|Global X, CC BY 2.0, flickr|
I think that the two main findings, as summarized by Farrell, are partially unsurprising, but also significant, especially as they reflect the nature of the mechanisms of social media feedback. Up/downvotes are the bluntest possible feedback indicators - no nuance, pure affirmation/rejection - and as such, lend themselves quite excellently to overinterpretation. My tweet gets 35 favorites, and I feel personally validated - irrationally, I know, but I still feel it. My comment on a Chronicle article gets 35 downvotes, and I feel hurt - again, irrationally, but the reaction is there.
Ceteris paribus, that mechanism alone would tend to trigger, I think, a strong challenge reaction. What motivates you more to take the time and energy to engage in public discourse - the prospect of competing with contrary views, or the prospect of preaching to the choir?
Now mix in the finer-grained feedback mechanism of comments. At first glance, you'd think these would have a mitigating effect on the blunt-impact up/downvotes. But then consider how poorly, overall, we communicate in the written word, especially in online contexts. Comments can be no more nuanced than a "like"; they often reflect a respondent's own agendas and priorities, which are orthogonal to any consideration of the original poster at all; and they are often embedded within a pre-existing conversational structure which tends to promote a spiraling escalation of reaction.
I think that Wong is right - the problem, such as it is, is that his position tends to get borne out over timeframes that are too long for individuals. Sure, redditors may trend toward higher-level discourse under Wong's laissez-faire model - over generations, maybe. And if trolls can game such a system in the short-term to predominate and push out the non-trolls, that long-term gameplan is a dead strategy.
But beyond the explicit point of the article, I was really interested in considering the idea of reddit as a government. The criteria for identifying geopolitically-relevant entities have varied over time, right? Family-based, early on; geography-based, mostly; sometimes ideology-based (hello, Roman Catholic Church!); nowadays, increasingly corporation-based. As online spaces continue to merge into our meatspace in socially, economically, and politically salient ways, why would we imagine that online communities would be exempted ex hypothesi from geopolitical relevance? (for fun, Cory Doctorow, "When Sysadmins Ruled the Earth" 2006). (for serious, interesting conversations around open-source governance - Doug Rushkoff's Open Source Democracy and spin-off/inspired projects like airesis.us, wematter.com, efficasync, the Shadow Parliament Project, and the broader "libre culture" conversation).