Please Support The Writing Show!

Jimmy Wales Talks Wikipedia


Jimmy Wales at Fosdem; cropped and touched up from Image:Jimbo-wales—fosdem-2005.jpg. Taken by Chrys; Rdsmith4 released his changes to the file into the public domain.

The above image is licensed under the Creative Commons Attribution ShareAlike 2.0 License.


January 2, 2006

Wikipedia, the free encyclopedia that anyone can edit, has been awhirl in controversy recently. This interview with the project’s colorful leader, Jimmy Wales, was recorded on December 5, 2005, after the John Seigenthaler incident and before Wales’ December 20th announcement that there will eventually be a “stable” version of the work.

The Writing Show (WS): Welcome to the Writing Show, where writing is always the story. I’m your host, Paula B.

My very special guest today is Jimmy Wales, an original creator and leader of Wikipedia, which as almost everyone knows these days is an international collaborative open-source free encyclopedia on the Internet. Welcome to The Writing Show, Jimmy. It is such a pleasure to have you here.

Jimmy Wales (JW): Thank you. It’s good to be here.

WS: Most Internet users are aware that Wikipedia is the free encyclopedia that anyone can edit, but I’d like to go into a little more depth about what Wikipedia actually is. Can you tell us a little bit more about its mission and scope?

JW: Yeah. I guess the most important thing that I would emphasize is that by free, we mean freely licensed. So free in the sense of GNU or in the sense of open source software so people can take our work and they can copy it, modify it, redistribute it, they can do all this freely, commercially or non-commercially. So that’s a really big part of what we’re doing. Then when people are working in Wikipedia, they can feel comfortable that their work won’t ever be made proprietary. It’s a gift from the Wikipedians to all of humanity, and that’s really a core value for us.

WS: What about the scope of the Wikipedia? Can you talk a little bit about that?

JW: Well, there’s a few things about that. First of all, Wikipedia’s an encyclopedia, so that’s a very broad scope, and one of our sayings is, “Wiki is not paper,” meaning we don’t have to restrict ourselves based on space considerations. There’s always more room available since it’s all electronic. But at the same time, of course, you can’t have an encyclopedia article about absolutely everything. We have rules like verifiability. You know, the information. If somebody wants to write an article about their cat, for example, that would just get deleted because there’s no way for anyone else to verify the information. So those are the kinds of things that we balance to figure out the overall scope of our work.

WS: I remember reading something about a subject having to be notable. Is that right?

JW: Yeah, notability is actually a very controversial requirement within the community simply because it’s so subjective. What’s notable enough? So what we prefer to do is more or less shy away from notability, just because it ends up being a pretty unproductive discussion, and focus really a lot more on things like verifiability, whether or not the information can be verified. That’s a much easier thing to decide rather than is it important enough. That’s a very tough argument to have.

WS: That’s really interesting. I wasn’t aware of that. So you could actually cover anything that is verifiable. Is that right?

JW: Yes, to some extent, although there are questions. That makes it sound easy, but what actually is verifiable? That’s the other, the next question. There’s a lot of different ways things can be verified, and so there’s no simple answer to this. But the complex answer is it’s a community process of discussion and debate and deleting things per the “votes for deletion pages,” and things like that. So it’s an ongoing process.

WS: What types of verification would you consider legitimate?

JW: Well, a simple example would be references to published books, academic papers, that sort of thing. That makes information verifiable. You can say, “I found it in this book.” An example of something that might or might not be verifiable would be something like a Web site about a band. So lots of little garage bands have very puffy Web sites about themselves that they made the Web site themselves, but you can’t find any reference to the band anywhere else. Not in any newspaper, not in any music sites, and you realize, Oh, this is just somebody who made a Web site and so the information that’s contained within that Web site is something that you really can’t verify.

WS: What about publishers? I mean there are all kinds of publishers, large and small, and I know some of the smaller ones could be a bit dodgy. What do you use to evaluate, say, books from somebody who’s either obviously got an agenda or is just a bit on the flaky side. Well, it’s not fair to say that, but you can kinda tell.

JW: Right. Sometimes it is fair to say that. There’s definitely kooks and crackpots out there who have self-published books and you have to treat those with great caution. One of the rules that we have in Wikipedia is no original research. So people are always coming out with their new theory of magnetism or something like this, and they want to put it into Wikipedia, and we just don’t allow that. And so in different areas the standards for what would count as a legitimate source will vary, just depending on the nature of the subject. So something in physics we’re going to want to have some reference to a mainstream physicist published in a mainstream journal or book. Whereas for other things, a lot of pop culture things, there are no academic references, and the only real sources are Web sites on the Internet, things like that. So it just depends. There’s no simple formula to answer this kind of question, really.

WS: And the contributors always cite their sources? Is that right?

JW: Not always. It’s strongly encouraged to cite your sources, but it isn’t actually necessary for every single piece of information that we have a source cited if it’s completely uncontroversial. If you say something like, “George Washington was the first president of the United States,” you don’t really need to go out and footnote that particular piece of information. That’s widely known and uncontroversial. Of course, if anybody challenges you on it and it’s a legitimate challenge, then yeah, you’re going to need to cite a source. And in fact, one of the techniques of editing that people use is if somebody’s putting in some questionable information that doesn’t sound quite right, we just delete it and say, “Hey, you can put this back in, but you’re going to have to give us a source.” So it just depends, again, on the specific context, and it’s really not any different from the kind of demand for references that you would have in any sort of traditional reference work. Some things you have to give a source. Some things are just obvious and uncontroversial, so you don’t need to, really.

WS: What about some of your other editorial guidelines?

JW: Well, I guess a lot of our guidelines have to do with behavior. They’re social guidelines which indirectly, they guide the editorial process. So we really try to strongly encourage a collaborative, helpful, friendly, thoughtful environment as opposed to an environment of edit-warring and fighting. And so that’s a big part of it. The wiki process, in and of itself, is something of a mutually assured destruction-type of process. In other words, if you write something that’s biased, it’ll just be deleted. And so everybody who participates has an incentive to try to write for the enemy, as we put it, or write for people who may not agree with you and try to phrase things in a way that’s as neutral as you possibly can because that’s the only way to write something that will survive the test of time.

WS: You mentioned encouraging an environment where everybody is neutral and working together. What specifically do you do to encourage that sort of environment and to help people understand how to write good articles that would be accepted by everybody?

JW: Well, I think a lot of it is, it’s just in terms of the most important and active members of the community setting the right tone, and we have a lot of internal rules having to do with behavior. So no personal attacks. If somebody makes a personal attack on the Web site, it can just be removed. And then if somebody does it repeatedly or too often, they can actually be blocked from editing for that. So it’s a lot of little subtle things more than anything else. It’s basically a…it’s an attitude of intolerance for bad behavior, I think is the way I would put it.

WS: When I was doing a little bit of research for this interview, it occurred to me that your guidelines seem to be a recipe for critical thinking. Would you agree with that?

JW: Yeah, I think so. Yeah, there’s an awful lot that goes into the development of an encyclopedia article. Yeah. A lot of what goes into writing a good encyclopedia article is all about critical thinking. It’s all about being very careful and citing your sources, phrasing things in a neutral way. So yeah, it’s absolutely that.

WS: So we have these guidelines: neutral point of view and documentation and all the things we just mentioned. How good a job do you think you’re doing in adhering to the guidelines?

JW: I think very good, for the most part. It’s a very big Web site, and very noisy and lots of things are going on. There are parts of the Web site that are better than other parts, and parts that are worse than other parts. But overall I think we’re doing a pretty good job of adhering to our stated guidelines.

WS: What do you do to enforce them? I know some of the entries get vandalized and….

JW: Yeah. Normally if somebody comes in and they start doing some vandalism, they’ll get one warning, but then they’ll just be blocked from editing. We block based on IP number, which is imperfect. People can get a different IP number and things like that. But it seems to do a reasonably good job of it. I guess the main thing is that there’s an escalation process. So once somebody is a community member, you’re not just a random person who shows up and starts vandalizing pages, but you’re a community member but you’re doing things that are not socially acceptable—you’re making bad edits or you’re picking fights with people—then in the English Wikipedia anyway, there’s the arbitration committee, which hears disputes among users and can ultimately block people from editing for short periods of time or longer periods of time up to a year. And it’s really a very huge and complex social system. It’s all grown organically over time within the community.

WS: What about inaccuracies? Maybe somebody’s not being malicious but things are just wrong. What do you do about those?

JW: Well, there are different kinds of cases. So one case there’s somebody who’s normally a good user just makes a mistake. They are reading a source and working on an article and they’ve misread the source or misunderstood it and gotten something wrong. And that’s no problem really. That’s part of the process of writing. The other case would be somebody who’s just persistently putting in bad information over a long period of time, and that person would ultimately be invited to leave the project.

WS: I know there have been some reports of inaccuracies that have stayed on the system for quite a while. There was the recent report of John Seigenthaler, assistant to RFK.

JW: Uh huh.

WS: How are you approaching those sorts of problems where there are so many articles—how can you patrol everything?

JW: Right. Well, so the first thing to realize is that the number of articles isn’t really the right measure for how hard it is to patrol because a lot of the articles just sit there and they’re not really edited, and so they’re no problem to patrol them because nothing’s happening to them. What really matters is the flow of new edits and the ratio between the flow of new edits and the number of active, experienced contributors who are keeping an eye on things. That ratio seems to be just fine. Everything seems more or less normal in that regard. In this particular case, what happened was that there was a new article created by an anonymous user. It was not well-linked from elsewhere on the site, and therefore when it slipped by the new pages patrol, who are constantly looking at things and deleting things, then it was, from there it never showed up in the view of the article of a…of editors who are very experienced in that area, the Kennedy era politics and things like that. So what we’re doing to try to fix that problem is basically we’re turning off today, I don’t know if that’s actually happened yet or just will happen later today, but we’re turning off the ability for unregistered users to create new articles in an attempt to slow down the flow of new articles to a pace such that the people who are monitoring new articles make sure to not let things like that slip through. There’s a cost of doing that though, because the vast majority of people who are doing work, even while not registered, do good work. And so we want to maintain that ability for people to edit without having to log in. So it’s a tough trade-off to figure out exactly the right way to improve the situation.

WS: How many people do you actually have monitoring the new articles?

JW: Hundreds. Yeah, it’s hard to say. People are always monitoring recent changes. Different people look at new articles when they come in. There’s a new pages tool where people can watch the new pages as they roll in. It’s kind of hard to count because sometimes people do it for five minutes and sometimes people do it for hours on end.

WS: I read something about a time delay. Was that the registration vs. non-registration that you just mentioned, or was that a different one?

JW: No, that’s a different…that’s another example of a tweak that we’re using. Basically, what happens sometimes is we’ll have a very high-profile page. The example I like to give is when the new pope was announced. That page was extremely high-profile. It was linked to from major news sources, it got tons and tons of traffic, but it was tons of traffic from people outside the community. And as a result, that page was getting vandalized. And so our response to vandalism is to protect a page, to lock it so that no one can edit it. Well, it’s really not good to have such an important page locked. We don’t like it when we do that. So we’re looking for a softer tool to deal with that situation. So the softer tool is instead of protecting the page, which is our traditional method, is to put the page into a state of time delay so that when people come in and vandalize, then we have ten minutes to catch it before it goes live on the site. Which we think from looking at our statistics that that would stop 99% of the type of vandalism from ever being seen by the average user. That will go into effect—I’m not exactly sure when. It’s kind of a tricky programming problem with our database design and things like that. So it’s taken a little longer to do than we wanted, but it should be coming within the next few months anyway.

WS: That brings up an interesting question. You were talking about a time delay of ten minutes, which is really short. And that implies that you have contributors who are there all the time. I’m really curious to know more about your contributors. Do you have any feel for what they’re like? Their demographics? What motivates them? That sort of thing.

JW: Yeah. In the core community, there’s several hundred people who are really, really active Wikipedia users. Those people are, and I travel all over the world meeting Wikipedians, so I have a pretty good sense of who these people are, they’re mostly in their late 20s and 30s, professionals, some graduate students, some professors. Lots of geeks who are working in Wikipedia outside their field of expertise. So the example I like to give would be of a math professor who works on Elizabethan history articles. And part of the motivation that people have for doing this is that in our modern world, people are strongly encouraged by the organization of society to specialize in some area, but they may be very well-rounded intellectually. So math professors really focus on a very, very narrow subset of mathematics and professional work, but they have broad intellectual interests and so really enjoy having the opportunity to interact with people intellectually on other types of subjects.

So that’s it. I’d say the main two motivations people have are the big picture charitable goal, the idea that we’re doing something to give a free encyclopedia to every single person on the planet. And then also just the fun of doing it. It’s just like lots of geek fun to get on and start working with people. So that’s why we do it.

WS: That’s really interesting. You mentioned people who specialize in one area that’s their profession, but they know a lot about something else and they don’t get the chance to express it. I really find that fascinating. I’m wondering, do you know if there’s any overlap, are any of your contributors people who also write for other encyclopedias like Britannica or World Book or any of the other encyclopedias?

JW: I have no idea who writes those encyclopedias, so I don’t actually know.

WS: I just wondered if anybody had mentioned it.

JW: No, I don’t really know.

WS: What do you think are the biggest misconceptions about Wikipedia?

JW: I think the biggest misconception is the idea that Wikipedia is something like a million people writing one sentence each. That it’s this mass phenomenon, when it in fact is written by and large by the core community. So something like 2% of the users do 75% of the work. And that 2%, that core community of the really active users is…in English it’s around, the last numbers that I’ve seen, it may be a little bit bigger since the statistics I looked at…was about 1750 people. So I think that’s really the most important thing that people should try to understand, is that it isn’t like random interactions like on eBay. It’s really more of a community of people who actually know each other. So it’s a lot more traditional internally than most people realize.

WS: And people don’t actually sign their articles, right?

JW: Right. But everybody, when you’re editing, if you click on the history, you can see everyone who contributed to the article. So it isn’t that people are editing in an anonymous way. Basically what, the way I like to put it is, instead of using a gatekeeper model, we use an accountability model. Every single edit that you make to the site is attached to your user name and people can actually look at your contributions and see every edit that you’ve made and judge the quality of your work. So it’s a very public, transparent process.

WS: Those user names aren’t the person’s real name, though. Is that right?

JW: Yeah, right, right.

WS: So they’re not really doing it for glory.

JW: Not necessarily for public glory, although there is a certain amount of, within the community it isn’t what the general public thinks or becoming famous from editing Wikipedia. But you do become famous within the community. And if it’s a community of people you respect, then that can be very valuable to you to say, “Hey, one of the reasons I do good work in Wikipedia is that other Wikipedians, who are people I admire and respect, appreciate my work.” And that’s a positive social benefit.

WS: Let me ask you a little about Britannica. Do you feel that Wikipedia is in competition with Britannica?

JW: In a certain sense of course we are. But in another sense we really aren’t. The issue is we’re a community. And so we’re just people who are doing what we do and we really don’t think much about Britannica either way. If their sales are increasing or their sales are decreasing, it really has no impact on us. We don’t care. So it’s really hard to say. I know that in Germany, the sales of Brachhaus are up 30% over last year. German Wikipedia’s, or I should say, Wikipedia in Germany is about 50% more popular than in English-speaking countries. Our per capita viewership is 50% higher in Germany. So it doesn’t seem to be hurting Brachhaus’s sales, so I don’t see why it would necessarily hurt Britannica’s sales.

WS: I was just going to ask you whether you thought that Wikipedia was a threat to Britannica, but I guess not.

JW: Yeah, it’s hard to say. It may just be that we remind people of the importance of encyclopedias and how cool encyclopedias can be, so they may go out and buy Britannica as well. Certainly their budget for buying encyclopedias isn’t, it isn’t a competition in that sense. If you were thinking of buying Britannica, you don’t have to say, “But I spent all my money on Wikipedia.” It’s free.

WS: And that actually segues into, sort of, into another thing I wanted to ask you about. How is Wikipedia financed?

JW: We’re financed with donations from the general public, is the vast majority of our donations are $50 to $100 range from the general public. We do get some larger gifts, and we get some support from corporations. Like Yahoo, for example, gave us some servers to use in our South Korean facility. But since it’s an all-volunteer effort for the most part, it means that it doesn’t really have such a huge budget. We have now three employees. We have our lead software developer, Brian. And we have my assistant in the office, and my assistant now has an assistant, so that’s three people. All of the rest of us, including me, are volunteers, so our budget for salaries is quite low.

WS: And I guess you have to pay obviously for, no Yahoo isn’t covering all the servers. You have to pay for your servers, right?

JW: Right. Exactly. So we spend a ton of money on new servers and we spend money on bandwidth and things like that . We do have expenses, but they’re not so bad. So far we’ve been very fortunate that the general public is willing to help us out. That makes all the difference in preserving our independence and letting us pursue our goals.

WS: Are you going to be doing any revenue sharing from any of these other projects like the German version being sold on CD and DVD by the German publisher, or any other you know….?

JW: Yeah, absolutely. The main thing that we really want to try to avoid is to have any advertising on the Web site, but there are other revenue opportunities in terms of licensing our trademark for sales of DVDs, for example, in Germany that’s being experimented with. And for example, has a software tool. They put out a Wikipedia edition of their software tool, and when people use that tool and visit the site, then they split the ad revenue with us. So that’s a way that they are trying to give back to the community and benefit us.

WS: Let me just ask you a question particularly for librarians. If a librarian is trying to help his or her patrons use Wikipedia, do you have any particular advice for them?

JW: For a librarian using Wikipedia?

WS: Not themselves using it, but helping their patrons use it.

JW: Oh yeah. Well, I guess…

WS: Anything special you’d tell them?

JW: Yeah, I guess the main thing is people need to understand that Wikipedia is very much a work in progress. That it is in many places very high quality, but that it’s, because it is an open-ended work in progress, there can be mistakes and errors that haven’t been caught yet. And that I would treat it as an excellent starting point to get some basic background information before doing further research.

WS: One of the librarians that I talked to for this project said that she had read, oh, there was some controversy about people leaving the community because they were feeling that they were attacked or something. Can you talk about that at all?

JW: I’m not sure of any specific incident, but there are people who do come and go from time to time. Some people do leave because they are being attacked, and they are being attacked because they’re being completely preposterous in their behavior. So we don’t feel too bad about that, although I prefer not to attack people. But if they can’t write in a clear and neutral and intelligent manner, then yeah, they probably should leave the project. Sometimes good people leave the project because they end up dealing with idiots until they can’t stand it anymore. And so we always have to balance a lot of different competing concerns within the community. We hate losing good people because we’ve been too slow in blocking a troll or a vandal who was messing up their good work. But on the other hand, we don’t want to be too quick to ban people if it’s just some ordinary editing dispute. And so finding the middle ground there is an ongoing process. And it’s messy and human and will never be perfect.

WS: Do you have any kids who contribute to Wikipedia? Is that a thing that kids could do?

JW: Yeah, we don’t normally specifically encourage contribution from children. It’s fine. We judge people based on the merit of their work, not on their age. We do have some very young contributors who do excellent work. We have one young user, 14 or 15, who has gotten dozens of featured articles on the site. He works on mostly things having to do with the British peerage and has just written volumes and volumes of fantastic stuff, very carefully researched. And other people help him, but he’s a great writer and so that’s great. But in general, I don’t know of any cases really of people much younger than 13 contributing a whole lot over a long period of time. Sometimes in pop culture areas. I don’t know exactly who it is who’s writing all these Pokemon articles, but I assume they’re younger than I am.

WS: How do you see Wikipedia evolving?

JW: I guess I think the main thing is we’re fully global, and so I think a big portion of my effort is going towards figuring out how to promote Wikipedia in more and more languages, because there’s huge amounts of work left to be done. The English Wikipedia is the hardest to predict because it’s the largest, and so as it grows, it’s always breaking new ground in terms of the social pressures, things like that. My belief is that we’re going to slowly but surely have a bit of a turn inward in the sense of once you have one million articles, which we’ll have just after the beginning of the year, the number of new, different topics, major topics that are not covered at all, becomes less, and it becomes more a matter of filling out the details of the things that you have rather than focusing on growth. And so filling in details, increasing quality, things like that are going to become an increasing focus within the community over the next few years, I’m sure. And that’s just a projection. That’s an easy thing to predict because that’s a projection of the trend that we’ve had for a very long time. More and more attention to how do we increase quality now that we’ve got X amount of articles. So that’s my prediction.

WS: Thank you so much for speaking with me today, Jimmy. I really appreciate it.

JW: Yep, very good. Yep, no problem. Thanks a lot.

WS: And, of course, as everyone knows, the Web site for Wikipedia is

Be sure to let us know what you think of Jimmy Wales’ interview, or any of our interviews, for that matter, for our new feature, Write to the Writing Show. I’ll be reading listener comments on the air. Email me at paula at compulsivecreative dot com..

I’d also like to invite you to dip into our archive, where you’ll find interviews with fascinating guests like Jim Cox of the Midwest Book Review. Jim explains what the book review process is all about and tells you how to get your book reviewed.

Just one more announcement, and that is that The Writing Show will now be posted on Tuesdays to give me and Alan the chance to actually experience weekends. I remember what that was like—I think.

Happy New Year from The Writing Show.

I’m your host, Paula B.