taking place in the world. This way of analyzing representations is inspired by science and technology studies (STS) and in particular by actor-network theory (ANT) (e.g. Latour, 2005).

STS has a strong tradition of studying representations of scientific objects (e.g. De Vries, 2007; Latour, 1999b). This paper draws on the STS tradition in analyzing the representation of the moral identities of violent computer games in the press. Using the principle of general symmetry (Callon, 1986) representation refers in ANT not to human interpretation, but to the processes through which phenomena are presented again (re-presented) as new arrangements of human and non-human components. The principle facilitates the study of the production of representations without ascribing prior categories of how agency is distributed in this process (Latour, 2005). Agency may be found in both human and nonhuman objects. On the basis of this the task is to analyze which objects or identities emerge as an effect of the relations observed in the data. It is an inductive, radical empiricist method, which, however, differs from empiricism in that it does not aim to build increasingly robust representations of the world. Instead it seeks to increase the differentiation and variation in descriptions of empirical data in order to develop an increasingly complex understanding of the world and of our representations of it. From this theoretical point of view the paper focuses on what moral identities violent computer games achieved in press coverage after the Emsdetten school shooting.

The paper firstly reviews the moral panic literature. Second, it presents the ANT approach to the study of the moral identities of violent computer games in the press as a result of representation. The empirical study reveals that the press coverage was organized in phases, and the four different identities that violent computer games achieved during the first of these phases are analyzed in detail. Finally, the paper presents an analysis of how one of these identities – violent computer games as an object of politics – came to be the most enduring identity of violent computer games in the press coverage following the Emsdetten shooting.

Violent computer games and media panics

Most of the literature on violent computer games can be found at the intersections of social psychology, aggression research and media effect studies (Anderson et al., 2007, 2010; Arriaga et al., 2011). This literature is heavily criticized for its limited methodological paradigm and questionable results (e.g. Barker and Petley, 2001; Goldstein, 2005). However, scholars of social psychology, aggression and media effects state confidently that the time when that criticism was valid is long past: the scientific evidence today shows unanimously a clear correlation between the use of violent computer games and an increase in the aggression of the players (e.g. Möller and Krahé, 2009).

Social science and media studies approaches to violent computer games generally seek to increase the complexity of data on which they base their studies, including children’s own views on violent computer games (Holm Sørensen and Jessen, 2000), the everyday embeddedness of gaming (Ito et al., 2009), and the content of games and the context of gaming (Williams and Skoric, 2005). However, compared to the dominant interest in the link between games and their users in the research on violent computer games, studies of how such games are represented in the press are rare.

A third line of research embraces culturally and historically oriented analyses of discourses on violent computer games (Otto, 2008; Pethes, 2004; Williams, 2003). This literature draws in general ways on debates circulating in, among other things, the mass media. However, this research does not analyze the press as a specific space of knowledge production. Instead it draws conclusions on general conceptualizations of violent video games in the historical periods in question, based on data from heterogeneous sources.

Research on press representations of violent computer games can be found in moral panic literature. The notion of moral panic originates in Young’s (1971) studies on the social meaning of drug-taking and Cohen’s Folk Devils and Moral Panics first published in 1972.

Cohen opens his seminal work on moral panics as follows:

Societies appear to be subject, every now and then, to periods of moral panic. A condition, episode, person or group of persons emerges to become defined as a threat to societal values and interests; its nature is presented in a stylized and stereotypical fashion by the mass media. (Cohen, 2002: 1)

An analysis is presented of Mods and Rockers, whose clashes at various places in Britain in the mid-60s were extensively covered by the local, national and overseas mass media. Cohen demonstrates how the ‘labeling’ (Becker, 1963) of Mods and Rockers by the press contributed to the formation of a negative image of the two youth groups, who became known in the public understanding as ‘folk devils’. Cohen understands moral panics as transitory and spasmodic mass-mediated public reactions to events perceived as deviant, and their function is to reestablish social order.

Over the following decades the concept of moral panic came to form the basis of a large amount of research whose methods and theoretical approaches vary as much as their empirical objects: mugging (Hall et al., 1978), school shootings (Burns and Crawford, 1999; Lawrence and Birkland, 2004; Lawrence and Mueller, 2003; Muschert, 2007a, 2007b; Springhall, 1999), drugs (Hier, 2002b; Young, 1971), films (De Coninck-Smith, 1999a, 1999b), computer games (Kline, 1999; Squire, 2002), new media in general (Drotner, 1999), and Internet humor (Kuipers, 2006).

While Cohen’s definition of moral panics as ‘a condition, episode, person or group of persons’ (2002: 1) points to social agents or events, Drotner (1992, 1999) emphasizes in her historical review of moral panics in Scandinavia, Britain and Germany that in the twentieth century most moral panics were caused by media such as popular fiction, film and computer games. Consequently she suggests the notion of ‘media panics’ as more precisely characterizing the phenomenon.

The move from moral panic to media panic was accompanied by a change in the character of the object of the panics. The objects of panic and the initiators of panic were the same in Cohen’s account, namely particular social groups. However, in media panics literature, what people panic about – the media – is not the same as what actually causes the panic, which is their ‘human companions’: porn-consumers instead of pornographic images, cartoon-readers instead of cartoon magazines, game-players instead of violent computer games themselves, etc. Consequently, we learn more about human perceptions and relations between social actors than we learn about what happens to the objects entangled in social panics. The priority of analyzing the social and human beings and their activities rather than the objects that are centrally involved in these activities is not specific to the media panics literature. Neither the studies of the effects of violent computer games, the social and media studies of the social embedding of such games, nor the cultural-historical analyses of violent computer game discourses discussed above regard violent computer games themselves as objects with moral identities.

Method and data

This paper investigates the moral identities violent computer games achieved in the press coverage following the Emsdetten school shooting. Following ANT’s emphasis on representation as re-presentation by way of mediation, the process of press coverage in which these representations are situated and through which they are produced are central to the analysis. It is thus an analysis of the specific identities violent computer games achieved due to the specific assemblages they were entangled in as press representations at specific points in time. These specificities are often overlooked when scholars study media panics as an expression of social conflicts existing elsewhere in society.

Based on the principle of general symmetry, the core question guiding this analysis concerns which were the phenomena that violent computer games became associated with in the press coverage. The analysis below differs from the moral panic analyses that are most often based on a collection of typical headlines and quotations taken from the newspapers to illustrate the perspectives conveyed by the press. It presents a sequential study of how moral identities of violent computer games came into being as an effect of the phenomena they became associated with. The understanding that the identities of violent computer games in press coverage are the result of specific processes of drawing things together first requires an analysis of the pattern in which the press coverage unfolded. As we shall see, the articles in the press coverage clustered temporally into phases. Following this analysis the identities that the violent computer games achieved in the first phase are then analyzed. While this analysis defines the identities as they were specifically situated in the first phase, the final analysis in this section looks at the way in which the identities were maintained over time, across the phases. It focuses on the identity that came to be the most durable throughout the press coverage: violent computer games as an ‘object of politics’.

The sample is taken from the four main German national daily newspapers: Süddeutsche Zeitung (SZ), Frankfurter Rundschau (FR), Frankfurter Allgemeine Zeitung (FAZ) and Die Welt.¹ The first two are centre-left, while the FAZ and Die Welt are conservative papers. Following Suler and Openshaw’s (1992) critique that moral panic literature draws conclusions from samples of incompatible types of newspapers, for example tabloids and broadsheets, this analysis is limited to the latter type. The period covered was from 21 November 2006 (the day after the Emsdetten shooting) to the 21 May 2007. All articles that contained the terms ‘school shooting’, ‘Emsdetten’, ‘computer game’ or ‘Killerspiele’² in the full text were collected for analysis from the sources presented in Table 1.

Table 1.

Data type and sources.

Newspaper	Search engine	No. of articles
Süddeutsche Zeitung (SZ)	Süddeutsche Zeitung ArchivLexis Nexis Total Research System	71
Frankfurter Rundschau (FR)	GBI-Genios PresseKontor	66
Frankfurter Allgemeine Zeitung (FAZ)	FAZ: Frankfurter Allgemeine Archiv	57
Die Welt	GBI-Genios PresseKontor	61
		255 in total

Violent computer games in press coverage

The six phases

When ordered quantitatively and temporally as shown in Figure 1 it becomes clear that the 255 articles were not distributed evenly throughout the period covered, but were clustered into six phases when there was a high frequency of articles, interrupted by periods with a few isolated articles or none at all. This section describes the six phases in terms of what elements were drawn together to form them and what patterns they took. The figures in parenthesis in the following text state the number of articles referred to.

Figure 1.

Number of articles in phases.

First phase: the Emsdetten school shooting

The first phase was triggered (Hall et al., 1978) by the Emsdetten school shooting and continued until 2 December 2006. Besides reporting on different issues related to the shooting and its possible causes (57), the press reports dealt with a variety of measures that might prevent further similar events (46). The measure most frequently discussed was a ban on violent computer games (36), mainly demanded by politicians (24), but also discussed by other commentators (17). Characteristic of this phase was its strongly centered focus on the Emsdetten shooting and the relation of the shooting to violent computer games.

Second phase: copycat declarations

The core period of the second phase started on 6 December, even though four articles had already been published on 25 November. The focus of this phase was the declarations of a number of students from all over Germany saying that they would commit school shootings (51). The late November declarations were reported in small notices among less important stories in the newspapers, and they did not at that time appear to belong to a phase. As several more copycat acts were announced in the following weeks, reports increased in number and a phase built up. On 6 December the newspapers reported that the Bavarian government was presenting a Criminal Code bill to ban violent computer games (7). This had been done several times before but without any result. Due to the fact that they appeared at the same time, the copycat declarations and this bill could be related. Reports concerning the two themes mutually reinforced each other and ensured a continuation of the coverage for another 17 days (65).

Third phase: school reopening

The third phase, after the New Year, had a different pattern. It concerned a variety of themes: the reopening of the Geschwister-Scholl-Schule in Emsdetten (5), the passing of the Criminal Code bill through the Bavarian parliament (2), the effectiveness of the German software rating organization USK (2), Internet crime prevention (2), and general reports on computer games and children’s use of computers (5). Each of these topics was only given marginal cover, and it may be hypothesized that most of them would not have found their ways into print had the others not appeared at the same time, allowing for a mutual interrelation to emerge, which developed into another phase consisting of these different reports.

Fourth phase: EU meeting

The focus of the fourth phase, from 15 to 24 January, was the first meeting of EU Internal and Justice Ministers which was held in Germany during the country’s leadership of the EU in the first half year of 2007. Newspapers reported on the ministers’ discussions of EU regulation of violent computer games (9). From 16 to 18 January, they related violent computer games to reports of a double murder committed by two 17-year-olds in the German town of Tessin (7). As was the case in phase 2, two themes were connected to form one phase due to their concurrent appearance. The EU meeting’s agenda had been decided on in advance, which allowed journalists to report on it in December (2), creating an expectation of further coverage. A month after the core period of this phase, the debate was followed up as the EU continued the discussion of the regulation of violent computer games (2). The phase was characterized by infrequent but continuous reporting which intensified when events occurred that could be related to the topic and thus allowed for a reinforcement of the coverage.

Fifth phase: immediate action program

The fifth phase was triggered by the announcement of an immediate action program by the Federal Family Minister (8). This initiative proposed new rating criteria for violent computer games and a revision of the Youth Protection Law (Jugendschutzgesetz). The phase continued with the Bavarian Criminal Code bill being heard and heavily criticized in the Bundesrat (Upper House) (7). Characteristic of this phase was its focus on the current political processes of regulating violent computer games.

Sixth phase: Virginia Tech

After a period of a month and a half when only a few isolated articles on violent computer games appeared in the press, the last phase of the sample began on 17 April, caused by a shooting at Virginia Tech University in the USA. This shooting was related to general debates on the causes of school shootings which mentioned, among other things, computer games (8), which then led to further discussions of a possible ban of such games (5). Furthermore, the US shooting was related to the forthcoming fifth anniversary of the school shooting in the German town of Erfurt (10), which had also been associated with violent computer games. Like phases three and four, this last phase consisted of a variety of events that were not in themselves directly related, but could be drawn together by newspapers to shape a phase because each of them was primarily concerned with the effects of violent computer games.

Patterns of phases

The description of the components of the phases shows different patterns in which violent computer games came to be related to other topics. This section discusses two different parameters according to which the patterns differ: first, references to violent computer games had different strengths in the five phases: only the fifth phase (immediate action program) had violent computer games as its main focus. Here, violent computer games were the dominant issue of the phase, supported by other topics. In the first phase (the Emsdetten school shooting) violent computer games gained saliency over the course of the phase. Computer games started out as only one among many other issues concerning the Emsdetten school shooting, but gained in importance to the extent that by the end of the phase it was the central issue, thus ensuring the persistence of the phase. Violent computer games were of lesser importance in phases two (copycat declarations) and four (EU meeting). Both of these focused on other events which in themselves were not of great importance in the news, but the issue of violent computer games added emphasis and importance to the reports on these events. Violent computer games could only enter these two phases by being related to reports on juvenile crime and on the first meeting of the EU under German leadership. In both phase six (Virginia Tech) and phase 3 (school reopening) violent computer games played very weak roles. The Virginia Tech shooting was extensively related to the Emsdetten shooting, but only a little to violent computer games. Due to the custom of the ‘press archives’³ connecting school shootings to violent computer games, their inclusion was meaningful and natural in the press. The school reopening was in itself a peripheral story, and although phase 3 revolved around the reopening, the phase did not have any clear single focus. It was composed of a variety of weak elements of which violent computer games was only one.

A further parameter according to which the patterns of the phases varied was the distribution of agency between newspapers and occurrences beyond the control of journalists and editors. All six phases were triggered by events taking place outside the context of the publishing houses.⁴ However, only phase 5 (immediate action program) was unambiguously defined as being about violent computer games by actors beyond the newspapers. The relationship of violent computer games to school shootings in phase 1 (the Emsdetten school shooting) was established by politicians and other non-publishing professionals but also by journalists reporting the issue. In the rest of the phases (2, 3, 4 and 6) references to violent computer games were introduced by journalists in one way or another, weaving issues on violent computer games into reports on other matters. This finding indicates that one of the reasons why violent computer games are discussed so prominently in the press is due to the ease with which many different topics can be ‘seasoned’ with references to violent computer games.

Combining reports into phases is a technique that helps newspapers with their basic task of producing new issues every day (Luhmann, 2000). Important topics and events taking place in the world happen more or less by themselves, but there are not enough important topics to fill newspapers. Violent computer games are often reported on only because they can be associated with other topics that either singly or in association with others form a phase.

Identities of violent computer games

After the discussion of the different phases, this section concentrates on the first phase (the Emsdetten school shooting), which comprised 38 percent of the sampled articles. The analysis looks at which identities violent computer games achieved by being associated in the patterns described above, and how these identities were shaped. Although the first phase was framed by a common topic the identities granted to computer games differed with the focus of the particular stories reported. This section presents four of the identities analyzed in the first phase. These four are selected because together they illustrate the variety of identities granted to violent computer games in just a phase.

Sebastian B.’s game

On the first day of the first phase violent computer games were associated with Sebastian B.’s character, as it was reported that he had played Counter Strike which is known as a violent first-person shooter game (11). Ten articles linked Counter Strike with information from the press archives concerning these games in which the player in a virtual scenario handles weapons and searches for enemies to shoot down. This indicates one of the reasons for linking the Emsdetten school shooting with violent computer games, and the history of press coverage of school shootings suggests that this is the usual way of reporting such events.

Several articles described a particular feature of the game that allowed the player to program his own game scenario, a so-called ‘map’. This feature was associated with Sebastian B. by way of the Geschwister-Scholl-Schule map (8) which he had allegedly programmed in Counter Strike in order to prepare himself for ‘shooting up the place’.⁵ Thereby a causal link was established between virtual and real violence and between Counter Strike and Sebastian B.’s shooting.

By drawing these elements together, violent computer games were identified as Sebastian B.’s games and thereby they were associated with a person who, in many ways, was presented as being deviant and who was recognized as the kind of person whom the moral panic literature would call a ‘folk devil’. By being linked to violent computer games his character also became related to the Youth Protection Law. Three articles connected the game with the ‘glorification of violence’ as a ‘youth-endangering’ feature. These notions stem from the German Youth Protection Law which puts forward the central criteria for classifying and banning violent computer games. Through the accounts of Sebastian B.’s character, violent computer games gained the identity of a causal factor leading to school shootings in particular, and to violence in general. It is hardly surprising for readers acquainted with the discussions of violent computer games that they appear in the mass media as one of the causes of school shootings. However, it is worth noticing that, contrary to the image conveyed by academic studies (e.g. Goldstein, 2005; Rutter and Bryce, 2006), this identity did not arise out of discussions based on scientific findings, but derived mainly from the press archives and by links to the Youth Protection Law.

The sports game

Apart from the identity of violent computer games as the cause of violent acts, during the first phase these games were granted a morally positive connotation of a sports game (8). The SZ presented an interview with a spokesperson of the World Cyber Games (the Olympics of online gaming) who emphasized that Sebastian B.:

was not one of our players. Because these games are about team play and tactics, about prudence really, and not about gratification of personal aggression. They are games of sports in which more than one million players participate, Counter Strike alone accounting for about four hundred thousand. Strategy and communication are paramount. Only co-operating teams have success. Aggressive loners have no chance. (Süddeutsche Zeitung, 21 November 2006: 10)

Here, violent computer games were identified by drawing Counter Strike together with team play and tactics, prudence, strategy and communication, with sports and with a large number of players. Other aspects reported on in the press in relation to violent computer games were separated out. Juxtaposing Counter Strike with sports released this game from connections to personal gratification and aggression, from individualists and loners. In the SZ this was achieved first by stating that Sebastian B. was not registered in the archives of World Cyber Games, and subsequently by contrasting the personality traits of Sebastian B. as a folk devil and the qualities of people playing violent computer games as a sport. In eight other articles of the sample a moral distance was created between the sports players of high moral character and Sebastian B.’s immoral use of violent computer games.

Through these and similar connections and separations violent computer games gained the identity of a tool in the hands of humans that could be morally legitimate or illegitimate, and accordingly could be used in violent computer games in legitimate or illegitimate ways. This is analogous to the way that the US National Rifle Association accounts for weapons as being, in themselves, neutral instruments, good in the hands of the good and bad in the hands of the bad (Latour, 1999a). The identity of violent computer games here was understood as dependent on the moral stance of its users. In this sense the identities of violent computer games as ‘Sebastian B.’s game’ and as ‘a sports game’ were the opposite of each other. The former was defined as the independent variable causing violent acts, while the latter was seen as a dependent variable contingent on the activities of the user. However, the two categorizations of violent computer games share the view that such games are autonomous objects, either influencing the humans engaged in them or being influenced by them.

Violent computer games as an object of politics

On the second day of the first phase, the identity that came to be the most dominant started taking form. Fewer articles than the day before discussed Sebastian B.’s life and background in relation to the school shooting. The press coverage broadened its focus away from Sebastian B.’s character by introducing more general themes concerning juvenile violence. Voices from the conservative parties the Christian Social Union (CSU) and the Christian Democratic Union (CDU) called for a ban on violent computer games (9), so that these games entered the scope of politicians and law-making, independent of Sebastian B.’s character and his games as the cause of the reported behavior, and equally distant from the ‘sports game’ and its heroic participants. Nine articles related the Emsdetten shooting to the previous school shooting in Erfurt four and a half years before. At the time of the Erfurt shooting, legislators had already been working on the reform of the current Law to Protect the Youth in Public Places (Schuster, 2003). As a result of the school shooting in Erfurt the law reform was speeded up, and the Youth Protection Law was introduced, making age labelling a precondition for release of computer games in Germany on normal market conditions. After the Emsdetten shooting this material from the press archives was circulated, together with conservative politicians’ calls for further legislation to limit the sales of violent computer games.

Further archival documents were drawn in that gave weight to voices associating violent computer games with a ban: ‘A year ago SPD [Social Democratic Party] and the Union [CDU/CSU] had already agreed on a ban’⁶ the SZ wrote, referring to the coalition agreement that presented the program of the SPD/Union government, which had been in office since 2005. In addition, a talk given five months previously by Uwe Schünemann (CDU) was introduced to the debate from the press archive. The Lower Saxony politician had warned that politicians should act as quickly as possible with a ban on violent computer games and should not wait for the next school shooting to occur. Reports of several other attempts over the previous years to introduce a ban on violent computer games were taken from the archives and given an extended life in the press coverage of the Emsdetten shooting (7).

FAZ also revived a parliamentary question posed in the Bundestag (Lower House) by the liberal party FDP that was in opposition to the SPD/Union government in July 2006, concerning the possibility of a ban on violent computer games. The FAZ noted that the government had stated in its answer that ‘the rules effective since 1 April 2003 have essentially proven their value’.⁷ The Emsdetten shooting allowed this answer to reenter the debate and its validity to be questioned in the light of the new events. This was supported by the reports of Uwe Schünemann’s speech, by the several unsuccessful previous attempts at further legislation, by the coalition agreement, and by the Erfurt shooting. Bringing these voices together, an argument against the government’s failure to act on violent computer games became stronger and stronger. While discussions about the need for further legal measures restricting violent computer games had been effectively terminated in 2003, after the Emsdetten school shooting violent computer games were again connected to politics and law-making.

On 24 November, nine articles announced that Bavaria planned to introduce a Criminal Code bill in the Bundesrat to ban violent computer games. The Federal government promised to take ‘drastic action’ against violent computer games. Three of the four newspapers reported on the advice of an official expert that was given to the Bundestag, which maintained that the German Basic Law did not prevent a ban of violent computer games. This expert advice had already been given in August 2006 but had been left unpublished.

Violent computer games were thus established as an object of politics. It happened through the weaving together of an extensive web of documents and actors: the coalition agreement, a parliamentary question and its answer, bills and expert advice, the Youth Protection Law, the Criminal Code and the Basic Law, along with comments, debates, announcements and speeches of politicians, parties and political institutions. As an ‘object of politics’ violent computer games were separated both from their association with Sebastian B.’s immoral behavior and with heroic sporting morality. The discussion of what sort of behavior violent computer games could cause or the way that violent computer games could be used morally and legitimately receded into the background of a discussion of action and responsibility. Violent computer games as a political object were thus defined as an object to which something should be done, and the question of whether something should be done became irrelevant.

As a consequence violent computer games could be distanced from their specific uses, and discussions of the games themselves – what they are and what they do – were not taken into account when they were discussed as an ‘object of politics’. This discussion drew on a vast archive of discussions and documents from the press, as well as on an extensive pool of political documents that could be introduced and rearranged in the press. Compared to this complexly woven discussion the identity of violent computer games as ‘Sebastian B.’s game’ and as a ‘sports game’ appears brittle and ephemeral, which it indeed was, as will be discussed further below.

Violent computer games as technical or civil object

Other articles related computer games to Internet platforms from which such games could be purchased (3). These platforms, it was argued, are beyond the reach of German legislation. The violent computer games were thus distanced from the sphere of politics and legislation and located in a technical area, which was conveyed as being inaccessible to politics.

Similarly, parts of the press coverage presented a call for more engagement and responsibility from parents, schools, sports clubs, youth facilities and the press itself (10), which a commentator summarized as ‘civil society’.⁸ It was implied that through social interaction these actors should watch over children’s media consumption and make sure they would not play games inappropriate for their age.⁹ Just as in the discussions defining violent computer games as an ‘object of politics’, the discussion of violent computer games as a ‘technical object’ and debates about them as an ‘object of civil society’, focused on what to do with these objects and less on the objects themselves. Similar to the political discussions these debates presupposed that violent computer games were harmful and should be regulated. When young computer players were drawn in, it was as potential victims of violent computer games, not as folk devils.

Multiple identities

Unlike the findings in the media panic literature, by applying the principle of general symmetry my analysis has shown that the morally illegitimate was not necessarily a human, but was sometimes an object − the violent computer game itself. Sebastian B. matched the characteristics of the moral panic literature of a folk devil, but this occupied only a very small part of the news coverage. Much more frequently (games as an ‘object of politics’ and as ‘technical or civil objects’) the games were freed from their connections to immoral humans and identified as immoral objects.

Similar to McRobbie and Thornton’s (1995; see also Hier, 2002a) critique of Cohen and Hall et al., my analysis shows that the press coverage did not have a monolithic character. According to Cohen one would expect violent computer games to be presented in a stylized and stereotypical fashion in the mass media. This was not the case. Neither were the four identities separate autopoietic definitions of violent computer games. The analysis of the press coverage as a whole, and of how the phases of press coverage emerged, shows that the multiple identities of violent computer games were related in the press coverage. They were related when the ‘sports game’ was defined in explicit opposition to ‘Sebastian B.’s game’. A different relation between identities was seen when the coverage defining violent computer games as ‘Sebastian B.’s game’ functioned as a spark to initiate the definition of these games as an ‘object of politics’ but without actually discussing the games themselves.

Durability of identities across phases

As has been shown above, each phase in the six-month press coverage gathered around specific topics. Within each phase violent computer games achieved different identities. Even though the sequence of phases jumped between loosely related topics, both phases and identities were connected over time, through the continuity of stories across the phases, which each conveyed a single identity of the violent computer games. The most enduring identity of violent computer games over the six months was violent computer games as ‘a political object’. This section follows this identity over the six phases and examines why this identity became so durable.

The identity of violent computer games as an ‘object of politics’ formed its own trajectory across the phases. In the first phase it was actualized in the discussions of the Emsdetten school shooting and banning violent computer games as possibly preventing further school shootings. The identity of violent computer games as an ‘object of politics’ was continued in the second phase in the reports on the Bavarian government’s announcement of a Criminal Code bill to bring the ban into effect. The third phase involved the passing of the bill through the Bavarian parliament. In the fourth phase violent computer games as an ‘object of politics’ reappeared in discussions of EU regulation of such games. In phase 5 it was continued as the federal Family Minister announced an initiative, and in reports of the hearing in the Bundesrat of the Criminal Code bill. Finally, violent computer games as an ‘object of politics’ was actualized in the last phase (Virginia Tech) through more general comments on bans.

In comparison to violent computer games as ‘sport’ and as ‘technical or civic objects’, ‘Sebastian B.’s game’ and violent computer games as ‘object of politics’ were more extensively based on archival material about discussions of violent computer games. They were not established as such solely as a result of the Emsdetten shooting, but existed prior to this event and continued to exist after it. Because these identities were, and are, created mainly out of repeatedly rehearsed archival material they are also particularly durable. The school shooting ‘folk devils’ (such as Sebastian B.), who are linked to violent computer games, often commit suicide as part of the school shooting, and accordingly the source of the press story soon dries up. For this reason the identity of violent computer games as related to folk devils tends to be short-lived, even if it recurs when new school shootings are committed.

Political negotiations and law-making are ongoing processes and as such they are enduring sources of copy for the press. Violent computer games as an ‘object of politics’ entered the press not just in one phase but repeatedly, as law-making and political initiatives were announced, discussed, decided upon, contested, followed up, reformed, etc. An enduring political process establishing and sustaining the identity of violent computer games as an ‘object of politics’ existed prior to and after the Emsdetten shooting. On the one hand, the Emsdetten school shooting allowed – or even required – this political process to enter the press coverage and to become mediated by it. On the other hand, the press coverage was also a medium for the political process, which could be intensified due to the sudden press attention.

Politicians are dependent on the press to report on their initiatives and to gain support and legitimacy in the same way as the press is dependent on politics for its continuous production of copy. Due to their sequential character (in contrast to single events, such as school shootings) political processes and the fabrication of news both gain from their mutual dependency.

Due to the mutual relationship between political processes and the press, violent computer games as an ‘object of politics’ became the dominant and most enduring identity of the press coverage following the Emsdetten shooting. None of the other three identities was entangled in a comparable continuous process.

Conclusion

Violent computer games appear in the press in clusters of articles forming relatively short phases of intense coverage on a specific topic. This topic is rarely violent computer games. As an object of the press, violent computer games are generally too weak to be the focus of a phase. Their function is more often that of ‘filling up’, adding to or ‘seasoning’ other reports. Material on violent computer games is extremely efficient for this. During the period analyzed in this study, material on violent computer games was drawn into phases of six different topics. They appeared either in individual articles or were inserted into stories on a large variety of different issues. Many of these articles, and three of the phases, did not have any strong thematic focus. Rather, they were collections of a variety of subjects such as issues of regulation, political arguments, discussions of the games themselves, debates on media effects, the reopening of a school, a former school shooting, the anniversary of a school shooting, diverse criminal acts, etc. Violent computer games did not, in themselves, gain much attention in the press, but they were repeatedly presented as ‘extras’ in other stories, without which these stories might lack substance.

The functioning of violent computer games as filling material may be one of the reasons why no consensus was reached concerning the identity of violent computer games in the press. The multiple identities of violent computer games were not separated or isolated from each other, but were defined through their mutual relations as oppositions, continuations, etc. However, they were related in ways that prevented any mutual influence and review of the identities. Therefore the style of press coverage contributed to the impossibility of violent computer games gaining a consensual identity in the press. The significance of the contrast between this conclusion and Cohen’s statement of the ‘stylized and stereotypic’ fashion of presentation in the mass media is to be noted.

Whilst there was a lack of consensus there was a clear hierarchy of the identities. The most durable was violent computer games as an ‘object of politics’. This identity was clearly the most functional for the continuous production of news in the press. No other identity had that potential, and no other identity was as continuously reiterated as the one of violent computer games as an ‘object of politics’. The style in which press coverage is produced was also crucial in another way for the identities created. The extensive use of the press archive in drawing in all kinds of speeches, parliamentary questions, political agreements, etc. contributed strongly towards making violent computer games as an ‘object of politics’ the most durable identity.

The theoretical terminology of ANT and the principle of general symmetry were applied to reach these conclusions. Following this line of analysis different results were reached from that in the moral panic literature, which gives a fixed sequence of episodes in the unfolding of media panics. By looking at the identities emerging out of the way that the elements are put together in the press, no linear sequence could be observed over the time covered by this analysis. Nor do the analyses support the view of press reports as consensual. My analyses do not suggest that this is because society is more pluralistic today compared to the times of the Mods and Rockers (see McRobbie and Thornton, 1995). Because the way phenomena are drawn together in the press has been analyzed symmetrically, and with the focus on the identities of the games and not on public understanding, as in the moral panic literature, rather different results are reached. The analyses presented here contradict the assumption that the way the press writes about violent computer games is founded on specific conflicts between societal groups (for instance, youngsters and adults). The different moral identities of violent computer games seem, rather, to have their roots in the way the press constitutes its coverage, and indeed in the events and processes that the press is dependent upon, such as politics. It is important also to note in the analyses above that there were many contingencies in putting together material in the press: for example, events taking place by chance on the same day and thus being related topically. The ANT analysis suggests that we should be careful about drawing conclusions about the underlying principles governing the identities of violent computer games achieved in the press.

The analyses and conclusions presented in this paper rest on a focused approach examining how diverse components are assembled in and through the press to create specific identities of violent computer games. The analyses allow us to understand the contingencies involved in creating identities of violent computer games. However, these analyses only hint at some of the principles of the production involved in generating violent computer game identities. Ethnographic studies such as those presented by Boyer (2005), Hannerz (2004) and Tuchman (1978) of the everyday production of press coverage of violent computer games would add to these insights about how journalists in practice draw different elements together to shape computer game identities in the press. A further limitation of the current study is its narrow focus on printed, non-tabloid press. This perspective has been selected in order not to mix incomparable data. However, it would be relevant to continue the study beyond newspapers to investigate how identities of violent computer games are shaped in other media. Recent studies (e.g. Cacciatore et al., 2012; Webster and Ksiazek, 2012) show that the Internet extends and pluralizes discourses in comparison to print. A study similar to the one presented here, but on online news, could reveal whether identities of violent computer games that are already multiple in the print media are even more fragmented online. Furthermore, the different characteristics of identities of violent computer games across online and print media should be investigated, as the different audiences of these media may overlap with groups of users and non-users of violent computer games. Due to its reliance on textual data the analyses presented here reveal nothing about how violent computer game identities are actually interpreted by news readers. Readers do not consume the total news coverage as analyzed in this paper. The findings of this study provide a foundation for investigating which identities – and combinations of identities – are actually read by media users, based on their media repertoires (Taneja et al., 2012). Finally, a very promising line of further research would be to take a media ecology perspective (Fuller, 2005; Thoburn, 2012), which would extend the perspective from the narrow view on news and investigate how the identities revealed in this study become part of – or fail to become part of – forming the politics of violent computer games in Germany.

Footnotes

Acknowledgements

I would like to thank Anna Salami for her extensive contributions to data analysis and graphics. I am furthermore grateful to two anonymous reviewers of this journal, whose advice greatly helped to improve the text.

Funding

This research was done during a fellowship in Germany funded by the Alexander von Humboldt Stiftung.

Notes

Author biography

Estrid Sørensen is Junior Professor of Cultural Psychology in the interdisciplinary science studies research group ‘Spaces of Anthropological Knowledge: Production and Transfer’ at the Ruhr-University Bochum, Germany. She is the author of The Materiality of Learning: Knowledge and Technology in Educational Practice (Cambridge University Press, 2009) and has published on cultural psychology, science and technology studies, learning, and materiality. Sørensen currently investigates how the effects of media violence in children has come to be known and enacted differently throughout society.

References

Anderson

Gentile

Buckley

(2007) Violent Video Game Effects on Children and Adolescents: Theory, Research, and Public Policy. Oxford: University Press.

Anderson

Shibuya

Ihori

. (2010) Violent video game effects on aggression, empathy, and prosocial behavior in Eastern and Western Countries: A meta-analytic review. Psychological Bulletin 136(2): 151–173.

Arriaga

Monteiro

Esteves

(2011) Effects of playing violent computer games on emotional desensitization and aggressive behavior. Journal of Applied Social Psychology 41(8): 1900–1925.

Barker

Petley

(2001) Ill Effects: The Media/Violence Debate. London: Routledge.

Becker

(1963) The Outsiders. New York: Free Press.

Boyer

(2005) Spirit and System: Media, Intellectuals, and the Dialectic in Modern German Culture. Chicago, IL: University of Chicago Press.

Burns

Crawford

(1999) School shootings, the media, and public fear: ingredients for a moral panic. Crime, Law and Social Change 32(2): 147–168.

Cacciatore

Anderson

Choi

. (2012) Coverage of emerging technologies: a comparison between print and online media. New Media & Society 14(6): 1039−1059.

Callon

(1986) Some elements of a sociology of translation: domestication of the scallops and the fishermen at St. Brieuc Bay. In: Law

(ed.) Power, Action and Belief. London: Routledge and Kegan Paul, pp. 196–233.

10.

Cohen

(2002) Folk Devils and Moral Panics. London: Routledge.

11.

De Coninck-Smith

(1999a) ‘Danger is looming here’: moral panic and urban children’s and youth culture in Denmark 1890−1914. Paedagogica Historica 35(3): 643–664.

12.

De Coninck-Smith

(1999b) ‘He wished nothing but good for the children; but he knew that many had been harmed by it’: film censorship in Denmark 1896−1922. In: Povlsen

Mellemgaard

De Conick-Smith

(eds) Childhood and Old Age: Equals or Opposites? Odense: Odense University Press, pp. 31–54.

13.

De Vries

(2007) What is political in sub-politics? How Aristotle might help STS. Social Studies of Science 37(5): 781–809.

14.

Drotner

(1992) Modernity and media panics. In: Skovmand

Schrøder

(eds) Media Cultures: Reappraising Transnational Media. London: Routledge, pp. 42−62.

15.

Drotner

(1999) Dangerous media? Panic discourses and dilemmas of modernity. Paedagogica Historica 35(3): 593–619.

16.

Fuller

(2005) Media Ecologies: Materialist Energies in Art and Technoculture. London: MIT Press.

17.

Goldstein

(2005) Violent video games. In: Raessens

Goldstein

(eds) Handbook of Computer Games Studies. Cambridge: MIT Press, pp. 341–357.

18.

Hall

Critcher

Jefferson

. (1978) Policing the Crisis: Mugging, the State, and Law and Order. London: Macmillan.

19.

Hannerz

(2004) Foreign News: Exploring the World of Foreign Correspondents. Chicago, IL: University of Chicago.

20.

Hier

(2002a) Conceptualizing moral panic through a moral economy of harm. Critical Sociology 28(3): 311–334.

21.

Hier

(2002b) Raves, risks and the ecstasy panic: a case study in the subversive nature of moral regulation. Canadian Journal of Sociology: Cahiers canadiens de Sociologie 27(1): 33–57.

22.

Holm Sørensen

Jessen

(2000) It isn’t real: children, computer games, violence and reality. In: Von Feilitzen

Carlsson

(eds) Children in the New Media Landscape: Games, Pornography, Perceptions. Göteborg: UNESCO International Clearinghouse on Children and Violence on the Screen, pp. 119–121.

23.

Ito

Baumer

Bittanti

. (2009) Hanging Out, Messing Around, and Geeking Out: Kids Living and Learning with New Media. Cambridge: MIT Press.

24.

Kline

(1999) Moral panic and video games. Paper presented at the Research in Childhood, Sociology, Culture and History Conference, Department of Child and Youth Culture, Odense University, Denmark, November.

25.

Kuipers

(2006) The social construction of digital danger: debating, defusing and inflating the moral dangers of online humour and pornography in the Netherlands and the United States. New Media & Society 8(3): 379−400.

26.

Latour

(1999a) A collective of humans and non-humans. Following Daedalus’s labyrinth. In: Latour

(ed.) Pandora’s Hope: Essays on the Reality of Science Studies. Cambridge, MA: Harvard University Press, pp. 174–215.

27.

Latour

(1999b) Circulating reference: sampling the soil in the Amazon forest. In: Latour

(ed.) Pandora’s Hope: Essays on the Reality of Science Studies. Cambridge, MA: Harvard University Press, pp. 24–79.

28.

Latour

(2005) Reassembling the Social. Oxford: University Press.

29.

Lawrence

Mueller

(2003) School shootings and the man-bites-dog criterion of newsworthiness. Youth Violence and Juvenile Justice 1(4): 330–345.

30.

Lawrence

Birkland

(2004) Guns, Hollywood, and school safety: defining the school-shooting problem across public arenas. Social Science Quarterly 85(5): 1193–1207.

31.

Luhmann

(2000) The Reality of the Mass Media. Cambridge: Polity Press.

32.

McRobbie

(1994) Postmodernism and Popular Culture. London: Routledge.

33.

McRobbie

Thornton

(1995) Rethinking ‘moral panic’ for multi-mediated social worlds. British Journal of Sociology 46(4): 559–574.

34.

Möller

Krahé

(2009) Exposure to violent video games and aggression in German adolescents: a longitudinal analysis. Aggressive Behavior 35(1): 75–89.

35.

Muschert

(2007a) The Columbine victims and the myth of the juvenile superpredator. Youth Violence and Juvenile Justice 5(4): 351–366.

36.

Muschert

(2007b) Research in school shootings. Sociology Compass 1(1): 60–80.

37.

Otto

(2008) Aggressive Medien: Zur Geschichte des Wissens über Mediengewalt. Bielelfeld: Transkript.

38.

Pethes

(2004) Spektakuläre Experimente: Allianzen zwischen Massenmedien und Sozial psychologie im 20. Jahrhundert. Weimar: VDG-Verlag.

39.

Rutter

Bryce

(2006) Understanding Digital Games. London: SAGE.

40.

Schuster

(2003) Das Jugendschutzgesetz. In: Hillebrandt

(ed.) Die Ordnung der Medienkontrolle in Deutschland. Berlin: Bundesarbeitsgemeinschaft Kinder- und Jugendschutz e.V., pp. 11–19.

41.

Springhall

(1999) Violent media, guns and moral panics: the Columbine High School massacre, 20 April 1999. Paedagogica Historica 35(3): 621–641.

42.

Squire

(2002) Cultural framing of computer/video games. Game Studies: The International Journal of Computer Game Research 2(1). Available at: www.gamestudies.org/0102/squire.

43.

Suler

Openshaw

(1992) When is a moral panic not a moral panic? Some difficulties in utilising the theory of moral panic in educational history. History of Education Review 21(1): 19–28.

44.

Taneja

Webster

Malthouse

. (2012) Media consumption across platforms: identifying user-defined repertoires. New Media & Society 14(6): 951−968.

45.

Thoburn

(2012) ‘Ceci n’est pas un magazine’: the politics of hybrid media in magazine Mute. New Media & Society 14(5): 815–831.

46.

Tuchman

(1978) Making News: A Study in the Construction of Reality. London: The Free Press.

47.

Webster

Ksiazek

(2012) The dynamics of audience fragmentation: public attention in an age of digital media. Journal of Communication 62(1): 39–56.

48.

Williams

(2003) The video game lightning rod. Information, Communication & Society 6(4): 523–550.

49.

Williams

Skoric

(2005) Internet fantasy violence: a test of aggression in an online game. Communication Monographs 72(2): 217–233.

50.

Young

(1971) The Drugtakers: The Social Meaning of Drug Use. London: Paladin.