Openness versus quality: why we're doing it wrong, and how to fix it
The views expressed in this opinion piece are those of the author only. Responses and critical commentary are invited in the comments section. The Signpost welcomes proposals for op-eds. If you have one in mind, please leave a message at the opinion desk.
Wikipedia's motto, from its very inception in 2001, has been "The free encyclopedia that anyone can edit". This emphasizes the openness of the project, which stands today as the greatest example of crowdsourcing on the Internet. With 3.8 million articles, the English Wikipedia alone stands as the largest encyclopedia ever. Yet despite our success, trouble is looming on the horizon. Wikipedia's model, though highly successful thus far, creates an intrinsic conflict between openness, allowing the greatest number of people to edit, and quality, aspiring to clear language and the highest standards of accuracy. Which is more important? In making Wikipedia more open, you risk ending up with poor information, poor writing, and rampant vandalism – turning the project into a big joke. In becoming more restrictive, you gain respect and accuracy but risk alienating users through complex policies, guidelines, and policy creep, which inevitably leads to editor fall-off.
The difference between the current number of editors and the number sought by the Wikimedia Foundation
I penned an essay on my experiences with Wikipedia back in December 2009. There I reflected on my early fascination with the project, my experiences with its mechanism, and how editors fit into the overall model of the project. At the time I was optimistic about the future of this project, saying, for instance, that "time in itself solves all problems".
Now, two years later, I'm not so sure. Editor retention is a worrisome topic, and one that has been at the center of the Wikimedia Foundation's Strategy Initiative since 2009. The difference between the Foundation's numerical goals through 2015 and what has actually happened is quite stark, and earlier this month the Foundation's Director, Sue Gardner, brought the issue firmly into the limelight again in her presentation at Wikimedia UK (see the video).
Let's take this statement as an axiom: Discussions on Wikipedia naturally lean towards stricter standards.
There are several reasons for this assertion. First is the growth of Wikipedia itself. In earlier times, editors were more concerned with plugging content holes and filling out red links than with specific, focused, well-cited, quality writing. Instances in which such quality was achieved were cataloged in BrilliantProse (note the name), an early version of today's Featured articles. As Wikipedia evolved, there were fewer content holes to fill, and editors began intensively improving articles. Processed by growth and parsed through instruction creep, BrilliantProse eventually became the featured articles we know today, in which high-quality prose is only one of ten criteria.
Second, we're self-conscious about how Wikipedia is perceived by the wider world. Regular editors spend many an hour laboring at prominent articles read by thousands of people every day, but find that outside Wikipedia their work is viewed as unusual. Some are even ridiculed by their peers, who perceive Wikipedia to be unreliable and poorly written. What are these editors to do? They return to their desks vowing to do better, and become protective of article quality. In discussions, some of these editors express the view that one good article is better than several rather poor ones. After all, we humans are social creatures; we seek to improve ourselves by improving our standing, and the standing of our work, in the view of other people. By writing better we hope to improve the public, outside perception of our work.
Third, it's very easy to increase standards incrementally because the repercussions for doing so take a while to appear. If you increase the quality requirements for a particular process, drama fails to occur. The standard becomes a little harder to reach, but the process generates better results. Examine any process following a major standards discussion and you'll see that the numbers will shift little in the short term. Short-term thinking, driven by idealization and natural growth, is at play here, as everyone accepts the current standards and interprets a bump in the upwards direction as a strictly positive thing. New editors have neither the credence nor the awareness to contribute to these kinds of discussions, thus involved parties tend to be veteran editors already familiar and comfortable with the standard.
Now let's look at the graph above, which shows that the active editing population hasn't grown, but has slowly dropped since 2007! Even more telling of this decline is the lack of activity in one of the most vital and most fluctuation-prone areas of Wikipedia—RfA. In November 2011 only two promotions took place; compare this with November 2006, for instance, when 33 candidates were promoted.
It's easy to miss, but every bump in quality we make is damaging to the new editor population. I like to think of Wikipedia as a tree, and editing as a ladder. You climb the ladder to reach different fruits ("articles"). Each time you make a process a tiny bit harder, you move the fruit a step higher up and add one more step to that ladder. That may not seem like such a big deal, but if you apply quality creep and repeat this process 10 times, you get one very daunting ladder—a ladder of such height that a new editor might say not even bother to climb it. The "low-hanging fruit" that should, in theory, account for this do not exist. The lowest prestigious piece of work a writer can achieve is a good article, and that too is daunting for a new editor inexperienced with Wikipedia's style and formatting. This, in a nutshell, is why Wikipedia is looking ill.
The Wikimedia Foundation, however, has been targeting usability as the core of our troubles. To this effect, they've redesigned the editing layout, softened the Wikipedia logo, introduced WikiLove, and done a host of other things to make Wikipedia a more comfortable place. While I agree that the Wikipedia interface should be friendlier, in key ways it's in a far better state today (I never liked Monobook and a new WYSIWYG interface would be much easier to use), in my mind the initiative has missed the point; thus far the Foundation has failed to address why people have been leaving, only giving them a more comfortable place to sit in during their stay.
But how do we fix it?
So you've read my rant and found yourself nodding at every sentence. Or you're completely opposed to everything I said and are already formulating an equally long rant on the talk page about why I'm completely and utterly wrong. Regardless of your take, we need to think about how we can kick Wikipedia back into the era of good feeling that my magical "five years ago" represents? Well, here's the kicker; we don't.
We've had a lot of time to develop and mature, and you may notice that in discussing why editor retention is falling, I never once explained why it is a bad thing—it really isn't. Now don't get me wrong, a growing population of active editors is always a good thing; but there's a limit to how far we can go, and in my mind, we've already passed it. The reason Wikipedia had such a rapidly expanding population in the first place was because we had so many content holes, we needed every hand available to keep our leaky boat afloat, and once our basis was established, there were plenty of idle hands eager to help. With the HMS Wikipedia now seaworthy, there is more quality content to write. With a limited pool of people with enough ability, interest, drive, and spare time to contribute such writing, a saturation limit develops, past which contributors are harder to find. And as standards continue climbing, this pool continues to shrink.
The Wikimedia Foundation should accept that there's a limit to how active a community can be, and that limit has been passed and distanced from. Openness and quality are a very real dichotomy, and one that has been around longer than the Foundation has, starting with the bitter split between Jimbo Wales and Larry Sanger. Although we can try various gimmicks to increase our credence among potential contributors, nothing short of creating a culture of forgiveness would push contributions back to the rise; and although Sue Gardner has often stressed the importance of not biting the newcomers, in a public system that sees a good deal of vandalism alongside legitimate contributions, a hard line is needed to keep out the trolls. A Wikipedia in which poor edits are reported upon with a shower of encouragements is an unmaintainable system. If editors have the will needed to maintain a presence in the system, they would appreciate real feedback more than petty encouragements.
The simplest thing you can do to reverse editor loss is this: whenever you come across a discussion on increasing the standards for a particular process, remember what it could mean for new editors, and pitch in by suggesting what it could mean for potential editors in the future. This would serve to remind people about the possible long-term consequences of their actions, regardless of immediate justifications.
Want the latest Signpost delivered to your talk page