Monday's rant on knowledge obsolescence and the valuation of technic to determine its long-term worth apparently struck a cord with folks--and not necessarily librarians (who have been dealing with or avoiding this question since Alexandria burned).
I hadn't thought much more about it until this morning, when I participated in a discussion about library policies on purchasing ebooks. Though I am by nature and training a printophile, I'm certainly not a Luddite, and the idea of providing access to core textbooks online is one I wholeheartedly embrace.
But the concepts of weeding and permanent retention are, if not totally alien, at least very new to the wild west that is the web. Lately, there have been many large-scale, highly-publicized efforts to bring some stability to digital resources, all of which are Very Good Things for those of us interested in keeping information accessible.
But what happens when the very nature of the digital objects you're trying to save, changes? Take ebooks: more and more of them are becoming what the library world has taken to calling "integrating resources." Which means, there is no 11th edition, 12th edition, 13th edition. There's just the current edition. Which is great for students and teachers. But what does that mean for libraries like ours, who have traditionally kept all editions of the classic textbooks? Older editions show us the development of medical knowledge in a given area, but also shed light on changes in medical education and dissemination of information. Once publishers stop putting out any print editions at all (which will happen, for sure), where will we go to track those trends?
This problem has already begun to plague us here in a very local sense: the last course catalog for the School of Medicine in the collections is from 1993-1994. Not long after that, the school went to online-only course descriptions and ceased printing catalogs. The School of Nursing still prints catalogs (Reason No. 8 why we love nurses), but anyone fifty years hence wanting to track changes in the medical curriculum here will need to go to other, more diffuse, sources.
OK, so let's say we decide that we can do some collection development of web-based resources and we've solved the problems inherent in maintaining those in usable formats (which are legion, in my opinion). What happens if the chain of ownership is broken? Curation ends; data sits, unmigrated for some period of time. Will a computer drive uncovered a thousand years from now still be decipherable? Sure, they just figured out the Antikythera Mechanism, but you can see that with the naked eye. Maybe I'm being pessimistic, but I think these considerations of what's worth saving are more important now than ever--the opposite of what you'll hear from a lot of digital gurus, who think that unlimited data storage will end the need for any kind of concerted weeding at all.
Can you imagine nothing ever getting tossed? That might be worse than the alternative....