The traditional question of whether viewing violence can make audiences more aggressive has been investigated by a variety of techniques. As social science, and psychology in particular, attempted to emulate the rigorous methods of the physical sciences, the question of television and violence was transferred to careful laboratory experiments. Inevitably, the nature of the issue placed practical and ethical constraints on scientific inquiry. A range of studies found evidence that subjects exposed to violent filmed models were subsequently more aggressive (see Bandura). Questions have been raised, however, as to what extent these findings can be generalized to natural viewing situations. What did participants understand about the task they were given? What did they think was expected of them? Can the measures of aggression used in such studies, such as hitting dolls or supposedly inflicting harm by pushing buttons be compared to violent behavior in real-world settings? Are these effects too short-term to be of practical concern?

     One strategy to overcome some of these problems was to conduct studies in natural settings such as preschools or reform homes. Children watched violent or nonviolent television over a period of several weeks and the changes in their behavior were monitored. Such studies resemble more closely the context in which children normally watch television and measure the kinds of aggressive behavior that create concern. Results, however, have been varied, and the practical difficulties of controlling natural environments over a period of time mean that critics have been quick to point to flaws in specific studies.

     From time to time, researchers have been able to capitalize on naturally occurring changes, gathering data over the period when television is first introduced to a community. A Canadian study compared children in  two  communities  already  receiving  television  to those in a community where television was introduced during the course of the study. Increases in children's aggressive behavior over time were  found  to accompany the introduction of  television.  A similar conclusion was drawn from a major study into the effects of the introduction of television in South Africa.

     An alternative to manipulating or monitoring group changes in exposure to violence is simply to measure the amount of television violence children view and relate it to their level of aggressive behavior. While many studies have found a clear association between higher levels of violent viewing and more aggressive behavior, proving that television caused the aggression is a more complex issue. It is quite possible that aggressive children choose to watch more violent programs, or that features of their home, socioeconomic, or school background explain both their viewing  habits and their aggression. Attempts to test these alternative models have involved complex statistical techniques and, perhaps most powerfully, studies of children over extended periods of time, in some cases over many years. Studies by Huesmann and his colleagues have followed children in a variety of different countries. They argue that the results of their research demonstrate that the extent of TV viewing in young children is an independent source of later aggression. They also suggest that aggressive children choose to watch more violent programs, which in turn stimulates further aggression. The research group gathered data from a range of countries, and these data indicate that the relationship can be found even in countries where screen violence is much lower than in the United States. A comparison of Finland to the United States found, however, no relationship between violent viewing and aggressive behavior in Finnish girls. This suggests that the impact of television has to be understood in a cultural context and involves social expectations about appropriate gender roles.

     Critics of these attempts to relate viewing and aggression have questioned both the accuracy of the methods by which reports of television habits and preferences were gained, either from parents or by retrospective recall, and the measures used to demonstrate aggression. In reviewing debates on research findings, it becomes clear that any study can be perceived as flawed by those taking an opposing position. However, supporters of the effects tradition point to the cumulative weight of research with different methodological characteristics; a meta-analysis by Paik and Comstock of more than 200 studies found a moderate effect of screen violence on aggressive behavior.

     Even among researchers who are convinced of a causal link between television and violence, explanations of when and why such a link is forged are varied. One of the simplest ideas is that children imitate the violence they see on television. Items associated with violence through television viewing can serve as cues to trigger aggressive behavior in natural settings. The marketing of toys linked to violent programs taps into these processes. Children are more likely to reenact the violence they have seen on television when they have available products that they have seen being used in violent scenarios. The challenge for social-learning theorists has been to identify under what conditions modeling occurs. Does it depend on viewers' emotional state (for instance, a high level of frustration) or on a permissive social environment? Is it important whether the violence is seen to be socially rewarded or punished? It has also been claimed that high levels of exposure to violent programs desensitize children, making them more tolerant of and less distressed by violence. Thus, children who had been watching a violent program were less willing to intervene and less physiologically aroused when younger children whom they had been asked to monitor via a television screen were seen fighting than those children who had watched a nonviolent program. Alternatively, high arousal itself has been suggested as an instigator of violence. The significance of such an explanation is that it does not focus on violence as such; other high­ action, faster-cutting programs may also stimulate aggression. It is evident that once focus shifts from proving causation to identifying processes, the characteristics of particular violent programs become important, because programs vary in many ways besides being classifiable as violent or nonviolent.

     The traditional violence-effects approach has been criticized as employing a hypodermic model, where the link between television violence and viewer aggression is seen as automatic. Such an approach not only ignored the complexity of television programs but also how responses to television are mediated by characteristics of viewers, by their thoughts and values. As psychology has become more concerned with human thinking, there has been greater interest in how viewers, particularly children, interpret the television they watch. Research has shown that children's judgments of violent actions relate to their understanding of the plot. This understanding in turn may be influenced by such issues as plot complexity, the presence and placement of commercial breaks, the age of the child, and so on. Rather than seeing violence as a behavior pattern that children internalize and reproduce on cue, children are seen to develop schematic understandings of violence. The values they attach to such behavior may depend on more complex issues, such as the extent to which they identify with a violent character, the apparent justifiability of their actions, and the rewards or punishments perceived for acting aggressively.

     It has often been feared that children are particularly vulnerable to violence on television because their immature cognitive development does not enable them to discriminate between real and fictional violence. In a detailed study of children's responses to television and cartoons in particular, Hodge and Tripp found that children could make what they termed "modality judgments" as young as six years old. They were well aware that the cartoon was not real. What developed at a later stage was an understanding of certain programs as realistic, building the links between television and life experience. Such research demonstrates a coming together of psychological and cultural approaches to television. Researchers interested in the structure of program meanings and in children's psychological processes can collaborate to increase our knowledge of how children actively interpret a violent cartoon.

     Another dimension of the television violence debate has been a concern that frequent viewing of violence on television makes people unrealistically fearful of violence in their own environment. Gerbner's "acculturation" thesis appeared supported by evidence that heavier viewers of television believed the world to be more violent than those who watched television less. Alternative explanations have been offered for these findings, with reference to both social class (heavy viewers may actually live in more dangerous areas) and personality variables. It has also been suggested that those fearful of violence may choose to watch violent programs such as crime dramas, where offenders are caught and punished. Again, viewers are seen as actively responding to violence on television, rather than simply being conditioned by it. Gerbner presents a valuable description of the violent content on television, differentiating between those who are portrayed as attackers and those who are the victims in our television world. Yet, Greenberg has argued against a cumulative drip-drip-drip view of how television affects viewers' perceptions of the world. Instead, he poses a "drench" hypothesis that single critical images can have powerful effects, presumably for good or ill.

     Traditional television-violence-effects research employed simple objective criteria for determining the extent of violence in a program. A feature of this approach has been the development of objective definitions of violence that have enabled researchers to quantify the extent of violence on our screens (80 percent of prime-time American television contains at least one incident of physical violence). From this perspective, cartoons are just as violent as news footage, and a comic cartoon like Tom and Jerry is among the most violent on television. Such judgments do not accord with public perceptions, and in recent years there has been an interest in discovering what the public consider violent. A carefully controlled study of audience perceptions of violence was conducted in Britain by Barrie Gunter. He found that viewers rated an action as more violent if the program were closer to their life experience than if the same sort of action appeared on a cartoon, western, or science fiction drama. He also found that ratings of violence were linked in complex ways to characteristics of the attacker, victim, and setting, and to the personality of the rater. This focus on what audiences found violent and disturbing and what they believed would disturb children has provided a rather different framework for considering issues of violence on television.

     Research for the Australian investigation of violence on television, in contrast to the U.S. Surgeon General's report, was not concerned with establishing causal links but on finding how audience groups reacted to specific programs. The aim was to improve the quality of guidelines to programmers and the information provided for prospective audiences. The research concluded that the most important dimension for viewers in responding to violence was whether the subject matter was about real life. The interest in public perceptions of violence on television has stimulated new research techniques. British researchers have asked their subjects to make editing decisions as to what cuts are appropriate before material is put to air. Docherty has argued that certain material, both fiction and nonfiction, can elicit strong emotional reactions, which he has termed "deep play." Individuals' reactions to a horror movie such as Nightmare on Elm Street appeared largely a question of taste. In contrast, a docudrama about soccer (football) hooliganism provoked polarized and intense reactions. Some viewers thought the violent material was important and should not be cut; others reacted with great hostility to a portrayal of violence that challenged their sense of social order.

     The issue of the appropriate level of televised violence arises not just with fictional violence but also with the televising of news footage. When terrorists attacked the United States on September 11, 2001, some of the televised images-particularly the  crashing  of the second airplane into the World Trade Center and the subsequent collapse of the two towers-resembled scenes from the disaster-movie genre, but this violence clearly generated an intense emotional reaction based on its reality, immediacy, and national threat. For a time, the significance of the story and the saturation media coverage overwhelmed concerns about the impact of the footage on the young and vulnerable. Generally, however, television stations attempt a balance between reporting what is occurring in the world and making the violence they cover palatable for the living room. Reporters have put themselves at risk attempting to film savage violence in a way that can tell their story but not overwhelm the viewers. The violence of the Vietnam War played out nightly in American living rooms, and this coverage has been seen as a major factor in generating the antiwar movement. Coverage of the Gulf War and the "war on terrorism," however, indicates how use of the media, especially television, has become part of policymakers' wartime strategy. Research on the role of the media in the Gulf War suggests that viewers  were often  happy  to be spared  the details of the war as  long as their side was winning. It is perhaps unsurprising that, despite concern expressed about the impact of such a violent crisis on impressionable children, the news image that evoked most anger and sadness in British children was on the plight of sea birds covered in oil.

     The televised portrayal of the war-the sanitized images of high technology, the frequently employed analogy of the video game, the absence of blood and gore-is thus relevant to the discussion of violence and television. The fact that the political debates about violence on television have focused so strongly on the potential harm to children may act to divert attention away from the way certain violence is censored in the interests of the state. An excessive focus on screen violence can deflect attention from the complex issues of state and interpersonal violence that exist in our world. Until recently, the potential of television to chal­lenge viewers to think about issues of violence has been largely ignored. A study by Tulloch and Tulloch of children's responses to violence in a series of programs has found young people more disturbed by a narrative about a husband's violent assault on his wife than the objectively more serious violence of a Vietnam War series. Their research has demonstrated clearly that the meanings children attach to violence on television is a function of their age. gender, and social class. Not only does this confirm other findings that relate the perception of violence to personal significance, it points to the potential educational effects of violence on television. Once the portrayal of violence is not seen as necessarily increasing violence, the ways programs can work toward the promotion of nonviolence can be investigated.

     However, despite attempts to broaden the debate about violence and television, the dominance within the United States of the media-effects tradition was again illustrated in responses to a series of headline-grabbing incidents of violence in American schools, particularly the deadly shootings in April 1999 at Columbine High School in Littleton, Colorado. At this time, new-media forms, video games, and the Internet were identified alongside television and film as causes of young people's aggression. The debates that followed the shootings, including testimony given to Congress, illustrated the gap between causal  and cultural  models. Grossman, a military psychologist, extended the desensitization approach to television violence to claim that video games are training children to be killers by helping to erode a natural reluctance to take human life. Taking a media studies perspective. Henry Jenkins argued that it was important to look at what young people did with the media, not what the media did to them. Young people are active media consumers, taking from media their own meanings. Some of these meanings are destructive of themselves and others, while some are a basis for creativity and positive social interaction. From this perspective, if we wish to engage with concerns about media violence, we need to foster young people's critical consumption of a diversity of media forms and content.