Delirium- It's not so much an Internet/paper distinction but a primary/secondary sources distinction. Errors being promulgated from one book to the next happens on paper just as it happens online. A personal webpage may well be a much better source for vital data than a biography written by a third party, and that in turn may be a better source than a book that briefly mentions the person in context. Also, of course, the Net has no barriers to publishing and the motivation for doing so is often much different than on paper. So when dealing with Internet sources we also need to consider the motivation and the nature of the source more closely, and the more suspect they are, the more important it is that they do their homework in terms of sources etc. But we should not be snobbish towards any source. There is a trend in the mainstream media to portray the Internet as a gigantic rumor mill with little redeemable qualities - it's the competition, after all. The reality is that its very openness has allowed vectors of research that were almost non-existent before. What the Net lacks are universal mechanisms of rating, public feedback and trust, largely because all attempts of building them (Alexa, epinions etc.) were centralized dot-com schemes and most open source people don't understand usability. As for certification, in the long run we will want to certify an article on several levels -- factual accuracy, quality of writing, comprehensibility etc. But as I said, let's move slowly towards the perfect scheme. Regards, Erik