Wikipedia, Citizendium, Frankenstein and the golem.

[image by fd]Everybody knows Wikipedia so no need for introduction. Let’s just say that of the many very different opinions I have heard about it – placeable everywhere in the range from most celebratory to destructive – there is a quote that I find very attractive:

Wikipedia should not work, but it does!

This quote, which I believe is attributable to one of the fathers of wikpedia, Larry Sanger, says much about the on-line encyclopedia:

    Wikipedia is an experiment. A crazy one. Most people would have thought that the wiki affair would have been a failure but in fact it is now after only few years the biggest encyclopedia available nowadays.
    Nobody really knows how and why wikipedia works. This is due to the fact that very little attention has been put in a meta-study about wikipedia itself. I believe the reason for that is that even the hard core wikipedians didn’t expect such an enormous growth in such a little time and so underestimated the importance of investigating the growth itself. This is a pity: it is an example of how not to run an experiment. You always want to have the situation under control and know any given time how the variables are affecting your system. For this wikipedia fails. There is not a systematic study of who is writing wikipedia (with one notable exception), of who is reading wikipedia, what is listed in wikipedia and, most important, how reliable wikipedia is (except a well known preliminary study).

So, now the monster wikipedia is alive. There is no way to control it or to stop it, says someone: the community became radical, self conservative, self proud and especially close to external inputs. For this reason the only way to have wikipedia evolve is to act radically and, in the best tradition of monster movies, try to kill it: this is Citizendium. (pull)Citizendium (CZ) should work better than wikipedia for one fundamental reason: people who write on CZ should be better. (/pull)The problem, apparently, is that the great degree of anarchism on which wikipedia is built has made life to easy to those who are having fun spoiling the party (trolls and vandals) and to those who do not possess enough knowledge of the scientific method to participate to a scientific opera (see for instance the eternal debate of creationist, political active people etc who often cannot distinguish empiric evidences from opinions). The proliferation of this group would scare off people who are not used or do not want to deal with them and this would lead to a disequilibrium rewarding people you don’t want and scaring away people you’d like to contribute. For this reason Citizendium will try to attract and give more power to those who, theoretically, you should expect to see writing in an encyclopedia, namely the academics. So, the new rules proposed for CZ are the following:

    1. no anonymity, no nicknames, only real name. This is not a game, we are serious here. Take your responsibility and qualify yourself for what you are.
    2. to become an editor, you need to qualify as an expert. Of course debate is going on on what exactly is an expert

Will these two basic rules change the whole system enough to guarantee credibility and at the same time allow substantial growth? I thought of joining CZ simply because I believe that the WP-CZ affair has become too important to just sit and watch what others are doing (at least when one has actually the ability to model it and try to change things). I must say, I think there are many structural flaws in the actual project but on the other hand this is what a blue print is for so I will try to give my share of opinion.

  1. Actual contents of CZ
  2. Internal vs. external references
  3. Vandalism and real names
  4. Be creative and innovative in providing contents
  5. Keep a schedule and monitor yourself

1. Actual contents of CZ
CZ’s first characteristics should be: be believable, be serious. We want readers to know they can rely on the kind of information the find on CZ. This is a *huge responsibility* and should be always taken as first objective and never forgotten. CZ’s plan is to offer credibility using expertise of people in the academy. Now, for this very reason I think we should recognize that this takes away from the project that kind of knowledge for which you’ll never find expertise in the University. That means that CZ should not and cannot be a fork or a mirror of WP in the first place because we all know that WP is, on the other hand, excellent in providing details about extra-academic contents. There are, I believe, dozens of pages about every recent TV-series on WP, describing everything from the plot to the psyche of the characters; there are dozens of pages with biographies of Hollywood film stars etc. This pages (which are probably an important core of WP) should not be included. They do not belong to the expertise we want to offer and they actually arm us: we have no control on them, they take resources (human and technical) and give to CZ no advantage compared to WP (meaning, editorially is a bad choice)I already saw on the pilot project a list of biographies of movie stars. Why would I, user, need to go to CZ rather than WP for checking out details about Britney Spears?
2. Internal vs. external references
The recruiting process is still in a very alpha phase. I didn’t like what I have read so far because it seems to brake one of the fundamental rules of community building which is: never mix up things belonging to the different worlds. I will make this point clear ’cause is very important, although it should be well known to all readers here. CZ, WP, the free masonry, this ML, are all examples of closed communities. We want to transform CZ into an extension of the real world but this is a very courageous action: history should have taught us that different people behave in different ways when they are up to different activities. I could be a great researcher but the worst teacher at the same time (or viceversa) although my job requires both – I could do wonderful creative job but being completely unable to expose myself in a convincing way in a conference (or viceversa). Credentials should be always taken according to the job you want people to accomplish: here you are “hiring” editors asking them a CV for “academics”. This is like hiring a gardener based on his CV as cook (well, not as much but almost). Now, historically, closed communities have solved the problem of sorting people in two ways: a) using internal-internal references (i.e. experience acquired within the system, like ‘ordos‘ for free masonry or post experience for a webforum) b) using internal-external references; (i.e. an extension of a) where expert users take the responsibility to guarantee for new users). CZ proposal is to use external-external references (i.e. the CV) and this is no good. This could be done to filter access of people on a first place (*although I believe is a bad idea*) but then should be immediately integrated to system a). It is true that with this system we may find again in a situation where a full professor may be turned down by a grad student but the bottomline is that sometimes this *should* happen. If this never happens then it means something is wrong.

3. Fighting vandalism using real identities
Fighting vandalism is another keypoint of CZ. This is again a big responsibility because we all know that characteristics of vandalism is to give strong power to a small group of people, meaning that lowering down vandalism to 25% is definitely not enough. Vandalism effects follow a non-linear curve and 10 vandals in a party of 10000 people can make enough damage to spoil the party to everyone. It has been proposed to adopt real name as a way to override that. This may work only if there is a systematic control behind this policy (meaning check IDs, which is impossible). Asking people to self-certificate their name, age and to provide an email account is a joke. We all know that people tend not to use their own data on the net. Moreover, some authors may want to keep anonymity for good reasons and it would be a pity to keep the door shut to them. For this I recommend the following: CZ will require credentials from all editor and verify them, then editor will have the choice of being out with a real name or not. It should be enough for CZ’s authority to be able to claim that “the CZ staff counts n full professors, n researchers, n TA, n different universities etc.”. Authors, on the other hand, will not have to provide real names but will have to be selected according to the points listed above (internal-internal references or internal-external) meaning for instance that people should not be able to start writing and deleting during their initial, test-phase. This will slowdown things but will offer the following advantages: drastically reduce the number of troll and vandals; give people an idea of personal growth, a concept that gives motivation. I am aware that all this is not very liberal for a wiki project but (pull)this is not just another wiki project.(/pull)

4. Be creative and innovative in providing contents
Other ways of providing contents could be used. I suggest two, as of examples. One, DO involve schools. Think of thousands of schoolclasses that are given the duty (and the honor) to work together, with their teachers, to contribute with a number of entries. These entries will be a *draft*, perhaps already pretty good, of the real CZ. This initiative could be very useful for kids, for their teachers, for CZ (also in terms of echo in the media). Second, cooperate with already existing free accessible scientific journals (and not only). There are a few already out there (PLOS, PNAS, Genome Biology to name some famous ones). Most peer reviewed journals have reviews on specific arguments that are thought for the general readers. Some of them could perfectly suit as CZ entries. Once again, some journals would be very happy to share their resource for instance having a monthly or weekly shared articled with CZ. It’s good for them in terms of publicity, it’s good for CZ. I am sure even more general journals like National geographic or Scientific American for instance would be happy to ‘donate’ some articles to be entered in CZ.

5. Keep a schedule and monitor yourself
Know yourself, philosopher says. I find the biggest mistake of WP was to be absolutely ignorant of itself. Who really writes wikipedia? How? What is the credibility of scientific articles vs. political articles, really? CZ should think of a way to track everything down from the very beginning. This is the only way to find and correct issues as soon as they arise, this is the only way to grow. I’ll stop now. Already too much. Hope I have been useful. Last thing: what CZ wants from experts it is not merely their theoretical knowledge but their ability to apply a ‘method’. A scientific approach in researching and exposing contents.