Faction

"I'll find a day to massacre them all, and raze their faction and their family..." - William Shakespeare, in Titus Andronicus

One good user-land definition of a faction is "those who disagree violently with others, but only non-violently with each other."

In wiki management, a faction is a mandatory clique to deal with alleged and collective identity problems that can't otherwise be sorted out without serious privacy problems, e.g. authorizing outing. The Wikipedia Red Faction is the most obvious declared public example.

The difference between a faction and an interwiki identity standard is that you yourself decide to assert a common identity with the latter, but with a faction, it would be others telling you "please go deal with "others of your kind" and come back when we can understand you, if ever." ;-)

Factions already exist as a group of users sharing a single account or using anonymous proxy services to reuse IP numbers and appear to be just one persistent troll or something. This can actually work better for some things than any permission-based model, but it gives an edge to those who have figured out how to do it reliably. It also makes it hard for any limits to be put on such activity.

So, there may be user groups that are answerable only to each other and not to the management, which only manages the conflicts between factions, rather like it also tries to stay out of conflicts between users and just play referee:

A faction competes with other factions, and has a tendency to view some things the same way - it may formally cooperate or list some values or principles. It has its own view of the glossary. Any funded trolls advancing a view implies there must be a faction with that view - maybe not one that Consumerium:Itself can see.

Often it is assigned a colour on the political spectrum.

Factions exist to acknowledge, limit, and channel various aspects of the self-interested fork problem. By anticipating factionally defined terms in the Consumerium License, we make it easier for factions to define their own Consortium license as a sub-license of our own parametric license.

Two factions are currently directly supported by the wiki software itself, those being developers and sysops. These have, as a consequence of the software itself, powers to label and identify others as vandals (those who damage or delete pages or insert erroneous assertions). There are also trolls (who annoy other users but may be doing so for legitimate political or social reasons, i.e. not simple vandals). It is rather hotly contested whether there is any one faction of trolls, whether trolls cooperate in factions already, etc..

One way or another, like it or not, there is reference to these collective entities in all large public wikis, and so there should be some formal support for creating arbitrary factions that actually represent the complexity of the social and political situation around the GFDL text corpus and the many point of view and user role differences involved. Factions would require some accountability, slowing down decisions on factionally defined terms like picket terms (like "pro-choice" and "baby murder" which you will never hear from "the other side"), and other political virtues and peacemaking tactics. Though it might just be one more thing to fight about.

Three models of dealing with this have been proposed, and one attempted:

• A GodKing who simply serves as the court of last resort, "master of truth and justice" (Michel Foucault's term for the role of the Western academic), and is probably trusted by, or is, "the owner of" a GFDL corpus access provider. This is used at Wikimedia. It has had some success in getting a lot of text contributed. It seems to be less successful in getting rid of systemic bias or removing irreponsible sysops. Also, few GodKings speak every natural language! Some suggest that a GodKing doing nothing is one of the best models, called lazy tyranny, and that it is inaction rather than action that has led to the success of Wikipedia so far.

• A faction system modelled on representative democracy as carried out in all democracies, where an edit stands or falls based on the willingness of some substantial faction of like-minded users to support it. These compete with other factions in a power structure to contain the more bureaucratic and police-like sysop power structure. This seems to have potential to simplify debate on extremely contentious issues in the same way that parties do so in countries.

This is the proposal that probably most suits Consumerium:Itself.

• A phyle system similar to that described by Neal Stephenson in his book The Diamond Age. As Metaweb is overtly an implementation of the Young Lady's Illustrated Primer from that same book, it seems entirely likely and useful that it would attempt to implement this variation of the faction. A phyle differs from a faction in various ways but mostly that it is defines an etiquette of its own and discourages informal interaction with those of other phyles, and typically takes revenge in extralegal ways if it is seriously offended (unlike a faction which is expected to work through some common bureaucratic or electoral process). See Metaweb:phyle for more details on this.

Because it implies Neal Stephenson's model, Metaweb will likely move in this direction, and attempt to model Stephenson's phyles as factions instead of letting them develop directly from politics as usual.

See Talk:faction for extensive discussions. It will be hard to agree on one definition of faction, so please review glossary in detail to see what you think of those generic ideas, and how a faction might define a lot of things differently.

See en:OurAnswer:faction for a similar faction system for a highly political project. Disinfopedia, also very politically focused, is declining for failing to have such a system.

See also en:Metaweb:faction. Consumerium and OurAnswer have more need to identify factions than Metaweb, so we may lead them in this regard.

See en:Recyclopedia:faction and en:Recyclopedia:direct_democracy for a proposal to deal much more effectively with community point of view bias.