Television Viewing and Violent Behaviour

Alison Oatey

Television violence and its effects on viewers has been a controversial issue for many years. Some viewers believe that there is an increasingly large amount of violence on television and this widespread public concern has "led to calls for stricter controls on the depiction of violence in programmes" (Gunter and McAleer 1990:92). Exactly how much violence is there on television though?

Many cultivation theorists have studied this, acquiring data in the form of content analysis. They agree on a definition of a violent act, for example Gerbner in his study used the definition, "an overt expression of physical force against self or other, compelling action against ones will on pain of being hurt or killed, or actually hurting or killing" (Gunter and McAleer 1990:94). This is an objective definition that can then be used to count the number of violent acts in whatever is being observed. Halloran and Croll (1972) used this technique to establish the amount of violence on British television in comparison with that of American television. For one week in April 1971, they observed the news, fictional drama, current affairs and documentaries on BBC1 and ITV Midlands and counted the number of violent incidents using Gerbner’s definition of violence. It was found that on average, 56% of British programmes contained some violence with four incidents of violence per hour. This was in comparison with American television which contained some seven incidents of violence per hour and where it was considerably more prevalent than on British television (Gunter and McAleer 1990:97).

Focusing now on British television and violence, we can analyse Guy Cumberbatch’s research on television violence in 1987. He looked at all types of television programme focusing on four separate weeks between May and September 1986. All four channels were reviewed, totalling 1412 hours of television (930 BBC programmes and 1146 ITV and channel four programmes). He found, using his own definitions of a violent act, that 30% of all programmes contained some violence with an average of 1.14 acts of violence per programme (Gross 1992:455). It was also found that there was much more violence on television after 9pm and that violence was rare in children’s television programmes other than cartoons. It has been questioned however whether the violence in cartoons should actually be classed as violence at all. "If cartoon violence is to be treated as more than a joke then cartoons do number amongst the most violent programmes on television" (Gunter and McAleer 1990:99).

It is clearly obvious from the research already done concerning television and its effects, that violence is quite prevalent on British television. Of more concern however is the effect that those can have on its viewers. Does the violence on television really increase the violent attitudes and behaviour in individuals? Violence on television can do one of three things. The first is make us more violent (Huesmann 1982), the second is make us less violent (Feshbach 1972) and the third is to have no effect at all (Freedman 1984, Kaplan and Singer 1976). Most evidence has supported the first argument namely that television violence does increase our own violent behaviour. There are four main effects that cause this violent behaviour when viewing it on television.

The first of these affects is arousal. It is believed that a violent programme increases levels of arousal and thus causes us to become more violent as we are not only excited by the programme but also agitated and nervous. This effect is not limited to violent programmes however, as a comedy programme may induce arousal in the form of amusement in the same way as a violent programme induces arousal in the form of anger. It is the type of arousal that occurs in the first place that determines the effect the programme has. Condry (1989:11) found that suspense programmes, comedy programmes and sport events (especially identifying with a team) are the most common programmes that cause arousal. Several studies have shown the effects of arousal on the viewer. Berkowitz (1967) used American College students in a laboratory situation. They were first antagonised and thus angered by the experimenter and were then allowed to watch either a violent or a non-violent programme. Some students were not allowed to watch anything at all. After this part of the experiment had been undertaken, the participants were given the opportunity to get the experimenter back for the behaviour displayed previously by delivering shocks every time they failed on a learning task. Findings showed that those people who had been angered and then seen a violent film, were considerably more aggressive than those not angered or those not angered but watched a violent film.

Liebert and Baron (1972) also found similar results. Children were shown a segment from a television show called The Untouchables and were then given the opportunity to either help or hurt another child through the use of button pressing. "These investigators found that the children who had viewed the aggressive programme, were far more willing to hurt another child than those who had not seen the programme" (Condry 1989:89). Those children were also significantly more violent in a free play session later than those who had not seen the violent television show. They were seen to play with aggressive toys and weapons.

This has, however, also been contradicted. In some instances, "there does not seem to be any strong relationship between perceiving a programme as violent and verbal report of emotional arousal" (Gross 1992:456). It appears that it is only if the violence appears realistic, that it is likely to cause arousal. " Before the age of nine, fantasy and reality are not easily distinguished and children tend to believe much of what they see" (Eron et al. 1983 cited in Van Evra 1989:86). This would link to findings that television programmes are seen as more violent, the more real they are. In documentary programmes and the news, violence is seen as more real, therefore we become more aroused than programmes that are in fictional settings (Gunter and McAleer 1990:100). It has also been stated that the closer to everyday life the violence is (i.e. the more real) the more serious we judge it to be and therefore the more aroused a person is likely to be.

The effect of television violence may depend largely on the person’s present cognitive state. If a person is already aroused, further television violence may cause no more arousal, however if the violence is related to what originally aroused them, the arousal may greaten. This evidence shows that the level of arousal is influenced by what the participant has seen on television, concluding that violence on television does in some ways affect our levels of arousal (Gunter and McAleer 1990:102). We cannot however necessarily say that it is just the violence that arouses us as other factors, such as cognitive state may come into play (Condry 1989:112).

The second effect that violence can have on its viewers is that of disinhibition. This is based on the idea that if we watch a lot of violence, we come to see it as "a permitted or legitimate way of solving problems or attaining goals" (Gross 1992:456). Berkowitz’s own study in 1966 suggests this. He asked subjects to write a written solution to a problem and a confederate then evaluated the solution giving the subject between one and ten shocks according to the accuracy of the solution. The subjects then watched either an aggressive film (a fight) or a non-aggressive film (a track race). They are then asked to evaluate a confederate’s solution to a problem with the appropriate number of shocks as had happened to them. It was found that those who had received the most shocks from the experimenter and had seen the most violent film administered the largest number of shocks. They came to believe that giving the shocks to the confederate was a perfectly acceptable way of releasing their aggression. (Gross 1992:453). This study has however been criticised quite extensively for its ethics. Gerbner and Gross have also found that viewers who watch too much violence on television rate the world as a much more dangerous place, therefore accepting that the violence on television is a portrayal of violence in the real world. This could quite easily lead viewers to accept that the behaviour on television is quite acceptable to be used in everyday life. This effect could have important consequences for young children who will accept this behaviour as being something quite acceptable in the real world.

The third effect that can occur in viewers who watch a lot of violence on television is desensitisation. In a sense this is linked to disinhibition as viewers who are prone to seeing too much violence on television, become used to it and therefore accept it as being part of our everyday life. The repeat exposure to it reduces our emotional response to it, increases acceptance and makes us more tolerant towards any subsequent violence that we see. A study by Drabman and Thomas (1974) clearly shows this. They studied eight-year-olds that watched either a violent or a non-violent video. Shortly afterwards they witnessed a real (staged) fight between two other children. Those children, who had just seen the violent film, did not tell an adult when they saw the fight. This was because they had been desensitised, they had got used to seeing the violence and therefore did not see it as being harmful or unnatural (Gross 1992:457). Cline, Croft and Courrier (1973) who exposed two groups of boys to a violent movie did a similar but more controlled study. The first group had a history of violent television viewing, whilst the other group did not. They found that the first group of children was somewhat less affected by the movies because they were used to experiencing violence and saw it as a part of everyday life. On the contrary to this however, " televised portrayals of violence are ‘cleaned up’ and ‘not true to life’" (Condry 1989:115). If this is therefore true, heavy television viewers should be all the more sensitive to real violence as they are only desensitised to the fake violence portrayed on television. "Violence on television teaches children to accept aggressive behaviour ‘as a way of life’ and tolerate this behaviour" (Van Evra 1989:98).

The final effect that occurs when watching violence on television is imitation. This is when the viewer is likely to imitate what they see on television; re-enact the behaviour observed. This is particularly likely to occur with young children who are unaware of behaviour that is correct and behaviour that is wrong. It is sometimes known as observational learning (Gunter and McAleer 1990:103). A real example of this is the Jamie Bulger case where the toddler was killed. It is believed that the two youngsters that killed him had been watching the film ‘Child’s Play 3’ and had then imitated the violence they had seen in the film. This consequently killed Jamie Bulger. If this theory is correct, it would appear that imitating violence from television can have some harmful effects. Bandura (1960) tested this theory of imitation using a laboratory experiment with a Bobo doll. The children were all asked to watch a video of adults ‘playing’ with a Bobo doll. The first group saw the adults hitting the doll, the second saw them hitting the doll and being rewarded for it and the final group saw them hit the doll and then be punished for it. The children were then one by one placed in a room with toys including the Bobo doll and a hammer. They were left for ten minutes and their actions were observed throughout this time. It was found that groups one and two imitated some of the actions of the adults using the Bobo doll, whilst those from group three did not show signs of violence towards the toys. It appears that the punishment they had observed with the adults had directly affected the way they responded to the toys. You could say that the punishment the adults encountered conditioned them to realise that this behaviour was wrong and therefore they did not imitate it. This study by Bandura has however been criticised greatly. It is firstly a laboratory experiment, which has many criticisms in itself. It is for example a very artificial setting, which means that no conclusive generalisations can be made from the study. It is also very unrealistic to use adults hitting a Bobo doll, as this is an unnatural action for an adult to engage in and doesn’t reflect the real world. These laboratory experiments often use small unrepresentative groups and are unable to satisfactorily show cause and effect relationships. "Their measures of television viewing and aggression tend to be so far removed from normal everyday behaviour that whether laboratory findings have any meaning in the outside world is something that can be debated quite strongly" (Gunter and McAleer 1990 cited in Gross 1992:457). Laboratory experiments can also only measure the short-term effects of violence. Although the children in group three of Bandura’s experiment did not immediately portray any violent or aggressive behaviour, it does not mean that no aggressive behaviour occurred at all. It may have been that following the study, these children may have become more violent and this was not seen directly after the study. The experimenter was therefore completely unaware of it.

Field experiments that are done in a natural setting are therefore likely to be more accurate (Condry 1989:87). They observe people in a subtle way and this means that the observers paradox is lessened. Parke et al (1977) used a field experiment observing male juvenile delinquents who lived in two separate cottages. The first cottage saw five violent videos over a period of one week whilst the second cottage saw five non- violent films. The aggressive levels of the delinquents were assessed before and after the study and it was found that there was a significant increase in aggressive behaviour in those from the first cottage that had seen the violent videos. The problem with this study however is that you cannot be sure that the only difference between the two groups was the programmes they watched. Using participants in this sort of setting also means that the sample was unrepresentative and therefore generalising to the general population proves difficult.

It has therefore been established that these four effects namely arousal, desensitisation, disinhibition and imitation can occur when watching violence on television. Other factors however may come into play that would mean that the results could not be conclusive evidence of television violence causing violent behaviour. The first of these is age. The more mature an individual is, the less likely it appears they would be affected. The level of comprehension an individual has is therefore vital. Younger children are less able to understand what they should and should not do and what the harmful effects are that may occur from watching television. A relationship between television violence and aggression has been observed in children as young as 3 (Singer and Singer 1981) although according to longitudinal data, the relationship is much more consistent and substantial for children in middle childhood than at earlier ages (Eron and Huesmann 1986) ( All images that children see, are seen to them as being real as they get very involved in what they are watching and find it difficult to differentiate between fiction and reality. Adults on the other hand are able to distance themselves (Van der Voort 1986 cited in Van Evra 1990:83). Van der Voort studied 314 children in Holland who were shown the following films; a full-length real crime, a drama, two adventure films, a fantasy film and cartoons. He found that 9-12 year olds judgement of what is violent was little different to adults (Condry 1989:80). Researchers believe that "there is a critical/ sensitive period during which violence has a maximal effect" (Van Evra 1990:84). This is believed to be between ages 9 and 12 when children’s perceptions and experiences change. Violence on television is likely to have most of an effect on the viewer up until this age although this is not to say that it cannot have an effect after this.

Another variable that could cause disruption in the results is gender differences. It is generally believed that less violent behaviour occurs in women than it does in men. Eron (1980) believed that this was largely due to there being fewer aggressive female models and therefore there was less possibility for imitation effects to occur. He then however found that the activity that the model was engaging in was more important than the actual gender of the model. Boys and girls were both found to be more aggressive after viewing a male model, perhaps suggesting that girls identify better with a male model than a female one. Eron concluded that susceptibility to violent behaviour begins at the age of three although by the age of eight he thought that girls had found other interests, learnt other ways of behaving and were therefore less affected (Van Evra 1990:85). The problem with this however is that only young children were looked at. Are there likely to be any differences related to gender in adults? It would perhaps seem less likely as gender differences are largely associated with the notion of imitation which primarily occurs with the young.

The amount an individual watches television in the first place may be important for establishing whether there is a relationship between watching violence on television and violence in the individual. Singer and Singer (1984) believe that the more an individual watches the television, the more likely they are to be aggressive, restless and have little belief in a ‘scary’ world (Van Evra 1990:90). This was concluded from studying elementary children. Belson (1978) studied 1565 boys aged between 13 and 16. He asked each of them through an interview technique, to recall their viewing habits over the previous years. It was found that the higher the exposure to television, the more they reported using violence under various circumstances (Gross 1992:458). We can therefore supposedly conclude that the more television is viewed, the greater the effect it has on the individual ( This study has however been extensively criticised because of its use of a retrospective interview. Looking back into the past can lead to unreliable and inaccurate results. The effect can also be a two-way effect. Heavy viewers are likely to be more violent yet if you are more violent you are likely to select more violent programmes in the first place. It may therefore still be difficult to establish a reliable cause and effect relationship. "Heavy viewers approve of violence more, use it to resolve conflict and perceive it as more realistic" (Huston-Stein and Friedrich 1975 cited in Van Evra 1990:91). The amount of television viewing in general is also important because violent behaviour can occur by programmes that arouse children. These programmes do not have to be violent (Clap 1988) (

Amount of viewing is also linked with socio-economic levels. It has been found that heavy viewers tend, in general, to come from more disadvantaged homes and have a high incidence of behavioural problems (Gernstein and Eisenberg). Van der Voort (1986) also suggested that these lower socio-economic people show more enjoyment and approval of violence, are more likely to identify with the television characters and have low achievements in school (Van Evra 1990:91). These factors make them more susceptible to being influenced by television violence.

Parental influence on a child’s viewing may also determine how affected they are by violence on television. Those parents, who are not concerned about the effects of television, will allow their children to actively watch whatever programme they want. This allows them to be more susceptible to violence as they may well choose violent programmes and consequently encourage their own violent behaviour (Van der Voort 1986). Huston-Stein and Friedrich (1975) however found that when parents disapproved of violence on television and limited the child’s viewing of it, it did actually cause a decrease in the amount of violent behaviour occurring in the child. It could however quite easily be counteracted because limiting the child’s viewing could cause frustration in the child (not being allowed to watch the same as others) and consequently make them as violent if not more than if they had been allowed to watch it from the start. "Encouragement/discouragement of aggressive behaviour by parents is a crucial variable in determining whether imitation occurs" (Messaris 1986). Rejection and low parental satisfaction may also lead to high levels of aggression. These are completely unrelated to viewing violence on television (Eron 1982).

Looking at all this evidence for and against this argument, it is possible to make several conclusions. Four specific effects of television violence occur from watching violence on television namely arousal, desensitization, disinhibition and imitation). These support the view that television violence does have an effect on its viewers, particularly it appears its younger viewers. A study by Williams (1986) is a good example of the overall effect. He compared three communities: Notel with no television, Unitel with one channel and Multitel with several channels. Six to eleven year olds viewing habits were observed in these three communities over a period of 2 years. It was found that in Notel, the aggressive behaviour increased after the introduction of television whilst there was no apparent increase in the other two communities (Gunter and McAleer 1990:114). "The more people are exposed to violent television drama, the more they are likely to be violent in their everyday lives" (Condry 1990:86). Other factors have also shown to be influential in this cause and effect relationship between television violence and violent behaviour. Such factors as age, gender, parental influence and amount of viewing contribute to how influential television violence is on an individual’s behaviour. Findings are still however inconclusive in this debate, although a large proportion of the evidence does appear to strongly favour the hypothesis that viewing violence on television does have an effect on a viewer’s violent behaviour. As a Washington Post article states "the preponderance of evidence from more than 3000 research studies over 2 decades shows that the violence portrayed on television influences the attitudes and behaviour of children who watch it" (Oldenburg 1992 cited at


Monday 30th November 1998