Saturday, August 26, 2006

African Languages Grow as a Presence on Wikipedia - New York Times

African Languages Grow as a Presence on Wikipedia - New York Times

By NOAM COHEN
Published: August 26, 2006

At the second annual Wikimania conference, held this year at Harvard Law School, there was what might be considered a quintessential Wikipedian moment: as Martin Benjamin, a researcher at Yale University, gave a talk about the Swahili dictionary he is creating online, Ndesanjo Macha was simultaneously sitting in the audience using a Wi-Fi connection and laptop to put the finishing touches on his Wikipedia entry, “Martin Benjamin,” in Swahili.

It was just the 1,025th article written for the Swahili version of Wikipedia (sw.wikipedia.org), the online, open-source encyclopedia founded five years ago by Jimmy Wales, and the fifth Mr. Macha had written that weekend in Cambridge, Mass.

In founding Wikipedia, Mr. Wales has said, he aimed to create “a free encyclopedia for every person on the planet in their own language,” a goal he has defined as having 250,000 entries in every language spoken by more than a million people.

But while larger Wikipedias, like those written in English (1,377,015 entries and counting) and French (348,243 entries), wrestle with questions of accuracy and vandalism, as well as the imposition of limits on who can create and edit entries, smaller Wikipedias face more basic questions: How do you create an online encyclopedia when few native speakers have access to the Internet? What use is an encyclopedia when literacy rates among a language’s speakers can approach zero? (This is not a problem for Swahili.) And who should control the content of an encyclopedia in a local language if not enough native speakers are moved, or able, to contribute?

If only native speakers of Swahili had contributed to that version, the Wikipedia might not exist at all. Though Mr. Macha, 36, who trained as a lawyer in Tanzania and is now director of the largest Boys and Girls Club in Greensboro, N.C., is a major source of entries, none of the other primary contributors grew up speaking the language: “One is German, one is in Texas, and one is in Canada,” said Mr. Macha, who was the only African — or African-American — to attend the entire three-day conference.

“They are all white, and to me it is very interesting — it shows that the world is not flat, that the world is still round,” he said. “We have allies, people who are willing to help us, but we need to be in charge of our own identity. When it comes to producing information, we don’t want to be dependent.”

Mr. Wales devoted part of his keynote address at the conference, which was held from Aug. 4 through Aug. 6, to the issue of African-language Wikipedias, of which there are now about 38, though many are mere shells waiting for articles to populate them. Swahili, which is thought to be spoken by as many as 100 million people, is the first African-language Wikipedia to have reached the 1,000-article mark — considered something of a tipping point toward faster growth in the Wikipedia world.

Far smaller is the version in Bambara, a language spoken by roughly three million people in Mali, in West Africa; as of the conference, there were more than 100 articles. Kasper Souren, 29, from the Netherlands, is responsible for most of them. Mr. Souren, who led one of the discussions at the Wiki conference, said he first came across Bambara as a volunteer in Mali in 2005. To get entries for the language’s Wikipedia, he said, he would introduce himself at a community center in the capital, Bamako, ask people to write articles in Microsoft Word and then pay “a buck an article.”

He then took the files they created and went back to an office and uploaded them onto the Bambara Wikipedia, being sure to credit the authors, even though they had never been on the Internet and thus could not have a Wikipedia user name. Mr. Souren learned enough Bambara to create the headers in that language; “nye” means article, for example.

Of the articles he posted, he said, “I can’t understand 100 percent of what they wrote, but I could estimate that it was right. It is a Wikipedia anyway, so I hope they can correct it.”

Most of the people who have heard about Mr. Souren’s decision to pay for entries lauded his goal but questioned his tactics, saying that they undercut the Wikipedia spirit, and that, ultimately, a Bambara Wikipedia would work only if there was a voluntary community to support it. Mr. Macha, who said he often felt that he was “speaking for a whole continent” at the conference, told Mr. Souren that “we don’t have to be paid to do it.”

Mr. Wales said that rather than pay for articles, the foundation that runs Wikipedia was likely to obtain a grant before the year’s end to sponsor a facilitator for an African-language Wikipedia, perhaps the Swahili one. “Writing encyclopedia entries sounds tedious, boring, but it actually is a very social activity,” he said. He said the facilitator would co-ordinate contributors among community leaders, bloggers, professors and graduate students and “jump-start” the growth of the encyclopedia.

To that end, among the most important connections made at the Wikimania conference was that between Mr. Macha and Dr. Benjamin, the Yale Africanist. Dr. Benjamin said that in the past Africanist professors have been reluctant to contribute to Wikipedias in African languages, either because academics look down on amateur contributors, are intimidated by the technology or find the process too time-consuming.

He said that sometimes entries have been mockingly shared via e-mail to point out errors, but that no one had offered to fix them. But there is talk of trying to change that attitude. He added that he thought that some of his students could begin contributing, too.

Mr. Wales contrasted the importance of this task with the relative success of Wikipedias in what he described as small European dialects, whether Luxembourgish, Cornish or Ladino.

“We are doing a good job in culture preservation,” he said. “We like that. We also think that is a secondary part of our mission. Almost no one who reads the Cornish Wikipedia reads only in Cornish. It is a little more critical or important in these other languages.”