Wednesday, December 14, 2005

Wikipedia at crossroads

I have always wanted to talk about my fascination with Wikipedia at some point. Having been a moderately active contributor for over a year and a half, and a content administrator for sometime, I have had experiences, both pleasant and unpleasant. I have marvelled at the way that a totally open encyclopedia is still able to make sense overall. Helping to build articles from scratch, and seeing them come up excellently is an exhilirating experience. Well-written articles not only are featured on the main page and gain superbly in Google ranks, but are replicated at various mirrors all over the web, and also available for citations and references in various academic works around the world. Believe me, it gives you the kicks to see your work get that level of footage. However, I have also been party to ugly disputes, which have left me high and dry. Building consensus and objectivity on politically sensitive articles is not a very easy thing, and unless you are very dogged, patient but also open-minded, there is a very good chance that you would lose your steam quite early.

In time, I lost touch with Wikipedia, but a recent controversy has rekindled my interest in the concept again. I am talking about the Seigenthaler incident which has indeed shaken Wikidom a bit. It all started when a prankster edited the entry for John Seiganthaler Sr. and suggested that he was at one point investigated for the assassinations of John and Robert Kennedy - a false allegation. Probably by suggesting that "nothing was ever proven" along with the insinuating lines, the prankster got away with the edit, escaping the eyes of the usually watchful administrators, who are constantly on the prowl to revert acts of vandalism. The edit went unnoticed and unchallenged for four long months, until a close associate of Seigenthaler fished it out and brought it to his attention. Some furious calls to Jimbo Wales, founder and the "benevolent dictator" of Wikipedia later, the editions were removed.

But, the damage was already done. Seigenthaler went on national television and other popular media to discredit Wikipedia as a "flawed and an irresponsible research tool". He also went public on USA Today, with a scathing attack, suggesting that Wikipedia is a tool that can be used to promote evil gossip, and also expressing discontent that there is no legal recourse to compensate for damage in such an environment. This had an avalanche effect of convincing other mass media players to pounce on to the story, and over the last fortnight, several newspapers and television channels covered the incident. Wales and Seigenthaler had an interview with Kyra Philips on CNN, not much later. There have since been unconfirmed reports of memos being circulated at places that work with information (like news agencies), not to trust Wikipedia as a reliable source of information.

Jimbo, on his part, took the unusual step of not only reverting the article, but also of deleting the edition from the edit history log, so that the damaging lines for all practical purposes, were "truly deleted". He has also announced a significant policy change of not allowing unregistered users to start new articles. When Jimbo Wales acknowledges the incident as a significant setback, it really things to perspective, and is a good indicator that Houston sees a problem. The incident has not only raised questions on Wikipedia's authenticity, and its legality, but also on if an open source encyclopedia is a fundamentally feasible concept.

In my personal opinion, the incident itself is a minor aberration, and there is no refuting the fact that Wikipedia is simply the most complete and largest repository of information or pointers to information on the web. The fundamental premise of Wikipedia is provision of information that will never be perfect in content, and it works under the best-effort . The difference between Wikipedia and other encyclopediae is that the former does not have systemic errors. There are no overpowering vested interests that would try to shape opinion one way or the other. In other words, the system being open, is designed to eliminate systemic errors. However, that comes with the cost of not being able to avoid random errors like the one that has caused the recent furore.

There is a reason why there is a big gaga over Wikipedia. Wikipedia takes the open source software model to the next level. The kind of knowledge base you need to attract is a million times more complex and diverse, that any restriction on who can/cannot contribute would have defeated the purpose. From the very begining, Wikipedia has strived to go for volumes - more information here than at any other single place in the web. And as long as you are a user with an internet connection, nobody is going to stop you from starting/editing any page on the web. You do not need to be a regsitered user. All you need is an IP address. If you are behind a proxy/firewall/dialup, that doesnt even say anything about you. Wikipedia has opened itself to such a diverse community of people around the world in close to a hundred languages, that if it works today as a credible source of information, it is nothing short of a miracle. And a working concept like that is a very credible brand ambassador for the entire open-software/free-content community.

Wikipedia, as a free content project, is by design nutured and sustained by criticisms. The single, largest promise that open-source/free content projects provides us is the refusal to go into the denial mode. We simply refuse to build cathedrals. If necessary, we should be willing to tear down anything that remotely seems like an edifice that cannot be sustained over the next few years. If necessary, we should start from scratch, change the way we have designed the system.

It would sound counter-intuitive, there are enough reasons for us to be happy about the Seigenthaler incident, because this is an oportunity to think of this and a few other issues about the way Wikipedia operates presently.

What are these issues? Is Wikipedia still workable? (Ok, you must have guessed that I do think it is workable ;-), but why?)

More on that in the next post(s)...

9 comments:

PS said...

You were the one who introduced me to wikipedia, long back.
Since then I've been a bigtime fan, edited/created a few pages, and I use it almost as much as I use google.
So this Seigenthaler incident was indeed quite sad.
I've been planning to blog about it for a while, will do soon...

PS said...

Further corroboration, this time by Nature
http://www.nature.com/news/2005/051212/full/438900a.html
It compares Wikipedia and Britannica and shows they are just as accurate.

sanchapanzo said...

RL,

Any idea who funds wikipedia?

When a newspaper does the same thing(i.e publishing faulty news), the affected party pounces on the newspaper to apologise and files defamation suit requesting astronomical amounts(dollars).
Now, if the same crime is committed by a portal(like wiki) who is going to be liable for the crime?

I think what this incident has done to internet forums is to 'make these forums accountable'.

Whoiscb said...

machi,..it is still pretty usable...so who cares...

LUCKY said...

Nice one RL.. great post... wonderful read.. shows ur involvement too..

KoPoS said...

Wasnt the previous incident an aberration and not a norm?

Anycase, Wikipedia is my online journal of knowledge it has never diasspointed me. You might want to post a link to recent comparisions between the Britannica and Wikipedia; where both were found to be equally good and bad.

Moreover, with community led projects as these, there is always a lower level of mistakes in the text because of policing by all the experts involved...

One of the links
http://www.freedom-to-tinker.com/?p=675

KoPoS said...

The official b vs w carried out :)
http://www.nature.com/nature/journal/v438/n7070/full/438900a.html

m. said...

hey!

ok... i've written something after reading what you've written, and have therefore linked to this post.

the argument continues eh :D

triple distilled said...

dude gimme your IM id da