I just finished reading an account of the long battle faced by Abraham Morganteler, a clinical professor at Harvard, to convince others in the medical community that testosterone treatment did not increase risk of prostrate cancer. Turns out that despite decades of teaching otherwise, the ‘basis’ for this claim was a pair of old studies, one flawed, another poorly reported, that most doctors had never read. When conference audiences and publishers gave Morgantaler’s initial contrary results a dismissive but not unscientific response, arguing that his sample was small (true), he gathered more data until this objection could no longer be levied. But the real breakthrough came when he visited the HMS library and in it’s darker recesses pulled out the original papers which had established the original recommendations of the medical profession. Surprise, surprise, the original 1941 piece was built on a sample of three, and two of these were confounded, leaving it to provide but a single data point. The second, a 1980s study, presented what appeared to be solid evidence in the summary which actually was not supported by a closer reading of the full text. And you thought medicine was a science, right? Of course, in many fields there exist ‘classic’ findings which become standard fodder for text books and which students and subsequently professionals never read in the original – such is the price of the explosion in publications. So while it would pay us all to check our sources, what we really need is a tool to help us mine the scientific record intelligently. In the absence of this, we might want to remind people of the importance of records, archives and the academic library.
V.M Daniel Regents Professor of Information
I’m in the process of starting a social experiment via the web, and am in need of someone that can write really well. The details of the project is laid out on my website at http://www.accreation.org. I’m looking for someone that can edit the content in a more easily understandable format. Please visit the site, and contact me if you or someone you know might be interested in participating in the project. Thank you.
PS. This is NOT a paid position; I’m looking for like minded individuals who are interested in making this prjoect become a success. Thank you.
While digitizing and indexing old scientific papers should make them more accessible in theory, in practice citation counts for most papers more than a few years old approach zero. (Well, at least that’s the case for my papers, so it must be true for everyone else as well.) Part of the answer must lie in education: we must not only create the water source and lead the proverbial horse to it, but also teach it to drink.
No one can disagree with the fact that the Libraries are always an integral part of the civilized social world, whether those are general libraries, Medical libraries, music libraries or any other. Even personal websites are a good source to keep the data intact.