The conversation at the taco stand
There's been a fair amount of talk about the conversation between me and Larry Sanger that led to the creation of Wikipedia. Here is my memory of it.
Background
Larry had been working for a while on Nupedia, a free on-line dictionary written by credentialed experts and following a strict review and certification process to ensure that every article was high quality and, above all, factually reliable. A lot of Larry's thought since I first got to know him back in 1994 had been devoted to how sources of knowledge can be reliable and unbiased. I believe he was always looking for foundational knowledge: something solid to build on, or at least as close to that as we can practically achieve. Nothing much had come of Nupedia, though. There were hardly any articles completed.
The previous several months, I had been participating heavily on two wikis: Ward Cunningham's original wiki and one called "Why Clublet", run by Richard Drake and Keith Braithewaite. I'd become interested in the way wiki pages would grow in directions that none of the contributors had anticipated, and sometimes with writing clearer and better than any one contributor was capable of. I was especially interested in how conflicting ideas could be explored in depth by separating discussion of them to different places. No back-and-forth head-on "debate", just parallel exploration of opposing ideas. I saw it work especially well when people focused on improving the quality of the writing and poorly when people tried to make the wiki officially declare their preferred view as the correct one. I had also been experimenting with Extreme Programming at work, and had discovered amazing synergy in pair programming. I liked how these structures enabled people working together to actually be smarter than people working alone--the exact opposite of "committee"-style collaboration, and also opposite from the kind of collaboration where each person "owns" a sectioned-off piece of the whole.
That night
On January 2, 2001, Larry and I ate dinner at the taco stand at 1932 Grand in Pacific Beach, San Diego. I believe this taco stand had no name, just the words "MEXICAN FOOD" written on a window. (I have a picture of it.)
The conversation started with Larry telling me about Nupedia's progress since the last time we'd talked. Not many articles had been completed, but he was optimistic about Nupedia's future. He wanted to speed it up, though, and said he'd been looking at technological ways to speed up editing. I said that I knew of a really neat tool for fast collaborative editing: a wiki. Larry hadn't heard of wikis, so I told him all about them at great length.
I suggested that instead of just using the wiki with Nupedia's approved staff, he open it up to the general public and let each edit appear on the site immediately, with no review process. Instead of preventing error and bias, I said to openly invite error and bias and make it very easy for people to correct them. My exact words were to allow "any fool in the world with Internet access" to freely modify any page on the site. Also, I said that on wikis, there are no "completed" articles, there is just endless chaos and conflict. I suggested that this might actually lead to better reliability and richer content than the careful, circumspect approach.
Larry raised some objections and we debated a bit. Couldn't people just vandalize the site? I said yes, and other people could then repair the vandalism. Couldn't total idiots put up blatantly false or biased descriptions of things, to advance their ideological agendas? I said yes, and other idiots could delete those changes or edit them into something better.
I told Larry about some of the conventions on Ward's Wiki, like running words together to indicate a page name. I mentioned that this required me to give the Ward's Wiki page about Aristotle the name "MrAristotle." Larry cringed at that.
Larry was not completely convinced by my answers to his objections about quality, but he definitely liked the way the wiki concept enabled a great increase in editing speed, and he was eager to try it. We went over to his apartment, and he tried to call his boss at Bomis, Jimbo Wales. Jimbo didn't answer, so Larry left voicemail. Larry and I talked about philosophy for a while, and roughly half an hour later Jimbo called back. They talked for ten or fifteen minutes. After the conversation, Larry had a big smile on his face. Larry said that he felt very optimistic that the idea would proceed, and that Jimbo was quite open to it.
Afterward
Within a couple weeks, Wikipedia had gone live, and Larry had posted the WikiPedia page on Ward's Wiki, inviting people to come contribute.
It's a rare thing to tell someone to do something exactly the opposite of what he's been doing and get a fair hearing. It almost never happens that someone actually takes the suggestion. But Larry listened to what I had to say, let his imagination engage, and ran with it. Back then, wikis were a very hard concept to "get", but Larry's mind began percolating immediately, and he got things started that very night.
It was extraordinarily fortunate that Larry was working for Jimbo. Jimbo had both the means and the vision to get Wikipedia moving. Jimbo had stoked my interest in the power of collective knowledge years earlier in a post to the MDOP discussion list about sports betting and how it sets the "line" more reliably and accurately than any individual bettors.
I'm not one of the founders
Some folks, aiming to criticize or belittle Jimmy Wales, have taken to calling me one of the founders of Wikipedia, or even "the true founder". I suggested the idea, but I was not one of the founders. I was only the bumblebee. I had buzzed around the wiki flower for a while, and then pollinated the free-encyclopedia flower. I have talked with many others who had the same idea, just not in times or places where it could take root.
In my opinion, Larry Sanger most certainly is a co-founder of Wikipedia. Larry came up with the perfect name and got the word out. He found people who were eager to contribute and got them writing articles in a matter of weeks. Larry crafted the policies that converted Wikipedia from a cool idea into a practical success: Neutral Point of View, No Original Research, and Verifiable Sources. He wrote most of the original help and policy files, showing people how to get started and what the project was all about. He did the day-to-day political work of persuading people to participate in accord with those policies, both in letter and in spirit (and took a lot of abuse for it). Without these and much more, Wikipedia might have ended up a chaotic jumble of people shouting each other down, like many pages on Ward's Wiki. It is the only case I know of where a philosopher applied his knowledge of philosophy to a practical problem to get great results, benefiting millions of people.
You have Larry to thank for "Be Bold". You have Larry to thank for getting Wikipedia started, no less than Jimmy Wales. The only people you have to thank more are the tens of thousands of people who actually wrote Wikipedia and are still writing it every day.
Response from Larry
Hi Ben, I hope you don't mind if I respond on your page; we can move it to mine, if you prefer.
You have certainly remembered some more of the details than I do, and there is one point where your memory diverges significantly from mine. You say, "I suggested to Larry that he make Nupedia into a wiki." But I don't actually recall any such actual suggestion. What I do recall is that you talked on for quite a while about wikis. I then said something like, "Wow, that's really interesting--that sounds like a piece of cheap (free), available software we could use to solve this problem with Nupedia." And then, I think, I described the problems we were having with Nupedia. And then I am very sure we talked at great length about how a wiki could be used to build an encyclopedia. I also remember that you explained and defended the idea of a wiki very well!
Also, you say, "Larry was skeptical at first." This is probably true, because I am skeptical of everything at first, even my own ideas. But you did not have to persuade me to think about it; what I distinctly recall is that the realization that "a website anyone could write on" could be applied to the encyclopedia problem was nearly instantaneous, and I was nearly instantly excited by the idea. So I might have been skeptical at first, but I was very excited, and you didn't need to persuade me to think it through.
I remember laughing at "MrAristotle." I still do!
Now, you say that I called Jimmy Wales when I got home, but to be perfectly honest, I don't recall you stopping by my apartment, and I don't recall calling Jimmy Wales, either. But if you have a clear recollection of that, then it must have happened. Perhaps I do have some vague inkling of that. Anyway, what I do very distinctly recall is that I wrote a document, one or two pages long (I think it was just one page long), describing how a wiki encyclopedia would work. I believe I sent it to Wales that very night, and he had a wiki set up for me to play with either Jan. 3 (the next day) or Jan. 4. I forget; might have been two days. I was never told who actually set up the wiki software. There was a sysop who could have done it, but Jimmy has claimed responsibility for it--whatever. It was very easy to set up UseModWiki, I understand.
Finally, let me comment on this:
- It's a rare thing to tell someone to do something exactly the opposite of what he's been doing, indeed the exact opposite of how he's been thinking and investing mentally for most of his life, and get a fair hearing. It almost never happens that someone actually takes the suggestion. But Larry listened to what I had to say, let his imagination engage, and ran with it.
This is disappointing, Ben, because there is apparently a few things about me that you do not understand, even though we have talked so much. :-) First, I had hoped that Nupedia would be more efficient. In fact, that was one of the aims I set myself with it; when it turned out not to be efficient, I again set myself to fix the problem. So, while Wikipedia was much more efficient than Nupedia, it was always part of the plan to create a truly efficient, productive system. Second, as anyone who observes me working knows, I am constantly trying to think of ways to make things work better. I "let [my] imagination engage"? Good lord, I have to restrain my imagination. In that regard, I am what people call (loosely) a "pragmatist": I am always thinking up and trying out new things. Moreover, I don't think that that is in any contradiction, or even any interesting tension, with my desire for a solid foundation for knowledge. Having reliable knowledge is one value; having lots and lots of knowledge is another important value, one that I have always held up. Besides, I think that imagination is one of the most important keys to knowledge, both theoretical and practical: that is the faculty we use to generate hypotheses and solve problems. With the Citizendium, I have retained what worked with Wikipedia, and gotten rid of some of what didn't work. In the long run, I expect that we will have more articles than Wikipedia, more content, and better content. If I didn't believe that was possible, I wouldn't have started the project.
And as far as this goes--"It was extraordinarily fortunate that Larry was working for Jimbo. Jimbo had both the means and the vision to get Wikipedia moving"--this does me a disservice, I think, Ben. You weren't actually present when it happened, but as those who were generally attest, I was the one who got Wikipedia moving in its first year. At the time, Wales was (as far as I know) working full-time as CEO as Bomis, and had relatively little to do with Wikipedia's actual origins. He was pretty hard to get a hold of, and didn't do that much work on or off the wiki. He, with his partners, had the means to get it going: Bomis could pay me and could pay for the servers and technical support. But Wales' "vision" was sketchy at best, limited to a very few, and vague, guidelines. The basic principles of the project were articulated by me, and I pushed them, day in and day out. Of course, many of the regulars also reinforced these principles very well--many of them rather more than Wales himself did. Of course, again, I don't mean to criticize you, Ben--I'm just tired of Wales getting so much credit for work, and "vision," he did not do or have.
Anyway, interesting write-up, Ben, and mostly true! --Larry Sanger (talk) 05:55, 20 January 2008 (UTC)
