The Citizendium one year on:
a strong start and an amazing future

October 30, 2007

Larry Sanger

It's been exactly one year since work on the Citizendium wiki started ramping up.  I said then that I was properly skeptical about our chances and that the project was experimental.  Well, no longer.  Now it's time to report the results of the experiment: we've made a very strong start and an amazing future likely lies ahead of us.

In the first several weeks of the Citizendium's existence, the project's chances were dismissed by the likes of TechCrunch's Marshall Kirkpatrick, BoingBoing's Cory Doctorow, and Clay Shirky--among others.  A lot of such Web 2.0 cognoscenti weren't just skeptical; they clearly disliked the idea itself.  It sounded too elitist for their taste.  (But we ain't elitist, really.)  It completely upset their notions of what Web communities are supposed to be like.  As Kirkpatrick put it, "Does the world need a Wikipedia for stick-in-the-muds?"

We shrugged and got to work demonstrating a better wiki model.  It launched publicly last March, boosted by an Associated Press story and other press coverage.  And as it turns out, a year after the pilot wiki was started, the project is actually exciting and refreshing--so that, increasingly and ironically, it is the received wisdom of the digerati that is looking dogmatic and hidebound.

Debunking some myths

There are some myths about the Citizendium floating around that might be keeping some people from getting involved.  We can't have that--so let's debunk those myths.

Myth: it's too hard to get on board the Citizendium. There's a long and complicated application process.

In fact, it's now simple and automated.  You fill out a short Web form, then a human being (a "constable") will respond--with a "yes" in the vast majority of cases--within a few hours, sometimes minutes.  All the constable has to do is press a button, and you're in.

Myth: the Citizendium is experts-only; it's an elitist project.

Outrageously false.  How many times do we have to say this?  We ain't elitist.  This myth does a huge disservice to the project, because it leads "non-experts" to think that the project isn't open to them.  It is.  In fact, we have roles for the general public, which may become authors, as well as for experts, which may become editors.  They work together very well every day in an open, bottom-up wiki project.  If you didn't know that was possible, we're here to show you that it is.

Here's a hint: just because we have a role for experts, it does not follow that the Citizendium is experts-only or elitist.  Particularly in an encyclopedia project, a role for experts isn't elitist, it's merely good sense!

Myth: the Citizendium is simply a revival of the failed Nupedia project.

Ridiculously false.  The only significant similarities that the project has to Nupedia are that we have a role for experts, and that we require contributors to use their real names.  But the differences are huge.  We are a cutting-edge, grassroots, open wiki, and we feature instant publishing; Nupedia had a fairly old-fashioned, top-down seven-step publishing process.  Anyone can start an article on Citizendium; articles had to be assigned by Nupedia editors.  After a year, we have over 3,200 "live" articles [Nov. 20, three weeks later: now 3,900] and nearly 5 million words; after a year, Nupedia had a few dozen articles.

The Citizendium was started with intimate knowledge of the strengths and weaknesses of both Nupedia and Wikipedia, by the person who engineered both of those systems.  Why on Earth would I revive failed systems?

Myth: the Citizendium uses old-fashioned, top-down editorial control, so it is going nowhere.  We can safely ignore it.

Wrong.  As much as some critics might wish this were true, it isn't.  The Citizendium is very much open and bottom-up and, as a result, it will become harder and harder to ignore, as our growth accelerates in the next year.

There is a crucial difference between the Citizendium and other expert wiki encyclopedia projects that have started recently, like Scholarpedia and the Encyclopedia of Earth: we invite the general public, we make no work assignments, and our progress, warts and all, is highly visible.  The fact that we require real names and that we have a role for experts doesn't change that!

Not only have we grown nicely in our first year, our growth is accelerating.  With nearly 5 million words and over 3,200 articles, we have tripled our article count since the conclusion of our pilot project last spring--and that is despite a predictable summer slowdown, and without the benefit of many press or blog mentions, as we had in our first six months.  In the fall, predictably, activity has started heating up again, and without any help from the press this time.

We have doubled our rate of article creation, from 7 to 14 per day, in the last 100 days, and quadrupled it since January.  [Nov. 20, three weeks later: rate in the last month has been over 20/day.]   This rate is almost certain to continue growing, because we have started a (so far very effective) recruitment push.  More people, more articles.

In short, not only have we been growing steadily, our growth is accelerating.  For more, see "The coming explosion," below.

Myth: who cares?  Even if the Citizendium is growing, there's still no point to it.  Wikipedia has already won.  Nobody can catch up.

This unfortunate attitude is confused on many levels.  Suppose we had said that about Encyclopedia Britannica when we were starting Wikipedia?

There's no doubting Wikipedia's present popularity and dominance, and we don't begrudge them their successes.  But those successes do not mean there is no point to the Citizendium.  After all, most importantly, we are rapidly ramping up to a level of reliability that, without an expert approval system and a more responsible governance system, Wikipedia will never be able to achieve..

Moreover, it is hardly as if Wikipedia's popularity means the Citizendium cannot find many able contributors; there are already many people at work on the Citizendium who simply would not consider working on Wikipedia.  The world is big enough to sustain two general encyclopedia projects.  The existence of one popular resource does not make all other resources pointless.

Finally, there is no good reason to think that the Citizendium will not grow at accelerating rates and, in time, have millions of articles itself.  Again, see below.  Moreover, I think that our model will prove to be far more attractive to more people than Wikipedia's.  I will not be at all surprised if, after some years, there are more active Citizens than Wikipedians.

Myth: the Citizendium has only 3,200 articles after one year.  This is a sign of failure, because Wikipedia had 20,000 after its first year.

There are several reasons why this is a faulty inference and comparison.

First, the first six months of the project were a private pilot project.  Wikipedia had no such pilot project.  So, a more meaningful comparison might be made in March 2008, after the Citizendium had been publicly launched for one year.  By then, I suspect we'll have something like 6-7,000 articles.

Second, our average article length is 1,173 words, while our median article length is 476 words--fairly substantial.  I don't have the statistics on what the typical number of words in Wikipedia articles was in 2001, but I do know it was substantially fewer. I suspect you can triple our article count if you want to use article count to compare our amount of content to Wikipedia in 2001.  I believe we also have many more images and other kinds of content than Wikipedia did in 2001.

Third, throughout our first year, it usually took at least 24 hours to get people on board.  But we've just added an automatic account approval system.  Getting on board is still not instant, but usually, accounts are approved within a few hours of being requested--and new people are contributing more, as a result.  This alone will accelerate our growth.

Fourth, our articles are far better quality than Wikipedia's were, and many of our articles are already better than Wikipedia's articles at present.  Here there is no contest whatsoever.  This, I hope you'll agree, counts for something.

Myth: learning how to edit the Citizendium, like all wikis, is too complicated for my poor, nontechnical brain.

Wikis aren't nearly as complicated as they might seem.  "Wikiwiki" means "fast" in Hawaiian--wikis are fast to edit, fast to update, and fast to learn.  All you have to do is get in the system, go to the page you want to edit (or if you want to start a new page, check out the easy way), and start writing your brilliant prose, just as you would an e-mail.  That is really all you need to know, to get started.  Really!  The rest you can learn "by osmosis" and in bits and pieces.  No one requires you to be a Wiki Master.  Some of us find Wiki Masters slightly annoying anyway--they're always fiddling with arcane code, and not adding content.  We prefer the content.  For that, no arcane code is needed.  It really is easy to dive in!

(Just kidding, you Wiki Masters.  We need you, badly, too.)

What we have demonstrated in our first year

We're doing well.

We have pioneered a way to use wikis that is new and importantly different.  Even more striking is the fact that ours is perhaps the best model yet for using wikis.  A lot of people don't realize this yet.  But they will--just wait--because this is all under-reported news.  Consider what we have demonstrated already:

We ask experts and the public to work together in an open collaborative project, and as a result, we've produced many long, meaty articles--in just one year.  (Nearly 5 million words.)

A project that asks experts to work side-by-side with the general public can survive, grow, and even accelerate its growth.  Making a meaningful role for experts in an open project really is a viable option for Web 2.0 communities, as I thought it would be.  That's news!

A wiki that requires real names can grow nicely and even accelerate.  Requiring real identities will not, in fact, doom an open online community to failure.  Actually, it's nice to know who you're working with.

Just as one would suspect, eliminating anonymous and pseudonymous contribution goes a very long way to preventing vandalism, uncivil behavior, and trolling.  We have had virtually none.  Yes, you read that right.  That's news, too.

Taking basic behavioral rules--like no personal insults--seriously, and putting rules enforcement in the hands of relatively mature, educated people, tends to make it easy to deal with disruption when it does occur.  Hooray for our constables!

Requiring that contributors "sign" an explicit social contract greatly reduces pointless debates with people who would argue for a radically different version of the project, allowing contributors to focus on "live issues" (not dead ones).

The Citizendium Editorial Council has passed six resolutions according to a version of parliamentary procedure that makes use of a mailing list, wiki, and Web forum.  As far as we know, this is unprecedented and has many interesting potential applications.  (We want to automate this, though--we're looking for someone who can code it up.)

Subpages can be used to organize a variety of info types.

We conceive of our purpose as extending beyond purely encyclopedic information into reference information of all sorts.  We are organizing various kind of reference info logically on "subpages," with all the subpages on a given topic making up one big "cluster."  For an example, see our Biology article (click the green "tabs" at the top of the page).  We've only just started with this--but so far, so good.

We've also had some accomplishments that aren't exactly pathbreaking, but they're still worth bragging about:

Nonprofit Web 2.0 projects can be started on a shoestring.

When we made the announcement of the Citizendium and secured the use of our first server free of charge from Steadfast Networks, we had a $0 budget.  We bootstrapped everything into existence.  Perhaps some people need reminding that large, active Web 2.0 projects don't necessarily require a huge amount of money and a half-dozen strategists.  We have gotten by with one full-time employee (me) and $40,000.  But it helps that I've been supported via speaking and writing fees, and frankly, we do need more money.  (More on that later.)

Many people are willing to support this sort of project with their labor.

If you needed proof that there are many people who are willing to put in many hours on a project like the Citizendium, then look at this post on the Citizendium-L mailing list.  There, I thank dozens of people for their contributions and a number of organizations for their support.

Our new initiatives

But enough boasting.  Other than the usual plugging away, what are are we doing now?

Subpages.  While we are still focused first and foremost on encyclopedia articles, we have opened our doors to other sorts of reference information, which we place on "subpages."  For an example, see New York City.  The Citizendium's subpages include the information normally found on good encyclopedia articles, such as Related Articles (example: civil society), Bibliography (example: Harry S. Truman), and External Links (example: airship), but in the fullness of time will include further bibliographic material (example: filmography of Joe Louis), almanac-like catalogs or lists of data (example: famous tennis players), image galleries (example: linguistics), timelines (example: Tony Blair), and more.

Core Articles.  For each of around 35 workgroup subjects, we are now making lists of 99 (or 198) top-priority articles to write.  We're specifically inviting people to come and start those articles, and have even started awarding "points" (redeemable for bragging rights).  We've only recently started this initiative, but it's growing  steadily.  (Why not have a look and see if you're inspired to write about one of those topics?)

Recruitment.  We've just started getting the word out about the Citizendium--we've sent calls for participation to only 10 mailing lists (recently).  Believe it or not, virtually all of our growth has been as a result of press coverage.  We have done very little of the sort of digital recruitment we used to get Nupedia and Wikipedia going.  Well, now that we have an automated registration system, we can handle a lot more applications.  So we've finally started seriously inviting them.

Eduzendium.  The Citizendium is the perfect venue for professors who want their students to do public writing.  It's perfect because most topics are wide open, and the project is managed in a way that will appeal to most professors.  Already, we have had a half-dozen or more articles contributed by students of Citizendium editors, as part of course assignments.  We hope to do serious recruitment for the program later this year and next year.

Fundraiser.  In November and December 2007 we'll be doing a fundraiser.  Our goal is $10,000.  Please help us toward this goal!  We hope to raise much more than that, and we know we might raise less--but we have not in fact done any fundraisers since an aborted effort in January 2007.  We badly need help from a full-time technical guru, and our full-time Editor-in-Chief (yours truly) is at the moment an unpaid volunteer, just like everyone else.  (My little family living off of my writing and speaking income, but this isn't much.)

Short-term plans

Within the next several months, we have a lot to do indeed.

Adoption of new license.  The Citizendium will, finally, adopt a license (GFDL, CC-by-sa, or CC-by-nc-sa).  A number of essays have been submitted to help us decide.  We've set November 15 as the deadline for making the decision.

Governance solidification and regularization.  Further development of many governance policies has been "on hold," as we have focused on other things.  At the same time that these policies are developed or reworked, we will do a "changing of the guard," meaning that people in positions of responsibility in the project may move about.  For example, our Editorial Council is likely to impose a requirement of a minimum number of edits in order to participate in the Council, and then several editors will exit and several newer editors will join.  To take another example, we will be establishing a Judicial Board. 

Expansion of subpages.  We've got a fairly elaborate plan for expanding and maturing the use of subpages on the Citizendium.

Advisory Board and Board of Directors.  The Citizendium Statement of Fundamental Policies provides that the Editor-in-Chief will appoint an Advisory Board which will approve a binding community charter as well the first Board of Directors.  I hope to choose the Advisory Board by the end of the year.

Adoption of a Citizendium Charter.  Shortly thereafter, my main task will be to draft the Citizendium Charter, with input from the entire community.  (No draft yet exists.)  This will supersede the Statement of Fundamental Policies.

Launch SharedKnowing (a mailing list).  While this discussion-and-announcement list is hosted by the Citizendium, it is a distinct service, and non-Citizens are welcome to join the list.  It is devoted to "Well-reasoned, polite discussion of the nature of online knowledge production communities, with special but not exclusive focus on community policy (production, governance, management) questions; 'the new politics of knowledge' broadly speaking. Though participation is by no means restricted to philosophers, we would like the list to have a more theoretical or philosophical focus, as opposed to being concerned with the specific minutia of specific communities (such as Wikipedia)."

Longer-term plans

In 2008 and 2009, what do we hope to do, in addition to growing at an accelerating rate?

Major MediaWiki improvements.  We badly need to make several improvements to our system.  One way or another--either through donations or with help from a technology partner--we hope to dig into these improvements next year.  For example:

• Convert the subpage-and-metadata system we've recently added to a built-in system.  This will allow people to change large amounts of metadata--and even do things like rate articles--by simply filling out a form.
• Build in the workgroup apparatus into the wiki system.
• Create a one-click article approval system, as well as a way to solicit approval or comments from the right editors quickly and automatically.
• Create an account management system that allows people to subscribe and unsubscribe to project mailing lists from one spot, manage various kinds of reminders, and designate themselves as "active" or "inactive," etc.
• Convert our talk pages into threaded forums.
• Add a public feedback system.  There has been some demand for this, but it's a non-trivial request.
• Display "thank you" messages from donors at the bottom of every page in proportion to the amount (and recency) the donor has given.

Search for technology partner?  If we do not soon receive sufficient funding to enable us to make the significant changes to MediaWiki that really need to be made, we might invite a special relationship with a technology company.  It seems likely that, if we achieve the success we hope for, the wiki software as configured for the Citizendium will be in some demand.  This could motivate a technology firm to supply us with the coding hours needed to make all the changes that we need to make; they then become the key service provider for the Citizendium configuration of MediaWiki.

The Citizendium in other languages.  Because an online republic actually requires a mature governance framework and an editor-in-chief, starting the Citizendium in other languages will not be very easy.  Still, it is something that we are committed to doing, or helping with, at least.  We will probably not have time to devote to this until 2008, however.  It will require considerable time and attention from the Editor-in-Chief and the new Board of Directors.

Independence from the Tides Center.  In December 2006, the Citizendium Foundation joined the non-profit Tides Center as one of their projects.  We did this only because Tides enabled us to accept donations immediately and assisted with administrative (office) details.  However, this is a temporary arrangement.  We wish to be our own, completely independent 501(c)(3) non-profit.  We will make time to do this once our Board of Directors is in place.

Launch new projects.  In 2008 or 2009, I will turn toward other, brand new content production projects on behalf of the Citizendium Foundation and integrated (as much as possible) with the Citizendium wiki.

Sanger turns to fundraising.  At some point in 2008 or 2009, I will move away from active management of the wiki--which at this point still seems necessary--and use more of my time for fundraising.  Given our fundamentals and success so far, I feel confident that we could be raising hundreds of thousands of dollars per year.  I worked on the project itself, however, to make sure that there is something to raise money for.  I don't regret the decision, despite our having a very small budget indeed!--Of course, if we have funds to hire a professional development director, I may not have to do this.

New editor-in-chief when funds are located.  Finally, in 2008 or 2009, consistently with the promise I made when first launching the project, I will step down as Editor-in-Chief and help guarantee the start of a regular, rule-governed, meaningful transition of management.  I want this position to be reasonably well-funded, however.

The coming explosion of growth

I want to make a prediction about the next year.  At some point, possibly very soon, the Citizendium will grow explosively--say, quadruple the number of its active contributors, or even grow by an order of magnitude.  And it will experience that growth over the course of a month or two, and its growth will continue to accelerate from that higher rate.  Yeah, maybe this is a little wishful thinking of my own.  But there is actually good reason to expect this; I am not merely trying to make a self-fulfilling prophesy.

Let me explain why it's reasonable to expect explosive growth in the not-too-distant future.

First, many people now know about us, but are watching and waiting before they get involved.  They're not early adopters; they'll join only after we're more proven or popular.  There are a lot of people who were motivated to make accounts (we have well over 2,000 "CZ Authors"), but very many haven't been actually motivated to start seriously editing the wiki.  (Over 200 accounts are used to make edits every month.)  I regularly find people online who say, "I support what you're doing, it looks neat, maybe I'll get involved, but..."

But what?  But they're not convinced we're a going concern, of course.  Why should they contribute to the Citizendium if it's always going to be small and unimportant?

This leads me to my second point. We will soon have developed to the point where a bunch of people can, all at once, prove to each other that the Citizendium is something really exciting.  The Internet is famously subject to "crowd" phenomena.  A news story, an endorsement from a famous person, or something more mysterious can cause massive migration to a new Web project--especially if there is an unfulfilled need for it.  When that happens, there is a sudden spike in activity--and those involved can observe the spike, and if the fundamentals are solid, that will feed on itself and lead to sustained growth.

Well, our fundamentals are extremely solid.  There is an unfulfilled need for an expert-guided, open, free encyclopedia under responsible management.  A huge number of people know about us, and they just need a little push to get involved.  When enough of them do, we will reach a tipping point--our visible growth will cause an avalanche of interest among our inactive supporters, who will then be convinced that, indeed, we really are a going concern.  And worth contributing to.

We're close to such a tipping point now, I think.  Here's a push, then.

The rate at which we have started new articles has actually tripled since January and doubled since July (the last 100 days).  In mid-January, just when we "un-forked," we were adding just 4.3 articles per day; in mid-July, we were at 7; and by mid-October we reached 14. We have tripled our creation rate since January and doubled it since July.

To put it simply, we aren't just growing; our growth is accelerating.

(Technical note: it is actually difficult to say precisely what our growth rate has been.  The total number of articles, including ones that are not "live," is currently 4,070, and the above growth rates include the growth rates for all of these, not just live articles. But the growth rate of live articles is probably even higher, because I think over the course of 2007, we have imported fewer and fewer unimproved Wikipedia articles.  Anyway, the increase in growth rate for "live" articles is also very probably close to 300% since January and 200% since July.)

Suppose that we continue to accelerate our growth.  This is not unreasonable.  The only question is how quickly we will accelerate.  If we were to continue to triple our article count each year, then we would break 100,000 articles by 2010, and one million articles by 2012.

Suppose we merely double our article count every year. Then we'll still break 100,000 articles by 2011 and one million by 2015.

Suppose we merely add 50% more articles every year.  We would break 50,000 articles by 2011, 100,000 articles by 2013, and one million by 2019.  Even this relatively slow pace would be well worth working toward, considering the other advantages of the Citizendium.

In short, if you assume that we will continue to accelerate our growth rate by at least 50% per year, you can expect us to have over 100,000 articles in about five years.  Frankly, accelerating our growth rate by 50% would be, by the standards of thriving and proven Web 2.0 projects--like the Citizendium--on the slow side.  I'd be willing to go out on a limb and say we'll do better.  I think we'll at least double our article creation rate every year.  So I think we'll probably have at least 100,000 by 2011, and one million by 2015.

And that's without any such "explosion of growth" as I mentioned earlier.  If we do reach a "tipping point" where loads of new people join all at once, we can increase our growth rate not by 200% or 300% in one year, but by 1000%, or more.

Of course, I don't claim to be able to predict what our rate of acceleration will be.  It's still possible that the project will, from here until eternity, muddle on creating 14 articles per day.  It's even possible that the project will simply collapse and our rate will go to 0.  I just don't think that these latter possibilities are at all likely.  Why?

The project's fundamentals are solid and growing stronger through motivated, diligent effort.  Most of our active contributors show no signs of giving up.  I'm not giving up.  Lots of new people are getting active, especially with our recruitment drive.  Getting on board is now quick and easy.  We become more and more credible, both as a productive community and as an information source, every day.  Citizendium articles are also rising in the Google rankings, which in time will create viral growth through the Google effect, as I argued last spring.  Finally, as I explained above, we might reach a tipping point sometime fairly soon, and increase our growth rate explosively.

A possibility you may not have considered

I've had a particularly fascinating idea in mind since before the Citizendium was conceived.  It is one of my deep motivations for starting the project.

Consider a possibility.  What would content shared in common look like if it were subject to open review and collaborative development from really large numbers of specialists, and other smart people with valuable input, from around the world?  Imagine particularly if versions of this content could be approved and displayed, while further work could continue, under expert guidance, indefinitely?

Given enough time and enough people, the results would surely be amazing.  The world has never seen anything like the picture I have in my mind's eye.  It is hard to predict for sure the quality of the content, but I suspect, based on my experience so far, that virtually every article created this way would, after some years, be wonderfully readable, yet also extremely detailed, perfectly representative of the range of expert opinion, and in general, magisterial.  What if we had hundreds of thousands of articles like that, on every subject?  A central storehouse of really reliable information would change the world, I suspect, in ways we can't even guess now.

If this possibility is amazing, it is even more amazing that it's within our grasp.  We're still in our infancy, but I see no reason to think that the Citizendium cannot seize this opportunity.  We are laying the foundations for it, and every day it seems more likely.

I suspect that even many rank-and-file Citizens (contributors to the Citizendium) do not fully grasp this potential.  Almost certainly, the vast majority of people who are casually tracking our progress haven't got an inkling of where the project might lead.  Those who do don't really care about what strike them as wild possibilities.  Perhaps it's a weakness of mine that I do care about wild possibilities.  But given two reasonable assumptions--merely time and further development along the path we've already struck out on--the outcome described seems not just possible, not just likely, but inevitable.

Of course, I could just have a surfeit of imagination.  Time will tell.  What I do know is that if we do have a good chance to create something so stupefyingly useful for humanity, we must try.

Do you agree?  Then join up and pitch in!

For more information on the Citizendium project, see the citizendium.org website.  You can easily join here.