Saturday, March 29, 2008

The Meaning of School Shootings: Part Four (and Last)

My apologies for the wait between parts Three and Four -- my personal life (GDC, finals, projects, etc.) got in the way, as did my desire to make sure that I adequately presented the arguments against my thesis (which meant I had to actually read a couple of books :-)). But we're here at last, and we'll finish off the argument and provide a suggestion for improving our situation.

Re-cap of Parts One, Two, and Three
School shootings are statistically insignificant as a cause of death and injury in America, but they garner an unjustifiably large share of our media when they happen. I contend that we are interested in them not only because "if it bleeds, it leads" on TV, but because we recognize them as aberrant phenomena and, as sense-making creatures, we want to know why they happen and what they mean.

We are addressing part of why they happen -- not the short and long-term triggering events (bullying, childhood abuse, mental illness, all mentioned multiply in comments) but the long-term enabling events which result in murders rather than brawls or other, less socially unacceptable, behaviour? Why shootings? How did we go from fistfights to "rumbles" with knives to guns?

Violent media imagery plays a significant part. Violent media imagery includes movies, TV shows, TV news, music, and video games (computer games, arcade games, and console games). Studies show clearly that consumption of violent media imagery correlates with later violent behavior and that there are at least two mechanisms at work: arousal (which reinforces violent behaviours observed and performed) and desensitization (which reduces barriers to violent behaviour). Statistical examination suggests that between 15% and 37% of the increased aggressive or violent behaviour can be explained by exposure to violent media imagery.

Arguments against violent media imagery promoting violent behaviour often actually support the idea: James Paul Gee promotes video games as very effective teaching tools and arousal is at the heart of his argument. Harold Schechter suggests that all popular media is demonized and that the US has always had extremely violent media in the form of broadsides, dime novels, public executions, and other "savage pastimes". But he acknowledges that children behave differently today consuming both passive and active violent media imagery and that the physical "roughhousing" that resulted from earlier media violence is now missing.

In my opinion, the link between violent media imagery and violent behaviour (both by correlation and by method) is well established. Violent media imagery effects (we'll see shortly why "causes" is not the right term) violent behaviour in two ways: it reinforces (via arousal and lack of contextual negative feedback both in story and real life) and it desensitizes (via repetition and contextual positive feedback both in story and real life). Note the common element of contextual feedback, which is covered extensively in Grossman: soldiers kill within a sharply defined context which requires (among other things) approval by an authority figure. Violent media imagery generally fails at "properly" (by which I mean in accordance with either military training or a desire to limit societal violence) contextualizing violence. This is a substantial difference between Schechter's "savage pastimes" and modern media violence: public executions, for instance, were clearly contextualizing non-state violent behaviour as inappropriate and subject to state sanction; even 1950's media like Daniel Boone (cited by Schechter as extremely violent by today's media standards) show a black-and-white view of violence (good violence by state actors and heroes in retribution for bad violence by non-state actors and villains like "Indians") that has context which is missing in much of today's violent imagery.

Viewed in this light, violent media imagery doesn't cause violent behaviour; it enables violent behaviour. As many commenters have noted, other causes (mental illness, abuse, bullying, ...) lead to violent behaviour: what violent media imagery does is make the response to those causes more violent than it would be otherwise.

One particularly frightening enabling factor is the use of video games (yes, here I am explicitly indicting video games as opposed to generalized violent media imagery) for training. The US military and law enforcement have used video games back to Duck Hunt to teach trainees in "shoot / don't shoot" choices (largely "shoot" for the military and "don't shoot" for law enforcement, but the techniques are very similar). As the military uses them, games of this sort can be considered conditioning tools. As the law enforcement community uses them, games like Hogan's Alley are some of the best shoot / don't-shoot training aids available, and are far cheaper than creating a real "Hogan's Alley" like the FBI's training facility.

There is another way these games work, however:

Fourteen-year-old Michael Carneal steals a gun from a neighbor's house, brings it to school, and fires eight shots into a student prayer meeting that is breaking up. Prior to stealing the gun, he had never shot a real handgun in his life. The FBI says that the average experienced law enforcement officer, in the average shoot-out, at an average range of seven yards, hits with approximately one bullet in five. So how many hits did Michael Carneal make? He fired eight shots; he got eight hits, on eight different kids. Five of them were head shots, and the other three were upper torso. The result was three dead and one paralyzed for life. I tell law enforcement officers about this when I train them, and they are stunned. Nowhere in the annals of law enforcement or military or criminal history can we find an equivalent achievement. And this from a boy on his first try. (Grossman & DeGaetano, Stop Teaching Our Kids to Kill, pg 4)

Video games are effective training devices. The US military wouldn't use them if they didn't work. Michael Carneal not only hit eight out of eight times on eight different targets, apparently all eight shots were into the "sniper's triangle" of upper chest and head. Impressive. Possibly unequaled. But still odd. Most inexperienced (and many experienced) shooters fix on a single target and fire until that target goes down, sometimes pulling the trigger continuously on an empty weapon.

The normal, almost universal response is to fire at a target until it drops and then move on to the next target. (ibid, pg 76)

There's only one school of shooting that teaches you to stand in one place and put one round into each target's head: video games.

Michael Carneal...never moved his feet during his rampage. He never fired far to the right or left, never far up or down....most video games each you to fire at each target only once, hitting as many targets as you can...And many video games give bonus effects ... for head shots. (ibid, pg 75-76)

Carneal is perhaps the most extreme example, but there are others. Wesley Schaefer in South Carolina, the Jonesboro, Arkansas shootings, and even Columbine have various links to video games ("obsessive" playing of video games, use as an explicit training tool, gaining tactical expertise, etc.). In none of these cases is the training link as explicit or as dramatic, and in none of them do video games cause the violence. But in all of them the training provided by video games enables the activity.

What's To Be Done?

Banning video games or violent media imagery is not on the table. Not only is it un-American, it probably wouldn't work. When you find a course of action which is both immoral and ineffective, it's best to look for other options.

Aside from attacking the actual causes of violent behaviour (mental illness, bullying, abuse, etc.), the best long-term solution is to properly contextualize violence, especially for young people. Soldiers returning from war do not have a significantly higher level of murderous behaviour than the population at large. Their training has enabled them to function in war without loosing warlike behaviour upon the rest of us, largely through dehumanization of the enemy and because of the strong requirement for authoritative orders before firing. These controlling elements are largely missing in most violent media imagery. In fact, much modern media glorifies the hero who breaks the rules in order to violently solve problems.

Consider the difference between Blade (the comic book, movies, TV series, and video games) and Grand Theft Auto: San Andreas. In Blade, the enemy is well defined and, in fact, inhuman (or serving the inhuman vampires). Killing a human in Blade: Trinity is clearly contextualized as a Bad Thing for a variety of reasons. Although GTA:SA is somewhat cartoonish in appearance, there's no dehumanization of enemies at all: they are fellow people, whether innocent civilians, enemy gang members, or even police officers. There is no clear distinction (in the game -- some players may provide one themselves) between civilians and combatants.

Let's be clear: the difference is not multiple forms of media versus video game. The difference is the contextualization of violence as acceptable against a limited class of enemy (one you're unlikely to encounter in real life, I might add) versus the contextualization of violence as acceptable against anyone. In 1950s and 1960s TV, violence was often properly contextualized (for the time: we need not discuss the inappropriateness of accepting violence against Indians or other class or racial grouping that was generally acceptable then) as acceptable when used against certain groups and unacceptable against other groups. We may lament the groups chosen then (or now) and we may consider this distinction irrelevant, but it is actually critical when you look back (see Part Two) at the five factors of Grossman's model for making killing acceptable to soldiers: Demands of Authority, Group Absolution, Predisposition of Killer, Total Distance from Victim, Target Attractiveness of Victim. Context figures prominently in Demands of Authority, Group Absolution, Total Distance from Victim, and Target Attractiveness of Victim.

Several parts of the 1954 Comics Code addressed this contextualization:

  • Crimes shall never be presented in such a way as to create sympathy for the criminal, to promote distrust of the forces of law and justice, or to inspire others with a desire to imitate criminals.
  • If crime is depicted it shall be as a sordid and unpleasant activity.
  • Criminals shall not be presented so as to be rendered glamorous or to occupy a position which creates a desire for emulation.
  • In every instance good shall triumph over evil and the criminal [be] punished for his misdeeds.
  • Inclusion of stories dealing with evil shall be used or shall be published only where the intent is to illustrate a moral issue and in no case shall evil be presented alluringly, nor so as to injure the sensibilities of the reader.

The 1954 Comics Code isn't a model I'd like to emulate, but you find ideas where you find them. An informal code for media intended for children (under R, certainly) might be useful.

Unfortunately, "proper" contextualization is culture and nation dependent. Consider Counter-Strike:Condition Zero, in which players play either anti-terrorist units or terrorist cells. From an American cultural perspective, proper contextualization would limit players to playing anti-terrorist units. Other cultural contexts might find more propriety in playing the terrorist forces (although they would undoubtedly be relabeled "freedom fighters" or "Warriors for God" or something similar). Contextualization requires moral choices that many liberals in America are likely to be uncomfortable making. Of course, some conservatives may find contextualization just as difficult: is it OK to use violence against people violating the rule of law in America? Regardless of your position on the political spectrum, there are moral elements present in determining a "proper" context for violence, and there are some positions (ultimate pacifism, for instance) from whom there is no context where violence is acceptable.

Ultimately, school shootings are meaningful because they are aberrant. They focus our attention on changes in our society that normally remain hidden: the pervasiveness of bullying and abuse, the increase in violent media images, the effectiveness of video games as training devices. What we choose to do with that attention and the knowledge that comes from it is the hard question. The harder we look at the system the more complex it becomes, and the more complex it is the less likely simplistic solutions (ban video games, demonize Hollywood) are to work. Complex solutions (reduce abuse by reducing poverty and rebuilding the family, reduce bullying by diversity and education, recontextualize violence as inappropriate in more circumstances) are harder to conceive and immeasurably harder to implement, especially when public policy ideas must be sold in six-second sound bites.

Finding a solution begins with understanding the nature of the problem.

James Paul Gee, Why Video Games are Good for your Soul, Common Ground, ISBN 186335574-X, 2005.
Dave Grossman and Gloria DeGaetano, Stop Teaching Our Kids to Kill, Crown, ISBN 0-609-60613-1, 1999.
Dave Grossman, On Killing: The Psychological Cost of Learning to Kill in War and Society, Back Bay Books, ISBN 0-316-33011-6, 1995, 1996.
Harold Schechter, Savage Pastimes, St. Martin's Press, ISBN 0-312-28276-1, 2005.

A Personal Note: when I set out researching this topic, I was of the opinion that violent media imagery was generally irrelevant to violent behaviour. I had worked extensively in video games and been constantly assured that we provided catharsis not conditioning. I ignored facts like the military using video games as training tools (both tactical training and conditioning). I've read some of Grossman's fiction and did not like it. I've followed the stories about Jack Thompson and find him odious, opportunistic, and overly sensational. I started writing this series (originally going to be a single post) with the idea that I would show that school shootings and violence are statistically irrelevant and have no underlying "cause".

But upon reading the source material, my opinion changed as my understanding of the mechanisms deepened. Imagine my surprise when I wrote this series instead. I have read more than a thousand pages on this subject (and have one major work left to read -- it turned out to be unnecessary for this series), and it turned me around 180 degrees. I feel like someone smacked me upside the head with a big pile of bumper stickers saying "if you can't change your mind, how do you know you have one?".

Thanks for reading.