Open Source Journalism: From Nupedia, to Wikipedia, to Citizendium

The evolution of collaborative efforts

Kevin Lim interviews Alex Halavais

Alex Halavais teaches Interactive Communications at Quinnipiac University, and he researches social media, particularly on blogging and wikis. He's particularly interested in open source journalism and scholarship. As such, he's done research on Wikipedia, as well as volunteered at the fringes of Nupedia and Wikipedia. Today he shares with us his views on the evolution from Nupedia, to Wikipedia, to Citizendium.

Kevin Lim: I understand that you did research on Wikipedia. Can you tell me about what you did?

A: A couple of years ago, I became interested in how good Wikipedia was at maintaining quality, so I changes some stuff--vandalized the site. Those changes were quickly detected and reverted, and although not a perfect test, it certainly affirmed my gut feeling that Wikipedia gets it right much of the time.

More recently, along with a collaborator, I've looked at a large sample of sites on Wikipedia to see how well it matches the coverage of other resources. Here, Wikipedia comes up a little bit short. When you have to rely on volunteers to write, those topics that tend to have larger fan bases also tend to be better covered. Unfortunately, some areas of expertise really don't develop the same sort of fan groups: the people who are interested tend to have a great deal of depth of knowledge, but there are not enough of them to draw on for that part of Wikipedia.

As it stands, Wikipedia is an outstanding resource by a number of measures. It's amazing that it has been constructed at all, and the kind of ad hoc volunteer structure that allowed for it is impressive alone. But even in comparison with some similar types of collections, Wikipedia often holds its own. It sits somewhere between the traditional encyclopedia and something a bit more ephemeral. As such, it is a bit of its own beast.

But much of the coverage of the "accuracy" of Wikipedia is, I think, overblown. Literate readers will take Wikipedia for what it is--a collaboratively edited document--and interpret it through that lens. The challenge now is to educate readers so that they are able to critically analyze what they find on Wikipedia, and indeed, the rest of the Web.

Q: What parts of Wikipedia do you spend the most time contributing to? (Is it a crowded or niche space?)

A: I tend to contribute when I find something wrong, and fix it, and I tend to do that anonymously. That is, if I am looking for information on some topic, and arrive at the page on Wikipedia and find that something is badly worded, or that there is some factual error, I will fix that. If I find that the information I need is missing, I will go back and fill it in once I discover it elsewhere.

So, I don't really have a strategy. The things I use Wikipedia for tend to do with topics of history or social science, and so that is where most of my edits end up being. With some rare exceptions, I don't write full articles.

Q: You also mentioned Nupedia. What is that about and do share a little about your involvement there.

A: Nupedia was, in some ways, the handmaiden of Wikipedia. I heard about the effort when I was doing my doctoral work at the University of Washington, and immediately volunteered. At the time, I was researching the idea of open-source journalism and I was an editor on the open source Web directory, DMOZ. It seemed like a natural direction. Since there was no editor for the Communications section of the encyclopedia, I volunteered to be the editor. I was rejected: after all, I didn't have a Ph.D. That stung a little, of course, but I kept with it as a peer-reviewer. Unfortunately, no article was ever published in Communications, and while there were some successful articles in other areas, I think Nupedia didn't get off the ground in part because it clung too tightly to the necessity of university credentialing.

Wikipeida was initially put together as a sandbox--a place where possibilities could be explored for articles in Nupedia. But it quickly took on a life of its own, and that momentum has continued to snowball over the years.

Q: Most netizens know about Wikipedia, and you've just explained what Nupedia is. What is your take on Citizendium and how is it different?

A: Citizendium attempts to fix the "authority problem" with Wikipedia. This is a real problem. Even faculty who are fans of Wikipedia note frustration with the "mob rule" that is sometimes felt on Wikipedia. It can be galling to see opinion of what is important or how to organize ideas thwarted by a small crowd.

I should note this isn't so much an issue of "facts." Facts really are important, but equally or more important is the order and structure in which those facts are presented, and the salience or importance of including particular facts.

Citizendium attempts to provide a structure for more strictly checking the accuracy and the structure of each article. It provides another layer of review; undertaken by "experts" who can best judge the quality of the article. For academic topics, that expertise is determined through the kinds of metrics that are usually found in the academic world.

The assumption--and I think it is a good assumption--is that traditional kinds of things like academic degrees are a decent way of estimating expertise in an area. While it may exclude some ardent and informed high school students and young people, it is likely that fewer people get through who are not in tune with the state of the art.

I worry about this approach for two reasons. First, a lot of people who look good on paper, and who hold positions at four-year universities, are not necessarily better encyclopedists than interested amateurs. Second, a far smaller proportion of tenure-track faculties are likely to volunteer their time to the project, I suspect. Finally, I'm curious as to how, exactly, decisions will be made regarding the expertise of any individual editor. These decisions ( e.g., tenure decisions) are hard enough to get right in the university setting. They are going to provide a lot of overhead for getting things off the ground, I suspect.

Q: You were quoted in the Chronicle of Higher Education in October of last year as saying that you didn't think Citizendium was going to work, that "it looks far too restrictive, and it costs too much effort to join and contribute." Now that it's firmly on its way, do you still feel that way?

A: I'm not ready to eat my words yet, although that quote really only got at a corner of what I thought on the matter. I still think Citizendium is going to face some of the difficulties that Nupedia did. It's a little like early aircraft design: there is a clear point at which a collaborative project takes off and gets off the ground. If it makes that point, then it will probably keep flying for a while.

Wikipedia took off by taking out a lot of the checks and balances that we expect to find in an encyclopedia. Citizendium is including those checks from the beginning, and I think that's an interesting model. I suspect that starting off loose and tightening up the ship as things move forward--much like the cyclical process of most programming projects these days--is probably a good model.

The truth is that Sanger has made clear he doesn't think of this as an elitist approach, and that there is a role for credentialed expertise in an open encyclopedia. Ironically, Wales seems to be approaching that view from the other side, recognizing the importance of stabilizing the content of Wikipedia. I think Wikipedia remains too loose, and Citizendium is still short on the momentum needed to launch it forward, but they are converging on a similar ideal.

I think it's good that Citizendium is there, and I think it is a very worthwhile endeavor. I just don't know that it's going to be able to gather enough excitement among its users to allow for the kind of mass of articles needed to make it appealing to a large user base.

Q: Finally, have you seen other forms of Wikipedia out there worth comparing Citizendium to?

A: Actually, I suspect that the best comparison is with Wikipedia itself. Over the last year, structures have emerged that allow for the checking of Wikipedia pages for inclusion in a stable version of the encyclopedia. The key difference is that there is no top-down approach to certifying the volunteers collaborating on these projects.

Perhaps Citizendium veers a bit closer to the closed wikis that corporations make use of on the intranets. Especially the requirement that real names be used, and the vetting process, feels like some of these approaches. In practice, many of these internal wikis have been quite successful, and so maybe Citizendium will be able to duplicate that success.