User:TeaDrinker/Heretical perspectives on (re)building Wikipedia's base of editors

Wikipedia’s demise has been predicted by someone at nearly every stage of the process. It is very much the project that no one expected to work, and when it started to work, the predictions changed to say it could never meet decent quality standards. When that proved to be incorrect, Wikipedia watchers predicted that it would never be taken seriously as a resource. So it is no surprise that a number of people are still predicting Wikipedia’s demise. However, this time, they may have a point.

Wikipedia’s base of volunteer editors has been declining since March, 2007. At the same time, Wikipedia has been growing rapidly; although the growth has been slowing somewhat, there is still substantial growth in the project. As such, a shrinking base of users is responsible for an expanding body of content. This is, by any account, cause for concern.

The goal, then, is to find ways to expand the base of active editors. Several approaches are discussed below, either focusing on editor retention or attracting new editors. Lacking good data on which one works, it seems like a reasonable strategy should be for Wikipedia to try out different strategies to see what works best. Some of the possible ideas are, of course, irreversible. These should be considered with more care.

The views expressed are my own, and are admittedly somewhat heretical in their leanings. They are included because my view is that they are all worth consideration, however there are some editors who disagree with each, and some are heavily discussed and rejected by the community. Nevertheless, they are included because they might still hold promise for the reasons discussed.

The problems
There are really only two issues relating to expanding the base of active editors. The first is editor retention, and the second is ensuring that new editors can seamlessly enter and share their work. These are more often stated to be in conflict than in fact. Both are keenly needed.

Retaining editors
There is going to be, as with anything, some loss of interest over time. However ensuring editors are not driven away has to be a major goal, if a large base of editors is going to be maintained. The largest problem, at least in the experience of this Wikipedian, is arguments and discussion which a volunteer can not reasonably walk away from. Most of us edit on a time-available basis. When Wikipedia has processes that demand time, this becomes stressful. While it may be impossible to entirely eliminate this, it is worthwhile thinking about how to limit it.

Major time demands often arise when multiple editors have (often reasonable) discussions about what goes into the article. Discussions of this sort are healthy when they are resolvable, but when the resolution is merely whomever has the most energy to expend, they cease to be about content or improving the encyclopedia. Wikipedia should focus on quick dispute resolution rather than universal adherence to process. Process is important, but not when it costs the project good editors.

Protracted, weeks long discussions turn into endurance contests in which the last editor standing gets the final say. These are particularly pernicious when one of the editors is not editing in good faith, such as editors paid by clients to never change their mind. Editors are over and over directed to "continue the discussion on the talk page" by experienced editors and administrators who would never dream of spending their time in such a fray. The default of argument by endurance is stressful and difficult for people who merely want to write articles. These protracted debates take Wikipedia from a "free-time" activity to a "must-complete" activity, and are detrimental to the most well-intentioned of volunteers.

Attracting new editors
New active editors are absolutely needed. Wikipedia needs new editors and new expertise to continue to function as an encyclopedia. New editors, however, face a daunting task of trying to become established in an environment which is fraught with challenges. New editors who merely want to help, and see a way they can, are routinely sent dozens of pages of guidelines and rules to read through. Then they are expected to edit very well on their first try, lest their work be deleted or reverted summarily.

They have to become familiar with Wikipedia markup language, the labyrinth of policies and guidelines, a really unfriendly comment and discussion system, and a culture which has its own unwritten rules. A new editor is thrust into an environment which is often driven by conflict and has its own prestige economy with no prestige and bereft of the tools needed to navigate those conflicts.

As democratic as Wikipedia is, established editors have a clear advantage over new editors in how they operate. New editors have difficulty even following the discussions on talk pages, much less are they knowledgeable of how to go about resolving conflicts. As Wikipedia has aged, conflicts have become more common. Attempts to soften the entrance have been nominal more than substantive—the equivalent of giving a high school student a sticker on a failing exam.

Lessening this barrier to entry while not compromising our editorial or encyclopedic integrity is absolutely vital.

Page deletion holding room
One of the most common ways for a new editor to enter the project is for that person to see something is missing. Usually, new editors want to build new pages for the project. This is fantastic, of course, from the perspective of building a base of new users. However it is sometimes difficult because people see Wikipedia as more of a collection of all true knowledge, not an encyclopedia of verifiable information. Thus many new users (quite correctly) have their pages removed rapidly from the project (along with a note saying their high school garage band “is not notable,” which only adds insult to injury.

There is nothing that can be done about the ultimate deletion of non-notable articles, at least if the integrity of the project is going to be maintained. However new page deletions happen rapidly and with almost no discussion, ensuring some good pages are deleted. They may be unsourced, they may be unformatted, but they are on a good subject and we have at least one (albeit new) editor who is interested in writing about it. Some guidance and we wind up with not only a reasonable article, but a new editor as well.

So the question becomes how to separate the good articles from the dodgy in a more careful manner. It seems reasonable to create a “new page holding area” where the article remains visible, but the editor can improve or work on it for a week. Instead of speedy deletions, most new pages could be sent to the holding area. It would be valuable to change the background color of the article or other signifier (more than just a banner, since those are so common they are ignored).

If this were an option, new pages could still be deleted outright (for example, copyright violations, vandalism, or attack pages). Questionably sourced pages, or pages with dubious notability, could be sent to the holding area. And articles which only have formatting or cleanup issues should be retained in the article space and tagged for cleanup.

This process is different from an article for deletion in that no person is necessarily arguing for deletion, only asking the editor to provide more information. It would be useful to also take steps to ensure the holding area does not become the “low quality” parallel Wikipedia, make it so that articles over a week old are automatically deleted unless someone has worked on them (or a trusted user such as an admin actively deletes it). If an article has been edited in the previous week by anyone, it sticks in the waiting area.

The point here is not to create more work for already overburdened volunteers, although it does create some new work. Instead, the point is to ensure that every new editor feels that they are treated fairly, without compromising the integrity of the project. The worst case scenario, which in my experience happens regularly, a bona fide expert tries to write an article on the project, only to have their expertly written (if poorly formatted and cited) article deleted. They conclude Wikipedia is not worth their time and give up.

Expert contact and review
Expert verification on Wikipedia is heresy, as has been demonstrated by numerous failed policy and guideline failures. Wikipedia welcomes experts, of course, but does not want to enter into arguments from authority (more on this below). As such, we do need to ensure that new experts are treated fairly when they enter the world of Wikipedia, just as we would like to do for any editor.

Providing a way for editors to contact their peers, Wikipedians who are similarly situated in research or academia, would be helpful to help new expert editors enter the somewhat mysterious and always complex world of Wikipedia. Of course, Wikipedia already has the “helpme” template; this suggestion is different. It is a confidential discussion, off Wikipedia, among people with (lightly) verified credentials.

Anyone with tenure on the project should have alarm bells ringing already. Off Wikipedia conversations is antithetical to openness, and verified credentials is elitist, conventional wisdom holds. However academia and scholarship do have those private routes of discussion as well. It encourages people to be open and discuss the issues they are having honestly. Furthermore, credentials exist for a reason. It is next to impossible for an anthropologist (for example) to distinguish the writings of a good physicist from a dodgy one without some credentialing. Certainly no anthropologist is going to trust a physicist to determine good anthropology from bad, provided it is cited and sourced. Nevertheless, there is a distinction, and Wikipedia should prefer the good over the bad.

In an ideal world, we could have anonymous experts in nearly every subject. However the fact is Wikipedia does not. Wikipedia, at its best, has botany experts giving advice to dueling social theorists, one of whom may be Randy in Boise. Wikipedia will never avoid such projects, but proper experts can be encouraged to do volunteer work if they can talk to their more-experienced peers about how to proceed, and they know that their hard-won credentials are valued.

Expert credentialing
Going a step further than the previous idea, it is perhaps worth revising the credentialing of editors. Once Wikipedia goes down the road of verifying the identity and credentials of editors (at their request), it is difficult to go back. Wikipedia’s view that editors should be judged not by their credentials but by their work is still a good one, most of the time. However it is worth noting that editors with credentials are disproportionately the sort of new editors Wikipedia would do well to encourage, and those potential editors worked on their credentials precisely because they see some value in them.

Editors who have advanced degrees, professional licenses, or other credentials would not be given the upper hand in discussions, except insofar as the discussants would be swayed by that fuller knowledge that one of the discussants is properly an expert in their field. Wikipedia should and must treat everyone fairly, but the ever-present Randy in Boise problem appears.

There are many problems, of course. Once credentials are verified, it is impossible to “unverify” them. In some instances, this will mean that good editors feel maligned because the “verified” editor has appeared and made pronouncements from the throne. This is a real risk, and the risk and benefits should be weighed carefully. However it is preferable to adopt this as a strategy rather than see an end to the project. Consideration of this approach should remain open.

Academic credit
Academia, in many ways, runs on Curriculum Vitae (CV) lines. A CV is an academic resume, but is not a summary, it is a history. Whether or not work will “add a line” on the CV is often the determining factor on whether that work gets done. Working on Wikipedia currently can not be easily listed on a CV.

Part of this problem is cultural. Wikipedia is seen as popular or non-serious scholarship (after all, much of it is written by non-experts), so academics are not inclined to give work on Wikipedia serious credence. The other problem is technical—it is difficult to both quantify and verify claims of Wikipedia authorship. However the rewards for facilitating Wikipedia work being academically valued are sizable. A large body of expert editors would become highly motivated to spend many hours improving the project.

There are two things that should be addressed, if this becomes a possibility. The first is the risk that this undermines the essence or integrity of the project. The second is how to accomplish this.

The risks, superficially, seem high. Wikipedia largely frowns upon paid editors because of the dual loyalties that pay induces. No longer do editors put the encyclopedia first, instead they put their pay first (or at the very least, allow it to influence their judgment). Superficially, writing for academic credit might seem similar. However there is a difference in that Wikipedia does not have an expected or preferred outcome when it comes to content. Academics are quite satisfied as long as the work is done, no matter the opinion expressed.

The second risk is that verification of identity may open a Pandora’s box of problems with regard to credentials and such. However some well known editors already use their full name as their user name, and it does not cause significant problems. The necessary identity verification process only allows some users to maintain their anonymity while editing, but still include the line on their CV.

The problem of implementation is three-fold. The first and largest obstacle is encouraging academics to accept Wikipedia work as a line on the CV. This can be done with concerted effort and essays in academic society journals and other media. A concerted effort directed in this area might sway some otherwise discouraged academics to visit and try adding some content.

The second problem of implementation is technical. Wikipedia needs to be able to verify and quantify the contributions of editors (on request) for inclusion in the CV. Most people reviewing a CV will have no idea how to value edit count (and that is not a great measure of contribution anyway). A metric of contribution will very much determine the sort of articles and edits which are generated, so this is a critical step. The verification step is important, but is more easily accomplished as a user-requested verification email. For instance, a person verifies the identity of an academic, but their name is kept hidden. If the Wikipedia user requests it, the identity information can be sent via email to any address they specify, but it is not available for public request.

The third problem, which is non-trivial, is acclimating academics to writing an encyclopedia rather than journal articles. Academics are more likely than most to own articles, more than likely, since they want to ensure their ideas are prominent. Nevertheless, they are also conscientious and capable of reasoned discussion, and informed discussion will undoubtedly be valuable to the project. Some academic editors, like any other group of editors, will be unable to differentiate their own opinions from what would make a better Wikipedia article. However this is a small price for a great benefit.

Improved comment systems
A substantial barrier to new editors is the rather unfamiliar system of comments used on Wikipedia. Experienced users can add comments and read comments with ease, however as the comment system is unlike most other forms of electronic discussion boards, it presents some difficulty for new editors. As such, it may be worth trying to update and modernize the comment system into something more like what people are accustomed to reading. This can follow the format of many discussion board posts, which allow the quotation of previous editors and have clear delineation of authorship (and time stamps, etc.). While there is some value to ensuring editors don't need to learn two systems of code, it is more valuable that the comments be understandable and comprehensible to everyone, especially new editors.

One approach would be to model the comments after many bulletin board posts. This would require a change in software, and is obviously not backwards compatible. But it might help ensure that new Wikipedians can participate in discussions in a meaningful way, rather than being relegated to the sidelines while the more experienced users hash out their ideas.

Another approach might be to incorporate a link on every article which allows inexperienced users to leave comments or even just flag the article in some way with a drop down menu ; these reports could go into either a separate per-article ticket system, or just get stuck onto the article talk page. In the process of submitting the comments, it would remind users that they are welcome to make the bold changes themselves, and possibly with a reminder about seeking talk page consensus for bold changes which are likely controversial. In many ways, this is just an extension of the Article Feedback Tool to make the transition from "there's something amiss in this article" to "new editor actively fixing the article" a bit easier.

Concluding thoughts
Nothing in this essay is guaranteed to work, nor is it a complete list of even the author's ideas on this. It is merely a list of suggestions for possibilities to try. There are many other people with many other ideas, some of them would seem entirely repugnant to me as well. Simple ideas which can be easily reversed should be tested out rapidly and with careful monitoring. Ideas which are more challenging to reverse should be tried out with greater caution, however they should be considered. Wikipedia faces an existential problem, in a way that was predicted by many but does now appear to be nigh.

Building a base of active editors must be a major goal for the long term survival of the project. It is not certain if any of these ideas will work, but it is certain that doing nothing will not work.