By a Bunch of Nobodies: A Q&A With the Author of The Wikipedia Revolution

Andrew Lih

Recently, and to the embarrassment of some major publications, a university student in Ireland posted a fake quote on Wikipedia to see how many people would trust it as fact. Several newspapers, including the Guardian, and major blogs published the quote without checking its accuracy.
Critics of Wikipedia say instances like this point to the continuing danger of taking the “encyclopedia that anyone can edit” seriously as an information source. Others call Wikipedia nothing less than a revolution in information. (Our own judgments have been decidedly mixed.)
These debates notwithstanding, Wikipedia’s popularity continues to make standard encyclopedias look as hip as buggy whips.
Wikipedia editor/administrator Andrew Lih, author of the book Wikipedia Revolution, has agreed to answer our questions about Wikipedia and what it means to society.

Are instances like the above-mentioned quote hoax a sign that the media is becoming too reliant on Wikipedia as a source?

I once visited the massive Reuters operation in Bangalore, India, which was taking over many of the fact-checking functions of the London newsroom. I had a good chat with the researchers there, who when they found I was writing a book about Wikipedia, were enthusiastic to say they were fans and users of Wikipedia, but were explicitly told that Wikipedia is not allowed as a final authority for their fact-checking. And that’s entirely appropriate and right.
Wikipedia should be the starting point of research, not the ending point.
To the prospective journalist: there is no better place to start researching a story than Wikipedia, and probably no worse place to stop and use as a final word. In short, don’t do it. Wikipedia has helped you get your research started faster; don’t ruin your experience by using it incorrectly.
That said, Wikipedia is looking into new mechanisms to help articles increase in quality without backsliding by implementing a new “flagged revisions” feature that will help editors identify revisions of articles that have been checked for facts and quality. German Wikipedia has already implemented such a system, with English Wikipedia starting to move forward with similar measures.

You wrote much of the book from China, behind its Great Firewall that blocks all or part of many sites, including Wikipedia. How did you circumvent the firewall for your research, and is it easy to do?

For the three years I lived in China, Wikipedia was blocked from access most of the time but was completely unblocked right before the 2008 Olympics.
The blocking of overseas Internet sites in China is actually fairly trivial to circumvent. The basic virtual private network (VPN) software that most corporations and universities use is not prohibited, so it’s rather easy to leap over the censorship to access foreign sites. (There are also other encrypted tunneling methods and proxy servers such as Tor.) While someone with a fair amount of knowledge can leap the Great Firewall, the general user population usually doesn’t bother since most of their surfing happens within China; and domestically there are rather stringent content restrictions that are not technical, but policy-based and enforced by the web sites, news sites, and portals that operate there.
So I like to say: even though getting over the firewall is easy, you don’t need perfect censorship to have effective censorship.

During your time as an editor at Wikipedia, what was your biggest issue with it? What was your favorite topic/entry to work on?

Having been an editor since 2003 and seeing it grow from a small corner of the Internet to become a dominant force in the Google search results page, my biggest concern has been the recent decay in editor participation and the shifting standards for what exactly belongs in an encyclopedia. While Wikipedia has thousands of smart and dedicated editors, the numbers are clear: there has been a dropoff in the number of active editors and new people joining the editing ranks since 2007. This may simply be the case that the low-hanging fruit of “ready knowledge” has been largely picked, and the deeper knowledge requires more work and expertise. But this is an issue that the community needs to look into immediately.
My favorite entries have been ones where the unique emergent nature of the crowd shows itself. Covering breaking news simultaneously from multiple sources is where Wikipedia truly fulfills its revolutionary role as a cross between the news and the history books. Over the years, the articles about the Asian tsunami, London bombings, Virginia Tech shootings, and the Sichuan earthquake have showcased how distributed unfamiliar users can work collaboratively to distill breaking news reports into well-referenced, coherent narratives. Because it has become so valuable a resource on the Internet, Wikipedia is widely linked to and retains a high Google ranking. Contrary to popular belief, there is no evidence Wikipedia has been blessed with special status by Google to push it up the rankings. It’s simply because it is cited and referenced so much on the Internet and people value its Neutral Point of View policy to deliver good quality articles, even if they are sometimes less than perfect.

As Wikipedia becomes more popular, is it becoming less radical, and how is the quality/quantity of the information on it changing as a result?

Wikipedians have taken to the Spiderman/Peter Parker warning: with great power comes great responsibility. When it was a young, radical project on the fringes of the Internet, it was more risque and allowed for more curious creations. But because it is now firmly planted in the top 10 most visited web sites on the Internet, the community has stepped up to realize the responsibilities that come with that much global influence.
The biography of living persons policy (BLP) is a good example. Information about living subjects is given extra scrutiny and needs to meet a higher standard for inclusion because of the issue of libel, and the potential harm done to the subject of the article. This has to be seen as a generally good, responsible action. This prevents a Richard Jewell-type case, where an individual’s life was nearly destroyed when he was falsely accused of bombing the 1996 Atlanta Olympics.
But this has also led to some articles that appear to pull their punches. Professional journalists often do publish information knowing that things will do harm to someone’s reputation or livelihood, but determine the public interest is more important. This is an issue that the Wikipedia community seems to still be struggling with, especially concerning articles about rape victims/accusers (Katelyn Faber) or unsought celebrity status (Levi Johnston). To be fair, it’s not an easy balance to strike in the newsroom, or in the virtual collaborative space of Wikipedia.

How is Wikipedia making the money it needs to support itself?

The people at the Wikimedia Foundation have expressed that the recent economic downturn has been “challenging” for fund-raising; but even in this financial climate, they have been receiving foundation grants and starting new initiatives. Also, small donations from Wikipedia users averaging around $20 per person during fundraisers bring in millions for the foundation, which is impressive.
One curiosity is that Wikipedia was established in 2001, and flourished in the wake of the dot-bomb. This has led folks to muse that perhaps economic downturns may be good for free content and open-source projects.

Is Wikipedia like society in how it governs itself? Does it survive by some sort of social contract?

People often erroneously point to Wikipedia as a great experiment in democracy. This is actually a misconception, since the mantras within the user community are “Wikipedia is not a democracy” and “Voting is evil.” The community values consensus and discussion rather than trying to set up California-style referendums. Wikipedians typically resort to binding votes after the failure of other options.
The social contract of sorts is to “assume good faith” when making first contact with users and strangers. That has served the community well, given that it’s going on its ninth year of existence which is a rather long life in Internet years.

Have you seen any interesting trends in what subjects/ pages are being edited most?

By now, most users of Wikipedia have realized the pop culture subjects in Wikipedia are particularly active and strong, as are the ones about breaking news and politics. These will continue to be the leaders in Wikipedia, since they are the most current and bring the most passionate users.

A while ago, Essjay, one of Wikipedia’s most prominent editors, lied about his background. What, if anything, did this do to Wikipedia’s credibility?

A prominent Wikipedia editor nicknamed Essjay claimed to be a tenured academic theologian who had to stay anonymous to protect him from trouble with his school. He was exposed in the end to not have any of those credentials, also lying to The New Yorker magazine about his background.
In this case, what’s interesting is despite his deception, the tens of thousands of edits he made and the community decisions he oversaw were, by all accounts, legitimate and useful. Even with much forensic investigation by community members who were skeptical about whether his fraudulent identity translated into fraudulent edits, they found nothing of note that was considered malfeasance.
This is perhaps why the biggest identity fraud in Wikipedia’s history has not created much of a crisis in community. From the very beginning, to borrow a sports analogy, Wikipedians “played the ball and not the man.” Being generally anti-credentialist, users evaluated the merits of each edit and not the particular personality behind it. So whereas the common lament about Wikipedia is that university professors participating in editing are given no more implicit authority than another editor who might be in high school, this is an ironic example of where that actually was a good thing.

What affects Wikipedia’s credibility the most?

How many users consistently refer to it and use it each day. To traditionalists who say nothing good can come of something written by the masses in a chaotic manner, they turn their noses up and decry Wikipedia’s model on the grounds that the model cannot possibly work. To this, the community jokes: the problem with Wikipedia is that it works in practice but not in theory.
Empirically, millions of people each day find Wikipedia’s articles useful and on balance highly accurate even if they are sometimes imperfect. As an engineer, scientist, and journalist, that makes me really interested in digging into this phenomenon.

Is Wikipedia edited differently in other countries/languages?

The spectrum of languages for which Wikipedia is available is fascinating, and the project puts the beauty and quirks of these different groups under one electronic roof.
One intriguing matter is simply displaying the proper written script for a language. The Serbian Wikipedia requires two writing systems (Latin and Cyrillic) even though the article content is the same. Kazakh Wikipedia is even more demanding and requires three different writing systems: Latin, Cyrillic, and Arabic. Fortunately, all these problems are helped greatly by some pretty clever software written by Wikipedia volunteers that can map between different scripts.
There is some evidence that in other Wikipedias, not all the policies and values set by the early English and German Wikipedias have mapped over well to other editions. Wikipedia’s Neutral Point of View policy does well when there is both rigor (by having many edits), and diversity (in having different viewpoints). English is so widely spoken that you get plenty of diversity in news and source material on the Internet. For Russian and Japanese editing communities, where the language speakers are relatively ethnically and perhaps ideologically homogenous, it’s often been noted that articles about geopolitical and historical issues are not as well balanced as in English Wikipedia. For Arabic Wikipedia, there is not a lot of Arabic source material to use as references, so demanding the same citation standard for Arabic Wikipedia is impractical.
These are notoriously hard issues to analyze without some real expert knowledge about the communities, the language, and the subject matter. Hopefully this will be an area for further academic study, as these systemic problems may be quite difficult to resolve any time soon.