Because we are human and thus, pack rats. We hold onto what we just might need, someday, just-in-case. Just-in-case what, I'd like to ask? Just-in-case that one bright-eyed student walks into your library, comes up to the reference desk and asks for that core title, written by a philosophy scholar sixty years ago and sitting on the shelf, never been used, just collecting dust. Frankly, we can do better. What use is that title sitting on your shelf for 20 years, when we should get rid of it, purchase materials users really want (maybe let them make the purchase requests?) and keep a constantly evolving and changing collection; as the needs of the community dictates. A "balanced" collection is really a fantasy; does anyone really have a balanced collection? I think its a philosophy that every library school student is taught and is supposed to emulate once they get a bibliographer gig, but it isn't practical and it isn't, well, to be honest, it isn't useful if the books are never being used. Why are we so afraid to weed and to order relevant, useful materials that students and faculty want? A lot of it has to do with our identities as librarians and territoriality. I'm the only one who knows what the user needs and I alone know what is best or relevant for the collection. We were the experts and our expertise was counted upon to find and acquire resources. This may have been accurate back in the day when materials were scarce, ILL was slow and your local collection had to be comprehensive, balanced and "just-in-case". In the age of Amazon, ILL turnaround of 3 days for books and 24 hours for journal articles, multitudes of electronic resources and databases, this isn't true anymore. So, why are librarians still afraid to weed?
Some of it has to do with faculty and the perception that what we are weeding is unique and invaluable and thus, shouldn't be discarded; in fact, much of it isn't unique or valuable, but commonly held and if published before copyright protection sets in, may be in the public domain and available in full-text online from one of several digital repositories. And yet, we still hang on to the books, not able to weed them from our collections, even though they aren't being used, most of them are irrelevant to the curriculum, research and scholarship occurring on our campuses, and duplicated in libraries all over the country. Our main weeding project this summer is the removal of 31,000 titles from book storage off-site, which is being re-purposed by our campus in the fall. Five years ago, these books were transferred from our main library to the storage facility, with the following criteria in mind; books published prior to 1959 (1979 for science titles) and had not circulated in the 18 years since our circulation system first went online. These titles would remain accessible through our OPAC, and users could request them through IDS. If an item was requested, it was brought back to the main library and the location was changed back to the main library in the catalog.
For this project, thanks to a great tool designed by our systems administrator (called the GIST De-Selection Manager) I'm able to run a de-selection report on the 31,000 titles in our off-site storage. The report, obtained by running OCLC numbers against a WorldCat API and several other APIs, returns holdings information for IDS Project libraries, NYS libraries which do OCLC ILL, full-text availability in Hathi Trust and Google Books, as well as Amazon prices (if available) and a recommendation to weed or keep based on a pre-determined conspectus ranking. What it does is make the weeding decision much easier for librarians; for a major weeding project, it streamlines a majority of the labor-intensive work in checking our holdings against WorldCat, title by title. Because the GDM also uses conspectus data, I can tailor our preferred conspectus holdings and the GDM will include this information in its final recommendation to keep or weed dependent on whether or not it is a subject area we want to grow or to shrink. While GDM is still buggy (since a lot of the items are pre-1959 and have no ISBNs, some of the APIs have trouble returning results since they are searched using ISBNs, etc.) it has enormous potential for freeing up the decision-making of librarians when it comes time to weed. With data in hand, I can easily make weeding decisions and feel confident to remove materials from the collection.
Its time to take the data and make evidence-based decisions; not subjective ones based on just-in-case. Having the data will go a long way towards feeling confident enough to make the decision to weed.