The Longest Now


Citizendium: failure to thrive, in search of peace
Friday July 23rd 2010, 8:59 pm
Filed under: chain-gang,fly-by-wire,metrics,wikipedia

After early months of interest and glory — peaking in a spike in mailing list traffic that was moderated for being too active — Citizendium’s growth all but shut down levelled off and has declined steadily since 2008.   Now it is looking for a long-term home.

I have mixed feelings about Citizendium.  I was excited about it in 2006 — at first blush, it offers a serious alternative for expert editors who want to contribute to free knowledge but feel unappreciated or unwelcome at Wikipedia.  And in general, compatibly-licensed alternatives to Wikipedia are a very good thing – the whole point of using free licenses is to encourage reuse.   But to succeed on the scale of its original dreams, Citizendium must overcome its insularity and make good on its core promise of quality.  Not unlike Wikipedia, it is currently known as much for its humorous highlights as for its best work.  And it faces the same problems with difficult and misguided editors — some who have quite solid credentials — only with a much smaller community to handle that workload.

I still hope for a proliferation of cousin projects, all competing to find the best way to spur collaboration around free knowledge.  There is so much to explore in the way of how to create welcoming communities for different audiences of writers and creators.  Community atmosphere, and a limitation in the types of knowledge that can be easily shared, are among Wikipedia’s major bottlenecks.   It is welcoming to a narrow[ing?] audience, and if this does not change it may face its own dramatic slowdown in participation – more joyful models are welcome.  (My recent favorite, in style, tools, and atmosphere: fotopedia.)

The questions that inspire Citizendium remain:  How can we expand collaborative production of educational works to topics that require rare expertise in a field?  How can we verify new works as quickly as they are produced, and how much does this speed depend on the commonality of the knowledge involved?  

Verification processes are time-consuming, as the slow but steady output of Nupedia, CZ, and even Veropedia show. Since 2006, CZ has produced roughly 150 verified articles (almost triple Nupedia’s output and, well, 150 more than Wikipedia’s own). The featured article and peer review processes on various language Wikipedias are likewise nororiously slow.

CZ and others try to accomplish this by raising the bar for personal credentials of contributors, and increasing the personal responsibility of a group of meta-editors for the quality of work in a topic.   Some common dilemmas:

Verifying expertise is difficult without it.
Experts face demand on their time from many projects.
Past expertise is no guarantee of future quality of work.
Professional reputation can be tied to a particular theory.

And a dilemma special to Wikipedia’s commitment to NPOV: experts often have strong opinions about which theories are right and wrong in their fields.   How can they contribute in peace to a discussion whose end result will not take a position on which is right?

Two thoughts:

  • One barrier to participation is the qualification expected of reviewers. We could learn much from how Law Reviews are published, I expect, since the field of Law is unique in depending on its students, still pursuing their degrees, to oversee and produce the most distinguished reviewed publications in the field.
  • Another is the inflexibility of a “yes/no” review system.  Less permanent and reversible ways to validate information can be based on guidelines for fine-grained citation and annotation, and a visible place for review and analysis of a text linked prominently from it.  Moreover a review process that formally places works on a spectrum of completion and verification can offer more useful and detailed information than a stamp of approval.

These ideas draw on work by the current assessment process used widely on the French and English Wikipedias.  I would be interested to hear thoughts from people familiar with law reviews and other large-scale review processes, or with the CZ verification process or that of other educational wikis.


9 Comments so far
Leave a comment

You know, S.J. Klein, you might wish it were true that CZ has “all but shut down.” This is a lie, and a damaging lie, because your saying it might actually cause other people to repeat it. The fact of the matter is that the Citizendium community has continued on in my absence with about the same number of edits per day, and about the same article output, as it did a few years ago. Sometimes it’s higher, and sometimes it’s lower.

Wikipedia’s “verified article” output? Zero. The number of CZ articles that are “articles” in the sense used by Wikipedia? Over 14,000. And our traffic continues to grow.

I know that my criticisms of Wikimedia’s vast porn holdings and truly disgusting drawings of child molestation–still guarded as of “historical interest”–have really cut you to the quick. But these criticisms have the virtue of being accurate. Your attempts to write our obituary do not.

Comment by Larry Sanger 07.24.10 @ 2:18 pm

A lot of Wikipedians–including me, and I think sj as well–genuinely hoped Citizendium would turn into a significant rival of Wikipedia, particularly after the decision to go CC-by-sa. Competition with a strong Citizendium would have been good for Wikipedia’s mission, and for Wikipedia itself most likely.

But while “all but shut down” might be a bit of an exaggeration, “failure to thrive” is a fact, and the graph of activity over time does seem to be a general downward slope since the early publicity and the period of open registration: http://en.citizendium.org/wiki/Image:Active_users.png .

Comment by Sage 07.24.10 @ 4:01 pm

Sage is correct. I still hope that CZ and projects like it become effective rivals to Wikipedia, pursuing the same goals with different ideas about how to cultivate community and how to focus creative energies. New project innovations, and learning from the mistakes of old ones like WP, are both important.

I revised my original comment: I thought the activity spike in ’07 had been higher, and the change has been gradual since then. I worry about any major project that fails to thrive, including some of the smaller Wikipedias. We must acknowledge and talk about the problems they face, if we are to overcome them.

In the case of smaller wikis, rather than demanding that each survive on its own, we might find better ways to connect communities through shared interests. Early wikis thought a lot about this level of community building, and how to work across community boundaries. Perhaps there is something to be recaptured from those ideas.

Comment by metasj 07.25.10 @ 11:49 pm

The more interesting accusation against Cz is that their policy of depending on expert credentials has not protected them from partisan point of view edittors. I had suspected this might be true but haven’t got the time to do the detailed research on Cz to prove it either way. I suspect Wikipedia’s policy of requiring a reference for any disputed fact (rather than having a previledged class of editor who can just claim “It is so because I say so”) will turn out to be the more effective policy in the long term.

Larry Sanger worrying about people repeating stuff is a bit 2-faced after the half truths he spread about wikipedia (and repeats here – Wikipedia does not have “vast porn holdings”) but this whole correspondence does have a little bit of the smell of the academic version of a flame war where cries of “your mother” are replaced by concerned musings about problems with “your methodology”. This is undignified and the two of you may wish to reconsider if you really want to continue this correspondence in public.

On the other hand – if you should feel the need to continue … .

Comment by filceolaire 07.26.10 @ 2:51 am

“their policy of depending on expert credentials has not protected them from partisan point of view edittors.”

Quite the opposite – cranks are good at credentials and status, because that’s all they have. So credentialism will eventually clog with the paper-qualified at best, and the homeopaths at worst. Citizendium’s pseudoscience output shows that the pure credentialist approach does not work – not all CZ contributors are cranks, but the cranks have gained and used authority to keep their articles pseudoscientific and a net subtraction in the sum of the world’s knowledge.

And this credentialism has actively repelled the actual academics and experts – because experts of substance care much more about their field than they do about credentials. And CZ actively driving away academics is well-documented, c.f. the two articles in THE. Dr Sanger continues to blame the malign efforts of Wikipedians for this, of course, rather than systemic problems in the way he set it up or the behaviours and attitudes of himself and his constables.

Larry, the Wikipedians really really wanted CZ to work. lt didn’t.

Comment by David Gerard 07.28.10 @ 8:02 pm

Larry, look at it this way by comparing a few things:

Samuel Johnson
http://en.citizendium.org/wiki/Samuel_Johnson
vs
http://en.wikipedia.org/wiki/Samuel_Johnson

No contest.

http://en.wikipedia.org/wiki/Elegy_Written_in_a_Country_Churchyard
vs
… You don’t have one or even one for Thomas Gray.

Same goes for Ode on a Grecian Urn, Kubla Khan, or Ode: Intimations of Immortality.

http://en.wikipedia.org/wiki/Ode_on_a_Grecian_Urn
http://en.wikipedia.org/wiki/Ode:_Intimations_of_Immortality
http://en.wikipedia.org/wiki/Kubla_Khan

Those pages are not only superior to what you have on the topic, but superior to the quality of pages you have on some of your best topics.

The problem with Citizendium is that you never bothered to recruit decent writers who are both experts and actually know how to write encyclopedic pages. Why would the guy who wrote [those articles] still be writing for Wikipedia even though the community got rid of him? What is it that Jimbo has that you don’t have, or that the WMF Board does that you don’t do?

Here is a little hint. Even though Jimbo is seen as a “godking”, he has actually bothered to praise, to support, and to encourage. The same with Sj and other board members. They recognize good content and bother to make an effort to encourage it.

The reason why you fail, even though I am 100% inline with every single philosophical view you have, is that you don’t even bother to be personable. Some could call it arrogance, but I think it is just a lack of realization. It would probably do you good to go through some of my best pages and copy them over. You have my expert credentials already so you don’t even need to worry.

Comment by Ottava 07.30.10 @ 7:57 pm

What Wikipedia has – and nobody else has, not Citizendium, not Wikinews, not any other project – is The Blessing Of Google. Why this is so, I don’t know, almost nobody does. But all discussions which do not take this fact into account miss the Leviathan in the room.

Comment by Seth Finkelstein 08.06.10 @ 12:12 pm

@ Ottava: I don’t understand how giving encouragements would significantly boost participation at CZ. I don’t like Jimbo for playing the cult leader. You can’t ever get in touch with him, unlike \Dr. Sanger.

Comment by C Park 09.14.10 @ 3:08 am

@Seth – Wikipedia has strong internal links, strong inbound external links, and a tremendous corpus of material. Those are always a recipe for high Google rank. I wasn’t aware that was a mystery!

Comment by metasj 09.19.10 @ 8:49 pm



Leave a comment
Line and paragraph breaks automatic, e-mail address never displayed, HTML allowed: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>




Bad Behavior has blocked 396 access attempts in the last 7 days.