Will Wikipedia Fail in 2010?

As Ars Technica reports, law professor Eric Goldman has been remarkably consistent with his predictions. Rather than backtracking on hyperbolic claims like it seems many investment or stock "experts" do, Goldman has continued to affirm that Wikipedia will eventually fail. Back in December 2005, he predicted the failure "within five years". Same time next year... four years. The year after that... well, he didn't write about Wikipedia, but don't think he's forgotten (1). He's been giving touring lectures on the subject, and if his predictions are true, then that means the failure is due in a little more than a year (2).

Goldman's basic point is that the seeds of Wikipedia's destruction were sown internally--that is, it is the entire concept it was founded on that will eventually cause the site to collapse. The "free encyclopedia that everyone can edit" model upon which the site was based worked for a while, but now, as the site has grown ever-more popular, that model is becoming outmoded. Many are the documented publicized errors and flops of Wikipedia--from the kid who pretended to be a tenured professor (3), to someone editing a biography to assert the subject was involved in the JFK assassination (4), to erroneously declaring Edward Kennedy dead a few months before he actually did (5); there's enough to write articles about (6) and probably a decent-sized book to boot.

In recent months, Goldman has toned down the rhetoric and opted for a more scholarly approach. He asserts that Wikipedia can either try and remain freely editable, or clamp down on that freedom and go for a more vetted approach, but both have problems. For example, to remain editable and fight back the vandalism and hoaxes, the site needs to grow new editors to replace the old ones. This is a problem, Goldman asserts, because Wikipedia has become more "xenophobic" and closed off to newbies, putting off would-be editors and shrinking the available supply of new and willing writers or janitors (1).

On the other hand, moves towards tightening standards might have a similar effect. Long-talked about, long-discussed, long-misunderstood, a featured called "Flagged revisions" could soon be making its way to some of the more at-risk articles, making it so that an edit has to be reviewed by a trusted editor before it goes live and appears to the general populace. The fear is that this might put of longtime contributors who say "this isn't what I signed up for", and by the nature of the revisions also disenfranchises new editors (1). Damned if you do, damned if you don't.

Some of these guesses are misinformed, however. Take for example flagged revisions. While it does impose another layer of work before an edit appears, it may actually open up articles to editing, rather than closing them off (7). For example, the english Wikipedia's main page is forever locked from editing by anyone save administrators using "page protection", which can be applied to block everyone save admins or just to block anonymous or relatively new users. But flagged revisions would possibly allow these previously marginalized editors to once again post to the main page; their changes would just be more easily scrutinized. Similarly, flagged revisions could replace a large portion of page protections as a simpler and more open way of handling possible harmful or junk edits.

There are concerns that flagged revisions could still hurt Wikipedia, but that's unlikely. For one thing the Germans have successfully had flagged revisions for some time, with no ill results. Further, there have been moves away from openess that have not hurt Wikipedia. For example, the above-mentioned JFK episode involving the biography of John Seigenthaler resulted in preventing anonymous editors from creating pages. But Wikipedia continued to grow by leaps and bounds (7).

For Goldman, flagged protection (and its companion, "patrolled revisions", which allows experienced editors to 'sign off' on other's edits as good) are a step in the right direction, but Goldman suggests it's not enough. Not only are more technical measures needed, but changes in Wikisociety are too. Many editor's contributions are quickly and discouragingly reverted anyhow (7).

On that score, Goldman has a point. Xerox PARC recently completed a study of editors and reversions on Wikipedia. They found that there were currently between 650,000 and 810,000 active editors of the millions of registered editors, and that editor growth had plateaued. PARC concluded that there was a clique of high-contributing editors who were responsible for most of what readers see; the smaller chums had their edits reverted at a much higher rate (8). What this means requires a reading of the findings and a bit of inside knowledge. The prevailing wisdom is that not only are veterans generally hostile or ambivalent about newcomers, but there's no easy way to learn aside from by doing and getting knocked down. Many "survive" the initial beatings and learn by osmosis; many don't and leave (9).

But does this mean that Wikipedia will fail by 2010? First off, while editorship has stagnated (and article growth has decreased) that isn't necessarily a bad thing, especially if the problems of content are fixed (see for a primer "The Future of Wikipedia"). It would take a highly unlikely precipitous drop of editors and viewers before an objective outsider would say the website has failed. Could the concept prove a failure, though? That is more difficult to decide. Adding flagged revisions was a long and arduous journey, and it's unlikely that the revisions will cover anything more than articles about living persons, for now; the community has shown itself somewhat resistant to dramatic change (7).

The ultimate result depends on what you define as success. If having more than three million articles on tens of thousands of topics is your idea, then Wikipedia has succeeded. If it's about having an accurate, clean, and generally high-quality product, then Wikipedia continues to lag. What does seem likely, however, is that Goldman, despite sticking to his guns, is wrong in this case.

Read David Fuchs' previous Wikipedia stories: "Memoirs of a Wikipedian" / "The Future of Wikipedia" / "The Giano Problem" / "The Media's Attitude Towards Wikipedia" / "Wikipedia's Treatment of Fiction" / "Wikipedia and the Death of Truth".