Wikipedia is approaching a crisis. In medicine, especially before antibiotics, bacterial diseases often resulted in a crisis, a turning point, after which the patient either recovered, or died. A rising fever and worsening of symptoms typically indicated to physicians that the patient was about to enter the crisis, and might either recover or die.

The fever of Wikipedia is the disputes that are not resolved readily and disrupt the Wikipedia community. Wikipedia currently has too many disputes, both great and trivial, that are contributing to its fever, and that are not being resolved in a timely manner.

The underlying cause of the crisis is the increase in the number of users of Wikipedia. The potential for disputes increases in non-linear proportion to the number of users (approximately as the square of the number of users).

When a sick human enters a crisis, he or she will typically either recover or die. No metaphor is perfect. An on-line community does not die unless its hardware is disconnected, but it can be weakened or become unable to fulfill its original purpose. In the case of Wikipedia, that purpose is the cooperative development of an expansible copyleft encyclopedia. If it does not recover from the crisis, then it will never achieve the respect it is seeking, and will be marked by permanently having holes in its coverage due to unending edits wars and flame wars.

It can recover from the crisis only by strengthening its "immune system". This should involve changes in several areas:

• Most obviously, dispute resolution.
• The roles of admins.
• Policy formulation.

Wikipedia currently has approximately four processes for the resolution of disputes:

• Discussion on talk pages, including the use of article requests for comments to draw editors to those talk pages.
• Mediation.
• User conduct requests for comments .
• Arbitration.

I have ordered them in a way that reflects my view of their seriousness, and I think I disagree with the current ordering.

Discussion on talk pages is of course the method that should be used to resolve disputes whenever possible. It may be slow if it is difficult to identify what is POV and what is NPOV, or if newbie editors need time to understand the NPOV concept, which is not an easy concept for many humans. Discussion can be stalled or can break down if civility is forgotten and there are personal attacks, or if problematical editors are disrespectful of consensus (e.g., and see differences of opinion as censorship or evidence of cabals). Civility can be set aside and give way to personal attacks in at least two situations. A few problematical editors are inherently uncivil and hostile. Also, some mostly reasonable editors may, in stress or due to loss of patience, develop personal conflicts.

Mediation should be the usual way that personal or personality conflicts can be dealt with, as a possible way of defusing conflict. It is not working very well. The number of members of the MedCom has not increased as the number of editors and thus of disputes increase. Also, the effectiveness and neutrality of two mediators have been compromised by allegations of misuse of administrative power, discussed further below. It is obvious that Wikipedia needs more mediators in order to resolve disputes more quickly. It also needs more awareness of and attention to alternative mediation procedures besides the formal MedCom.

The expansion of the number of mediators, and possibly the number of mediation procedures, needs to be done quickly.

Also, the introductory statement on the Requests for mediation page is unwelcoming and discourages the use of mediation, which should be instead encouraged. It states: "Please read the information in the Wikipedia:Mediation, before asking for formal mediation. You must follow the preliminary steps laid out in Dispute resolution if you have not already done so. You may want to use the less formal (and more open) Request for comments instead." In emphasizing the "formality" of mediation, it suggests that user conduct RfCs, which are also a formal process, should be used instead.

User conduct RfCs, with a certification threshold, should not be seen as an "open" or "informal" process, or as an alternative to mediation. They should be used not in cases of personal conflicts, but only for problematical editors. (There is currently an edit war in progress about the language to this effect.) They should not be seen as a preliminary or alternative to mediation. They may be being used as such because mediation has too long a backlog, and sometimes a long backlog is not satisfactory (such as when a page has had to be protected due to revert wars). Some experienced Wikipedians think that the entire user conduct RfC process is non-functional. It certainly should not be seen as a step toward mediation, but as a more serious step than mediation.

There has been a proposal to increase the number of members of the ArbCom to 24 at the end of the year, and to have cases heard by smaller panels. I think that both of these measures are needed, and that both should be taken immediately, without waiting until the end of the year. The present ArbCom resembles a Supreme Court (in the United States or one of its states) that is the only court in existence, but in which all the judges vote on whether to hear any case, and then all the judges decide every case. I would strongly suggest that cases should be accepted more quickly, and should be heard by panels of either three or five arbitrators. This would mean that less total arbitrator time would be spent on hearing any given case, so that more cases could be adjudicated. (I am sure that some Wikipedians will disagree, and will say that the number of cases arbitrated should be kept to a minimum. I think that the detriment of disputes not resolved outweighs any detriment of resolving cases by arbitration when necessary.)

Improvement of the arbitration system needs to be done quickly.

With more than 500 administrators, there will sometimes be disputes involving their conduct. The number of such disputes could probably be minimized by a few reasonable guidelines as to what are considered normal, intermediate, and extraordinary actions by admins. Normal actions by admins include blocking an IP address for vandalism, blocking an IP address for 3RR, and blocking a user for 3RR. Intermediate actions should include blocking a user for vandalism or blocking a user for personal attacks. Simple vandalism is only rarely committed by signed-in users. Some editors are too quick to label the deletion of paragraphs as vandalism when it is more typically a revert war that should be dealt with by discussion. Similarly, personal attacks, unless extreme, can be dealt with in less drastic ways than by blocks. Blocking another admin, for any reason, even 3RR, should be considered an extraordinary action. It indicates that at least one admin has made a mistake that is a misuse of administrative power. Admins should know not to commit 3RR in the first place, and should know to do the arithmetic properly before treating a revert war as 3RR.

It should be clear that when an admin blocks another admin, they need to work out their issues quickly, or go to mediation, or go to arbitration (and blocking makes it difficult to work things out quickly).

There needs to be some way of removing administrative privileges more quickly than is presently possible. Streamlining arbitration will help. Other methods should be discussed.

Policy changes in Wikipedia are normally made by consensus. That is how it should be. There are times when consensus needs to be reached quickly rather than slowly, either in order to solve an immediate and pressing problem, or to discourage the recruitment of meat-puppets.

Upgrading the mediation and arbitration systems needs to be done quickly, whether by consensus or by top-down leadership (by Jimbo Wales or the Board of Trustees). If it is not done quickly, the crisis may break unfavorably, with a departure of a critical mass of constructive editors, leaving a weakened on-line community. Robert McClenon 19:38, 27 August 2005 (UTC)

This is an excellent analysis, and don't think I haven't forgotten your suggestion at the RFArb. After a regrettable multi-week delay, I have begun to turn my attention to revamping Mediation. Please continue to offer advice like this! Uncle Ed 11:18, August 28, 2005 (UTC)

I'd be interested to know what you think of the proposal I am working on (although it was certainly not my original idea) at User:Ryan Delaney/sandbox as an improvement for the arbitration process. If you have any thoughts, please feel free to leave them on the talk page there. In particular, I addressed this idea of increasing the number of arbitrators, and why I feel it would not be a sufficient solution to these problems. --Ryan Delaney ^talk 23:17, 8 November 2005 (UTC)

I think that another problem of wikipedia is that though there are wikiprojects no one sits down and evaluates the encyclopedia's weaknesses and organizes work loads for prject members to rectify them. --Ian 02:05, 14 February 2006 (UTC)