Violence and the mass media

Michael Kunczik
A summary of theories and research

About the author

Professor Dr. Michael Kunczik, Institute of Communications, Johannes Gutenberg University, Mainz, Federal Republic of Germany, has researched mass media effects,
theories of mass communication, international communication, mass media and social change, media economics.

A summary of theories and research

The author thanks the Singapore ComNet Centre AMIC (Asian Mass Communication Research and Information Centre) for its support in identifying Asian studies and findings on the theme of violence and the media.

Copyright 1994 by Friedrich-Ebert-Stiftung
Published by the Media and Communication Department of Friedrich-Ebert-Stiftung (FES)

Godesberger Allee 149, D-53175 Bonn, Federal Republic of Germany
Telefax: (02 28) 88 36 00

Translation from German and proofreading by Diet Simon, Mathiaskirchplatz 23 E, D-50968 Cologne
Edited by Gunter Lehrke and Dr. Heinz Dieter Bauer, FES
Typeset by Gerd Ernst, FES
Printed by satz + druck gmbh, Düsseldorf
Printed in Germany 1994


Preface 5

1. Introduction 7
2. Clarification of terms. Personal and structural violence 17
3. On the quality of discussion about the effects of media violence 21

4. On the historical dimension of the discussion about the effects of depictions of violence 37

5. Findings of content analyses 55
5.1 Structure of violence depictions 55
5.2 Functional content analysis 64

6. On the effects of portrayals of violence 69
6.1 Effects of the mass media 69
6.2 Field or laboratory? Remarks on methodology 74
6.3 Catharsis and inhibition thesis 75
6.4 Thesis of cognitive support 80
6.5 The alleged proof of harmlessness I: The studies of Stanley Milgram and R. Lance Shotland 83
6.6 The alleged proof of harmlessness II: Exploding a myth 88
6.7 The stimulation thesis: The Wisconsin studies 90
6.8 Learning theory I: The experiments of Bandura 95
6.9 The suggestion thesis 102
6.10 General excitation 106
6.11 The habituation thesis 109
6.12 Justification of crime 113
6.13 Selected field studies 116
6.14 Cultivation analysis 129
6.15 Meta-analyses 139
6.16 Sex and violence 140
6.17 Learning theory II: A frame of reference for sorting many research findings 152
6.18 Problem group analysis: the questioning of clinical psychologists and psychiatrists 159

7. Mass media and society: functions of violent depictions 165
7.1 The "no effect" hypothesis and the anomie issue 165
7.2 Introduction of television to hitherto TV-less areas 168
7.3 Television entertainment as merchandise 176
8. Reporting real violence 183
8.1 Why so much violence? Criteria of news selection 183
8.2 Landslide effect and reciprocal effect: reporting demonstrations and sports events 185
8.3 Legitimation of violence 190
8.4 Mass media and terrorism 191
8.5 Discussion of reality-TV and the problem of secondary victimisation 192
8.6 The decision making dilemma of the journalists 196
8.7 "Positive" results of reporting violence: satisfaction with one's own situation 197
9. Summary 199
Bibliography 203


We are observing, or think we are observing, violence increasing. In society. In the media. All over the world.

There is an obvious connection between the occurrence of violence in society and the thematisation and depiction of it in the media. But what is the nature of this connection? What causality is involved? Do the media merely mirror the real violence in society? Or do they cause it or contribute to it?

The author's task was to investigate these questions. We believe he did it thoroughly. The result is a comprehensive overview of research on the subject.

It shows that there are still blanks on the map of empirical violence research, but also enough substantiated findings to help de-emotionalise the discussion of the media and violence.

As the media become ever more important in the countries of the socalled Third World, so is discussion of these issues increasing. This book aims to help the orientation of journalists, publishers, broadcasting heads, scholars and politicians in that discussion.

This debate is also - and not least - about defending freedom of the press.

Dr. Heinz Dieter Bauer

Media and Communication Department

1. Introduction

Discussion of the effects of depictions of violence by the mass media has a long tradition (cf. Chapter 4) in which the arguments and theories have very often been repeated over hundreds, even thousands of years. Likewise the discussion of the commercialisation of the media, i.e. the relationship between art and/or culture and commerce can be traced back hundreds of years. The flood of publications and debate of the possibly socially damaging effects of the dissemination of depictions of violence is of unbroken topicality worldwide. In fact it has probably never been as broadly based as now. In the U.S.A. and also in Germany it has become the starting point of movements in society with the aim of fighting violence in the media. In December 1993, for example, President Clinton saw cause in a speech he gave in Los Angeles to call on the film industry to exercise self-restraint in the depiction of violence. He was facing the task,he said, of pulling a whole generation back from the precipice.

The 1993 killing in Britain of little James Bulger by two ten-year-old boys, which attracted worldwide attention, was also blamed on the media because the father of one of the boys sentenced had hired a horror video before the killing. The German media reported: "The murder of the child happened like in a horror video". Judge Morland voiced his suspicion that the two children could have been incited by video films to emulate their unspeakable barbarism. But Superintendent Albert Kirby who headed the questioning of the murderers said that there was nothing to indicate this. The deed is almost impossible to explain but the parallel between the way it happened and the plot of the film is at least clear. Three weeks before the murder the father of one of the killers had hired a horror video. In it a doll was brought to life, a kidnapping depicted and the doll-child murdered by two boys in a ghost train, its face being mutilated. The film was, in fact, still to have been broadcast on a satellite channel in November 1993. But minutes after the two murderers were sentenced Members of Parliament demanded immediate action against horror videos. Later the Sky TV company cancelled the film which had been shown twice previously and as a reason cited the comment of the judge. British politicians are now against demanding stronger controls of horror videos.

Most probably the British child murder is not the only one of its kind. In a survey of German psychiatrists (cf. Chapter 6.18) a youth psychiatrist doing forensic work reported a similar case. Two girls aged about 12 who were often left to themselves (socalled "key-children") and saw a lot of violent videos in their free time decided to "try out" what they had seen. They murdered a small child from the neighbourhood. The girls said they carried out the killing because they "wanted to try out how that is in real life". Similar horrifying effects of horror videos are also reported from Japan. In August 1989 27-year-old printer Tsutomu Miyazaki admitted to murdering three young girls and making videos before disposing of the bodies. According to the "Straits Times" (AFP, 11 September 1989) about 6,000 videos were discovered in his home. "One of the videos found in Mr. Miyazaki's room was A Flower of Blood and Flesh, a Japanese horror film which graphically depicts a young artist dismembering a young girl and painting a picture with the blood of the corpse." According to Tokyo correspondent Peter Hazelhurst Miyazaki had murdered three nursery school pupils (all under the age of five) in a gruesome ritual similar to scenes from horror movies. Professor Keigo Okonogi (Psychiatry at Keio University) commented on the case: "... in the mind of Miyazaki, the real world and the fantasy world were probably exchanged. I guess he thought that his victims were dolls or characters in movies he had seen." (Cf. "Straits Times", 18 August 1989).

The Miyazaki case was not the only one of its kind in Japan. Under the heading "Crimes forcing Japanese to examine pornography" the "Asian Wall Street Journal" quoted extensively from a police report about a number of teenage boys admitting to abducting, raping and torturing a high school girl before killing her and encasing her body in concrete. The killers told investigators they got ideas for most of their actions from an adult video. In this context the newspaper quotes a member of a women's action group: "Adult videos serve as textbooks for rape."

Portrayals of violence in the mass media and their possible negative effects have become the subject of worldwide intensive public discussion. A few examples, chosen at random, are presented here, though the list could go on endlessly. A seminar on "TV & Violence" was held in Sri Lanka in 1984 (cf. Report of the Seminar conducted by the Sri Lanka Rupavahani Corporation in collaboration with the Sti Lanka Foundation Institute, 14th and 15th of July 1984), at which the theme was, however, treated quite generally and more or less philosophically, which is also due, however, to the then state of research. The seminar did set itself quite specific aims, such as: "Examine a possible relationship between societal violence and television violence in Sri Lankan society. An attempt was made to examine the crime rate of the country, particularly the homicides and grave crimes and establish whether there had been an increase in this category of crime since TV was introduced and to what extent, if any, this increase was related to TV viewing." T.E.N. Goonatilleke spoke on "Facts about media and crime in Sri Lanka" but was not able to present data: "A study has not been done in Sri Lanka to ascertain whether the mass media including television, video cassettes and films depicting sex, crime and excessive violence and the sensationalisation of the gruesome and/or successfully executed crime, in fact, tends to motivate adults and youth, particularly, those criminally inclined" (1984, 56). Goonatilleke argued: "The average, educated, law abiding and well adjusted adult male may be able to view any television programme whatever its content without it being harmful to him." (60). It was further argued that contrary to most beliefs, the social impact of TV in Sri Lanka was not anywhere near what most people imagined it to be. Almost at the same time, namely 1985, Rajeshwar Dyal stated in an article "Censorship, sex and violence in Indian films", published in the Press Guild Annual: "Violence forms an essential part of Indian films these days."

Politics has also been sensitised to the theme worldwide. For example, according to "The Jakarta Post" (8 September 1993), the Indonesian minister of education and culture, Wardiman Djojonegoro, laid the blame for the upsurge of youth delinquency, particularly schoolyard brawls, on violence on television. The minister was quoted as saying: "Every day, our children are being subjected to films filled with violence." Parents were not taking enough care of their children and had left them widely exposed to the negative excesses of television: "The only thing they (the children) digest are the violent scenes portrayed on television." There was also intensive discussion of media violence in Malaysia. Its information minister, Datuk Mohamed Rahmat, was quoted by "The Straits Times" (22 January 1992) as saying that Radio Television Malaysia, TV3 and the National Censorship Board had decided that sex, violence, horror and other undesirable scenes would not be screened on television. The minister said: "Violence for the sake of violence will be totally wiped out so there will be less action on television." The minister told the paper Radio Television Malaysia would cut down its telecasting of Hong Kong-made Chinese serials depicting violent scenes because "we fear some people will be influenced by these scenes". The "Sunday Times" (11 April 1993) reported that from May 1993 Radio Television Malaysia would stop screening programmes depicting violent scenes. The information minister told reporters that there would be more shows reflecting "healthy family values".

Television violence is also on the public agenda in Singapore. Police there questioned about 50 teenagers who had been involved in violent crimes about their television viewing habits. According to "The Straits Times" (17 March 1993) most were found to enjoy violent films and get a kick out of watching people being beaten up or killed. Law and home affairs minister, S. Jayakumar, summed up the findings as "very worrying". The Singapore Broadcasting Corporation, for example, was criticised for the violent series it produced and telecast (e.g. Angel of Vengeance). A Mr. Goh Choon Kang was quoted with his complaint that SBC productions were becoming more and more like the Hong Kong series in respect of violence. "Singapore is not Hong Kong, and SBC is not TVB or Asia TV. We have different social environments, values and I think different thresholds of violence. What is acceptable in Hong Kong may not be equally the case in Singapore."

Televised violence was also debated in the Singapore parliament in 1993. "The Straits Times" (24 March 1993) reported that in the recent budget debate on estimates for the ministry of information and the arts a number of members of parliament voiced concern that censorship standards were on the decline. According to the paper some were particularly worried about the growing violence on television. The minister in charge, B.G. George Yeo, rejected the accusations: "Some people say there is no clear nexus between TV violence and crime. But if there is no clear nexus, we must always err on the side of safety." He added that he had instructed the chairman of the SBC to do an independent audit of SBC's censorship standards and that the panel had been satisfied with the level of censorship applied to the treatment of violence in programmes.

"The Straits Times" (24 July 1993) sees a growing concern in Asia about the level of violence on television. But according to a survey released in July 1993 young people seemed inured. The highest level of concern was reported from the Philippines where more than half of those questioned said there was too much violence. This was followed by Japan where 42% considered the level of violence too high. But in Indonesia 64%, in Malaysia 62%, in Hong Kong 54%, in Singapore 52%, in Thailand 49%, in Korea 48% and in Taiwan 44% of the respondents considered the level of portrayals of violence on television to be about right.

But also important to the discussion about the possible effects of violence portrayals is the fact that in no other area of research into media effects have as many studies been done as in this one. One can assume that there have been more than 5,000 but also that they vary enormously in quality. That indubitably makes this field, at least as far as the quantity of publications is concerned, one of the best researched in media effects research.

Before the concept of violence is clarified, a few terms will be defined and a few examples of emulation of media violence given without any claim whatsoever to completeness. Socialisation is perceived to be the process a child undergoes in acquiring the values, norms, customs, attitudes and behaviours valid in a certain society to become a full member of it. The socialisation process is a human being's second birth as a socio-cultural personality. The importance of socialisation is made clear by the famous sentence that every culture is only 20 years, i.e. one generation, away from the condition of total barbarism since the newly born "little savages" know nothing about culture. The transfer of culture happens in learning processes restricted not only to childhood but which basically are lifelong, with learning from models predominating. The passing on of information through learning from models guarantees the survival and continuity of social structures through the relatively frictionless transfer of culture. Whereas the main part of the socialisation process has hitherto been ascribed to the parents, many authors suspect that their importance in children's socialisation is curtailed by television, video and other new media (e.g. computers and computer games). It is argued that television is a new third parent exercising considerable importance because the direction in which a personality develops is largely set in childhood. Learning is understood as the ability to behave differently to before due to experiences in certain situations. Learning describes relatively lasting changes not determined by maturation processes in personality development.

To many authors there is no question whether television violence has negative effects or not. One believes from personal experience that such effects occur. Further empirical research is therefore not considered necessary since its findings could only be banal. Examples of empirically examined "banalities" are found inter alia in the American Soldier Study done in the US Army during the second world war. One examined, e.g., whether educated soldiers displayed more psycho-neurotic symptoms than less educated soldiers - even though the psychic instability of the intellectual compared to the robust psyche of the average person was known. One examined whether men from rural backgrounds felt better during their service time than those from towns - although everyone knew that farmers could take physical strains better than townsmen. With great exactitude one addressed the question whether soldiers from the southern states could cope better with the climate in the South Sea islands than soldiers from the cool north. A lot of trouble was taken to find out whether soldiers were keener to be repatriated to the U.S.A. while the war lasted or whether their applications would only increase after the German capitulation - as if there could be people who like getting killed. The findings were astounding: uneducated soldiers were more neurotic than the educated, townsmen felt no worse than soldiers from rural backgrounds, soldiers from the hot American southern states could adjust no better than those from the north to the tropical climate of the South Seas, the repatriation applications rose after the German capitulation (on the examples cf. Stouffer 1949). The conclusion to draw from such findings for the violence theme is that even apparently self-evident things require empirical investigation.

Unfortunately there are considerable flaws in the research techniques and data interpretations of many authors. It is claimed, e.g., that the danger of depictions of violence on television was scientifically verified because since brutal television programmes had been broadcast crime had risen (cf. Chapter 3). That not only infers without questioning the existence of a wave of violence and crime but correlations are wrongly interpreted. By correlation is meant the purely formal connection between two variables (e.g. television consumption and violent acts). Correlation coefficients inform about the frequency of joint occurrence of two variables and contain a general and a specific factor. The specific connection is between the two correlated characteristics, whereas the general one points beyond this to other characteristics which also correlate with the two characteristics.[1] A cause-effect relationship cannot simply be derived from correlation coefficients. If this is forgotten one can prove what one wants - say, that clergymen drink especially much alcohol or commit especially many rapes, because in areas with high alcohol consumption and a high rape rate one generally finds a lot of clergymen; i.e. both characteristics, e.g. alcohol consumption and number of clergymen, frequently occur together. To conclude from this that clergymen are drunkards probably appears as grotesque to everyone as the insistence that the sun rises because the cocks crow. But if by the same procedure it is proved that television violence causes crime and aggressiveness to rise, the approval of the public is great.

In addition there are those journalists who are not exactly diffused with seriousness who appear to be out to stir fear in the public. Thus in 1993 the illustrated magazine Bunte widely circulated in Germany incited totally unfounded fears of the alleged dangers of the film Jurassic Park by distorting the data of a study. The magazine, not read by intellectuals, reported under the headline "Classification: Dangerous" on a medically accompanied ECG experiment with two (!!) 12-year-old filmgoers. "An extremely high strain on the children's heart" had allegedly been measured. The doctor accompanying the experiment who evaluated the ECG curves described the article as "completely fabricated". According to his analysis there were no alarming heart frequencies nor any dramatic overexcitement during the film. The two test persons were slightly excited only before and after being in the cinema. Perhaps the journalists were also excited because they believed that such spectacular "reports" could increase circulation.

Nor can the existence of a crime and violence wave just be assumed - often the good old days entreated when everything is supposed to have been better but really was not - because mostly a large part of the increase in violence and crime documented in official statistics is not really an increase at all because, e.g., certain offences may have been reported more often or the unknown rate gets smaller through better equipment of the police, etc.. [2] In other words the rise in crime indicated by crime statistics is often no more than an artefact. Crime statistics often tend more to reflect public worry about crime and the trigger coming from it to solve more than its actual extent.

It is undisputed in literature that time and time again there have been emulations of criminal behaviour shown or described in mass media. Thus the American psychologist Robert Liebert (1974) reports about a boy who killed his family by secretly mixing ground glass in their food, something he had seen on television the previous evening. Also proof of the activating significance of deviant mass media models is a wave of bomb threats against American airlines in the third week of December 1966. The cause was the broadcasting on 13 December 1966 by the NBC television network of the film Doomsdayflight spelling looks wrong, should probably be 2 words despite angry protest by the pilots' union; in the film a mentally disturbed man makes a bomb attack on a plane. The first threat came in during the broadcast. In the week following there were as many as there are usually in a month. These examples can be regarded from two points of view. One as the clear proof of emulation of criminal television models, the other that obviously not everyone emulates criminal television models but rather only pathological individuals. Nearly all television viewers who saw Doomsdayflight felt not the slightest inclination to make bombing threats. The question whether deviant models offered by the mass media are imitated indubitably requires a more differentiated exploration.

The press regularly reports individual cases of imitation crimes the truth of which is often difficult to check. Thus in Malaysia in 1983 four children are said to have tried to emulate the flying prowess of the comic hero, Batman, by trying to fly out of a window, with fatal consequences. A study done by the Consumers' Association of Penang (CAP report 1984) about The effects of TV violence on Malaysian children reveals that at least four Malaysian children fell to their death trying to fly off buildings in the TV style of Superman. The reports goes on "that 12 other children were injured whilst trying to fly and another five year old child was seriously injured by a moving car which he told police he was trying to lift". The CAP report says further that the children watch television for an average of two to three hours a day, which is more than the time they spend at school. Most of the programmes show violence. A content analysis comes to the finding that Popeye and Superman are the worst culprits of television violence. The favourite TV programmes of the children were Greatest American Hero, Battlestar Galactica, Mickey and Donald, All New Popeye Show, New Adventures of Superman, Eight is Enough (a programme which according to CAP can be recommended for children's viewing), Ultraman and Buck Rogers. The most popular TV characters were ranked in this order: Greatest American Hero, Starbuck/Apollo (Battlestar Galactica), Popeye, Mickey and Donald, Superman and Ultraman. The children questioned gave the following reasons for liking the characters: the Greatest American Hero could fly and fight; in Battlestar Galactica there is plenty of action/fighting, it is nice to see the shooting; Popeye can punch very hard; Superman can fly, he can fight, he has power; Ultraman can fight the creatures, destroy the monster. In response to the findings of its study the consumer association demanded action from the government, schools and parents. It called for guidance of children in their media consumption and: "When it comes to fantasy and reality they need to be told the truth. They should know that in real life people cannot fly and that violence is very wrong because it hurts people."

In Germany children replayed scenes from a televised Western in which one of the protagonists lets himself be led to the gallows and freed at the last moment by his companion; in this case, too, a child died. According to another press report a 50-year-old woman is said to have suffered a fatal shock watching the Alfred Hitchcock thriller Psycho. Conversely - also according to a press report - an English court believed a shock could have a salutary effect. A 14-year-old girl who had set fire to a scouts' home was sentenced to watch a horror film showing an old woman dying in flames.

"The Straits Times" (2 April 1988) reported that nine year old Thaveesak Rojanapitakcheep, lying unconscious in a Bangkok hospital, was a sad reminder of the negative influence the mass media can have on the young. "The boy apparently tried to hang himself to imitate a mass suicide he saw in a television news report." Occasionally examples are cited which are more curious than anything else. "The Straits Times" (8 November 1993) reported the following ones as examples of how media change behaviour: "Randeep Singh Kalra, 13, says he is often shocked out of afternoon naps when his nine year old brother Mankirath takes flying leaps on to his stomach. Mankirath is merely imitating his heroes in cartoons, Chinese martial arts serials and wrestling matches. Six year old Sandeep Singh Sidhu was three when he fractured an arm trying to break a fall the Teenage Mutant Ninja Turtle way."

But imitation deeds have been triggered even by supposedly harmless comics like Mickey Mouse. In Norway a crime was copied that had just appeared in an issue. Copying a coup by the Beagle Boys, the imitators put a notice on the night safe of the Lörenskog savings back that customers should exceptionally please use the letter box above it to deposit money because the safe was out of order. Business people followed the notice and placed some 200,000 Norwegian Crowns in day's takings into the unsecured letter box which the thieves easily broke open. Should one therefore ban Mickey Mouse? In Germany a television thriller in which three men raid a passenger ship also triggered an emulated crime. Two weeks after the thriller was seen by about 17 million people two men got away with DM 60,000 in a similar attack.

In a suburb of the French town of Bordeaux two 13-year-old schoolboys tried to extort 20,000 Francs in late August 1992. They sent the director of a company three letters threatening to kill his son if he did not pay. According to AFP the police said the boys gave themselves away by their childish handwriting, gross mistakes in what they wrote and an awkward way of expressing themselves. At the place they had stimulated for the ransom to be handed over they had been received by the police. They told the police they had wanted to try out "whether extortion scenes from television thrillers also work in reality". A similar crime happened in Germany. A 14-year-old school boy stuck up a bank with his father's revolver and got away with DM 50,000. He maintained to the police the raid had been a test of courage. He had wanted to be better than the bank robbers on television, who always got caught.

A reader's letter printed in August 1993 in the German news magazine Focus belongs on a different level. It referred to an article about shoplifting in a previous issue which described the thieves' tricks in detail. "Many thanks for the great tips on shoplifting," the letter to the magazine said, "there were some I didn't know."

Most recently (1993) the animated characters Beavis and Butt-head have been much talked about because they are not only detestably ugly but also downright evil and violent - e.g. sticking pencils in eyes, stuffing a poodle into a washing machine and then vomiting on its corpse, killing with a chainsaw a giant grasshopper chirping for mercy. Worried teachers see a stringing together of acts in bad taste. Beavis and Butt-head, which are broadcast for about an hour a day on MTV in the U.S.A. have become the channel's most popular programme in that country. There are said to have been emulated acts already (cf. Newsweek, 11 October 1993). Children are said to have followed Butt-head's suggestion to stick a firecracker up a cat's anus and light it. An innocent cat had paid with its life. A Californian lottery millionaire is reported thereupon to have organised a telephone hotline to channel the protest. Firefighters also blame the comic figures for two fires in which a small child was killed and a house burnt down. But there is no proof that the comic figures triggered the acts.

In Germany there was a violence commission whose task was to find out the causes of violence and which was also to examine the possible effects of the media in this connection. In respect of acts emulating media violence it was argued on the basis of the research findings made that depictions of violence directly trigger violent behaviour in only few watchers and the emulated acts by people inclined to be violent anyway are probably not the actual problem of violence in the media (Schwind and others 1989, 93). One has to agree with that because it is more important that attitudes and behaviours can be learned from the media which under certain conditions influence behaviour and can heighten the aggressive behaviour repertoire. The psychologist Henning Haase (1984, 25) has expressed this fittingly, arguing that the findings of psychological laboratory experiments were greatly over-interpreted and problems inherent in the research design were ignored. He says it is possible but not very likely that the factors potentially relevant for the effects of violence depictions isolated in the laboratory, e.g. stimulus similarity, readiness to act out aggression, etc., will interact. Haase writes: "If they come together in one person and one situation then this is a most rare occurrence in relation to the entire population, regrettable individually but a marginal problem for society as a whole." In such an argumentation it must be considered, however, that in terms of millions of viewers, when does the marginal problem become a societal problem? This is already the case if effects were to happen at the per-thousands level.

©Friedrich Ebert Stiftung | technical support | net edition mv&ola | August 1997