3. MISINTERPRETING REALITY

All the political arguments for restricting media because of a purported link between media violence and real violence are based on studies finding a correlation between the two phenomena. But correlation is simply two things happening in proximity, at the same time, in the same person or people. One of those things does not necessarily cause the other. The alarm clock ringing at six a.m. does not cause the sun to come up. In fact, determining when correlation can be read as causation is a crucial and controversial issue in every science. "Causality is very hard to prove," explains Carole Vance, professor of anthropology and Director of the Program for the Study of Sexuality, Gender, Health and Human Rights at Columbia University's Mailman School of Public Health.

"Correlation is a first step, like a red flag, of a possible relationship that's worth investigating. After that, many research designs that go beyond correlation are organized. These better studies feature prospective, longitudinal designs and designs that try to avoid various biases that can produce apparent but mistaken causality. Only after many, many studies have been done, by different investigators, using different designs, with many arguments about possible other explanations for the relationship, is causal relationship even plausible."

After several studies, the evidence of a correlation between media and violence is still weak. Therefore, a causal relationship isn't plausible. The body of data is compromised in other ways, too. Studies that find a "null" effect-that is, neither a positive or a negative effect-tend to be published in obscure journals, if at all, and are excluded from reviews and analyses. That skews the "average" effect upward." Pointing to what he has called this body of "pathetic" evidence, Toronto's Freedman cautioned his colleagues not to leap to conclusions:

"Some of those who read the available research carefully may conclude that the effect probably exists. Others will find that they are unable to make a reasonable guess, and still others will be led to think that watching TV violence probably does not affect aggression. But the research has not produced the kind of strong, reliable, consistent results that we usually require to accept an effect as proved."

Henry Jenkins, director of the Comparative Media Studies Program at the Massachusetts Institute of Technology, explained the Columbine shooters this way to the U.S. Commerce Committee in a hearing on youth violence:

"Far from being victims of video games, Eric Harris and Dylan Klebold had a complex relationship to many forms of popular culture. They consumed music, films, comics, video games, television programs. All of us move nomadically across the media landscape, cobbling together a personal mythology of symbols and stories taken from many different places. We invest those appropriated materials with various personal and subcultural meanings. Harris and Klebold were drawn toward dark and brutal images which they invested with their personal demons, their antisocial impulses, their maladjustment, their desires to hurt those who hurt them.

"So far, most of the conversation ... has reflected a desire to understand what the media are doing to our children. Instead, we should be focusing our attention on understanding what our children are doing with media."

Inconclusive and controvertible data, much of which does a crude job of describing a complex and poorly understood social process, should not be the basis of highly consequential public policy.