Menu Home End users Hacking Humour Interviews Opinons Reviews All articles Issues Books
Free Software magazine
Gary Richmond [opinions]

Google's Knol: it's a Wikipedia Jim but not as we know it [wikipedia] [google] [citizendium] [squidoo] [helium]

The launch of Google's Chrome has created a frenzy of online activity (just Google it and it will return in excess of fifty one million results), including mine. and already the world and his wife has been busy publicising tips, tricks and hacks. There is absolutely no doubt that Google is very serious about its new baby. They hired no less than four Firefox developers--Ben Goodger, Pam Greene, Darin Fisher and Brian Ryner. Enough said. It wasn't dreamed up on the spur of the moment as another speculative product of the Summer of Code. Can the same be said of Knol? What is it, how does it work and more importantly, does it conform with the principles of free software and is it a serious challenger to Wikipedia?

What's in a name?

Knol? Unfortunately, every time I see or hear it I am transported back to Deely Plaza or the Texas Book Repository

Shall I compare thee to a Knol? Hmm, perhaps not. Wikipedia sounds just right. Memorable and serious but not too serious. Of course Wikipedia is now an established "brand" and it has a big headstart on any competitor. Just like Google's own search engine. If it is going to position itself ever to rival Wikipedia perhaps they should be thinking about a more pithy name. Knol? Unfortunately, every time I see or hear it I am transported back to Deely Plaza or the Texas Book Repository. However good the product, many have been done for because of poor marketing. This is perhaps a quibble. If a product is good enough it might survive uninspired marketing.

Knol is short for "unit of knowledge" and anyone can create their own knol. If there is one thing which characterises the free software model it is that it is collaborative, whether it is a wiki, a distribution or the development of a piece of software--or Wikipedia. Of course there is nothing to stop you acting as a one man band, but knowledge is often advanced and improved when the effort is voluntary and collective. There the resemblance with any kind of Wikipedia-killer ends. Knols are individual creations and the creator of a knol retains ownership of their creations and controls any moderation. That is ensured as you have to log in via your Google account to edit anything. Hacking aside, anonymous editing is verboten.

Wikipedia is not anonymous anyway. A Caltech student called Virgil Griffiths devised software called WikiScanner which reveals, via IP addresses, who edited what on Wikipedia. Some worrying stuff. Virgil is no shrinking violet or wallflower. He doesn't pull his punches and gave a speech at the Hackers on Planet Earth conference titled "Wikipedia: you will never find a more wretched hive of scum and villany" available as am mp3 download (third article from the bottom of the page). That's telling 'em Virg! He will also be releasing WikiGanda, another piece of software, this time to reveal edit wars between two factions. All the tools are available at wikiwatcher.

Wikipedia without links would like a dog without a bark, but you look hard for them on a knol. You may also find that there will be multiple knols on the same topic by different authors, possibly containing disparate accounts. That might be a reaction to the furore caused by uncontrolled editing in Wikipedia which threatened its integrity. Alas, Knol goes to the other extreme and at the same time creates multiple "locked duplication" which is not conducive to a one stop shop like Wikipedia. If this is the basis for Knol, then Google's baby is in trouble.

It's the advertising

Since Google is relatively friendly to free software and the principles of open source, it is unlikely that they do not understand the philosophy and practice of the model. So, why have they designed Knol as it is? The mantra of Google may be "do no evil" but it is nothing if not remorseless and sufficiently flush with funds to launch all sorts of speculative projects and have most of them fail. If just one succeeds...

The answer to the above question lies in one word: advertising. People who write for Wikipedia, whatever else their reasons, do not contribute for profit. Google's business model is advertising. Plain and simple, so it is not exactly surprising that people who write knols will have the opportunity to share contextual advertising revenue. You don't have to be prescient to see that this model could encourage a commercial bias towards knols whose content would be favourable to revenue-generating advertising. The problem with that is that it may skew the type of article you can find and if it is obscure or highly technical chances are you'll find yourself back at Wikipedia. Perhaps that is why Cedric Dupont, a Google product manager, has described knol as "a supplement" to Wikipedia.

a lack of reciprocity concerns the fact that all outbound links on knol are nofollow

That said, I have no doubt that Google eye Wikipedia with considerable envy when they view the page hits and boy, would they ever like to monetise that kind of throughput. To that end, what is to stop any knol contributor taking content in Wikipedia and simply copying it wholesale? Well, knol has been released under the Creative Commons by attribution (CC BY) licence which means that knol content could be use in Wikipedia but this appears not to be reciprocated. Wikipedia content cannot be used in knol as knol does not include a copyleft licence. Another and equally important lack of reciprocity concerns the fact that all outbound links on knol are nofollow which means that while you can link to knols and improve their Googlebot authority, knol does not pass this authority onto other sites. That surely infringes the whole ethos of Creative Commons. This is a moment to remind Google that they have committed to "organize the world's information and make it universally accessible and useful" and the problems with nofollow and Creative Commons militate against that intention.

If you are determined to use it one enterprising person has created a kind of knol tracker service to keep abreast of knols being added with Knol RSS Featured Feeds and archival searching. Some of the content is so trivial ("how to unblock your toilet", "super soft plastic baits for fall") that describing it as a wikipedia would be ludicrous. Blogs wold be a more honest description. If Google think they will grab more market in search results where reference content is concerned by offering knols then they need a reality check. Besides the problem is, to some extent, self created. Google had a problem with spam some years ago and they solved the problem by realising that most searchers rarely ventured beyond the first three results, and they decided to promote Wikipedia. This has been amazingly good for Wikipedia's page ranking. Udi Manber, Google's Vice President of Engineering, hasn't bothered to conceal the fact that the purpose of Google knols is to become the first entry on Google's search results pages, but I just do not see it dislodging Wikipedia anytime soon. However, it might only be a matter of time before a term like "knoljacking" becomes commonplace vernacular.

Google may have the secret algorithm for page ranking, but if it uses that in conjunction with knols to skew results it may get its fingers burned by consumer reaction. Perhaps at that point users would switch to another search engine, Jimmy Wales' nascent search engine for example. Or, you could try Citizendium founded by Larry Sanger, the co-creator of Wikipedia. Sanger claims that Wikipedia has achieved great things but that he wanted to create a dependable encyclopedia. Like knol, Citizendium requires you to sign up for an account using real names and create a culture of peer-reviewed responsibility. And due to the generous nature of the Wikipedia licence Citizendium forked its content and ported Wikipedia's entries (in excess of one million) though this was later deleted in favour of a building an encylopedia from the bottom up--without prejudice to any one incorporating information from Wikipedia. It's ethic is for non profit and it attempts to merge "open source hacker culture" with the culture of academia, teaching the latter about the collaborative ethos as outlined in Eric Raymond's "The Cathedral and the Baazar". Unlike Google knols, Citizendium has no ulterior motives to do with page ranking or monetizing content by way of advertising. As they say on their own website:

Not in it for the money? Why not? Because that's the only way the Citizendium can thrive as a project that is at once reliable, guided by experts, and maximally welcoming to the open source community. We believe that volunteers--both from academia and from the hacker world--will refuse to contribute to a Wikipedia-style knowledge project if it merely lines the pockets of profit-making enterprise. Besides, the best way to ensure freedom, independence, and neutrality of information is to make sure that the information does not depend on any particular vested interests--that the content itself is ultimately in the hands of a responsible online "republic of the mind."_

That could serve as a definition of any free software project or distribution. "A republic of the mind". I like that. Of course there is no real point of comparison between Wikipedia, Citizendium and Knol. Knol is better described as a library rather than an encyclopedia, but in its present state of stock I wouldn't expect to find too much. There are many empty shelves and they may never be filled. It is early days of course and Wikipedia didn't emerge into the light of day as a fully stocked encyclopedia either. Even so.

Knowledge wants to be free but it's not always democratic

Given the fecund hothouse of ideas that is Google I find the knol format, frankly, dreadful. No central source of information, links to adsense which could act as an incentive to contribute articles of less than original quality and authorial subjectivity. This latter feature is directly at odds with one of Wikipedia's central tenets: Neutral Point of View (NPOV). This is "non-negotiable" and the standard for all articles but knol permits authors to submit personal opinions and control their articles and that is the very antithesis of community development of knowledge. Of course community is a very overused word in this context. In reality the bulk of contributions and edits is the preserve of the few but you can become one of the few. It's up to you. Otherwise, as Wales himself has acknowledged, it is very much a case of the 80-20 rule. That is, 80% of the work done by 20% of the users. Google's knols only make sense if the are really conceived and viewed as a library composed of individual books representing a personal credo. It might be more accurate to compare Knols to Squidoo, Helium or WikHow.

Wikipedia isn't perfect either, but if you are not contributing you can't very well complain. Total democracy is an illusion. Even the Debian project has to have "leaders". There will always be those who make a disproportionate contribution for whatever reason, even when it is by default or accident. If you are a parent (I am not) you will be familiar with reverse delegation from your children. They want this, they want that and they promise the earth in order to get it but of course you end up doing it all yourself. Sound familiar? The same seems to apply to the world of free software projects and a wikipedia. No matter how painfully and exquisitely democratic they attempt to be, an individual or group of people end up being at the core of a project. There is a limit to consensual leadership. Even the editor of Freesoftware Magazine agrees with that--and he will be editing this article.


Knol is a stinker. It is misconceived and badly executed

Google is a lot of fun. For all the valid criticisms about security, privacy and potential monopoly the internet is probably a better place for Google being there. It could have been worse. It could have been a Microsoft product and the needle on the fudometer would have gone through the red and off the scale. Google have the resources, technical and financial, to launch an armada of projects. They can afford a high failure rate. For every Google Base there is Google Chrome, for every Orkut there is the Android phone. You have to speculate to accumulate. However, Knol is a stinker. It is misconceived and badly executed. Google's motives have nothing to do with the disinterested pursuit of knowledge for its own sake and everything to do with monetizing content and pageranking. Free software would be ill served by it. So, Mr. DiBona, I'm afraid that it's a matter of drag and drop the knols into the recycle bin. If you don't know how you could always look it up in Wikipedia. Just a thought.

I don't have a crystal ball and if you cross my palm with silver I couldn't guarantee to predict the future but I will stick out my neck and make one prediction: if Google knols raise a real and serious challenge to Wikipedia then I declare publicly that I will abandon GNU/Linux for a year and run Windows Vista on my aging desktop computer with its 400MHZ Celeron processor. So there.


Verbatim copying and distribution of this entire article are permitted worldwide, without royalty, in any medium, provided this notice is preserved.