Why the Citizendium Will (Probably) Succeed

March 2007

Larry Sanger

This essay argues that the Citizendium is feasible.  The wiki is in public view, or will be within a few days.  You can join now.

1. So far, so good.

The Citizendium pilot project wiki got under way privately at the start of last November.  In the intervening months, we have steadily grown to some 1,100 "CZ Live" articles--that's approximately how many articles we have done significant work on.  A fairly large percentage of these, I believe well over half, are either original articles or have been significantly changed from Wikipedia sources.  We have steadily added authors and editors in this period, so that we have 820 authors and 180 editors (some of whom also have listed themselves as authors).  Our activity has grown from 100 edits per day in the first month to over 500 prior to launch.  Every day, a large variety of people from many fields sign on and do some work.  This is all in a period in which the project has been visible only to those who have applied to the project.  In addition, while it has received a fair bit of press, we have done very little in the way of recruitment--but with good results when we have.  More aggressive recruitment is our trump card, which we haven't played.

Some will take this progress report to show that we are a roaring success; others will take it as evidence of our impending doom.  In fact, the progress report shows merely that the fundamentals of the project are sound, many basic doubts are now dismissible on the basis of solid experience--and little more than that.  It shows that that experts can be quite good at wiki-style strong collaboration; that they can work well together with the general public; that a wide variety of people have a substantial desire to work on this sort of project; that a largely collegial and pleasant community can be built on principles of the use of real names and gentle expert guidance; that, so long as we avoid wide-open self-registration as we tried for about three weeks, this sort of project can be free of vandalism.  In short, there are no "gotchas"--nothing that makes me think this project can't work--and quite a bit of good news.  We are obviously a long way off from "unseat[ing] Wikipedia as the go-to destination for general information online," as our first press release said we're trying to do.  But our progress does show that we're well justified in moving out of a pilot project phase and into a new "beta project" phase.

The question on everyone's mind, no doubt--and which determines one's willingness to work on the project at all--is whether we will thrive in the long term.  In this essay, I want to advance what are, I think, some strong arguments that we will thrive in the long term.  I will also respond to a passel of ill-founded doubts.  What will this argumentation establish?  No guarantees; it would be silly to make guarantees, because future human behavior is hard to predict.  I expressed no small skepticism about our prospects myself last September, and I am still skeptical, but now less so.  So I would say that the arguments here should establish that involvement as an "early adopter" is well justified.  In other words: dive on in, the water's fine!

2. The Google effect.

It is worth reviewing why Wikipedia grew so well, because I think we will grow in a similar way and for similar reasons.  At some point, Google started spidering Wikipedia, that is, it indexed the whole wiki and started serving up pages among its search results.  The first time we noticed this, I think sometime in the spring of 2001, we saw a spike in traffic as well as in activity on the wiki.  With the new people on board, the rate of article production increased.  The next time Google spidered the wiki, more pages were indexed, and we got even more traffic.  It was an enormously productive feedback loop.

Obviously, not all wikis enjoy this Google effect; many die on the vine.  There are, I theorize, at least four requirements to enjoy the effect.  First, there needs to be a reasonably large fund of content to begin with.  Second, this content needs to be spiderable (by Google) and readable (by the general public).  Third, registration needs to be fairly easy and open.  Fourth, the project itself needs broad appeal to users who are also potential participants.

(You might think that the search results have to be fairly high up, as well.  Well, in fact, that doesn't seem to have been the case.  I remember being quite excited, in 2001, when a Wikipedia article appeared on the first page of results--and that was certainly some time after the Google effect had kicked in.  Our success seems to have had a lot more to do with the sheer quantity of pages indexed by Google.)

So let's see how the Citizendium stacks up against these requirements.

First, we may not have millions of articles, but as of this writing we do have about 1,600 pages in the Main namespace, and about 1,100 "CZ Live" articles; that's enough, I think, to begin a positive feedback loop with Google.  And there's no reason to think we won't continue to grow at least as fast as we have been growing.  Indeed, after opening up, we should grow, I hope, at least somewhat faster.  (That's the conservative prediction.)  That means that, if we need a few thousand more articles before the Google effect kicks in, we merely need to wait a few months.

Second, we should have launched into publicly-viewable and -joinable beta project by the end of March.  Google will be able to spider the wiki then.

Third, getting on board will typically be as easy as filling out a Web form.  Right now, users need only send an e-mail that, since a biography is required, might take five or ten minutes to prepare.  It is not necessary to have any special qualifications to get on board.  Most applicants are given a username and password within 24 hours, and many within just a few hours.  Whether this is "easy and open" enough to support the Google effect is debatable; I think it will be, but time will tell.

Fourth, the Citizendium has, I think, an immediate and broad appeal to many readers who are also potential writers.  The appeal to readers is obvious.  Finding factual or "encyclopedic" information about general topics is one of the main things people use search engines to do.  This no doubt is why we click on Wikipedia links so frequently: regardless of how dodgy the information might be, it does, after all, purport to be accurate information, which is what we're looking for.

If we add reliability to this basic, winning formula, the appeal to readers increases hugely.  I suppose the reason Wikipedia articles are as attractive to search engine users as they are, is simply that they sum up a lot of information.  That implies a high signal-to-noise ratio.  But if an entry has been overseen by experts--that is, if the project as a whole is evidently devoted not just to boatloads of information, but boatloads of credible, expert-vetted information--then it becomes much more attractive.  Imagine if Britannica were somehow (magically) to produce 1.5 million articles in English of the same average length as Wikipedia articles, and imagine that it made those articles free for all to view.  People would obviously turn to the Britannica articles first--because their first concern is credible information.

In addition, the potential appeal of the Citizendium to contributors is sizable, and growing, as I will argue.  This is important enough that I want to develop the argument at some length.

3. The latent demand is sizable and growing.

The latent demand among potential contributors for something like the Citizendium is tremendous.  This demand has been created, in general, by the availability of so much information on the Internet and the lack of few easy, effective ways to pick out what's credible.  More particularly, the demand for the Citizendium has been created by Wikipedia's problems.  With each new Wikipedia scandal, there is a growing outcry: "Can't we do any better than this?"

This outcry is loudest among professors, teachers, and librarians, who with increasing alarm have observed their charges using Wikipedia uncritically, as if it were just like Encyclopedia Britannica--only free, and bigger.  Every plugged-in student and researcher in the world has been given a giant "encyclopedia" that, despite lacking authoritativeness, is just so darned useful that it seems inefficient to consult anything else.  Wikipedia isn't going away, either.  Therefore, those professors, teachers, and librarians have every reason to root for and support the Citizendium.  Once it looks to them like we're a going concern--which, arguably, we already are--there's a good chance that increasing numbers of these information professionals will join us and recommend that others join us.  And once we have enough of the educators of the world on our side, we'll have an unstoppable momentum.

The news media will probably help as well.  They have already given the Citizendium some much-needed publicity; many of our early contributors have come via news articles.  The reasons for the press interest are obvious.  Having reported on Wikipedia's many problems, they understand its drawbacks, and they themselves are professionals and so naturally appreciate the value of professional involvement.  So journalists naturally think that Wikipedia could use some competition.  That's us.  The Citizendium, organized by the same person who organized Wikipedia, is perhaps the most viable free alternative to Wikipedia under development.  It has been growing respectably in its private pilot project phase, and is now launching into public view.  The story seems compelling, and it will become only more so as we grow.  The result will be that news coverage will probably continue to send many new people our way.  This is very important since only a tiny fraction of our potential contributors have even heard of us, let alone visited the site and considered joining.

Not all of the attention we've received, however, has been positive.  We have plenty of fans in the Blogosphere, but also a good many detractors.  A lot of the negative posts only help prove some points we've made and help establish us in our (quite desirable) niche more firmly.  Too often, these posts are poorly-reasoned, written in ignorance of basic, easy-to-find facts, and exude contempt for anyone who would even suggest that experts be given a special role, or that Wikipedia needs competition.  Know us by our opponents.  To the extent to which our opponents reveal themselves to be closed-minded, more open-minded people will want to know what we're all about.  The more that the dogmatists spout off, the more potential allies will rally to our cause.

A good number of disaffected Wikipedians have joined us.  Our increasing activity will bring over even more.  These are frequently the sort of people we want.  After all, our natural contributors like the idea of Wikipedia.  They love the ease of contribution, the instant visibility of their work, the sense of shared purpose inherent in strong collaboration, the gradually improving quality, and so on.  They love working with Wikipedia's many excellent contributors.  Despite all that, they even more strongly dislike having to deal with its many problem users--disrespectful, immature, ideologically driven, or  unstable people, that administrators are unable to rein in.  Indeed, if the many complaints are to be believed, such people are to be found among Wikipedia's administrators.

So there are a lot of good reasons to think the Citizendium is filling a demand for a new alternative, and that that demand is growing.  So, given what I said earlier, there's an excellent chance that we will enjoy the same Google effect that helped Wikipedia to grow.

4. Objections and replies.

A lot of doubt about the viability of our enterprise has been generated over the last six months.  As I hope to make clear, these doubts are generally poorly founded.

Objection.  You can never overtake Wikipedia.  It's growing at a staggering rate and has a head start.  You'll never be able to catch up.  As long as Wikipedia remains so much huger, why should anyone contribute to the Citizendium?

Reply.  First of all, it is a huge mistake to think that as long as Wikipedia remains bigger, nobody will see a need for another resource.  Plainly, our many supporters and growing roster of contributors see the need.  Second, we can become more useful and more reliable than Wikipedia with fewer articles.  Success is not directly tied to quantity of information--and many of our writers implicitly understand this.  But, third, the real question is how many people will want to contribute to the Citizendium after a few more years, once we've grown more and the project has been better publicized--once the word has gotten out better to our potential contributors.  This is an empirical question.  If you ask me to give an answer a priori, I'll hazard a guess that, in the long run, there will be more people who will want to contribute to a free encyclopedia under our rules than under Wikipedia's.  That's just because our system is likely to be more civil and pleasant and actually focused on the work of creating a credible encyclopedia.

Objection.  Citizendium articles right now are often short, or derivative of Wikipedia, or in other ways unimpressive.  Wikipedia is actually of higher quality than the Citizendium!

Reply.  Well, make sure you make the correct comparison.  You should have seen Wikipedia after its first few months!  The quality of its articles, at the time, was laughable.  Besides, in our new "article checklist" that tracks various statistics, almost half of our articles are either approved, developed, or developing articles, which means they are beyond the (very short) "stub" stage, and they are not merely copies of Wikipedia articles.  That is way better than Wikipedia was after its first few months.  Our most active editors tend to put enormous amounts of effort into relatively few articles, with excellent results.

Objection.  You have nearly 1,000 contributors signed up, but (as I write this, just prior to launch) you are dancing around 500 edits per day.  Surely you should have more than that.

Reply.  There are two effective replies to this.  The first is that very many people who have joined us did so simply out of curiosity--to see what we're all about while the project is still in its pilot project phase.  I have no idea what percentage of our registrants of which this is true, but it's probably sizable.  Second, there's the old 80-20 rule: 20% of your participants will make 80% of the edits.  And that rule has got it about right in our case.  I'm told that about 250 people have made 10 or more edits to the Citizendium pilot project wiki, while total number of accounts making edits is over 1,200--which must mean that there are a lot of people who made edits without actually adding the "CZ Author" tag to their user pages.

Objection.  Professionals don't work without compensation.  They require either personal credit that can be used in tenure and advancement committees, or else money.  You're offering neither.  How can you possibly expect to get enough experts to make this a viable project?

Reply.  Wikipedia itself has collected quite a few actual experts (i.e., people who would be invited to be editors here, if they were to apply).  How were they motivated to do their work, particularly when the discovery that they're contributing to Wikipedia would, if anything, be more likely to harm their careers more than help them?  It seems a lot of people, including a lot of experts, are strongly motivated either to show off their knowledge or to teach; it's fun, or fulfilling.  Wikipedia and the Citizendium are outlets for this laudable behavior, and the Citizendium is likely to become increasingly more attractive to experts than Wikipedia.  Moreover, it's likely that success in writing decent articles will lead to more success at recruiting editors, who will want to have their say about topics that we have not quite gotten right, by their lights.  It's also worth pointing out that in the few forays into recruitment we have done, we've had decent success.  Therefore, if we really feel that lack of expert participation is a concern, the problem is easily solved by doing more directed recruitment.

It is an empirical matter whether we'll collect enough active editors to be able to create a large enough collection of approved articles.  We won't know until we try, and try we will.  I have to admit that it would be astonishing, really, if we found enough experts to approve on the order of millions of articles, which is our goal.  It's more likely that we'll have a decent set of approved articles which is always a fraction of the total number of articles we are working on.  This, at least, would be a decided improvement over Wikipedia, and it's something I am hopeful we can achieve.

Objection.  Anonymity is really the main reason for Wikipedia's growth.  Unless you are open to people who refuse to give out their real names, you're doomed to be small and irrelevant.

Reply.  Anonymous contribution is not the main reason for Wikipedia's rapid growth--virality is.  And virality doesn't require anonymity.  The "six degrees" friend websites, some of which have worked quite well because they are viral, generally make use of real names.  The Citizendium is devoted to the proposition that we can grow a large community of named, responsible individuals virally.  We think it's worth a try.

Objection.  The Citizendium is doomed, because it's "credentialist."  You're using credentials to give people special rights in a wiki encyclopedia project?  Nobody can take that seriously.

Reply.  This is a uniquely Wikipedian objection, and it is little better than wishful thinking.  I gave a speech debunking the underlying view.  You are free to disagree with whatever threatens absolute, pristine, radical egalitarianism.  But even if we were philosophically wrong, which we aren't, does it really follow that we won't be able to find enough participants?  Of course not, that's a total nonsequitur.  Since a lot of people do like the notion (correct or not) of an expert-led wiki encyclopedia, it is not unlikely that we'll be able to get enough contributors.

Objection.  (From Clay Shirky.)  "The costs of certifying experts and insuring deference to them--the costs of creating and sustaining the necessary social facts--will sandbag the system, making it too annoying to use. ... The Citizendium project assumes that the desire of ordinary users to work alongside and be guided by experts is high, but everything in the proposal seems to raise the costs of contribution, relative to Wikipedia. If users do not want to participate in a system where the costs of participating are high, Citizendium will simply fail to grow."

Reply.  When Shirky originally wrote this, I fully intended to reply, but I got sidetracked by doing what was apparently impossible.  Now, some six months later, we're in a better position to evaluate his argument.  I note two facts.  (1) A few hundred people did participate in our system over the several months of the pilot project.  (2) The Citizendium has grown rather nicely, particularly considering that it has been a private pilot project for which we did rather little active recruitment.  Since Shirky's conclusion looks false, where did his argument go wrong?  I think it's this assumption: "The costs of certifying experts and insuring deference to them...will...mak[e] it too annoying to use."  In fact, it hasn't been that hard for people to send in e-mail applications with biographies and supporting Web links.  Lots of people have done so.  We're going to be semi-automating this process, too, so that constables can approve new applications with the press of a button.

Moreover, the cost of "insuring deference" to experts isn't as high as Shirky thought it would be.  Our constables have had to do very little indeed, to my recollection, to "insure deference" of authors to editors.  This might have something to do with the fact that we require our contributors to endorse a Statement of Fundamental Policies that says, basically, that editors have certain privileges.  But I think it has more to do with the fact that people who are committed to the elegant expression of expert opinion tend to have the common sense and politeness necessary to ensure that they can collaborate with others very productively.  One of the things that I personally was a little surprised at was just how well our editors took to collaboration.  For some reason, many people just assume that professionals just won't be so good at wiki-style collaboration.  But I've repeatedly observed that those editors who tried it just got it right away.  Perhaps the reason they get it is that successful collaboration is all about being collegial, and true professionals are naturally collegial and reasonable.  And, to come back to the point, when you've got such collegial, reasonable people serving as editors, it's not hard to defer to them when necessary.

Objection.  (From Cory Doctorow, citing Clay Shirky again.)  "...as Shirky shows, an expert-focused Wikipedia would likely devolve into interminable pissing matches over who was and was not qualified to be called an expert, because expertise isn't a measurable quantity, but rather something that is socially constructed."

Reply.  It hasn't so devolved yet.  While we've had many a polite dispute, I'm not sure I can recall a single "pissing match" between editors over expertise.  This isn't surprising to me.  Most experts are pretty comfortable in their expertise; they don't have to prove it to anyone.  It seems Doctorow assumed that editors would often try to settle disputes by citing their credentials, as Wikipedia's Essjay infamously did.  Few actual tenured professors would say things like, "This is a text I often require for my students, and I would hang my own Ph.D. on it's [sic] credibility."  True professionals rarely say such things, and that Essjay did should have been a dead giveaway that he was a fraud.

Objection.  Won't experts and authors be endlessly at war?  Isn't the idea of giving experts a "gentle oversight" role in an otherwise open Web 2.0 "bazaar" a recipe for social disaster?

Reply.  If this were true, we would have already seen some inkling of it.  But the fact is that we've seen very little author-vs.-editor conflict.  Disagreements tend to be editor-on-editor and author-on-author, and actually, we have seen very little acrimonious conflict, period.  Experts and people who want to work with them tend to be boringly, yet refreshingly polite, which is how we like it.

Objection. What about dealing with difficult users, whether editors or authors?  The typical personal attacks and other disruptive behavior will inevitably drive off editors, who have a low tolerance for such nonsense.

Reply.  This objection assumes that we will have many similar social problems to Wikipedia's.  This is unlikely and, in fact, we have had all such problems well in hand.  There are several excellent reasons to think that we won't have as many difficult users, and that they'll be easier to deal with.  Here's a brief run-down: we require people to use their own real names, which tends to make them behave better; we have strict rules against abuse and disruption, and constables aren't afraid to ban people because of it (we've already done so in a few cases); the project's editors set the tone, which has kept things fairly collegial; and contributors are required to endorse a Statement of Fundamental Policies, which requires would-be disruptors to recognize the authority of editors and constables.

Home