For over a century violence in the media has been framed as a “problem” by social commentators. The arrival of a new medium (from the early tabloids to comic books, from cinema to the Internet) has typically been accompanied by a wave of concern about its potential for “exposing” an audience to representations considered undesirable by policymakers and moral guardians. It is no accident that it is the mass availability of these texts that has typically been of concern: the “mass” audience being conceived as endangered and dangerous, vulnerable and victimizing. These concerns have also shaped academic debate. As a result, work on violence in the media has been concentrated in social science disciplines, particularly in the US, and the work of arts and humanities scholars has been marginalized in public debate and policymaking.
One of the difficulties of summarizing this work is that there is no clear consensus as to what violence is. Much work in the field has defined violence as a physical action (a punch, a slap, a shooting). One of the benefits of such a definition is that it makes it relatively easy to determine how much violence there is in a text or group of texts. Content analysis has been a key method for analyzing violence in the media, particularly on television (although “television” has itself become more difficult to define, and hence sample, as channels and modes of delivery have multiplied). Such studies have measured violence and tracked changes in the amount and nature of violence on television over fixed periods. The most influential of these studies in the US is the Cultural Indicators Project begun by George Gerbner and his colleagues in the 1960s (Gerbner et al. 1994).
The findings of these studies are easily extracted by policymakers – typically in sensationalist terms – but such statistics cannot shed light on the meaning of violence, with the emphasis on action and the inevitably limited sample period diminishing important differences in generic context. For example, it is difficult to quantifiably differentiate between the shooting of an abusive husband by his wife in a long-running soap opera, and the shooting of an incidental “bad guy” by the hero in an action movie. Clearly, viewer engagement with these acts may be very different, however: in the first instance a viewer may have spent years in the company of these characters, whereas in the second they may know (and care) little about them. Broadening the definition of violence to include explicit verbal threats and talk about violence has given a more complete picture (Seawall 1997). However, less explicit behaviors such as stalking, a staple of the horror genre, and nonphysical injuries (e.g., social alienation or fear) remain difficult to account for in this way (Parks 2003).
Research that has considered viewers’ own definitions of violence is a newer development. This work has suggested that viewers are more likely to define as “violent” situations that they consider unjust or socially undesirable, while sometimes failing to recognize as “violent” (and hence socially undesirable) texts they enjoy (Morrison et al. 1999). This draws attention to the fact that all definitions of violence, including those used by researchers, are ideological. This is only a problem when the ideological assumptions are not made explicit and a consensus is assumed as to how “violence” is defined and for what purpose.
Gerbner’s work also sought to investigate whether people who watched more television “cultivated” a television worldview in relation to violence. This exemplifies a broader trend in thinking about violence in the media as a social problem, namely a concern with the putative effects or impacts of these representations. Empirical research on the effects of media violence has a long history, dating back more than 70 years to the Payne Fund Studies, conducted between 1929 and 1932, which investigated the influence of movies on children’s emotions, attitudes, and behaviors (Jowett et al. 1996). Effects research is based on the assumption that the effects of the media can be isolated, measured, and/or predicted. Such research is typically conducted in a laboratory environment where media consumers are “exposed” to a media text in controlled conditions, and the consumers’ subsequent behavior is measured. Notably, the research participants are rarely questioned directly about their experiences and understandings of the text: rather, the model relies on disguising the nature of the experiment from the participants.
Research conducted by Albert Bandura and associates in the early 1960s provides a prototype for research in the field (e.g., Bandura et al. 1963). Bandura’s work investigated the possibility of imitative aggression, focusing specifically on children. The typical design involved showing a child a film of adults behaving violently, most famously by hitting a doll. The children would then be placed in a room with a similar doll and their behavior would be monitored to see if they too would behave aggressively (e.g., by hitting the doll). Bandura, and others who followed him, also considered whether the mood of the child before the play session would impact on the likelihood of their behaving aggressively: for example, if the child had been angered or frustrated by an adult beforehand, would they be more or less likely to hit the doll?
Experimental research of this kind has been widely criticized. Key criticisms are, e.g., that sanctioned behavior in a laboratory setting is no indicator of people’s willingness to behave violently in the real world; that the research pays little attention to the media texts studied, using clips or even material generated for the experiment rather than the commercial media; that it removes consumers from context and offers little choice about consumption practices; that it conceives of audiences as inadequate and fundamentally “different” than the researchers; and that it uses unrepresentative audiences. The effects paradigm has nevertheless retained a hold in US scholarship. Although the media have changed, altering the modes of engagement for consumers (e.g., with developments in gaming technologies), the question for effects researchers has remained broadly the same: does x cause, or increase the likelihood of, y? Yet even if we accept the premise of this research, there is a striking lack of consensus as to the sum of its findings, with the evidence often being marshaled to prove opposing viewpoints (compare Freedman’s 2002 meta-analysis of the literature with Kirsch’s 2006 account).
The enduring influence of the effects paradigm can in part be explained by its continuing high profile within news media, policy reviews, and even courtrooms wherever a simple answer is sought to the question, “Do representations cause violence in the real world?” This question is frequently posed in the aftermath of otherwise apparently inexplicable acts of real-world violence: the murder of 2-year-old James Bulger by two 10-year-old boys in 1993 is perhaps the most notorious example in the UK; the 1999 massacre at Columbine High School has a similar status in the US. Scrutiny of the evidence in these cases typically concludes that a causal link cannot be proven (Buckingham 1996; Barker & Petley 2001; Boyle 2005).
This has led critics outside the effects tradition to a further question, namely, what is to be gained by making this link? Feminists, noting that the vast majority of alleged copycat cases involve male perpetrators, have offered one explanation: blaming the media is a way of not blaming men, of ignoring the connections between violent men and ideals of masculinity that permeate the culture more broadly. Such an approach makes it more difficult to isolate, as “problems,” individual media texts, as it establishes continuities in cultural representations from the Bible to competitive sports and fairytales as well as the usual suspects of horror, pornography, and violent video games (Boyle 2005).
Moving beyond the effects tradition, researchers – primarily, though not exclusively, in arts and humanities traditions – have begun to examine how people make sense of these contentious texts in the context of their day-to-day lives. Research on the audiences for a wide variety of texts – from controversial films such as Cronenberg’s Crash (Barker et al. 2001) to children’s television dramas (Buckingham 1996) and reality-crime shows (Schlesinger et al. 1992) – has pointed to diverse understandings and interactions. For example, Barker et al.’s work on Crash highlighted the extent to which viewers’ responses were shaped by another media text: a daily newspaper’s moral crusade against the film. This research demonstrated that viewers understand that individual representations of violence are embedded in a wider cultural context, that they make informed choices about their media consumption, and that they are engaged in an active process of making meaning from what they see.
Investigating media in the context of complex lived experiences, including personal experiences of violence, adds a further dimension. Research by Schlesinger et al. (1992) and by Jenny Kitzinger (2004) has identified the importance of media representations of domestic and sexual violence against women and children in allowing victims/survivors to recognize their own experiences and so to escape violent situations or seek support, challenging the assumption that violence in the media is necessarily undesirable. In a very different context, scholars of the Rwandan genocide have identified how radio (a hitherto rather neglected medium in violence research) was used to incite Hutu violence (Kellow & Steeves 2006). Indeed, close textual analyses of news media and their portrayals of violence are increasingly central to the violence debates, demanding an attention to the specificity of both textual practice and socio-historical context (Weaver & Carter 2006). Clearly, just as violence in the real world is multifaceted and complex, so, media researchers are coming to understand, are the ways in which media consumers make sense of, use, and interact with media discourses of violence. One size does not fit all.
It is not possible to prove, conclusively, that a media text causes a human being to behave in a particular way, although this question has dominated the first century of debate on violence in the mass media. As the debate moves forward, not only are the media texts and modes of delivery different but the types of violence of most concern to social commentators in the west are changing in response to the post-9/11 socio-political context. Questions about how and why we use, make sense of, and produce violence in the media at this historical juncture require an attention to the specificity of both text and viewing context.
- Bandura, A., Ross, D., & Ross, S. A. (1963). Imitation of film-mediated aggressive models. Journal of Abnormal Social Psychology, 66(1), 3–11.
- Barker, M., & Petley, J. (2001). Ill effects: The media/violence debate. London: Routledge.
- Barker, M., Arthurs, J., & Harindranath, R. (2001). The “Crash” controversy: Censorship campaigns and film reception. London: Wallflower.
- Boyle, K. (2005). Media and violence: Gendering the debates. London: Sage.
- Buckingham, D. (1996). Moving images: Understanding children’s emotional responses to television. Manchester: Manchester University Press.
- Freedman, J. L. (2002). Media violence and its effect on aggression: Assessing the scientific evidence. Toronto: University of Toronto Press.
- Gerbner, G., Gross, M., Morgan, L., & Signorielli, N. (1994). Growing up with television: The cultivation perspective. In J. Bryant & D. Zillmann (eds.), Media effects. Hillsdale, NJ: Lawrence Erlbaum, pp. 17–42.
- Jowett, G. S., Jarvie, I. C., & Fuller, K. H. (1996). Children and the movies: Media influences and the Payne Fund controversy. New York: Cambridge University Press.
- Kellow, C. L., & Steeves, H. L. (2006). The role of radio in the Rwandan genocide. In C. K. Weaver & C. Carter (eds.), Critical readings: Violence and the media. Maidenhead: Open University Press, pp. 112–127. (Original work published 1998).
- Kirsch, S. J. (2006). Children, adolescents, and media violence: A critical look at the research. Thousand Oaks, CA: Sage.
- Kitzinger, J. (2004). Framing abuse: Media influence and public understanding of sexual violence against children. London: Pluto.
- Morrison, D. E., MacGregor, B., Svennevig, M., & Firmstone, J. (1999). Defining violence: The search for understanding. Luton: University of Luton Press.
- Parks, L. (2003). Brave new Buffy: Rethinking “TV violence.” In M. Jancovich & J. Lyons (eds.), Quality popular television. London: British Film Institute, pp. 118–133.
- Schlesinger, P., Dobash, R. E., Dobash, R. P., & Weaver, C. K. (1992). Women viewing violence. London: British Film Institute.
- Seawall, M. (ed.) (1997). National Television Violence Study. Vol. I. Thousand Oaks, CA: Sage.
- Weaver, C. K., & Carter, C. (2006). Critical readings: Violence and the media. Maidenhead: Open University Press.