The sum of knowledge?

The largest and perhaps the most daring Web 2.0 project is the Wikipedia. This online encyc­lo­pedia is free and it is created through con­tri­bu­tions by its users. You or I write and submit articles on a subject of interest, and then other users, rather than offi­ci­ating editors, add to and correct those articles. With millions of articles already, and about 16,000 active users in any one month, people are clearly keen to add to this common pool of knowledge.

On the philo­sophy of the site, founder Jimmy Wales quotes the fraud­u­lent quiz winner who became an editor of the Encyclopedia Britannica, Charles Van Doren, who said: “Because the world is rad­ic­ally new, the ideal encyc­lo­pedia should be radical, too … It should stop being safe — in politics, in philo­sophy, in science.” The project depends on offering its users more trust than any tra­di­tional website would allow. Long-​​serving Wikipedian Angela Beesley says, “Wikipedia exists to provide a globally avail­able, free (as in freedom, as well as money), encyc­lo­pedic (veri­fi­able and unbiased) resource to everyone in their own language. I sub­scribe to this goal and I also enjoy working with people who share it with me.”

It started in 2001 as an adjunct to Wales’ previous project, an expert-​​written encyc­lo­paedia, Nupedia. Nupedia had stumbled at the first hurdle and only amounted to 21 articles after a year’s work. Then Wales and his editor Larry Sanger heard about wikis. Wikis are online content man­age­ment systems designed to allow col­lab­or­a­tion. They put together a wiki site and told the Nupedia mailing list about it. Wales hoped that the new site might allow him to gather a few articles for the Nupedia, but it quickly overtook the expert ref­er­ence. After one month of opening the new site to user con­tri­bu­tions, it had 600 articles. After one year, there were 20,000. In June 2006, it had more than 4,300,000 user-​​written articles in over 200 lan­guages, including more than 1,200,000 in the English-​​language version. Plans to produce a paper version of the German-​​language edition of the Wikipedia, which has a ‘mere’ 417,000 articles, are estim­ated to involve 100 volumes of 800 pages. By com­par­ison, the print version of the most com­pre­hensive edition of the Encyclopaedia Britannica com­prises just 26 volumes. According to Nielsen Netratings, Wikipedia is the seventh most popular news and inform­a­tion site in the world, way ahead of any other encyc­lo­paedia or printed news­paper site. In June 2006, tracking site Alexa (www.alexa.com), said Wikipedia was the six­teenth most visited site on the web among its users.

Running an encyc­lo­pedia “anyone can edit” is prone to dif­fi­culties. The person con­trib­uting an article and the people editing it may not be experts in a par­tic­ular subject. They may not be very good writers. They may, indeed, be mali­cious and aim to con­tribute libel­lous inform­a­tion or deface existing articles through their editing. In November 2004, former Encyclopedia Britannica editor Robert McHenry led the charge in an article entitled, ‘The Faith-​​Based Encyclopedia’. Observing a number of inac­curacies and internal incon­sist­en­cies in an article about Alexander Hamilton, McHenry attacked the ability of Wikipedians to maintain quality without qual­i­fied experts, peer review or guiding edit­orial prin­ciples. And, in any case, he noted, inac­curacies aside:

…the article is what might be expected of a high school student, and at that it would be a C paper at best. Yet this article has been “edited” over 150 times. Some of those edits con­sisted of van­dalism, and others were cleanups after­ward. But how many Wikipedian editors have read that article and not noticed what I saw on a cursory scan? How long does it take for an article to evolve into a “polished, present­able mas­ter­piece,” or even just into a usable workaday encyc­lo­pedia article?

In a further article, McHenry disputes the notion of the com­bining an encyc­lo­pedia with any kind of ‘wisdom of crowds model’:

One simple fact that must be accepted as the basis for any intel­lec­tual work is that truth – whatever defin­i­tion of that word you may sub­scribe to – is not demo­crat­ic­ally determ­ined. And another is that talent, whether for soccer or for expos­i­tion, is not equally dis­trib­uted across the population…

The cri­ti­cisms started to get worse. John Seigenthaler was extremely dis­tressed to learn that, “John Seigenthaler Sr. was the assistant to Attorney General Robert Kennedy in the early 1960’s. For a brief time, he was thought to have been directly involved in the Kennedy assas­sin­a­tions of both John, and his brother, Bobby. Nothing was ever proven.” He’d never heard such alleg­a­tions before and was extremely dis­tressed to hear about them, pub­lishing the story of his attempts to track down his traducer in USA Today. Other cri­ti­cisms of the poor quality of the writing in the Wikipedia and factual inac­curacies found in some entries made by Nick Carr, the author of Does IT Matter, led Jimmy Wales to note on the com­munity message board:

…the two examples he puts forward are, quite frankly, a horrific embar­rass­ment. [[Bill Gates]] and [[Jane Fonda]] are nearly unread­able crap.

Why? What can we do about it?

The site intro­duced a greater level of edit­orial control, locking some subjects from editing by any but well-​​established con­trib­utors. It has also insti­tuted greater controls on the quality of edits and con­tri­bu­tions. But the encyc­lo­pedia was to receive a robust defence from unusual quarters. In December 2005, UK sci­entific journal Nature pub­lished the results of a test of the accuracy of Wikipedia. They randomly chose articles on sci­entific topics, along with their cor­res­ponding entries in the Britannica. These were sent for review to acknow­ledged experts in the fields chosen: “The exercise revealed numerous errors in both encyc­lo­pae­dias, but among 42 entries tested, the dif­fer­ence in accuracy was not par­tic­u­larly great: the average science entry in Wikipedia con­tained around four inac­curacies; Britannica, about three.” The sub­heading for the ensuing article cham­pioned the online ref­er­ence, saying, “Wikipedia comes close to Britannica in terms of the accuracy of its science entries”. Subsequent reports about the study either hailed the Wikipedia as being just as good as its print ancestor, or cri­ti­cised the meth­od­o­logy of the research and Nature’s con­clu­sions. The Encyclopedia Britannica cer­tainly believes that the study’s methods and con­clu­sions are incor­rect. The debate still rages.

In some respects, though, the level of accuracy cur­rently attained by Wikipedia is irrel­evant. Its user figures show that it is good enough for most people to trust its results. Certainly good enough for many people to avoid paying a sub­scrip­tion fee to expert-​​edited encyc­lo­pedia sites. And the alleged lack of accuracy in the Wikipedia is some­thing that is being con­tinu­ously worked on. To choose a random example, the article about Sir Isaac Newton was edited more than 50 times in June 2006 alone. Maybe that fact seems worrying – more than 50 short­com­ings in a single article spotted in a single month – but others may see it as reas­suring. More than 50 people a month are combing through this article and refining and per­fecting and adding to it. The speed at which Wikipedia can be updated may be a weakness in some ways, but it is also cer­tainly a strength, says Wales: “There is a small scandal going on in Germany. One of the ques­tions on the German version of ‘Who Wants to be a Millionaire’ was wrong. The show had ref­er­enced an answer on the German version of Britannica, which was wrong. It was wrong on Wikipedia as well, but we were able to update it immediately.”

If you will forgive the broad state­ment, the Wikipedia is already the largest single source of know­ledge ever created, dis­counting aggregate sources such as lib­raries. Yet at the same time, nobody has been paid to write articles for the service. The authors and ad-​​hoc editors of articles do this without acknow­ledge­ment – only a user name on the history page of the article iden­ti­fies the author, and hundreds of people may have changed that author’s content since it was written. There isn’t even any advert­ising on Wikipedia – it is paid for by user donations.

The Wikipedia is clearly a very special and unusual project and quite dif­ferent to the average Web 2.0 startup. However, the point about what people are doing with the web remains. People are very clearly prepared to combine their know­ledge and work at making the web a better place for them­selves and others. Going through the Isaac Newton article for the 1000^th time to check the facts and add a little extra is quite serious work; work for which there is no reward except the personal sat­is­fac­tion derived from having con­trib­uted. The people who con­tribute to Wikipedia – and the hundreds of other user-​​led sites avail­able – are clearly not “surfing” the web. They are making it.

Update: The Atlantic Review has an even longer and more com­pre­hensive article outside of their paywall here.

Share this post:

Possibly related:

• Encyclopaedic Knowledge
• Seeking Answers
• Three Cheers for Twonks