Keywords

1 Introduction

To date, conflict research has struggled to agree upon a uniform definition of war. Whereas scholars agree that wars are a specifically severe and intense type of violent conflict, there is no unified way of clearly distinguishing the two. Following Wallensteen (2015), a conflict can be understood as “a social situation in which a minimum of two actors (parties) strive to acquire at the same moment in time an available set of scarce resources” (pp. 17–18). In this, resources should be conceptualized in a broad sense ranging from physical resources like money or gold, to territory or political power. Moreover, conflicts do not automatically emerge from an existing scarcity of resources, as they rather occur from the perception of conflicting aspirations of different actors (Meyer et al. 2018). Conflicts thereby can be differentiated based on several characteristics such as the involved actors (states, non-state actors or both), the issue of dispute (e.g. ideology, autonomy, resources), the type of conflict (interstate, intrastate, substate or transstate) or the level of conflict violence (for more details see: Heidelberg Institute for International Conflict Research 2019). In trying to distinguish conflicts from wars, scholars have mainly followed three approaches (cf. Heidelberg Institute for International Conflict Research 2019).

First, especially during the first half of the twentieth century, research has differentiated conflicts from wars based on the existence of an official declaration of war between the conflict parties (e.g. Wright 1951). Since this approach excludes a large variety of conflicts which are commonly considered to be a war, e.g. civil wars, this approach is widely regarded as outdated (Heidelberg Institute for International Conflict Research 2019). Second, quantitative definitions have described wars as conflicts with a high number of casualties. The Correlates of War Project, for instance, defines a war as a conflict with 1000 battle-related death in a given conflict year (Sarkees and Wayman 2010). This, however, means that during the course of a violent conflict the definition of said conflict can change between being a “war” and being a “conflict”, which seems counterintuitive. Finally, other scholars have advocated for a qualitative definition of war suggesting that the quality of conflict actions separates conflicts from wars. Here, the Heidelberg Institute for International Conflict Research (2019), for instance, proposes that a war is a violent conflict in which at least one conflict side uses a substantial degree of physical force against individuals. Moreover, the quality of conflict actions is also judged by the severity of used means (e.g. what kind of weapons are used) and the severity of the consequences of the use of violence (e.g. the number of casualties and refugees). Due to this multitude of definitions, scholars have widely avoided the label war coverage and mostly use the more general term (violent) conflict. To keep with this research tradition and since it is a nearly impossible task to retrospectively differentiate between research that analyzes a conflict and research to address a “proper” war, the terms war and violent conflict will be used interchangeably in the remainder of this chapter.

Despite the on-going debates about the definition of war, it is safe to say that we live in an age of conflicts. Following data by the Uppsala Conflict Data Program, the number of violent conflicts reached a peak after the year 2014 that was only matched by the early 1990s and resulted in a total number of 160 different conflicts in 2018 (Petterson et al. 2019). The analysis of how these conflicts are covered thereby draws its relevance from the fact that conflict is often mediated since most wars take place outside of people’s direct sphere of experience (Baden and Meyer 2018; Cottle 2006). Consequentially, war coverage can influence the perceived relevance of a conflict, the predominant interpretation of conflict events, the public’s attribution of conflict roles (e.g. victim, perpetrator or hero) and the public support for conflict interventions (e.g. Althaus and Kim 2006; Edy and Meirick 2007; Entman 2004; Iyengar and Simon 1993; McCombs and Shaw 1993). Moreover, violent conflicts have a very high news value as they fulfill many of the criteria that guide media attention (Eilders and Hagen 2005; Galtung and Ruge 1965; Harcup and O’Neill 2017). Violent conflicts are relevant and negative events that might endanger the wealth, freedom and safety of many people. In addition, as conflict events involve individuals (e.g. soldiers, victims, perpetrators) and their actions (e.g. fighting, looting or rescuing), they tend to be very personalized occurrences. Finally, even though conflicts often emerge from complex cultural, historical or political backgrounds, they “usually consist of clearly identifiable events or episodes such as battles, invasions or peace negotiations” (Jungblut 2020, p. 4). As a result, war coverage has been a central component of news since the emergence of the mass media itself (Knightley 2004; Wilke 1995).

2 Theoretical Approaches and Frequent Research Questions in Content Analyses on War Coverage

Content analytical research on war coverage mostly focuses on two main research interests. They either analyze (1) how independent the media is from political influences or they examine (2) how (different types of) media cover conflicts. Each of these research interests thereby is relying on their own theoretical approaches.

The largest share of studies on war coverage seeks to analyze the independence of the press during violent conflicts (e.g. Aday 2010b; Althaus 2003; Bennett 1990; Hallin 1989; Robinson et al. 2009). In this, scholars try to find out whether or under which conditions journalists act as the “faithful servant” (Wolfsfeld 1997, p. 69) of the government helping them to legitimize conflict interventions and to increase the support for their conflict policies (e.g. Baker and Oneal 2001; Eilders and Lüter 2000; Glazier and Boydstun 2012; Hayes and Guardino 2010; Kutz 2014). Moreover, research also addresses the question of whether and under which conditions journalists act as a watchdog and whether critical news coverage can undermine the government’s war-efforts or even influence their decision-making (e.g. Althaus et al. 1996; Bloch-Elkon 2007; Hallin 1989; Livingston and Eachus 1995).

Research on the relationship between media and politics has developed a variety of theoretical models that can be divided in three different groups (Brüggemann and Weßler 2009). Models of ‘media omnipotence’ argue that since we live in a media society, politicians that seek media attention have to adopt to the media logic to have any chance of entering the news discourse (Strömbäck and Esser 2014). The paradigm of ‘media omnipotence’, however, has received substantial critique. The main argument here is that in conflicts, media and politics are not equally powerful opponents, since political actors can at times restrict the media’s access to information (Brüggemann and Weßler 2009). Models of ‘political dominance’ postulate a substantial influence of political actors on the news coverage (Brüggemann and Weßler 2009; Cottle 2006). One prominent example of these models is Bennett’s (1990) Indexing Hypothesis that suggest that the spectrum of opinions expressed by relevant political elites determines the spectrum of opinions in the news coverage. Models of ‘political dominance’ has been criticized because they run the risk of viewing the media as passive information transmitters. These approaches often imply “that the action only occurs on the side of politics” (van Aelst and Walgrave 2016, p. 510). The third perspective can be labelled as models of ‘political-media interdependence’ (Brüggemann and Weßler 2009). These models suggest that the media’s independence from political impact is influenced by contextual factors like characteristics of conflict events, variations in the political environment, professional journalistic routines, characteristics of the media system, and cultural variations (Brüggemann and Weßler 2009; Jungblut 2020; Wolfsfeld 2011).

The second main area of research analyzes which conflicts are covered in the news (e.g. Hawkins 2002; Zerback and Holzleitner 2018) and how the media portrays them (e.g. Deprez and Raeymaeckers 2010; Entman 1991; Wolfsfeld 1997). Studies on the visibility of conflicts mainly rely on the agenda setting paradigm (McCombs and Shaw 1972), while research on how conflicts are covered oftentimes apply the framing approach (Entman 1993) to identify media bias in war coverage. As such, this branch of research extends the perspective of media’s independence from political influences by analyzing the relevance of other influences (e.g. media logic, editorial line, ethnocentrism, journalism culture) on the way conflicts are covered. As a result, scholars have addressed the question of whether the news is biased towards one of the conflict parties (e.g. Deprez and Raeymaeckers 2010; Jungblut and Zakareviciute 2019; Sheafer and Gabay 2009; Wolfsfeld 1997), how ethnocentrism affects conflict coverage (Baden and Tenenboim-Weinblatt 2018; Entman 1991; Wolfsfeld et al. 2008) and whether the coverage of the same conflict differs cross-nationally (e.g. Dimitrova and Strömbäck 2008; Kim et al. 2007; Sheafer et al. 2014) or across different types of media outlets (e.g. Aday 2010a; Carpenter 2007).

3 Common Research Designs and Combinations of Methods

Research on war coverage has applied a diverse set of methodological approaches and combinations of methods. Similar to differences in the theoretical framework, the predominant methodological approach largely depends on the main research interest. Studies that analyze the interdependence or power relation between media and politics often do so by examining which voices and views can be found in the news coverage of a violent conflict. In doing so, they often make use of quantitative content analyses to unravel whether or to what degree official or elite statements dominate the media debate (e.g. Althaus 2003; Althaus et al. 1996; Eilders and Lüter 2000; Hallin 1989; Hayes and Guardino 2010; Robinson et al. 2009). A second set of studies applies different forms of input–output-analyses comparing how the government talks about war to how the media covers it (Glazier and Boydstun 2012; Jungblut 2020; Kutz 2014; Sheafer and Gabay 2009). A third and final branch of research on the independence and role of media in war combines content analysis with survey data to unravel under which conditions the news coverage can influence the public opinion on conflict interventions and thereby undermine or legitimize the government’s conflict policies (e.g. Aday 2010b; Baker and Oneal 2001; Bloch-Elkon 2007; Gershkoff and Kushner 2005; Iyengar and Simon 1993).

Studies that analyze the visibility of different conflicts (e.g. Hawkins 2002; Zerback and Holzleitner 2018) or that examine how conflicts are portrayed in the news (e.g. Aday et al. 2005; Bennett et al. 2006; Wolfsfeld et al. 2008) predominantly make use of quantitative content analyses. In this, research that is focused on how conflict events are framed in the news mostly applies deductive frame analyses (e.g. Deprez and Raeymaeckers 2010; Dimitrova and Strömbäck 2008; Entman 1991). Here, frames are first derived either from the existing literature or based on a small set of texts by different conflict actors and coded in the news coverage hereafter. Inductive framing analyses are less common in research on war coverage (e.g. Jungblut 2020; Jungblut and Zakareviciute 2019). In this approach, frame elements (problem definition, causal attribution, treatment recommendation and evaluation) are coded individually before reoccurring combinations of frame elements are identified by means of cluster analysis or latent class analysis (cf. Matthes and Kohring 2008).

In general, recent studies more and more apply comparative research designs, for instance by identifying reoccurring patterns in the coverage of different conflicts (e.g. Jungblut 2020; Baden and Tenenboim-Weinblatt 2018) or by comparing the news coverage of the same conflict across a set of media outlets (e.g. Sheafer et al. 2014). Moreover, research increasingly makes use of computational social science with the majority of studies either relying on dictionary-based approaches (e.g. Baden and Tenenboim-Weinblatt 2017, 2018; Jungblut 2020; Tenenboim-Weinblatt and Baden 2018) or using automated grammatical analyses (Sheafer et al. 2014; van Atteveldt et al. 2017). Finally, there is also a recent trend towards focusing on visual war coverage (e.g. Dobernig et al. 2010; Jungblut and Zakareviciute 2019; Schwalbe 2013).

4 Main Results for Relevant Variables and Constructs in War Coverage

Content analytical research on violent conflicts is very rich in quantity, yet it mostly consists of individual case studies (Baden and Tenenboim-Weinblatt 2018). In this, a significant part of existing research is focused on a few very salient conflicts, such as the Gulf War of 1990 (e.g. Althaus 2003; Iyengar and Simon 1993), the Kosovo conflict (e.g. Eilders and Lüter 2000; Kutz 2014), the Israel-Palestinian conflict (e.g. Deprez and Raeymaeckers 2010; Dobernig et al. 2010; Jungblut and Zakareviciute 2019; van Atteveldt et al. 2017; Wolfsfeld et al. 2008) or the post-9/11 wars in Iraq and Afghanistan (Aday 2010a, b; Bennet et al. 2006; Carpenter 2007; Hayes and Guardino 2010; Kutz 2014; Robinson et al. 2009). Despite this shortcoming, the rich quantity of existing research allows to identify some common characteristics and trends in studies on war coverage:

  1. 1.

    The visibility of different conflicts: Research that is mainly based on the agenda-setting paradigm (McCombs and Shaw 1972) has addressed the question which conflicts are salient in the coverage, which conflicts are only briefly mentioned and which are completely ignored. While most studies only point towards differences in the visibility of conflicts (e.g. Hawkins 2002, 2011), Zerback and Holzleitner (2018) also try to identify determinants of a conflict’s media visibility, i.e. the quantity of news coverage on a given conflict. For the case of Germany, they show that greater geographical distance to a conflict leads to less reporting. In contrast, the media visibility of a conflict is increased if the own armed forces are involved, if one of the conflict parties possesses nuclear weapons, and if the UN or EU imposes sanctions on one of the conflict parties. For political proximity, economic power and the number of conflict deaths, however, the study could not find a significant effect on the amount of media coverage.

  2. 2.

    The independence of the press: There is a vast body of research that aims to unravel if the government is able to influence the way the media characterizes violent conflicts. These studies oftentimes make use of the above described Indexing Hypothesis (Bennett 1990) that assumes that the spectrum of opinions reflected in media coverage is an index of the spectrum of opinions of relevant political elites. As a result, if the opposition does not question the government’s policies, reporting is one-sided and in line with the government’s position. These studies thereby measure the correspondence of elite and media discourse by analyzing the media’s framing of conflict (Bennett et al. 2006; Eilders and Lüter 2000), by comparing the opinions expressed in Congressional Records to the one expressed in the media (Althaus et al. 1996; Zaller and Chiu 1996), or by analyzing the sources that are present in the news coverage (Althaus et al. 1996; Hayes and Guardino 2010). Overall, the majority of empirical studies from the USA and Europe point to a general validity of the hypothesis (Bennett et al. 2006; Eilders and Lüter 2000; Zaller and Chiu 1996). There are, however, also some valid concerns of an uncritical acceptance of the Indexing Hypothesis. Althaus and colleagues (1996), for example, demonstrate that in the absence of critical domestic voices, journalists often use foreign sources to present opinions contrary to the dominant political position.

  3. 3.

    Bias in war coverage: Studies that aim at identifying media bias in war coverage demonstrate that war coverage largely applies an ingroup-specific perspective (Baden 2014; Sheafer et al. 2014). In this, the ingroup perspective is echoed repeatedly and often uncritically, whereas outgroups are marginalized and discredited (Baden and Tenenboim-Weinblatt 2018). Research, for instance, shows that the quantity of new coverage on events that are congruent to the ingroup perspective tends to be larger than the amount of coverage on similar events that contradict the dominant ingroup perspective (Entman 1991). Moreover, war coverage largely tends to apply frames that are in line with the dominant ingroup perspective or with the perspective of conflict actors that are perceived as similar to us (Entman 1991; Jungblut 2020; Sheafer et al. 2014). Finally, this is also reflected in explicit evaluations of the conflict parties. There is some evidence that reporting on outgroups is generally more negative, and in some cases even dehumanizes the opponent (Baden and Tenenboim-Weinblatt 2018; Liebes 1997).

  4. 4.

    Visual conflict coverage: Research on visual war coverage mostly builds on the framing approach and either deductively measures pre-defined visual frames (Dobernig et al. 2010; Fahmy 2010; Schwalbe 2013) or inductively tries to identify how conflict is framed by measuring the occurrence of frame elements and aggregating them to re-occurring frames by means of cluster analysis (Jungblut und Zakareviciute 2019). Research points towards two overall trends in visual war coverage. First, visual coverage tends to focus largely on the suffering of civilians. This is especially true for wars in which the own country is not directly involved. In asymmetrical conflicts, visual reporting is thus often biased in favor of the “weaker” conflict party (Dobernig et al. 2010; Jungblut and Zakareviciute 2019). Second, similarly to the textual conflict coverage, there is an ethnocentric perspective in visual war reporting. Thus, visual frames are emphasized that support in-group perspectives and legitimize government policy, while perspectives critical of the government tend to be sidelined or neglected (Fahmy 2010; Schwalbe 2013).

5 Research Desiderata

To be able to advance existing knowledge on war coverage, future research has to address at least four existing challenges. First, there is a need to move beyond the currently predominant case study designs and increasingly conduct comparative research that systematically compares how different types of conflicts from different regions of the world are covered in different types of media outlets in the context of different journalistic working environments. In doing so, studies will be able to identify generalizable patterns of news coverage and determinants that shape how conflict is covered, for instance by influencing the power relations between the press and politics or by posing a challenge for journalistic research. In this, research should also increasingly examine conflicts without a major Western involvement, since these conflicts are widely outside the scope of existing scholarship so far.

Second, existing conflict research contains a strong focus on a few moments of relative escalation (Baden and Tenenboim-Weinblatt 2018; Jungblut 2020; Wolfsfeld 2004). Consequentially, research on war coverage needs to extend its scope towards other conflict phases such as peace processes and situations of fragile peace. In doing so, studies will be able to analyze media’s role in peace building, during times of transitional justice (cf. Golčevski et al. 2013) and in (re-)escalating conflict situations. Even more so, studies should also try to incorporate a longitudinal perspective to unravel how a conflict is covered during different conflict phases (Fröhlich et al. 2007).

Third, there is a need to stronger focus on visual conflict communication. Since conflicts are negative and at times spectacular events that happen outside of the audiences’ direct sphere of experience, they are often described as highly visual occurrences meaning events in which visuals can be especially relevant and influential for the audience (Fahmy 2010; Jungblut and Zakareviciute 2019). As such, visual conflict coverage tends to capture the “drama that words cannot always convey” (Dobernig et al. 2010, p. 90) potentially leading to strong emotional effects such as distant suffering (Konstantinidou 2008). While content analyses on visual conflict coverage is rare, there is some anecdotal knowledge of the impact of strong emotive visuals on public opinion (cf. Dahmen et al. 2018), for instance the raising of the American flag on Iwo Jima (Spratt et al. 2005) or the image of the execution of a captured Viet Cong officer by the chief of the South Vietnamese National Police (Bailey and Lichty 1972). By focusing stronger on the visual aspects and on multimodal conflict communication meaning the interplay between textual/verbal and visual elements, research will be able to get a deeper understanding of how conflicts are covered and how conflict coverage evokes emotions.

Finally, future research on war coverage should more fully embrace the potentials of computational social science. In doing so, studies will be able to identify re-occurring patterns in war coverage across time and, for example, demonstrate how significant changes in the geopolitical landscape such as the end of the cold war or 9/11 changed the way conflicts are reported. Moreover, computer vision, i.e. automated analyses of visual material, might be a promising approach to analyze large visual corpora (cf. Peng 2018) and therefore get a deeper understanding of visual war coverage.

Relevant Variables in DOCA—Database of Variables for Content Analysis