The increasing popularity of on-line discussions has given rise to an argumentative neologism that may be more widely applicable: “trolls.” Trolls commit an inappropriate move in an argument, saying something unreasonable that derails the discussion. (I recall analogously in my highschool biology class we learned to ask the teacher, Mr. Houghton, about living through the London Blitz in order to steer the conversation away from the work at hand.)
These unpleasant people are not trolling the web in the sense of carefree fishing, or surfing, but today Mike Elgan, who bills himself as “the world’s only loveable technology writer,” suggests that trolls are seeking something, namely attention. That quest does not particularly distinguish trolls from the rest of us, but it does explain the behaviour as depending on that exclusive or predominant motivation.
Trolls are argumentative, and they may be either deliberate and malicious or inadvertent and well-intentioned. Egan’s distinction, borrowed from Matt Honan, between deliberate and inadvertent trolls corresponds to Walton’s distinction between fallacies that are sophisms and those that are paralogisms.
Yet Elgan points out that those who are well-intentioned and argumentative (the academy is so full) are not always trolls. Passionate advocacy frequently may be trying but it need not be ugly, and it is often beautiful and worthwhile.
How then do we identify trolls? Might this be a species of fallacy that can be identified as deviating from an otherwise acceptable form of argumentation, that is to say forms of advocacy? Perhaps we could articulate the appropriate critical questions (using the Walton / Tindale model of fallacies) for identifying such trolls.
If the desire for attention is the cause of the misstep, then what is the missing (or side-lined) motivation that would be appropriate? How ought we to be motivated? That is a central question of argumentation theory, and answers include resolving disagreement (pragma-dialectics) and developing understanding (epistemology). Fabricated disagreement and errant claims thus would be paradigmatic troll moves, but that is only to say they are fallacious.
If trolls fit no particular pattern of fallacious reasoning, they may nonetheless indicate a new need for fallacy instruction: preventing trolls from derailing discussion. On-line trolls have made available for instructors a new wealth of examples of fallacies. Students should also learn that the fallacies approach to argument evaluation may be a good defense against trolls, a way to defuse a diversion by naming the problem. While this rhetorical power of the fallacies approach can be misused, it can also be valuable in dealing with trolls.
RAIL 
A good post Cate, and timely. By an amazing coincidence, trolls have been on my mind of late also. It seems to me that trolls inhabit certain domains more than others. Domains where, perhaps, actual expertise or command of facts is less important than, say, the ability to draft clever put-downs, or engage in rhetorical grandstanding seem to me to be more “troll-friendly”.
Also intriguing to me is the possibility that some trolls–while almost certainly seeking attention as you suggest–are also guilty of terminal failures of Grice’s rules too. Though I would defer to my friends more schooled in pragma-dialectics on this point, I think it possible that trolls fail the most fundamental test: they don’t seem to follow the cooperative principle at all. This is because they’re never actually speaking to their interlocutors, but to those they seek to entertain, or perhaps those in the audience who already share their point of view. If the audience is divided just right in terms of its adherences, the case could even arise where multiple trolls, catering their constituencies within the audience, create a verbal event that would resemble a dialectical exchange to outsiders, but would in reality be little more than co-located, synchronized monologues.
Fascinating stuff, trolls. Perhaps there’s a paper in this?