Literature Review II

This is part two of my literature review, continuing from where part one left off. My previous post ended with thoughts on trolls as "actors" and their victims as the "audience" in a performance where the "actors" are the ones deriving entertainment.

I have found a case study to illustrate this. Paul Baker, in his article Moral panic and identity construction in Usenet in the Journal of Computer Mediated Communication (2001 v. 7, i. 1) argues that online identity exists as a process resulting from the interactions between a person, the “actor,” and others, the “audience.” So then, a person constructs his/her identity, but others contribute to that construction via their perceptions of that individual. Their contribution is manifested in the manner in which they treat that individual. Baker discusses a case study in which a troll attacked a Usenet community for fans of the television show Melrose Place, alt.tv.melrose-place. The topic that sparked the conflict was the troll’s anti-gay comments, in which he verbally attacked the target group, gay men. Baker defines six strategies that the other users utilized in order to resolve the moral panic created by the troll, “Macho Joe.” The last one, exposure (or “outing”), entailed all of Joe’s activities on all groups being made public, revealing contradicting messages. Following his exposure, the users realized he was simply provoking them, his identity within the group was changed, and the conflict was dropped. “Macho Joe” claimed to continue verbally attacking the target social group, gay men, in the alt.tv.melrose-place community in order to elicit the reactions of the politically correct people there, which he found amusing. His real targets were the other users, not gay men.

Baker also discussed "moral panic," in which a community encounters a dilemma, to which they collectively attempt to find a solution. After arriving at an acceptable one, there is return to normalcy. In this case, the community's uprising against the troll Macho Joe is the moral panic.

If Macho Joe's motivation to attack the alt.tv.melrose-place community truly was simply for the entertainment of watching the users' reactions, then he is similar to so many other trolls who claim to "do it for the lulz." But are they really trying to create situations of moral panic?

At this point I would like to make a distinction between "flaming" and "trolling." Trolling involves setting “bait” for unwitting people while flames are openly antagonistic attacks on a person or group of people. A troll is harmless to the person who does not take the bait. They are two distinct forms of "deviant" behavior on the Internet. I place "deviant" in quotation marks because what is and is not "deviant" on the Internet is not always clearly defined. As Janice Denegri-Knott and Jacqui Taylor note in the article The Labeling Game: A Conceptual Exploration of Deviance on the Internet in Social Science Computer Review, Special Issue: Deviance and the Internet: New Challenges for Social Science (v. 23 i. 1, 2005), "deviant" behaviors on the Internet are conceptualized mainly in terms of similar real-life behaviors (p. 94). Strictly defined, "deviant" indicates "not normal" or "departing from the norm" (from Dictionary.com, accessed February 25, 2009). But if a behavior that is defined as deviant in real-life (such as illegal downloading, to use Denegri-Knott's and Taylor's example) yet it qualifies as normal on the Internet, is it still deviant (note the distinction between "illegal" and "deviant")?

And when it comes to trolling, what is the real-life equivalent? This remains unclear, as does trolling's classification as "deviant" behavior. In some communities, trolling is very commonplace, especially in those communities mentioned in my previous post where trolling is used as a means to solidify group identity through shared knowledge/injokes. In other places... not so much. Trolling would likely be considered deviant on a cancer victim support group on Facebook, for instance. Denegri-Knott and Taylor have a workable definition: "a nonconformance to a given set of norms that are accepted by a significant number of people in a community or a society" (p. 97). However, this definition is very subjective and difficult to measure. Denegri-Knott and Taylor suggest that it can be addressed with the social identity deindividuation effects model, aka SIDE, in which the norms people follow depend on whether they identify themselves more in terms of the group or as their own individual selves. Stronger identification with the group leads to stronger influence of the group norms (p. 101). This is all well and good, except that a unique set of norms will develop in each different online community.

What role does anonymity play in trolling behavior? Less accountability and fewer repercussions are two aspects of Internet anonymity that are often attributed as enablers, if not motivations, to troll. These fall within John Suler's categories of dissociative anonymity, invisibility, and minimization of status and authority, which (along with three other factors) he considers to be related to the online disinhibition effect, which he describes in the article The Online Disinhibition Effect in CyberPsychology & Behavior (v. 7, i. 3, 2004).

The research on troll behavior is patchy at best. To begin with, it is difficult to operationally define. Tepper and Donath give us two, contrasting definitions of trolling, which are really two different behavior patterns occurring in different contexts. There are likely more distinct behavior patterns that fall within the vague category of "trolling." Next, is trolling deviant behavior? Assuming that it is deviant at least some of the time, when, where, and what type of trolling is deviant? Furthermore, what purpose does trolling (all types) serve? To reinforce group identity, to identify outsiders, and to entertain oneself have all been suggested as some of the motivations behind trolling. It is clear that a more systematic investigation is needed in order to first define trolling in its various manifestations and then to probe its function and meaning within online social interaction.