Video Games and Violence, 1: A Rhetorical Aside
Steven Johnson's most recent book, Everything Bad is Good for You, has sparked some controversy over at his blog, largely over the use and interpretation of statistical data relating violent video games to subsequent violent behavior. This particular debate is motivated by general concerns over violent and inappropriate portrayals in video games, typified by Hillary's recent letter condemning the excesses of Grand Theft Auto: San Andreas.
The participants in the conversation over there have already mentioned a number of studies, and I should defer to the insights provided therein. I don't do quantitative studies, nor do I do psych work, so I can't comment on the credibility of the literature in question. At least not at the level of methodology. But I can comment on those rhetorical constructs that antedate those studies and that thus provide the interpretive framework. In order to make just a cursory remark, let me take one of the big meta-studies out there as a guide. In a comprehensive review of the literature concerning the effects of media violence, W. James Potter summarized the major findings regarding the relationship between violent portrayals and violent behavior, arguing that there are ten “empirically established laws” of media violence. These laws are: (1)exposure to violent portrayals in the media can lead to subsequent viewer aggression through disinhibition; (2) the immediate disinhibition effect is influenced by viewer demographics, viewer traits, viewer states, characteristics in the portrayals, and situational cues; (3)exposure to violence in the media can lead to fear effects; (4) an immediate fear effect is influenced by a set of key factors about viewers and the portrayals; (5) exposure to violence in the media can lead to desensitization; (6) an immediate desensitization effect is influenced by a set of key factors about viewers and the portrayals; (7) long-term exposure to media violence is related to aggression in a person’s life; (8) media violence is related to subsequent violence in society; (9) people exposed to many violent portrayals over a long period of time will come to exaggerate their chances of being victimized; and (10) people exposed to many violent portrayals over a long time will come to be more accepting of violence (see p. 26). While Potter argues that “after more than five decades of research . . . [we learn] that the strongest immediate effect is [that] exposure to violent portrayals in the media increases subsequent viewer aggression,” his review suggests that these “laws” highlight the vast complexity at work much more than they establish any simple set of causal relationships (see p. 42).
Research shows that while there may be an immediate effect of aggressive behavior after a single exposure to mass mediated violence, people who expose themselves to violent portrayals have more aggressive tendencies (Potter, 1999, 27-8). Similarly, some of the outside “key factors” which mediate the effects of violence in the media include: viewer demographics (including intelligence, ethnicity, and personality type), aroused states of the viewer, degree of identification with a character, the amount of punishment and reward in the portrayal, level of justification in the portrayal for the violence, the realistic nature of the violence, prior real-life experience, humor, and the list goes on (Potter, 1999, pp. 29-39). In short, while these laws may seem to conclude that violence in the media does in fact cause many negative and positive effects, there are many mediating circumstances which makes this causal relationship more complex. Due to the complexity of many of these findings, we need to pay pretty careful attention to the rhetorical understanding of violence, of behavior, and of what, precisely, constitutes a violent portrayal. Individuals who, after looking at the studies identifying a link between violent portrayals and subsequent violent behavior are not only mapping causality onto a particularly difficult-to-chart terrain, they are also relying on some initial assumptions that may not withstand interrogation.
Ian Vine warns, for example, that while media’s influence is undeniable, “the ghost of Behaviourism which still haunts the laboratories of media psychologists can be blamed for mechanistic fairy-tales about how audiences process messages. . . . [T]heir methodologies reveal assumptions which are both obsolete and deeply insulting to the intelligence of the ‘subjects’ who participate in their investigations” (Vine, 1997, p. 125). Vine argues that cause and effect relationships are more dynamic than those posed by media violence pessimists, and emphasizes the need for research to establish whether media’s influence on aggressive behavior is a result of a viewer’s “real inability to resist some potent powers of peripheral persuasion” or in poor media literacy skills (p. 143). As Vine asks it, the most fundamental question is this: does “the fault [lie] in our living rooms and classrooms, rather than on our television screens” (Vine, p. 143)?
One potential answer is, quite stubbornly, yes. It's not an eithor/or; the answer doesn't rest with one or the other, but rather with the articulatory practices of both (and more) of these environments. The point isn't to negate the potential relationship between the viewer and the depiction of a violent act, but rather that this question is not as interesting or as pertinent to scholars interested in the potential impact of violent media portrayals, since these sociological discourses already function so as to articulate the meaning of a violent act. In other words, if there wasn't already an agreement as to what qualifies as a violent act, there could be no potential agreement as to a relationship between an initial portrayal and a subsequent behavior. This linkage, which is always a negotiation based on study assumptions and the workers who code the data sets, may explain some the complexity that makes drawing any conclusion regarding a causal connection so difficult. This is a rhetorical problem, or at least a gap in what communication scholars call the "co-ordinated management of meaning," and it's not something that gets solved easily.
There is another, larger problem with these debates though, this one also rhetorical, and also explicitly comparative. It is the problem of mimesis, a problem that underlies all of these "video game violence is bad" narratives, and yet receives far too little attention. I'll try to get to that tomorrow.
