At 08:14 AM 1/12/2008, Anthony wrote:

Wikipedia has needed a constitution for a long time now. Top-down or bottom-up, that's the only way to stop "this" from happening again. The board seems to have rejected the top-down approach, as has the arb com. But then, the community seems to have rejected the bottom-up one.

There are two classic approaches to the structural problem: Anthony has named them top-down and bottom-up. Nearly all coherent organizations of any significant size are bottom-down, such that most of us think "organization" means top-down, and we know what problems top-down organizations have, so we reject "organization" entirely. Or others, quite correctly as well, see organization as necessary, but, again, from long habit, think it must come top-down, hence the call for "leadership."

This problem was actually solved over sixty years ago; a peer association was created that was, in its functions, bottom-up. There is no top-down organization, formally, though outsiders might think there is, out of the habit I mention above. The organization is Alcoholics Anonymous, and it was structured by Bill Wilson -- as the "leader" -- to avoid what had made prior attempts at temperance organizations fail. AA functions at the local meeting and personal level, the national organization has no control over local organizations ("intergroups"), and intergroup has no control over individual meetings, and individual meetings may have "leaders," but the AA Tradition is "Our leaders are but trusted servants, they do not govern." Also from the Traditions: "For our group purpose there is but one ultimate authority-a loving God as He may express Himself in our group conscience. " While many will immediately object to the "religious" aspect of this, consider when it was written and the context, it's actually irrelevant: notice how the AA group knows what "God" wants: its group conscience. Not some scripture, not some dogma, not what the leaders say. In AA, group conscience means consensus. Meetings will go a long way out of their way to satisfy *every* member with a decision. In the end, of course, they *might* decide to simply vote, but it is considered undesirable, damaging to group unity, only to be done when necessary.

At the national level, there is a General Service Conference consisting of delegates elected by -- if they follow Wilson's advice -- supermajority (2/3) from regions. If no supermajority can be obtained after repeated attempts, the delegate is chosen by lot from among the top two. The thinking is that this increases diversity. This, of course, is a very simple system, what Wilson thought of more than sixty years ago -- and I may be ascribing to Wilson what was actually the product of wide discussion in the fellowship --, and there are other possible ways to create a democratic and representative national structure. But the national structure doesn't control the local *at all*. Rather, it advises the nonprofit corporation, AA World Services, Inc., the most significant function of which is to publish literature and to handle a few other national and international coordinating functions. AA World Services, Inc., is a traditional nonprofit, with a self-elected board, legally. However, the Conference essentially nominates candidates for the board, and the board traditionally accepts them. What has been set up is a legal structure where the corporation is legally responsible for its own actions, and is free to follow its own opinion, but it has strong motivation to follow Conference recommendations.

Why? Well, there is another Tradition: AA "organizations," from meetings on up, don't accumulate assets beyond what is called a "prudent reserve," which is basically enough to shut down gracefully if the flow of donations stop. These service organizations within AA are continuously dependent upon the ongoing support of their members, and their members *do* have options. Don't like a meeting? Start another, and starting another meeting is trivial. AA grew in this way, actually. It's part of why it rapidly became almost the only show in town, competition to AA is tiny; and the strongest reason is that there is practically no reason for it. Tradition 3: "Our membership ought to include all who suffer from alcoholism. Hence we may refuse none who wish to recover. Nor ought A.A. membership ever depend upon money or conformity. Any two or three alcoholics gathered together for sobriety may call themselves an A.A. Group, provided that, as a group, they have no other affiliation." (Other affiliation means some defining characteristic that would exclude some alcoholics. Exceptions have been made, such as women's meetings, but even those have been hotly debated.) What this means that, in theory at least, there is no top-down dogma imposed on meetings. In practice, I'm sure, AA members are human and there may be places where some meeting which was promoting "controlled drinking" might have trouble getting listed. But I don't know what actually happens if that is tried.

My point is that AA is *highly* structured, but the structure is bottom-up. And AA has been phenomenally successful. I'm not an alcoholic, but study of AA led me to develop the concept of "Free Association," which is a peer association that essentially follows the AA traditions, organized for any purpose. There are actually countless such, for the Free Association concepts are pretty much the default for small peer associations. It's only when they grow beyond a certain point, or when they formally organize, that they become something different. So the problem reduces to the well-known problem of scale in democracy.

Attempts to maintain direct democracy as an organization grows run into the well-known problem on Wikipedia. The problem is not voting, in itself. The problem is deliberation. If everyone can participate in deliberation, and without structure to organize deliberation, the traffic becomes overwhelming as the organizational size grows. This discourages participation, and, for a time, there is some homeostasis due to participation bias. Only those who care participate; however, this, then warps the deliberation away from the general community. Town Meeting governments often make decisions in Town Meeting that are then, when they must be by law submitted for ratification by the electorate, rejected. Somehow the direct democracy of Town Meeting did not actually represent the town. It's participation bias, plus, as well, poor communication between Town Meeting and the general citizenry.

I have a solution. I invented it something like twenty years ago, but did not publish it. However, it's been independently invented in various places around the world, I've encountered at least five independent inventions. On Wikipedia, it's called [[Liquid democracy]], but liquid democracy was a narrow concept dealing with elections and choice, whereas my own concept, delegable proxy -- which is structurally the same thing, at least in some version -- has been replicated by others, and there are a few small organizations using or planning to use it. In a Free Association context, DP would not make binding decision, it is an *advisory* network, and the advice flows in both directions, so the proxy network that DP creates essentially becomes a full "nervous system" for the organization, with afferent and efferent "nerves," and every "synapse" being an intelligent filter for information traffic. I'm not going to detail how it would specifically apply to Wikipedia, but I can see fairly clearly how it could solve the problem of participation bias, how it would create ad-hoc assemblies that directly or indirectly represent everyone, yet which can deliberate issues and make recommendations and measure consensus through the active participation of a relatively small group.

Contrary to what one might think at first, this creates no bureaucracy. It requires no action on the part of WikiMedia Foundation. It requires no general consensus. It would be, in fact users doing what they already have the right to do: communicate with each other, choosing whom they communicate with.

As there is little or no cost associated with it, and it is, in fact, designed to make communication efficient, and participation is fully voluntary, what's the problem? What is stopping this from happening? Two things. One is inertia. Few see the benefit, few will follow the analysis of the problem sufficiently to understand what this would do and why it would be desirable. Another aspect of inertia is a general despair over there being any solution at all to the problem of scale. The other obstacle is that if the proxy networds actually start to form, they will destabilize the existing power structure -- or, more accurately, those with excess power in that structure will fear this -- and action will be taken to stop it. I was shocked to discover how, on Wikipedia, voluntary associations of Wikipedia members that did not require anyone else to do anything, that were aimed at helping people with Wikipedia process, were deleted and salted. (I'm talking about Esperanza and the arbitration advocates, whatever they were called.) There was no significant allegation in the deletion reviews that they were harming anyone. Indeed, the deletion arguments were that they were inefficient; surely that would be a matter for the members, since all the cost of the inefficiency was born voluntarily by the members, and if no members carried it, nobody else was harmed. The lesson I took from this: if there are to be developed proxy structures for Wikipedia members, they must be independent of Wikipedia, not vulnerable to interdiction by deletion of the organizing mechanism. And, again, this would be taking an example from the AA experience. Members communicate directly with each other, they do not need any support or structure or permission from AA World Services, Inc., from Intergroup, from their local meeting. All those simply facilitated their meeting and do not control, in any way, the direct communication.

If anyone is interested in pursuing this topic, write to me directly, and I will organize a mailing list to help develop it. As far as possible general applications -- some of you who have come this far might easily recognize that Wikipedia would be a relatively minor application compared to some other possibilities -- again, I welcome communication. I'm not about to exhaust myself trying to "get" the Wikipedia community to adopt these concepts, it will or it won't, and if nobody is interested -- hey, the motion failed for lack of a second. I just think Wikipedia will be a *lot* more successful, in the long run, if some of us look at this, and it only takes a few. The problem of scale will get worse, unless the community develops a coherent means of addressing it, or, alternatively, the Foundation steps in and exerts control. Watch.

Wikipedia Anonymous First Step: We admitted we were powerless over Wikipedia. :-)

(This means individually, *not* collectively.)