Erik Moeller revealed how Wikipedia wants to expand its reach and bring even more participants into the fold.

However, it doesn't bode well that he seems to have attracted about 18 audience members at a conference that purportedly drew 500 attendees.

Step 6: PROFIT!!!

Jimmy Cricket would be proud of Erik "Wikipediot" Moeller.

It's been fascinating to watch the Foundation utterly squander the money they got for strategic planning. It's taken them three months to come up with the conclusion that a strategic plan should be a plan involving, in some way, a strategy.

Their money would have been better spent on a dozen copies of Strategic Planning for Dummies (and, yes, that book really does exist.)

Something tells me you don't mean the Irish comedian.

Something.

He touched on governance, and he's aware of the problem of, say, a new editor writes an article and WHAP! a deletion tag appears on it within minutes, but ... he is talking about scaling up by a huge factor in terms of editors actually interacting with the project, and shows no clue that he has a concept of what kind of governance could handle the scale. The fact is that with a few thousand editors reasonably active, the structure is overwhelmed.

Good ideas, to be sure. Lipstick on a pig? As long as the governance issues aren't addressed, yes, that's about what can be done. The numbers of active editors levelled off at the beginning of 2007, yet I'm sure new editors are arriving, and the arrival rate may have even increased (has it?). What's happening is that they are leaving at the same rate, burned out with a process that makes them push the boulder up the mountain, only to see it roll down again, and over and over. The level of wasted labor in active articles is astonishing, and it is only that the project placed no real value on this labor -- free, right? -- that it seemed that this wiki structure, raw, was a great thing, once Wikipedia moved beyond the stage were stubs were welcome and almost never deleted unless blatantly inappropriate. It works in small communities, for sure, but doesn't scale. Wikipedia is still, in certain ways, small-scale; that is, the community working on each article is relatively small, usually. However, as the overall scale grows, so too will the community interested in each article.

And every new article faces a phalanx of "interested" Recent Changes patrollers. If they don't dislike the article, they leave it alone -- and it doesn't end up on their watchlist. If they dislike it, for whatever reason, they'll tag it. As the number of Recent Changes patrollers increases, fewer and fewer articles will survive, unless somehow RCPs include active inclusionists, and I know what has happened to those: they burn out, if they aren't harassed off the site or banned. Yet, of course, RCP is still inefficient enough in the other direction that tons of drek survives....

I see no current trend on-wiki that could possibly address the problem. Initiatives have been started that might have done something, and they were shouted down. There are no leaders who are able to organize the community, and no structure that allows leaderless, efficient, decision-making.

There is a solution, at least one, but it's going to take a few people willing to, first of all, recognize the problem, second, study and develop solutions, and, third, implement them regardless of the screaming and shouting; implement them, not to control everyone else, but simply to make their own work more efficient and effective, exercising their own natural discretion and power. Whatever group starts this will outcompete other groups, it will be able to edit circles around them, and, after it's realized that it's not stoppable, it will be imitated. The organizational technology to do this is known, but little applied.

In reality, it's a formalization of what already happens on a small scale, but by being formalized, it becomes reliable and scalable, and much more efficient.

Clue: off-wiki structure, independent, efficient, as inclusive or exclusive as the participants in each element of it want to make it. ArbComm had better get ready for it, because they are utterly unable to stop it. The EE mailing list was only found and may be sanctioned because it was not defensively structured and cellularized. The hopes of Coren that by sanctioning these editors he will somehow stop the trend are false hopes, doomed to failure. Repression doesn't work, in fact, unless those repressing have enough power, and the power of ArbComm even as a whole is severely limited, it depends on voluntary cooperation, and voluntary cooperation cuts in both directions. He's very clear: he wants them to stop talking to each other.

I don't think that's going to happen. What I've suggested to the EE list is that they open it up. It seems they are actively considering it. What they were doing would have been fine if it had been open. They *talked* about doing stuff that would not have been fine, but, turns out, they did little of it. It was talk. Imagine, now, thousands of mailing lists on thousands of Wikipedia topics, where editors discuss articles and news in the field (most of the EE list is news and commentary on news). The list can regulate itself, it can "ban" editors who are disruptive, without having to refer to ArbComm or anyone else, and someone who doesn't like it can "start their own damn list." And much more becomes possible.

There are some real howlers in this thing.

Early on, he says the existence of Facebook "proves" that it's possible to get people to "interact on that scale." I mean, yeah, sure, if all your users are doing is uploading pictures of themselves, and then clicking on things and chatting. It proves absolutely nothing about an attempt to build a huge database of supposedly "neutral" general-reference content.

He later says there's "no reason why editing a wiki article should be harder than Google docs," a ridiculous claim on its face - Google docs are mostly private or intended for specific audiences, you don't have to worry as much about security or the user interface or any number of other things. A minute later, he admits this, essentially contradicting himself.

In effect, the "300 million" figure is a scare tactic. (Near the end he refers to it as a "straw man," which suggests that he, too, doesn't know what the term "straw man" means.) He doesn't have any new ideas; he seems to think they should just expand on their old ideas. "We need more red links," as if the number of "notable" topics that can be written about in an encyclopedic fashion is infinite. "We should have more wikiprojects," as if you can just pull entire areas of study or general interest out of a hat.

Then, it gets better: They should "build more stable and reliable physical spaces" for people to get together and personally "interact." What are they gonna do, stand behind the guy who's actually at the keyboard and cheer him on? "Go, Erik, go! Revert that trollish vandal! Awesome!" If there's a more obvious indication that these people are running a cult, I'd like to see it.

Later: "It seems bizarre that universities are still such passive institutions," in effect saying that college classrooms should be replaced by collaborative websites. That one would belong in the "careful what you wish for" category, I should think.

And: "The Wikimedia community is fast at assimilating good ideas."

Towards the end, he says that "promoting every brand in the Wikimedia universe is a losing proposition," which is to say that secondary wikis like Wikisource and Wikibooks and Wikiquote are not worth promoting. I believe he's said this sort of thing before (on WikiEN-L maybe?) - he seems to believe that the WMF should concentrate its recruitment and promotional efforts on Wikipedia, perhaps even exclusively. He's probably right, but it's a retrenchment, not an expansionist "strategy."

And finally: "You may think Facebook is an evil, proprietary, privacy-invading company, but whether they are or not is irrelevant." The important thing for Erik is that they're "a huge community that is doing a lot of things incredibly well" - specifically referring to the ease by which people can "notify" everyone in their friends list of something that's important to them, "with just two mouse clicks." He then goes on to admit that doing this on Wikipedia requires talk-page spamming, and that this is discouraged, but he apparently doesn't seem to think that replacing or augmenting talk pages with Facebook-like social interactivity could be seen as a bad thing - when it should be obvious that the only way for this to be acceptable to the core/dominant user sub-community is to ban all the "you're not here to build the encyclopedia!" types from the site, leaving them with about 10-20 percent of their current core/dominant user base.

Still, I'm sure they'd have no problem rebuilding their numbers from there, what with all those cool "friend-notification" features.

Erik once proposed renaming all the subsidiary projects to names containing the word "Wikipedia". So Wiktionary would become Wikipedia Dictionary and Wikisource would become Wikipedia Documents (or something of the sort). This approach might actually be better from a branding point of view, but it was met with massive scorn from the communities of these projects, many of which pride themselves on not being Wikipedia.

Erik, fundamentally, has no interest in Wikipedia's (or Wikimedia's) ostensible goals; his interest is and always has been self-aggrandizement. Erik wants to be Important, and the more people are involved in Wikipedia the more important he is. So it is little surprise he pushes the social networking aspects of the Wikimedia projects, because it's that, more than anything else, that turns his little reward-crank. Whether they actually generate any content doesn't matter; all that matters is that he have worshippers and followers to gaze upon his importanceness in awe and wonder.

One of the board candidates tried to sell me this "call everything Wikipedia" idea for my support. It did affect my voting, but not in the desired way.

When Newsweek was researching a story about the Wikisource Wiki Bible project, Matthew Philips spoke with Jay Walsh and somehow "Wikipedia" ended up used in the article God’s Word, According to Wikipedia, despite it having nothing to do with Wikipedia and being contrary to the "No original research" pillar of Wikipedia. Wikisource is not even mentioned, and googlers and bloggers ended up thinking it was about another project.

Yeah, my first guess would have been this.

So, assuming I've read all this correctly and the short version is that "off-wiki structure," currently in the form of secret mailing lists and so on, will become more open and such groups will eventually be major forces on WP... I could certainly see that happening, but frankly, I still believe that secret groups will always have an (unfair) advantage over open ones, even if the open ones are always larger (due to less barrier-to-entry).

The one positive thing Wikipedia has going for it in this regard, which I doubt will change any time soon, is that most "normal," down-to-earth people realize that spending hours on end futzing about with WP articles (and the WP "community") is a bad, or at least unconstructive, thing for them to be doing. The public perception of Wikipedia as cultish, insular, and hostile effectively limits their growth - in fact, they probably reached that limit long ago, and we're only now starting to see the effects of their having reached it. They're not going to eliminate those perceptions, because they're not just perceptions, they're inevitable realities - those are just the ways people behave when presented with a system that does what WP does.

Besides, the effects are mostly good - at least they're talking about things like social responsibility, human decency, really helping parents protect their children from stuff they're too young to be seeing, etc., as if some of them are really seeing the value in those things.

Anyway, when I suggested that the "300 million" figure is a scare tactic, I'm referring mostly to the effect it's supposed to have on people like myself. Ultimately, what I really mean is that they've gotten some experience now regarding what deep-pocket funding entities want to see, and they realize that those entities are looking for Big Ideas™ and Big Ambition™. Philanthropists often don't bother looking past the surface and charting trends - sometimes they just look at the prospectus.

Well, you should understand that I work with two concepts, together: Free Association concepts which are modeled after Alcoholics Anonymous and which are, in fact, a practical implementation of old anarchist principles, and Delegable Proxy, which is a formalization of natural human organizational structure as it happens in small groups, but, because of not being formalized, it's swept away by other kinds of structures. So let's look at the problem. Wikipedia depends, in theory, on consensus, but it never developed scalable methods of finding consensus. The problems of scale meant that it had to attempt to prohibit canvassing, because if you make decisions in small groups, with all participants being peers, preferential canvassing can easily overwhelm whatever spontaneous and uncoordinated support exists on the other side. The no-canvassing rules are made necessary by the ad-hoc voluntary committee structure that WP uses. But, of course, this then leaves these "virtual committees" highly vulnerable to accidental combinations or other situations, such as watchlists, that can attract preferential support or opposition. It creates a strong participation bias. None of this would matter if decisions were made as they are supposed to be made in theory: by cogency of arguments, per a neutral closer. But ... in reality, closers are affected and often follow the majority, and not just because of statistical likelihood, nor because the majority is presenting cogent arguments, I've seen plenty of closes, following the majority, where there was no cogent argument, just a lot of votes with no evidence.

This is not the way one would intelligently design a system; it grew that way because of what worked, well enough, when the scale was small. When the scale is small, nobody would object to someone telling a friend that an issue of mutual concern was being discussed!

How to negotiate consensus when the scale becomes large? The basic method that is known to work is to break up discussion into small groups. These groups may be neutral, or they may be biased, "caucuses," I call them. Then people representing the groups negotiate on behalf of them, and, often, the discussions go back and forth. At every point, however, actual discussions (or debates) are small-scale. These "meetings" can be mailing lists, and some would be open and some would be closed. But with many people cooperating, the decisions that slip under the radar, the countless abuses of authority by certain administrators, the silent AfDs that pass unnoticed by anyone who knows the topic, all that would become much more rare. With many eyes, much more can be seen and analyzed and handled, far more efficiently.

Mailing lists alone won't do it. Secret lists won't do it, perhaps. They might, but all secret lists can do is to help a caucus get coherent about its position and share labor. Piling in to a discussion isn't sharing labor, so the existing system, which does respect votes, in reality, doesn't become efficient. Sure, with the existing defacto usages, a secret cabal has an advantage, but only under some circumstances, and it's pretty unstable, because the advantage can only lead to higher vote counts and ability to revert, and those can backfire, they get reversed when wider attention is brought, and those who tag team revert may end up being sanctioned, even without revelation of the sekrit cabal mailing list. If they are actually in the minority. If it's majority pov-pushing, they are pretty safe, as far as I've seen. Which is a problem in itself. Majority pov-pushing isn't seeking consensus, it isn't neutral. It's merely a little better (usually) than minority pov-pushing. But it's harder to address.

What will do it is the usage of mailing lists to seek consensus, whether in small groups or large. Large-scale consensus can be negotiated on mailing lists, with associated polling tools, but list traffic must be controlled when the scale becomes large, and that's an issue of its own. How does one control traffic without introducing bias? There is a way, but it's beyond the scope of this post.

Suffice it to say that external mailing lists are one possible way that can be used to negotiate Wikipedia consensus, and consensus is powerful, and the power is intrinsic. External lists cannot be controlled, the oligarchy cannot prevent people from discussing topics; if they try, they will create serious opposition to their own power. The Eastern European mailing list is very small, a handful of people, so it's fairly safe for the oligarchy to attack it as a threat to its power. But such a group could easily be a hundred times as large (if the focus were broader).

The concepts behind this are designed to work under highly repressive conditions, Wikipedia cannot withstand organized discovery of consensus off-wiki, nor, if it wakes up, would it want to.

It's possible that good structures could be organized on-wiki, though the software isn't designed for it. How much of the overall structure is off-wiki and how much is on-wiki is something that would develop, and it's not a crucial question at all, except that I prefer distributed independent networks for security reasons, they are far more difficult to centrally control, except through voluntary cooperation, in which case it isn't "control" at all.
That's right. So a system that leaves them disempowered if they don't do that is one that can't work in practice. I'd expect that when I talk about finding consensus, you might easily imagine large numbers of people discussing stuff. Nope. Small numbers. Representing, in effect, large numbers. But without the weight, the heaviness of that, without the centralization of power that corrupts judgment.
My comment generally is that it's the system that is the problem, not the faces, the people. You can get rid of the people but if the system doesn't change, they will be replaced with new faces that generally behave quite the same. The system generates the behavior it fosters and rewards, interacting with natural human propensities.

It is not the goal of the project, per se, that generates the behaviors, though that's an aspect. It's the particular structures that grew up in the early days of Wikipedia, that "Wikipedians" became familiar with and which many consider responsible for the growth, the goose that laid the golden egg, which they don't want to lose. They don't notice that the egg is rotting, they deny that the smell is coming from the egg which is rotting because they are sitting on it way beyond their time.

Some aspects of the system are indeed responsible for the level of success that Wikipedia enjoys, but some aspects are also responsible for the problems, and the system is burning out editors, and has been for some years now. Those that remain are often the least functional, the least suited for a true collaborative project that needs consensus to fulfill it's basic mission of neutrality. The Ponzi scheme is nearing collapse, if something isn't done. As Milton has pointed out, the content isn't at risk, it's easily mirrored and is mirrored, but the community that maintains it and expands it is at risk.

Something happens in projects that create a higher goal than the welfare of the community running the project, it happens in many nonprofits. As soon as one places a higher value on the "goal," that beautiful creation that is imagined, than on the health of the community, people start to dehumanize opponents to their own view of the goal. We see the argument over and over, it's about text, right.

No, it's about human knowledge, and human knowledge requires human beings and we are human beings. If you elevate knowledge above being human, well, that's the classic position of Satan. That, in a nutshell, explains a great deal about Wikipedia, how events there can become, for those who become attached to anything about the project, so amazingly .... evil. It's because it is! Contempt is the classic trait of Satan, and Wikipedians, too easily, fall into contempt for anyone who doesn't agree with their vision of good Wikipedian behavior. How can one possibly work on content with those POV-pushers, those fringe nut cases, those nationalist fanatics, those who want to fill the project with fancruft, those who actually supported the RfA of an editor who had previously evaded a block. Can't get more shocking than that, eh?

At Wikipedia Review, the relief is that one can actually talk about the insanity of Wikipedia as-is, with others who understand. WR doesn't have decision-making processes, but it's already useful sometimes to bring some matters to the attention of the WP on-wiki community. It's way too unfocused to serve as anything more than a good discussion group, though, because it has no means of measuring consensus. That could be fixed, you know. But it's not necessary for WR to function. For all I know, the trustees here (the actual oligarchy) are arrogant assholes who arbitrarily suppress what they don't like. Or not. Or some of this and some of that. Sometimes you get lucky, and when you don't, you move somewhere else....

Mailing lists can be started in minutes, to deal with any specialized topic, and they can be maintained for the long term, keeping editors involved long beyond what WP can do, mailing lists are push content, and with good practices, they can keep people a very long time. That incident at nl.wp with editors showing up to vote who hadn't edited for a long time, how did that happen? Quite likely, mailing lists. Not possible to prevent .... so use it.

tl;dr

Thanks, Kohser, for taking the time to tell me that you did not read my message because it was too long for you. Is it the case that you read none of it at all, or that you only read a little? Is it too long because you don't have time, too long because you aren't interested in the topic, too long because you are an arrogant asshole who doesn't think he has anything to learn, or what?

If it's that you are not interested, then please let me know, because then I won't waste time editing messages down for your benefit, and you can skip them.

I suspect Mr. Kohs is correct in suggesting that nobody read this far, but since I did:

It sounds like you're talking about the ostensible, stated goals, as opposed to the real, underlying goals of WP overall. I'd be willing to admit that many people become interested in WP because of the stated goals ("hey, let's assemble lots of free knowledge!"), but the percentage of people who stick around and become heavily involved primarily for that reason seems to be rather small, maybe something like 10 or 20 percent. The rest are basically gamers, socializers, and various agenda-driven types. (Note that I'm not talking about "dabblers" or casual users here...)

What I was really referring to, though, was something much more fundamental: Any wide-coverage, heavily-visited, large-search-engine-footprint information reference that can be edited by anyone, anonymously, is going to eventually be dominated by a smallish group of heavily-vested users whose organization and behavior makes the site cultish, insular, and hostile. If you replace individual users with interest-group cohorts (in the form of mailing lists or what-have-you), you might improve a few internal processes, and you might even get better articles, for all I know. But you're not going to prevent the creation of dominance hierarchies consisting of biased and agenda-driven individuals, and you're not going to have a system for which "vandalism" isn't a recruitment strategy.

In fact, I'm not sure there's any way to prevent those things and remain "free" and "open" - I suspect not, in fact. What I do believe is that if content quality and responsible management were of any interest to the WMF whatsoever, they wouldn't be talking about "scaling up" at all, they'd either be talking about "breaking it up," at least administratively, or locking the whole thing down.

Warning: Tome.

Sure, I thought so. But I also thought it was a good thing that some of these issues were being talked about; what's a shame is that the people and expertise that might actually be able to design structure to handle the scale and accomplish the core goals of the project seem to be entirely missing, the thinking is that of a web designer and someone focused on web business, i.e., how to make a site successful.

Core to the project would be building an effective and true dispute resolution process that seeks true consensus of the knowledgeable, with broad consent beyond that expert core on any topic, rather than simple decision-making. From a web business perspective, you just want a decision, and you don't want everyone to waste time making it when they could be out there fixing typos and adding references and deleting whatever they think non-notable, or non-verifiable, or Bad.

So a shallow web business perspective will miss what is needed, and what is needed will actually seem like a negative thing.
Actually, he made a very good point here, Somey. He was talking about the cumbersome wikisyntax interface, and how difficult it can be for some to learn it. I think you missed his point. It's a purely technical one, i.e., the solution is software and the goal simply ease-of-use, so that facility with a new language (wikitext) isn't a warping factor in determining who can edit. Now, you could make the point that the wikitext barrier filters out, to some extent, the incompetent, but it also filters out lots of people who would be expert on topics but who simply aren't interested in learning a new language.

Further, wikitext isn't that hard to handle for basic editing, it gets hairy with references though, and where do new users fall down most? I still find some of the more sophisticated referencing schemes to be way too complicated, and, brilliant design, you don't see how they will actually appear when you save a section, unless you edit the whole page, which can then create a major nuisance if you make a mistake. The work he's doing or planning is good work, it could improve the project, and it's among the things that WMF should support, but it's also quite naive, taken by itself.

Straw man, Somey. No, that's not what they will do. They will interact as human beings rather than editorial personas. That alone would resolve maybe half the disputes, and make it much easier to resolve the rest. However, what restrains the development of these "stable and reliable physical spaces" is active repression from the core. I knew from the moment I saw what happened to Esperanza. Esperanza had its faults, but participation was voluntary, and it should have been up to the participants to determine its future; instead, when it hit a flat spot and some of the participants thought it a waste of time, it was crushed. Because Esperanza depended on on-wiki structures for most of its communication, with only the "board" or whatever they called it being off-wiki, it was vulnerable to such disruption. Esperanza convinced me that true reform could only come from off-wiki structures that are efficient enough and effective enough to attract at least some kind of loose affiliation from a few editors.

To be sustainable and to grow, these structures must provide some immediate value to participants. They must attract an initial core that understands the goals and which will persist in spite of opposition.

What I've seen with ArbComm is that it seems to have an instinctive "kill" mentality about "cabals." It's considering the desysopping of an admin for what amounts to very little more than subscribing to a private mailing list and making some supportive comments there. The only use of tools alleged was a semiprotection of an article that was, in fact, the target of tag team reversion by IP and a new account that appeared just for that purpose, and when semipro wasn't enough, the article then was full protected, and an account involved in edit warring there was blocked, by uninvolved admins.

I was rereading the decision made in my case. One of the sanctions I'd forgotten about. I'm not to get involved, period unspecified, in disputes where I'm not an originating party. I.e., no neutral intervention. They included "informal dispute resolution." And I'd been fairly good at that, and that kind of work was a major goal of mine on-wiki. WTF? There was no evidentiary basis for this sanction. What's the purpose?

The purpose is to stop interference with the administrative cabal. I'd been good at it, intervening when what admins would do, or even had done, wasn't likely to be supported by broader consensus.

There was more to the decision that was without foundation in evidence. Bottom line, the majority wanted to get rid of me and my work. It made them feel uncomfortable. The cabal issue got mentioned indirectly, I'm ordered to not make allegations not supported by evidence. The only ones allowed to do that are arbitrators and plaintiffs against people like me! This was based on the cabal allegations, apparently, there was an FoF.

But I defined "cabal" carefully, and demonstrated that what I defined existed, the evidence I provided was overwhelming, and readers of WR know there are cabals, it's obvious. So, again, WTF? Well, it was claimed that I "really intended" to claim that there was reprehensible collaboration., that my careful restriction of my meaning was a smokescreen, and so I was reprimanded for claiming what I did not claim, and then not proving it with evidence.

This kind of behavior by ArbComm is a sign of Rule 0 violation. Thou shalt not mention the nudity of the emperor. But, of course, Rule 0 cannot be stated, because it mentions what it prohibits mentioning. So anyone who mentions or refers to that nudity, or who supports anyone who mentions it, must be punished, but the actual offense can't be mentioned, so many others will be invented. "Trolling." "Disruption." "Personal attacks." (I'd claimed that a named set of editors were mutually involved, such that their calls for my being banned weren't from "uninvolved editors," as WP:BAN had come to specify. Because I called the set of editors a 'cabal," and even though the same basic set of editors had been called a "cabal" in the media, this was then considered a "personal attack," and it went so far that the evidence pages were blanked after the case closed. Given the kinds of allegations I've seen over and over in ArbComm cases, strange, eh? Not really! ArbComm was demonstrating how Wikipedia works. At it's best! (Well, not really. At its best it works quietly, out in the calmer reaches, where people make a few corrections that they notice, or add a fact or maybe a reference. Or an article, though, increasingly, the core will delete it immediately instead of working to improve it. They can't afford the time, too much work.)

Back to the speech:
Well, perhaps badly stated. The community should be fast and efficient at vetting new ideas. Fast adoption is a hazard, actually, and the bulk of what many of us experience at repression is normal social filtering to prevent premature adoption of change. The Wikipedia problem is that the bypass mechanisms, that allow new ideas to be examined in small groups, and thoroughly discussed there, then expanded non-disruptively to collect views missed in the small groups, is missing. There is no coherent and efficient way to experiment with new ideas in ways that would be harmless if they don't work. WP:PRX was explicitly an experiment, to see what would happen if there were on-wiki structures that would collect expressions of trust to allow delegable proxy to be played with. It was rejected, though really it wasn't something that required permission, per se, and the attempt was made to completely delete the proposal, very unusual for something that had been discussed. If not for a sober long-term editor who correctly saw that there was something off about the fuss, it would have been deleted, the !vote was overwhelming for Delete. How is it that such a minor and harmless proposal attracted such a massive and rapid rejection? A rejection that showed no understanding of the proposal itself, that actually contradicted it? It's easy. Rule 0 violation, it threatened the oligarchy, because it could make the basis for oligarchical power on Wikipedia, participation bias, visible. It would empower occasional editors.

Certainly there are obvious dangers with the proxy concept, but the proposal assigned no power to the proxies, no policies were violated, the proposed structure was only for information, and no bureaucracy was being set up, no new standards to follow, and the claim that it would be paradise for socks was thoroughly bogus: puppet masters would never have their socks name them explicitly! That only makes sense in a structure where voting power is controlling, and there are no mechanisms to detect sock puppetry. And neither condition applies to Wikipedia in any way that would be enhanced with delegable proxy; if closers started to pay attention to proxy counts, the result would likely be lower editor comment counts in contentious discussions. Which would be an improvement, making them more readable and the average comment probably more cogent. (How DP would work isn't likely to be well understood by someone who hasn't put some time into considering it. Loops would form, you would not have huge structures with one person at the top who then decides everything. Most people will not understand DP until they see it working.
Yup. And so much for the "encyclopedia that anyone can edit." More accurately, "anyone" means "anyone who will be a subservient drone who only does what the core wants him or her to do, making no trouble, following all guidelines and policies, written and unwritten, unless the guidelines and policies protect the editor, in which case they are then violating the "no-wikilawyering" policy, which is vigorously enforced.

Direct community networking, on Wikipedia, can be done with existing tools, though they certainly aren't designed for it. There are operating examples that are tolerated, because they are centrally controlled, such as WikiNews. But anyone could start their own wikinews, but if it expressed a factional view, it would be crushed. There are other devices, but the biggest problem with on-wiki networking is that it requires login, for most users, and people who aren't fanatics don't log in enough.

This is what is legitimate about the fear of direct networking: the effort is to protect the small-scale discussion model. If small groups of editors form and seek consensus, Wikipedia works. It's considered that if a member of a faction recruits participants, small discussions will be overwhelmed.

The level of denial is very high, because the small-discussion model is already terribly broken whenever a cabal becomes involved, and it's broken even outside that through sheer quirkiness or lack of necessary participation. It's not that small-scale discussion is the problem, it's actually part of the solution. (Large-scale discussion is never deliberative, it's impossible, not if everyone can comment. It's possible with open discussion (in terms of access), and everyone being able to vote, and with representational weighting of votes, but that is highly experimental and not familiar to people. Hybrid representative/direct democracy.)

Real-world small discussion groups can limit their own participants, to keep the group on-focus toward finding agreement instead of endless struggle. ArbComm wants me to stay away from dispute resolution. What was I doing? Well, I'd see a problem developing where what appeared to me to be good-faith editors were falling into tendentious conflict. Often one or both of them had already been blocked. I'd point out the risks if they continued, and invite them to seek agreement on a page in my user space, and then I would guide the discussion to attempt to get them to fully listen to each other, to understand the arguments of the other side well enough to be able to repeat them accurately. (Classic technique is to ask a disputant to state the arguments of the other side so that the other side would say, "Yes, that's right!") The ultimate goal was to find agreement, if not on results, then on method to reach results. It worked. Editors who had been at each others' throats ended up cooperating. That's what ArbComm has prohibited.

In my user space, I had, in theory, quasi-administrative authority. I could revert without limit. I could ask disruptive editors to stop, and then seek enforcement if needed.

I could do all the same, without any question of my auithority, off-wiki, with email or a mailing list for larger disputes. And if I acted as an arbitrary dictator, trying to force one party to accept the position of the other, they'd simply ignore me and do something else. If I run a discussion group abusively, the rest can, having direct contact information for each other, form their own group. This kind of process, if certain conditions are met, can't remain abusive with participants who realize they have alternatives. That applies to Wikipedia as a whole, by the way, and the problem is only that participants don't realize they have options, or when they do, and they set up the options naively, they can be crushed.

On-wiki structures, as I've mentioned, could accomplish these goals, but .... they are then vulnerable to oligarchical control. To the oligarchy, that might seem a good thing. But it isn't, if the goal is consensus. Failure to understand the importance of consensus is the classic error that oligarchies fall into. If an oligarchy serves consensus, it will be invulnerable; instead, unenlightened oligarchies serve their own limited conceptions of their own or public welfare -- they are not good at distinguishing them -- and when they fall, in the real world, the fallout can be a complete disaster for the oligarchy.

The solution isn't to eliminate oligarchies, it is to do two things: educate them and make them responsible to a larger process that advises the entire community without creating a new oligarchy, since the larger process doesn't concentrate power in a corruptible fashion (more accurately, it will respond to corruption by identifying and isolating it, so that corrupted power is quite limited). The oligarchy continues in the executive function, with very substantial discretion.

All cult leaders want/need more unpaid workers as all cults suffer high drop out levels.

Ascendance within the cult, especially to a financially remunerative position, largely depends on one's ability to do so ... and in doing so the ability to detach from any conscience factor.

Projection of the cult's would be self-image, rather than an ability to analyze the reality and empathize with the unpaid and disempowered workers, are key components.

(As a minor aside, the way the Pee-dia - or Mediawiki - handles images is pretty shit too and acts as a disincentive to donations/participation IMHO. The bar is too high for non-tech users, as even is new page creation. They need a "ebay-like" dummy's front end for ordinary users AND the techie, old school back end for old hacks).

Wasn't trying to be an ass. Was trying to teach Abd about common principles of success on message boards. Except in very rare cases, shorter is better. Provide links to external sites if you wish to feed the hungrier reader. And, vary your post with bold text or bullet points, because that helps to

• hold
• the reader's
• attention

I'm not the first to advise you that your "wall of words" technique is both out of place here and lacking in efficiency as a communication device.

If you choose to take this advice by lashing out at me, that's your deal. I'm not bothered at all by it. If you want to write a blog-length article, why not do so on Akahele.org, where readers are more conditioned to take the time to read a longer-developing argument?

Well, if you did, he's not correct. Q.E.D.
The former, plus the collective agreement with that goal, to the extent that such collective agreement exists. The "real, underlying goals of WP overall" involves a lost performative. Strictly, "goal" is a characteristic of a sentient being. WP overall is not a sentient being, it's a phenomenon, an effect, and the effect is only the product of the collective activity of many people, each one of which has goals. We can talk about an overall goal, but only by imputation from the effect, and because people can act in ways that actually frustrate their own goals, the "goal" of a group that we would impute from the collective effect of their efforts may be very different from what is common about their individual goals. In the presence of consensus mechanism, it becomes possible for collective "goals' to form and to be coherent with the individual goals; it's an interaction, because the process shifts the individual goals and the collective goal into voluntary conformance and unity.
That's right, I'd say. However, this is what trustees are for. Trustees are charged with the original goals, they are chartered for that, it's their duty. Their job is to intervene when a group composition strays from original purpose, without actually forming a consensus for that change. This would be the theoretical role of the WMF, but it doesn't have the expertise, it may not even recognize the problem. That might change, I can't predict if it will or not. You never know until you try, for sure.
I agree, unless there are mechanisms in place which interdict this in order to maintain the original goals. But the core original group was phenomenally naive about this, they imagined that restraint would arise naturally, from the sheer beauty of the original goals. The problem is not anonymous editing, per se, the problem is lack of structure that can handle this when the scale becomes large. "Structure" is not a synonym for "central control," though the most commonly-known structures involve such. To get down to basics, the human nervous system is a collection of cells, each operating according to its own nature, and it is not centrally controlled. In a healthy individual, it operates through a kind of consensus, with strong and dangerous decisions only being made with high consensus, immediately necessary decisions being made by majority rule with lots of caution and room for revision, and sometimes experimenting with minority suggestions where it seems safe enough to try them out.

Trustees cannot be anonymous, probably, because they should be personally responsible. Trustees hold the keys. In FA/DP theory, there will be, for any substructure, a trustee who controls a particular "meeting." Should there be trustee failure, where a trustee abuses the power, the recourse in FAs is generally not to contest the trustee, but to bypass him or her, to start another "meeting."

Large nonprofits form in FAs, or even for-profit corporations, but the two, the Free Association and the corporations, remain quite distinct and independent. They cannot control each other, but it is in their common interest to cooperate. The FA is broad and diffuse and does not own property; but, collectively, its resources dwarf every other player. If it can find consensus, it is practically unrestrained. The problem then reduces to one of how to find consensus in large FAs.

Revolutionaries have often realized that the problem is awakening "the people," but historically, they didn't really want to awaken the people, because they imagined that the people are ignorant. Rather, they wanted the people to follow them as "vanguards of the proletariat," but they did not want to allow the people the power to restrain them, so they never set up mechanisms whereby true consensus could be found among the people; the result, in the end, because one repressive oligarchy was merely being repressed by another, supposedly better because of its noble intentions, as self-conceived, was simply a change of faces. The old oligarchy unseated was also under the belief that what was best for it was also best for the people, that they were the most qualified to lead.

Hence FA/DP theory suggests that unseating the oligarchy should not be the goal. The goal should be the development of scalable consensus-finding mechanisms, and the application of these to real situations. Practically any situation will do! Applying this to Wikipedia sets up cognitive dissonance, if done properly. People will react to cognitive dissonance in many different ways, but over time, positions will shift. The dissonance is that the oligarchy believes that it is serving consensus, or that consensus supports it. Yet, functionally, it acts to oppose the development of broad consensus, out of its fear of loss of control to the "great unwashed," what it sees as the naive mass of editors. Does anyone really think that most editors are deletionist? Has the WMF ever done any decent polling to determine what the bulk of editors, by various definitions, think about deleting articles about minor topics?

(Possible polling technique: restrict the poll to registered editors with email enabled, simply because deeper polling would be intrusive and very difficult. To go deeper would require polling of the general public, much more expensive. Provide the sampled editors with examples of AfDs and ask for opinions about the results. The questions would be looking for absolute opinions, they would not be asking the editors to determine if the results were in conformance to guidelines, because that's circular, in fact, guidelines simply representing the consensus of the core, when they express their ideals. The so-called "non-negotiable policies," in their simplest and most-publicised form, though, might be stated. Then the results would be analyzed from various perspectives. It would be interesting to weight the results by edit count. The goal is information, not deletion decisions! What does the community really think? Sometimes the broad community will be naive, it doesn't have sufficient experience to understand the range of problems encountered, the reasons for the operating practices. But in that case, the problem is to educate the broad community, to bring the actual practices and the opinions of the broad community into alignment, so that there is unity of purpose between the "vanguard" and the "proletariat.")
We don't know what we would get, not in detail. I'm sure there would be better articles, how much better, and how broad participation would be in outside structures, I don't know. The easier it is made, the easier it is to find the structures, the easier it is to participate in them, to find congenial places, the more effective it will be. TANSTAAFL. It can all be done badly, but external structures, done badly, will fail. The site won't depend on any single one of them, many could fade away, disappear, with no enduring effect and very little loss, if any. What this kind of external structure involves is simply people discussing what interests them; only at what I call a "top level" does anything else appear, and for a very restricted number of people.

Oligarchies and hierarchies will still exist. There will still be cabals that attempt to impose an agenda (specific or more subtle, as with the present administrative cabal). However, cabal power will be reduced, not increased, because there will begin to be what might be called the "community cabal." The problem with present political systems, most of them, is that only specific interest groups are represented, and larger interest groups, such as political parties, tend to, for their own power, form fixed positions with respect to specific interest groups, so as to win their support. There is no "party of the people," only parties which may use slogans like that for recruitment. A party of the people would be neutral! Or it is just one more special interest group in disguise, typically one interested in its own power.

A true "consensus party" would be maximally inclusive, and non-destructive, I predict. It would see social problems as being the result of lack of attention, not the "bad guys." Sure, there are bad guys, but it is defective systems that give them to power to do damage, and that, in fact, encourage the "badness" by rewarding it. The "consensus party" would not run candidates, the very concept is contrary to the principles. (I think Demoex, the Swedish party that is the only known attempt to apply delegable proxy in a political context, made a mistake by running a candidate, and they have been far less effective than they could have been, as a result, because they defined themselves as competitors with existing, and widely-supported, politicians, and they elected a rubber-stamp candidate, which is a direct attack on the foundations of deliberative democracy. Doomed, I'd say, unless they shift their approach.)

It would simply make sure that information and trustworthy analysis was available, and it need not fix on one kind of analysis, its open structure would allow for differing analyses to coexist. The Wikipedia problem is a generic one, and its solution would benefit society as a whole. That's what makes it all worthwhile to me. It's an example. An experiment, if you will, though everything I did was designed to be consistent with Wikipedia purposes, i.e. a neutral encyclopedia of "all human knowledge." I was indeed using Wikipedia for my own purposes, but those purposes were -- at least I believed that they were! -- fully consistent with the stated purpose of the project, intended to better implement it.
I'd say that, paradoxically, both you and they are correct. The solutions will involve partial opening up and partial lockdown. Flagged revisions, if more generally applied, is a partial lockdown. The problem then reduces to how flag privileges are determined, and better definition of standards. Clear standards reduce disruption, and they do not prevent IAR transcendence of rules, but they protect editors who follow the standards, and the bold ones who follow IAR need only be able to justify the effort as good-faith (and be responsive to rejection of the IAR claim.)

I edited Cold fusion during the one-month "community ban" that cabal pile-in at AN/I was able to accomplish (and which i accepted to minimize disruption). Because I was banned, I reverted myself, with "self-revert per ban." I was still blocked by WMC as a result. The edit itself was purely technical, attempting to fix a broken reference. It was the kind of edit that, when ScienceApologist made them while topic-banned by ArbComm, were clearly accepted by the community, and WMC had stated that to block someone for such an edit was "stupid." My edit was an IAR edit, clearly. So, when i was blocked, I did not even put up an unblock template, it was only a 24-hour block, anyway, and when I went to RfAr over WMC's insistence that, after the 30-day community ban expired, he could still continue to ban me, I did not complain about his block of me. However, in typical fashion, ArbComm assumed and reacted as if that were my complaint! My original filing was concise and clear, and everything only became confused when the cabal piled in, pouring in "evidence" and argument in the Workshop, and I began to respond to that. My point is that if I violated a rule (Later expressed as "A ban is a ban, it means "all" edits, not just bad ones") it was incumbent on me to understand that I might be sanctioned temporarily, pending review.

However, I also argued, both before I was personally involved, when the one being examined was ScienceApologist's, that self-reverted edits would ordinarily not violate any ban at all, because they created no disruption, no mess to clean up. They could be ignored; but, in fact, a self-reverted edit is the fastest and most efficient way for a banned editor to make a suggestion, and disallowing them, I guarantee, prevents improvement of the project, closes and keeps closed a means whereby banned editors can have a gradual path to reintegration, because self-reverted edits are useless to the banned editor if they are ignored. That there could be exceptions, i.e., self-reverted edits that are disruptive in intent, doesn't change this, and because of a ban in place, an editor can easily be sanctioned for actual disruption. Self-reversion creates a situation where ban enforcement is actually useful instead of being a pure waste of time. An admin looking to see if an editor has edited may see a self-reverted edit, and it's practically one click to accept it and bring it back in.

When PJHaseldine was topic-banned, from the topic where is may be one of the world's foremost experts, it was shortly after the SA flap over his pushing the edges of his ban (his intention had been actual disruption, and he was looking for edges to challenge, whereas in editing the Cold fusion article, I was just trying to fix a damn link! I invented self-reversion as a technique for him to use if his intention were really to help and not to disrupt, and I cleared it with Carcharoth before proposing it. He rejected it, and it's obvious why. It involves acknowledging and cooperating with a ban, instead of defying it, and defiance was his intention.) So I suggested self-reversion to him. He took it up, and the result was cooperation between him and the editor who had sought for him to be banned! It worked. And when I was then blocked, later, for a self-reverted edit, that editor brought my block to AN as a question. Was this okay? And the cabal piled in, with other strict-rules editors, and said, no, not okay, a ban is a ban is a ban why is there even any question? Brilliant, guys....

The overall point? Mature systems will have a kind of change that is a "proposed change." It won't be implemented until it is seconded. It won't be in the article as viewed by default, but it will be visbile to anyone who wants to see such proposals. It is possible that the "seconding" privilege will be restricted.

Citizendium, by the way, has a reputation of requiring editors to be recognized or qualified experts. Not. Anyone can write a Citizendium article, but it's classified as a proposed article, and only becomes a full article, whatever they call it, after being vetted by an expert who signs off on it.

Structure. The flat Wikipedia structure, it's in the encyclopedia or it isn't, doesn't work. The existence of History helps, but there is no organization of history. It's practically impenetrable, with articles that have seen serious controversy. Carcharoth noted that Cold fusion could use a FAQ that would explain why the article is the way it is. However, the only editor actually interested in creating such a FAQ, which would be a consensus document, necessarily, or it wouldn't work, was me. And they topic-banned me. For reasons that weren't made clear at all, the evidence cited was contradictory to what ArbComm has elsewhere made clear.

Hmmph.... I can now join the crowd of editors who have experienced similar responses. I like that.

Abd, you seem a lovely chap - so I wanted to echo (with a smile!) Greg's advice above - I'm a pretty invested sort of wiki observer, prepared to trawl diff.s and read talk pages - but your posts tend to defeat me around the 2nd or 3rd para - I just can't read them!

I think greg's idea of putting thoughts on a blog and flicking a link here might make them a lot more readable.

You aren't bothered being lashed? Sorry. I only do lashing with consenting adults, who want to be bothered and like it.

Seriously, of course you are right. But I'm not writing for everyone, and this isn't polemic. I can write polemic, I'm actually good at it, but I have to be highly motivated, it's a lot of work. I write a post here, takes an hour. To boil it down might take another hour. Will this increase the effectiveness of what I write? Depends. If I have a specific goal, perhaps to convince some set of people, the writing is then polemic, not discussion. Discussion is not for everyone. And part of what happens here is what I'd call banter, not discussion. Banter should be exactly what you describe, Greg, and polemic usually is similar, i.e., shorter is better. With discussion, shorter is not necessarily better.

Thanks for the suggestion about akahele.org. Eventually, I'll do that, I expect. Right now, I'm mostly experimenting with little initiatives, plus I have tons of Other Stuff to deal with, like setting up the conditions for low-energy nuclear reactions in my kitchen, and, as well, instrumentation to show that, indeed, that's what's happening. And doing that on a shoestring, to boot.

As to efficiency, it's obviously more efficient if I could present you with prose -- or poetry would be even better -- that would explain what I have to say in words such that everyone would immediately get it. That is, more efficient for you. It might be impossible for me, and approaching it so incredibly inefficient that, quite simply, I don't have the time for it, period. What I want to do requires a core who understand it, and a core begins with one person, not many, I know of no historical examples where the early core was massive and where the vision was also true. It always starts, from what I know, with two people. Often the second of the two is much better at explanation, but I do know of an historical exception to that.

Let me make this prediction. After perhaps a year from first contact with these concepts, a Wikipedia editor will decide to start asking questions instead of relying upon me to come forth with some condensed, FA/DP for the Compleat Idiot, piece of polemic. The question will be answered, and will become a FAQ that will then serve as a design guide for implementations. Perhaps by that time there will be some demonstrations in place, perhaps not. There are a few people who have sufficient time in contact, so this might happen at any time, but most of these people, that I can think of, have retired from Wikipedia out of frustration or whatever, or are active at only low levels.

I should actually be writing much less, not in terms of shorter posts, but in terms of overall time I spend. I might even disappear, for the most part. Wikipedia is not, properly, the center of my life, and I already spent way more than I should have working on the one project, and other projects suffered.

This post, while quite long, was worth reading. It was very informative, and it proposed some solutions which could be quite useful. Implementing the ostensible goals of the project in regards to its "consensus structure" is just not possible when done in the framework of Wikipedia-style discussion. The "goals" of the project seems good in theory, but Wikipedia needs to change the way it operates socially. The majority of editors are not unified under those goals, and they need to focus on consensus-made decisions through informal association of editors. There are large social problems in the structure of Wikipedia, and open debate is needed on how to resolve those issues. Even more significantly, however, is the way Wikipedia relates to people who have nothing to do with them, and their consideration for people, instead of their project. Many members of Wikipedia do not realize that the world doesn't revolve around them.

That aside, the social organization of Wikipedia does not reflect an accurate method of attaining consensus on the way they address issues. Vetted contributors are trusted more so than others, which in itself is not a bad thing, but Wikipedia has no system of accountability for the implementation of decisions. Nor do they possess any type of trustee system, like the one that Abd explained above, to ensure that the project is going in the right direction. The consensus system is not flawed, the way that Wikipedia sees "consensus" is. Solutions should be debated, but what everyone is aware of is that a problem is present, and that it needs to be fixed, even if that takes completely destroying Wikipedia and starting over again.

Any presentation where Erik "the molester" Moeller is talking about a Big Hairy Audacious anything should be looked upon questionably.