Will the Stanton usability grant stop Wikipedia community atrophy?

The recent Stanton Foundation grant to improve MediaWiki’s usability hopefully will lower the barrier for computer novices to get started on Wikipedia editing. This comes at an opportune time: we recently learned that the size of the Wikipedia community has not only stopped growing exponentially, it actually has been gradually shrinking since early 2007. The most likely causes of the decline include:

  • lack of “low-hanging fruit”
  • lack of new potential editors who are just discovering Wikipedia
  • Wikipedia’s scope gradually narrowing to mirror that of traditional encyclopedias (a.k.a., deletionism run amok)
  • Wikipedia’s occasionally expert-unfriendly culture that turns off those with the most to contribute
  • a Wikipedia culture that gives little priority (or even respect) to activities focused on the community itself rather than the encyclopedia
  • the natural decline in participation of early community members; according to Meatball Wiki, users of any online community generally say GoodBye after between 6 months and 3 years unless that community is connected to their offline lives

Usability improvements, it is hoped, will open editing opportunities to people who are scared off by the intimidating and sometimes overwhelming markup that appears when one clicks “edit”.

Whether or not this will halt or reverse the decline in editing activity on English Wikipedia is tied up with several conflicting currents of thought in the community. As Liam Wyatt and Andrew Lih have been pointing out in recent Wikipedia Weekly podcasts (66 and 68 are both very astute discussions), the standards for what is and is not valuable content have been shifting consistently towards the convential encyclopedia definition of valid topics. Quirky lists, small organizations that don’t meet the ever-harsher notability standards, obscure books and concepts, anything ScienceApologist finds to be an illegitimate invocation of scientific authority, anything deemed too ‘mere news’, and, increasingly, simply anything that wouldn’t be found in tradional encyclopedias–these are candidates for deletion.

The implications of deletion trends for community health are not entirely straightforward. Overzealous deletion leaves a sour taste in the mouths of many editors who have spent a lot of time adding the kinds of content that now gets deleted regularly. Some leave because of it, or lose their enthusiasm. On the other hand, a lot of what gets deleted is simply weak, unsourced content; removing it the article pool means that new editors will not base their own contributions on such bad examples. Deleting content on the borderline of notability, or better yet, downright notable and significant topics, also replenishes the supply of low-hanging fruit. If someone thought a topic deserved an article, someone in the future may think the same thing and recreate it in better form. Citizendium recognized the advantage of redlinks early on, and decided to start from scratch rather than from a Wikipedia dump.

And while about two-thirds of those polled want to see Flagged Revisions implemented, the other third think it would be too much of a dilution of the “anyone can edit” ethos. Although I’m in favor of Flagged Revisions, it’s not clear to me whether it would improve or worsen the problem of commnity atrophy. It’s a question of balance: some people are drawn in by ‘instant edit gratification’, while others are turned off by the perceived free-for-all nature of Wikipedia and assume their contributions would simply be swept away in the chaos. So the lure of stability might or might not outweigh the immediate thrill of seeing one’s edits go live. (I suspect the waiting, and the tacit acknowledgement of good work when someone approves a newbie’s edit, would do more to draw in new users to the community than the instant, impersonal status quo.)

So how would improved usability shake things up? On the one hand, it might spark a wave of naive article creation followed immediately by a wave of deletion of new content produced by newbies with no grasp of the community’s standards. If someone can’t figure or won’t figure out how to use basic wiki markup (says the cynic), how can we expect them to use proper sourcing and adhere to Wikpedia’s core policies of NPOV and Verifiability? Lowering the barriers to entry might just exacerbate the us-versus-them mentality of deletionism. On the other hand, maybe a host of new users would integrate well with the community and restore some of its past vitality while pulling the philosophical center back a bit from the deletionist brink. (Of course, it’s an open question how much usability improvements could actually affect the influx of new users; the difference might be rather small, if lack of tech savvy is highly correlated with other factors that make people unlikely to edit.)

As Erik Zachte has pointed out (in the earlier version of this post), many Wikipedias are still growing; English Wikipedia is not the be-all, end-all. It is not clear whether each language will follow a similar pattern in the rise and peak of community (accounting for number of speakers, connectivity, and economic issues) or whether different languages can develop sufficiently different Wikipedia cultures to avoid the failings of English Wikipedia (or perhaps generate unique problems of their own).