Wikipedia:Village pump (proposals)/Proposal to require autoconfirmed status in order to create articles

The following discussion is closed. Please do not modify it. Subsequent comments should be made in a new section. A summary of the conclusions reached follows.

Executive summary: In this broadly attended discussion, more than two-thirds of those expressing a clear "support" or "oppose" opinion supported the proposal to limit article creation to autoconfirmed editors, either as a trial or on a permanent basis. Non-autoconfirmed editors would either need to submit new articles to Articles for Creation or create a userspace draft, preferably using an improved Article Wizard. However, because there is a vocal and passionate (albeit minority) opposition and there are empirical uncertainties about the exact impact of the proposed change, the preferred approach for proceeding is to conduct a trial of a definite duration, followed by the autoconfirmation requirement being turned off for a period of review, after which the community can decide for or against long-term implementation. The exact duration of the trial and review periods, and the preferred criteria for review, should be clarified prior to the trial commencing. A trial of 3-6 months followed by a review of 1-3 months seems to be the most common view in discussion thus far. In addition to these conclusions regarding the primary topic of the RfC, the discussion also showed consensus for making (unspecified) improvements to the Article Wizard and giving more attention to the Articles for Creation process, regardless of the ultimate outcome of the main proposal. Extended close discussion below.

This RfC received substantial community input, with over 500 editors participating. As with any large and well-attended discussion, there were comments about secondary issues and tangents, ambiguous statements, even contradictions. Nonetheless, there was considerable clear commentary for and against the main proposal of the RfC. One supporter of the proposal created a summary table that classifies the views of the participating editors. I conducted my own review of the discussion independently of this, but I believe the results are broadly similar. (Where I have noticed differences in spot comparisons, I was more conservative about not assigning a clear "support" or "oppose" value to ambiguous comments.) The overall result: more than two-thirds of the participants expressing support or opposition were in support. Depending on how some of the less clear comments are allocated, support might be as much as 70%, but I have chosen to avoid potential over-interpretation of such comments. Either way, the result is consistent with the Village Pump discussion that preceded the RfC, in which just under three-fourths of about 60 participants expressing a clear position supported the idea. (Most, but not all, of the participants from the earlier discussion followed through to the RfC.)

In addition to the discussion on the primary RfC page and the prior Village Pump discussion noted above, there was extensive talk page discussion and some discussions spun off onto subpages. The topic has also been raised at least twice at User talk:Jimbo Wales. I have also received comments regarding my prospective closing on my own user talk page and via email. While I have read all of this and it has informed my evaluation of the consensus, I will not attempt to summarize all of these numerous discussions. Some of the participants have already attempted to summarize the common arguments for and against the proposed change. I will not repeat all of those. What I will focus on from the primary discussion is two key points of disagreement, one philosophical and one empirical, and one point of agreement. I believe there are also some secondary points around which rough consensus can be inferred, which I will discuss at the end.

The philosophical disagreement is that many opposers see the proposed restriction as contrary to the "anyone can edit" principle, a core Wikipedia principle that is affirmed in multiple locations. Supporters argue that limiting article creation to autoconfirmed users leaves "anyone can edit" intact: anyone (even unregistered users) can still edit most pages, and those who wish to create new articles directly instead of through AFC will only need to achieve the relatively low hurdle of autoconfirmation. They also point out existing restrictions on editing that are accepted by the community: requiring account registration to create pages, and semi-protection of some pages. (A minority of the opposers – clearly not a consensus, even among the oppose positions – believes that these pre-existing restrictions are also contrary to "anyone can edit" and should be abolished. Similarly, a small and clearly non-consensus portion of the supporters called for even more restrictions than what was initially proposed.)

The key point of agreement, also philosophical, is that both sides want to encourage participation by quality contributors. One of the comments made to me was about the recent Wikimedia Foundation resolution on openness. I believe many participants on both sides would generally support the positions expressed in that resolution. But both sides view their own preferred approach as encouraging quality contributors and making Wikipedia more inviting to them, which is the key empirical disagreement. Supporters of the change believe that the current approach drives away new editors, because when they are allowed to create new articles without prior editing experience, the result is often for their articles to be deleted, which dispirits them and ends their involvement with Wikipedia. By channeling prospective article writers into editing established articles, they will become more familiar with Wikipedia norms before creating an article, and as a result will have a better overall experience. Some would-be article writers won't bother, but supporters believe these will mostly be the creators of vandalism and attack pages, who will not be missed. Opponents, on the other hand, believe the proposed restriction will simply drive potential editors away: upon realizing that they cannot immediately create a new article, most will simply give up on Wikipedia. Those that persist will primarily be dedicated vandals and POV-pushers. In effect, the two sides have mirror image expectations about the effect of the proposed change, and as suggested by Jimbo, the absence of empirical evidence prevents us from knowing which of these expectations is accurate.

The desire for better empirical evidence and sensitivity toward the strongly expressed concerns of the opposition – one of the more popular opposition views said this change would "kill Wikipedia" – led to discussions about conducting a trial. Although these discussions drew less participation than the primary proposal, discussion was robust enough that I believe some conclusions can be taken from it. Most who discussed the idea of a trial generally supported having one, with some supporters directly conditioning their support for the proposed change on the idea of it being a trial. Discussions about possible details of the trial were less conclusive, but a couple of common themes did emerge. Most crucially, the trial should be for a strictly defined time period, with a firm understanding that the feature will be deactivated at the end of the trial and not reactivated (if ever) until the results are reviewed and discussed by the community. The messy end of the recent pending changes trial was cited as an example of what to avoid. A trial period long enough to gather substantial data (at least three months) was the most common preference. The need to gather "before and after" data sets and have some prior notion of what to review were discussed, but no specifics were decided. Further planning is needed to refine these items.

Finally, some "secondary" items were discussed that had broad support. Almost everyone who commented on it seems to think that the Article Wizard can and should be improved. There were also repeated concerns about making sure that the Articles for Creation process gets more attention so it does not become clogged and proposed articles get the improvements they need. Participants on both sides of the discussion agreed on these points. There were also a number of ideas proposed that I cannot describe as having consensus, but which did have some support and may be worth discussing on their own merits. These included trying to make New Page Patrol less confrontational through better notifications and/or grace periods for new articles, a suggestion to limit new articles to titles redlinked from existing articles, and a suggestion for automatic creation of user sandboxes. Finally, there was noteworthy support for the view of User:Malleus Fatuorum, which cautioned against neglecting established editors in our efforts to improve the experience of new ones, a fitting point on which to end this somewhat TL;DR summation of an even longer discussion. --RL0919 (talk) 02:10, 27 May 2011 (UTC)[reply]