The Utopian Pseudo-Science of Media Effects

On the 20th of April 1999, Eric Harris and Dylan Klebold embarked on the then most high profile school shooting of all time. Donned in black trench-coats, and armed with shotguns, explosives, and semi-automatic handguns, the pair murdered twelve students and one teacher before committing suicide. Twenty-four other students were injured in the incident. The agony of that tragic day was quickly followed by a wave of moral panic, and the public search for answers was couched in a predictable litany of finger pointing and recrimination (Moore, Glynn, Czarnecki, Bishop & Donovan, 2002). For many, media violence was the culprit; however, as we shall see, this was far from clear. What is undeniable was that two mentally ill teenagers had run amok.1

It is true that Eric and Dylan played a popular type of violent computer game, know as first-person shooters. They were also part of a Goth subculture that listened to “degraded” music, and were the victims of schoolyard bullying. Furthermore, FBI investigators pointed to pre-existing psychopathology: Eric was a psychopath, and Dylan was depressive (Cullen, 2004). So of course, public outrage was directed squarely at Marilyn Manson, and his “antisocial” music.

Such public outrage is fertile territory for moral entrepreneurs. Politicians create the perception of action by picking easy targets. Violent media is imbued with the causal power to harm society, and this arms moral authoritarians in their quixotic quest to control culture for the sake of “at risk” sub-populations. Whether it is teenagers, or naïfs who have not had their consciousnesses “raised,” the at-risk groups are always someone else, someone less intelligent and less wise. While moralists enjoy the thrill of placing themselves above others, the underlying problems are neither understood nor addressed.

Media researchers, however, are unlikely to attribute the Columbine massacre to the effects of violent media. Eighty years of research, and hundreds of studies, have put strict upper limits on media effects (Huesmann, 2007). The pre-eminent researcher in the field, Dr Rowell Huesmann, has built his reputation on proving the causal hypothesis. When critised for overstating the case, Huesmann conceded: “Nowhere have we ever indicated that media violence is the only or even a major cause of violence among youth.” (Huesmann & Eron, 2000).

This deflection is intended to sustain the rectitude of the researchers’ life work; however, media researchers frequently offer explanations for social behaviour that extend far beyond the external validity of their studies (Freedman, 2002). The activist researcher proceeds from conclusions to evidence, making the pernicious effects of media violence “data-proof”. Something must be done! Media policy debates have always been marked by the attempt to control degraded public appetites (Miller, 2002; Trend 2007), and violent movies and video games have recently become the lightning rod for moralistic discontent. Unfortunately, the causal hypothesis is weak.

Taken at face value, less than half of media violence research supports the causal hypothesis (Freedman 2002), and to derive this generous figure, one must confuse correlation with causation while ignoring other methodological flaws. The measured effect sizes are so small and inconsistent that they are easily consumed by alternative explanations such as arousal, and experimental participants responding to the demands of the researchers. In the quest to prove their case, researchers overlook the limitations of psychological experimentation, and the statistical models that undergird it. If a media violence researcher were to study traffic, they would be liable to conclude that you should not cross the road, because on average it is dangerous — missing the complex interaction between the attention of both driver and pedestrian, and countless other variables that could contribute to an accident. Instead of questioning the integrity of small effects sizes, media violence researchers frequently exaggerate them, and a canard has developed that the effect size is about 0.3 (Freedman). Even granting this canard, the most important deficit in media violence research is that no media violence study has credible external validity: generalising from results to society requires a leap of faith. For the true believer the problem no doubt lies in “masculinist” science and “value-free” objectivity. However, at a certain point social scientists must ask themselves why their results are so poor. (For further reading, see: Ferguson & Kilburn, 2010; Fischoff, 1999; Freedman; Pinker, 2002; Savage, 2008; Trend, 2007).

Social scientists sometimes see themselves as part of a cultural revolution to transform society — drawing a thread from themselves, back to the civil rights movements of the 1960s and 1970s, and further in history to other liberal transformations. There is a certain grandiose excitement to casting down the shackles of conservative thinking, and “fixing” society. Alas, there’s a long history of the damage wrought by righteous zeal, and the consequent reckless tear through societal afflictions. The true believer thinks they are saving society by indoctrinating others into their moral system; however, in reality they are planting the seeds of suffering. It is an age-old formula.

One popular avenue for indoctrinating students into the causal hypothesis is “content analysis”. While interesting in-and-of-itself, content analyses almost always beg the question, a logical flaw not limited to student assignments. The NTVS was the largest content analysis ever conducted, employing 300 researchers who quantified violence in 10 000 hours of television, and concluded that “40% of violent incidents were initiated by ‚Äògood’ characters who are likely to be perceived as attractive role models.” (NTVS, 1998, p. 1). An examination of the causal hypothesis is conspicuously absent from the solemn proclamations of the long-term ill effects of media violence. The NTVS researchers identified violent actions in media, and then proceeded on the basis that these violent actions adversely affect behaviour. After-all, “a multi-billion dollar advertising industry is built on the premise.” (Freedman, 2002). Missing from the NTVS report — and content analyses in general — is why we would choose to watch violent media at all.

Social scientists are “effects” researchers, and generally approach the topic of media violence with this disciplinary bias. That is, like a stamp collector collecting stamps, there’s a tendency for social scientists to “collect” experimental effects from studies, while eschewing the demanding task of synthesizing grand theories. Instead, effects are arranged in an argument for a particular moral position and call to action. Social constructionists want to build a case for re-engineering culture as a panacea for violence, gender bias, discrimination, and other forms of collective suffering. Culture is the cause and behaviour is the effect. Furthermore, because culture can be “changed”, and human nature cannot, reference to human nature is often misconstrued as arguing for the inevitability of violence, and other ills. But culture does not exist apart from human nature, and cannot be understood without reference to human nature (Harris, 1998; Pinker, 2002; Turkheimer, 2000). Misunderstanding cause and effect goes hand-in-hand with misunderstanding action and consequence, and social constructionists are perpetrating their own curious brand of violence. In this space, nothing is what is seems.

For example, perpetuating the myth of media violence plays into the hands of the moral entrepreneurs who took advantage of the Columbine High School Massacre. Riding on the coat-tails of more rigorous science, the media violence myth also creates confusion among professionals. For example, when sentencing Jon Venables and Robert Thompson for the murder of Jamie Bulger, the trial judge denounced media violence as a causal factor (Freedman, 2002). These two ten-year-old boys kidnapped two-year-old Jamie, tortured and beat him, and left his body on train tracks to be run over. According to the judge, Jon and Robert were influenced by a TV program that featured kidnapping and murder despite the fact that the police report indicated that the boys never saw the program. Perhaps its mere existence was to blame, as if the TV producers has dreamed the murder of Jamie Bulgar into existence.

Professional associations, such as the APA, AAP and APsychA, have whole-heartedly endorsed the causal hypothesis, and do far more societal harm than isolated incidents. Sommers (1994) details how in the 1980s, the boards of professional academic bodies were systematically overtaken by ideologically pure radical left-wing thinkers. More than ever, science is peddled as a political and moral tool, and ultimately this can only damage the reputation of science and undermine its ability to shape public policy when it is really needed. For example, global warming deniers assert that scientists are totalitarian ideologues, and that anthropogenic global warming is the product of a liberal education system. Such assertions are infuriating because they contain a grain of truth. By carelessly using the reputation of science, media effects researchers become talking points for right-wing demagogues in their assault on intellectualism. By endorsing a half-baked theory as fact, the APA and its ilk are reducing the credence for evidence based public policy, and alienating politicians who must sometimes wonder why scientists think they should be taken seriously.

Another serious side-effect is unwarranted parental stress. The AAP offers unrealistic parenting advice that is bound to raise the anxieties of already over-worked parents — advice that paediatricians are meant to administer in a professional capacity (American Academy of Paediatrics Committe on Public Education, 2001). The lead author of the policy document took umbrage at the suggestion that the causal hypothesis was not proven, and defended his work with appeals to authority, and the suggestion that deniers are somehow in the pocket of big media or suffering from some sort of pathological denial (V. Strasburger, personal communication, October 4, 2010). Beliefs are often protected via appeals to authority; however, in the case of media violence, those authorities have not answered the hard questions themselves. Instead we see minimisation (Anderson, Ihori, Bushman, Rothstein, et al., 2010), and denial (Heusmann, 2010).

Academics from other disciplines see violence in different ways. For example, criminologists have long been skeptical of the pernicious effects of violent media simply because the case has not been made (Savage, 2008). Perhaps criminologists are aided by professional distance from the debate. Academics who study media outside of psychology are well aware of the complex interaction between media and viewer. Violence is not a one-way “transmission” to the viewer, but rather, the viewer engages in a process of interpretation that accepts, rejects, ignores, or reinvents the intended message of the media producer. Thus the experiences and beliefs of the audience are important factors in the interpretation of media, and children are not the passive tabula rasa sketch pads that social constructionists want us to believe (Harris, 1997; Pinker 2002). Human nature has complex propensities and proclivities, and is at least as important as sociocultural factors in explaining behaviour (Turkheimer, 2000). Young children spontaneously construct their own hierarchies, are fascinated by acts and displays of power, relish in hero discourses during unstructured play, and universally identify altruism as a hero’s defining trait (Dyson, 1996; Harris, 1998; Tucker, 2006; White & O’Brian, 1999). Boys in particular hold physicality as a male preserve, and consider it important in delineating social hierarchy (Dyson).

This is often attacked as some form of extreme essentialism or “biological determinism”; however, the opposite is indeed the case. Stating the importance of both biology and culture is a middle point of view that is backed by a vast array of evidence. There are no main-stream theories that remove culture from the human story, but there are plenty of main-stream theories that remove biology. Social constructionists let their theories drift on an amorphous cultural super-organism that shapes a comparatively structureless mind. Consequently the things that one does not like about themselves — the shadow — can be projected onto society. And so, the super-organism is merely a circuitous approach to saying that racism, sexism, and violence are bad. One does not need eccentric theories of human nature to do that.

The media industry argues that the free market is the final arbiter of what audiences want, and the public has overwhelmingly voted with their wallets for more media violence. It is easy to dismiss media industry arguments and first amendment whining as self-serving, but there is always the possibility that they may be correct.

The imaginary beginnings of humankind were marked by violence that is unprecedented in modern times — even after accounting for two world wars (Keely, 1996). Like it or not, people are thrilled by sex and violence, separate or together. The media industry is on a hedonic treadmill in its quest to out-titillate, and the purpose is to sell ever larger audiences to advertisers. Each year sees more violence porn — aesthetic and unrealistic acts of power and mayhem — and cynical young audiences are still awestruck and come back for more. Given the paucity of evidence for the causal hypothesis of media violence, it is doubtful that youths are being harmed by having their brains buzzed by primitive fears and desires. The real tragedy of media violence has been overlooked: in the sustained effort to engineer a predictable emotional response from the audience, the breadth of human experience has been truncated by trite cliches. Media conglomerates count money, not meaning. To be fair, intelligent human stories continue to be told between sexy outfits and gratuitous chest thumping. Media producers such as HBO combine intelligent writing with shocking immorality, in shows such as The Tudors and Six Feet Under. And the violence and crime rates continue to come down despite the most sophisticated and depraved media industry in all of history.

The real statistics on violence are rarely talked about. They are an uncomfortable fact for media violence researchers, who usually minimise the role of media violence before quickly moving on — having it both ways. It is human nature to defend one’s intellectual creations, but academics have an ethical responsibility to engage in a basic level of academic discourse — if not to themselves, then to the tax payer who subsidises them, and the students who trust their expertise.

Although the academy is famous for its ivory towers, every media researcher is sure that those lofty egotistical heights lie elsewhere. Having impugned objectivity for the sake of pushing political action, media violence researchers have made themselves moral bastions who battle the pernicious and self-secret effects of society’s myths and lessons. They will change our environment for our own good, because they have the “insight” into our predicament and want to shape us into better people. How much of this sentiment is externalised discontent? Political preference has a biological interaction, and people automatically sort themselves into roughly two groups: liberal and conservative. No group can claim a monopoly on wisdom. Where conservatism can stray toward the status-quo, liberalism is prone to this type of externalisation. Blaming our problems on others. If society is to blame, then it must be fixed so that others do not have to suffer in the same way.

The simple truth is that violence can only be understood with reference to both culture and biology. This should pose no threat to the project of reducing societal violence. What is real is real, and violence rates continue to decline, so we’re doing something right. A “results focused” approach would be prudent: understand why violence rates are spiraling down, and push harder. There’s no space for moralism there. Placing ourselves above others is just one of the endless ways that the we double cross ourselves. Ignorant actions have unintended side-effects, and this is exactly what we see in the Utopian pseudo-science of media effects. No doubt, the media violence “debate” will continue, as it always has, handicapped by an inability to question. Hopefully psychology will experience a paradigm shift that places significant distance between politics and science before the tax payer grows tired of quixotic quests. An urgent need if we’re going to develop a deeper understanding of why we are sometimes violent, and what we can do about it.