At the same time, we can hardly ignore the fact that TV does feature aggressive and violent behaviour. One commentator notes that by the age of 14 the average American child has seen 11,000 murders on TV (Harris, 1989). In fact, studies have shown that violence is much less prevalent on British TV than on American TV (Gunter & McAleer, 1990). However, the type of programme matters: there's more violence in cartoons than in many other fictional programmes, but children do discriminate between cartoon violence and more 'realistic' violence. NeverthelEss, violence is commonplace even on British TV.
There has been a considerable amount of research into inter-relationships between the viewing of violent films, videos and TV programmes and aggressive behaviour by the viewers of such material, in particular the behaviour of children. My words were carefully chosen in that description. More commonly, research is framed as being concerned with what are called the 'effects' of television. This perspective represents the dominant paradigm in TV research. In its crudest form the relationship between children and television is portrayed as a matter of single cause and direct effect, which puts this kind of research firmly in the behaviourist tradition: based on what's sometimes referred to as the 'magic bullet' theory. Approaches have become more sophisticated in recent decades, stressing such complicating factors as the variety of audiences, individual differences and the importance of 'intervening variables'.
The early survey work in the 1950s by Wilbur Schramm and his colleagues in the US and by Hilde Himmelweit and her colleagues in Britain are remarkably cautious compared with many later studies. Both present children as active agents rather than passive victims, unlike most of the research in the 1960s. Both Schramm and Himmelweit suggested that the effects of television violence vary according to the personal and social characteristics of viewers, and according to how violent acts were portrayed. Sociological research has in fact tended to stress longer-term changes in behaviour and the enmeshing of television with the rest of social life, whereas psychological research has tended to focus on short-term changes in behaviour, treated in isolation in the laboratory.
The most famous psychological studies of children and aggressive behaviour are Albert Bandura's Bobo doll studies, which are now widely regarded as early research classics in the field. These were experimental studies in which children of nursery school age observed a playroom in which an adult was hitting, punching, kicking and throwing a large inflatable doll. Particular actions were used (such as using a hammer and saying 'Pow... boom... boom') which children would be unlikely to perform spontaneously. The children were then observed as they played alone in the playroom with the doll for 10 to 20 minutes. A control group of children was allowed to play with the doll without observing the aggressive adult behaviour. As one might expect, the children who witnessed the adult aggression performed similar acts; the others did not. In a series of studies, Bandura and his colleagues have shown that children display novel acts of aggressive behaviour which they have acquired simply through observing someone else engaged in these acts.
In a later version of the experiment (1965), the children were divided into 3 groups. One group went straight into the playroom. The second group saw the model being rewarded for aggressive actions before they went in. The third saw the model being punished. Those who saw the model being punished showed significantly less aggression that those who saw the model rewarded or who saw no consequences. This suggests that seeing a model punished leads to less learning of the model's behaviour. However, after all the children had played in the playroom with the doll, they were offered rewards to behave in the playroom like the adult model had done. In the first stage of the experiment the consequences for the adult affected the children's behaviour. This second stage showed that they had in fact learned the behaviour because they were able to perform it. So those children who had seen the model punished had still learned the behaviour but would only behave like that if offered an incentive.Bandura suggested that we should distinguish clearly between the acquisition of aggressive responses and the performance of aggressive acts: observation of modelling is sufficient for aggressive behaviour to be learned, but reinforcement is necessary for aggressive acts to be actually performed.
Bandura felt that there were three main sources of aggressive models: the family, the sub-culture and the mass media. Of these sources, research has concentrated on the mass media, and in particular on television violence. The conclusions from such studies range from Howitt and Cumberbatch (1975: vii) who argue that 'the mass media do not have any significant effect on the level of violence in society' to Comstock and Lindsey (1975: 8), who state that 'the widespread belief that... the evidence suggests a causal link between violence viewing and aggression is correct'.
Bandura did in fact do Bobo doll studies replacing the direct observation of an adult in the playroom with a filmed version. However, later studies have used video material which is more typical of interpersonal aggression in broadcast television programmes. Liebert and Baron (1972) conducted an experiment using real television programmes, in which they measured the willingness of children to hurt another child after watching a programme. In a laboratory, children were shown either a race track or an aggressive programme and then allowed either to facilitate or disrupt another child's game. They could hurt the other child by pressing a button to make the handle hot which the child was holding. The children who had seen the aggressive programme were significantly more aggressive than those who had seen the non-aggressive programme. This was particularly the case with boys. In addition, when the children were later observed at play, those who had viewed the aggressive programme showed a stronger preference for playing with weapons and aggressive toys than did the other children.
Similar results have been found in most experimental studies. They suggest that the more violence is viewed, the greater the likelihood of aggressive behaviour. However, apart from ethical objections one might raise, such experimental studies have major limitations in terms of their artificiality . They have been criticized for a lack of 'ecological validity' since they were concerned with strange behaviour in strange settings:
Long-term field studies do not suffer from the same limitations as laboratory studies. In one study a 'natural experiment' focused on the introduction of television into a small Canadian town, and looked at the level of aggression in the community before and after its arrival (Joy et al, 1977). Other researchers (e.g. Phillips, 1986) have looked at the influence of heavily publicized cases on the rates of suicide and murder.
In other field studies, people have been exposed to violent or non-violent television programmes over the course of days or weeks, and their behaviour in their natural environment has then been observed. Parke et al (1977) found that viewing violent films such as Bonnie and Clyde and The Dirty Dozen lead to an increase in aggressive behaviour. However, Feshbach and Singer (1971) found that adolescent and pre-adolescent boys at a residential school were more aggressive if they had a diet of non-aggressive TV programmes than if they had watched aggressive programmes. Watching the programmes seemed to be cathartic, harmlessly discharging pent-up aggressive feelings. This study, has, however, been found to be flawed, and an attempt at replication did not produce the same findings. In any case, catharsis requires more intense emotional involvement than may be typical in normal home TV viewing. And the catharsis theory does not square with evidence that more aggressive children prefer to watch aggressive programmes, and are more likely to do so than children who are less aggressive (Chaffee, 1972).
There are, of course, problems with any research method, and the field study is no exception. Feshbach and Singer, for instance, trying to compare diets of 'aggressive' and 'non-aggressive' TV with the boys already mentioned had to give in to the demands of both groups to watch Batman, which by their standards was counted as relatively aggressive. Studies of adolescent boys in residential homes are also unlikely to be very representative. However, to suggest as some critics have done that the studies are consequently worthless is typical of how aggressive rival researchers can be!
Another problem with field studies is that they are far more complex and expensive to set up than laboratory studies, which means that we don't have so much evidence from such sources. This consequently skews findings in favour of laboratory studies. Laboratory experiments are more likely to find positive effects than other methodologies, partly because of the narrowness of their focus, and partly because lab studies reporting 'null findings' are much less likely to be published in the academic journals. In the academic world no news is not good news! So lab studies may tend to exaggerate effects.
Another technique in television research is the survey, but these are not of much use in studying young children. Longtitudinal studies, or studies over time, can of course involve any kind of mix of techniques, but have special advantages in testing causal hypotheses.
Various hypotheses have been offered to describe processes of influence which violent TV might have on children's behaviour. All I can do here is to refer to some of these proposed processes briefly. No single process is likely to offer an adequate explanation.
Modelling/Imitation: Social learning theorists (such as Bandura) emphasize the 'observational learning' of particular kinds of aggression from a 'model'. Those who employ this argument see film and TV characters as models from whom children learn behaviour which may be imitated in everyday life. Unless they had seen the film The Deer Hunter the American teenagers who killed themselves with randomly loaded revolvers (as in the film's grissly game of Russian roulette) might not have done so. In such cases, simple imitation of media violence is widely cited as the reason for violent behaviour.
Symbolic Modelling is a variation on this process, whereby watching violent programmes may be a factor in encouraging violent behaviour which is not directly imitated but which has been generalized from the specific behaviour demonstrated in the media. Identification In another modified version of the imitation theory, it is argued that viewers tend to adopt the aggressive behaviour of characters only if they identify with them and if the character's behaviour is seen to be justified. Obviously people are more likely to imitate the behaviour of an attractive model than a less attractive one, and empathy is likely to heighten this tendency. Vicarious Reinforcement: If violence is 'reinforced' by being seen to 'pay off' for the aggressor this may promote its acceptability to the viewer compared with violence which is punished or unproductive. We have already referred to one such example in one of the Bandura studies. Both vicarious reinforcement and identification may mean that aggression by 'goodies' can be more of a problem with children than that of 'baddies'. Criticism of aggressive acts either within the programme itself or by co-viewers tends to reduce imitation by children.
Perceived Reality: Another variation of modelling which lacks an agreed label is imitation subject to the degree of perceived reality. Some studies (eg Feshbach 1976) have shown cartoon violence to have less negative influence on children's behaviour than more realistic violence.
Instigation/Arousal
According to this variation on modelling, those who are already in a state of high emotional or physiological arousal (which may itself have been influenced by TV) are more likely to engage in aggressive behaviour in response to watching a violent incident on TV than are others. Leonard Berkowitz found that if viewers of a violent film were made angry or frustrated before they watched it, they expressed more aggression than those who were not already angry or frustrated. We may also note that heavy viewers tend to be less emotionally aroused by violent TV itself than do light viewers (Gunter & McAleer, 1990).
Related to the influence of arousal, some commentators refer to 'reinforcement' in a general sense, meaning that TV violence has relatively little independent influence on behaviour, but tends to reinforce any aggressive attitudes and behaviour which may already exist. Some also refer to this as 'preobservation reinforcement'. The more general argument of reinforcement is that the more aggressive tendencies are reinforced in this way, the more likely it will be that they will produce aggressive behaviour.
Sensitization
Sensitization is a sort of reverse modelling, whereby viewers react so strongly to some extreme example of realistic violence that they are less likely to imitate it. Ethical considerations mean that it's not much studied. Where viewing of violence is 'light' sensitization may be more likely than desensitization.
Catharsis (or 'symbolic catharsis'): As we have seen, Seymour Feshbach (1955; Feshbach & Singer, 1971) has argued (rather differently from all the varieties of modelling theory) that fantasy violence can have a cathartic effect on viewers, defusing latent aggression, and reducing the possibility of aggressive behaviour. People often report feeling better after watching a really scary film. Note that this theory, which suggests that aggressive behaviour television may not have harmful effects, is often singled out for attack, as in the case of the study already referred to by Feshbach and Singer. A serious objection, though, is that the content of TV programmes may be partly responsible for any pent-up aggression or anxieties in the first place! Certainly there's no doubt that TV characters enter into children's dreams, and TV-inspired fantasies may not only inspire nightmares, but may also perform a valuable role in developing defences against real or imagined vulnerability. Another version of catharsis theory is that watching violent programmes decreases levels of arousal, leaving viewers less prone to aggressive behaviour.
This related theory formerly advanced by Leonard Berkowitz suggests that people are naturally aggressive, but that they normally repress this aggression. Heavy viewing of violent TV weakens their inhibitions and leads them to feel that aggression is acceptable.
Desensitization
The notion of desensitization involves the argument that heavy viewing of violent TV over time conditions viewers gradually to accept violence as normal, dulling their sensitivity to aggressive behaviour in everyday life. The conditions of ordinary TV viewing may encourage us to relax and enjoy violent images. Arousal declines as the viewing of violence becomes routinized. Drabman & Thomas (1984) found that children of 8 to 10 shown a video of aggressive behaviour took longer to intervene in apparently real life violence between two younger children they were left in charge of than children who had not seen the video. However, such studies are still artificially lab-based and do not explore children's own thoughts and feelings. The origin of such theories is again in the behaviouristic tradition of 'behaviour modification'. Observations suggesting densitization may in fact have been observing the development of children's defenses against anxiety.
TV viewing may influence not only behaviour but also attitudes and beliefs...
Value Reinforcement: Whilst this doesn't have a technical label, this refers to the theory that TV programmes may reinforce certain values about the use of violence (rather than directly influencing behaviour). Programmes where violence is used frequently to settle disputes reinforce the value that aggressive behaviour is acceptable.
Cultivation Theory: George Gerbner and his colleagues in the USA argue that the most significant effects of TV violence are ideological rather than behavioural. Gerbner sees TV as a modern 'opiate of the people', serving as a tranquilizer which legitimates the current social order. He has shown that there is a correlation between TV viewing and viewers' estimations of the frequency of violence in the everyday world. 'Heavy viewers' are more likely to mistrust other people and to experience fear and insecurity, and therefore to support stronger forms of policing and social control. However, Gerbner makes no allowance for the variety of individual interpretations or for the kind of programmes involved. It may be that more fearful people are drawn to watching more TV. And other social and personality factors may counter such cultivation.
Much of the research evidence tends to suggest that over a long period, 'heavy viewing' of violent programmes increases at least slightly the likelihood of a disposition towards aggressive behaviour amongst children and adolescents (eg Huston & Wright, 1982; Pearl, Bouthilet & Lazar, 1982; Singer & Singer, 1981). Nevertheless, any firm conclusions are hotly disputed (e.g. Social Research Unit, 1983; Milavsky et al, 1983). Many critics argue that the case is unproven.
There are several special reasons for contradictory findings:
Even if we can define violence or aggression on TV or in the behaviour of viewers we need to be aware that individual interpretations and special circumstances are important. What is mildly assertive for an instigator may seem like rampant aggression for someone on the receiving end. Simulated violence in a fantasy drama may be perceived quite differently from violence in a live news item. And we'd need to know how 'realistic' viewers thought the action was. More holistic studies are essential. We haven't discussed, for instance, the way in which youth subcultures may use TV as a way of acquiring or maintaining a sense of group identity, not simply as a source of models to be imitated.Even the pro-effects evidence that exists does not suggest that watching violent programmes inevitably and immediately leads to people behaving violently, and acknowledges the importance of other factors.
In the UK, Barrie Gunter has made allowance for the ways in which viewers themselves differentiate between and evaluate various forms of violence on TV. He found that viewers themselves made allowance for:
Gunter focused on adults. However, there is some research which suggests that children are capable of making similar allowances for the contexts involved. Other factors may also come into play (Gunter & McAleer, 1990). For instance, children's self-esteem may be an important factor in their defining of what is 'violent'. Though research findings differ, children's experience with the various genres - cartoons, drama, news and so on - obviously affects their perceptions of degrees of 'realism'. Programmes perceived by children as realistic are watched with more involvement, more emotion and less detachment than those seem as fantastic (such as cartoons). Cartoons are seldom seen by children as being violent at all.
Many theories about children's behaviour and the influence of TV are in the behaviouristic tradition: where the emphasis is on the passive learning of habitual behaviour through conditioning. They tend to ignore the active meaning-making that children engage in, and the variety of meanings which they construct with TV (Dorr & Kovaric, 1980). If you write an essay on the topic here's where you can legitimately criticize experts without yourself having an extensive knowledge of the field which can only come with the investment of more time than you have as an undergraduate student:
Daniel Chandler 1992