For a long time, I’ve used the term “delusionist” rather than “sceptic” to describe those who reject mainstream science on global warming. In general, the term “sceptic” is inappropriate for members of this group, since their position is hardly ever based on a willingness to look sceptically at evidence without reliance on a preconceived views. Rather the dominant characteristic is wishful thinking based on perceived political implications. The gullibility with which so many delusionists parrot the latest talking points (“Hockey stick broken!”, “Global warming on Mars”, Warming stopped in 1998″ and so on) is clearly incompatible with any kind of scepticism.

Given the volume of evidence that has accumulated on the issue, only an adherent of some very strong form of scepticism could reasonably remain undecided. Such a sceptic has now appeared in the form of Adam Shand, an Australian program on global warming “it’s only an assumption” that summer is warmer than winter. I imagine he gets great prices on ski holidays, by going in January.*

Of course, once you’ve gone this far in scepticism, why not go the whole hog? Radical scepticism provides the perfect argument for rejecting action to mitigate global warming – if we have no reason to believe in the existence of the external world, then trashing it can’t be a problem, can it?

* Northern hemisphere readers can make the necessary adjustments.

I once said that work on social software formed the experimental wing of political philosophy. I said it to a room full of geeks, not philosophers, by way of exhorting them to consider social ramifications of seemingly technological choices, such as “If you have a point system for good behavior, people will behave to optimize points, not to be good.” (cf. John Quiggans’ post on grades.)

Behind the basic point of this throwaway line, though, is something that has been puzzling me for some time. Like all groups with shared pursuit of shared goals, mediated groups need governance, which is to say rules for losing. It has to be the case that at least some participants in a group are willing to regard not getting their way as both legitimate and acceptable, or the groups would simply fork with every non-unanimous decision, and dividing groups with powers of two in the denominator would atomize even huge collectives after a handful of such decisions.

And so, several years ago, I began reading classics of social contract theory. After the initial excitement of seeing the similarities between Federalist Papers #10 and the Slashdot moderation system, though, I bumped into two key ways in which the arrangement of constitutions didn’t fit with the sort of rules for losing that are essential on the net.

The first is the concern, in recent centuries, about reining in majoritarian tyranny — preventing 50.1% of the polity from simply voting themselves into a permanent advantage over the other 49.9%.

This is something of a concern online, but its also clear that the really novel threat to group action in mediated fora is the tyranny of the individual. Even in systems not constructed around consensus, one or a small group of people determined to upset the proceedings can do enormous damage.

The second is the concern, at the center of the debate since Hobbes, about how leaders are to be legitimated, and under what circumstances, if any, they can be removed and replaced. This concern seems to stem in large part from physical and political facts — to a first approximation, each person is a citizen of one and only one country, and can’t readily switch citizenship should they object to the policies of that country. In the troika of exit, voice and loyalty, much political theory assumes that exiting is off the table for most people.

Online, though, inflexible one-to-one mappings of member to group are rare. One can contribute to Apache _and_ Linux, comment on MeFi _and_ BoingBoing, and so on. Indeed, the two most normal cases of governance on the net are the cabal (there is no cabal) and benevolent dictatorship, as with Linus and Linux or Guido of Python, whose acronymed title, BDFL, stands for “Benevolent Dictator For Life.” What keeps these dictators benevolent is precisely that membership in various groups is non-exclusive, and switching allegiances is under the user’s control, with no analog for rules of state.

So what I want to ask of the collected wisdom of CT readers is this: what one or two works would you pick, from any discipline, that best illuminate the group governance issues we see on the net, as different from political thought about the real world? (Mine would be Exit, Voice and Loyalty, and Logic of Collective Action.)

I believe that Peter Tatchell is planning to run for the Greens in Oxford East against Andrew Smith at the next election. What he ought to be, of course, is a rather dreary backbencher who held a minor position in Blair’s first and second governments, but quietly resigned in the lead up to the Iraq war. At least, that’s what he’d be if the Labour Party hadn’t decided to make him something else. Jonathan Derbyshire has a very fair and accurate account of Labour’s more minor but nevertheless spectacular own goals of the eighties.