Yesterday afternoon, I spent some time in the stacks of my library looking for books to remove from our collection. The euphemisms used to describe this process make me laugh: librarians weed books from their collections or deselect them. Yes, this means that sometimes books are recycled or thrown away. In my case, the books we weed are shipped to a used book service to sell or recycle as demand suggests.

Removing books from a library collection often courts controversy, but there is good reason to think critically about a library collection and to selectively remove materials.

The number and type of books removed from a library collection is dependent on a few things:

The type of library. My library serves a small, public liberal arts college. We are not a research library, and we exist to serve the curriculum of our college. We try not to buy books unrelated to our curriculum, and we rely on faculty to select books that are the most relevant. One result is that books related to discontinued programs are candidates for weeding. A few years ago we finally removed a lot of books that had been used by a library studies graduate program at my school that closed in the early 1980s. We recently closed several programs for budgetary reasons and there may be a point in the future when it makes sense to weed some books on those topics.

The size of the library. We don't have a large library and it is full to the brim with students working on papers and projects. With students constantly clamoring for space, we need to be strategic in how we use our limited square footage. Many of our print journals are now in closed storage, for example. And just yesterday, I spotted 9 shelves that held 18 different editions of a six volume print reference book. Do we need all 18 editions? No. Let's keep a couple and remove the rest.

The type of book. Many of the books I pulled off the shelves yesterday were how-to books for specific computer programming languages and applications from the late 1970s to the mid-1990s. These books become out of date pretty quickly.

Books are for use. "How-to" style information about programs that no longer exist are not useful. I held on to one particular book about statistics from but withdrew the 1986 companion volume about how to use Minitab 1 on a DOS IBM computer.*

The continued accuracy of the information. Scientists learn new things everyday that render previous books and articles on a topic out-of-date or simply incorrect. Yesterday I pulled a book off the shelf about how to conduct radiometric dating published in 1954. There have been major advances in the topic in the past 60 years and we have more up to date information available on the shelves.

How often a book is used. While usage is not the only criterion for removing a book from a collection, it can be a useful datapoint. Let's say that I pull a book off the shelf with the intent to remove it from the collection because it is woefully out of date. If a usage report suggested that the book was still getting a lot of use, I might return it to the shelf or find an updated book on a similar topic. Alternatively, we may hang on to books that are seldom used because our copy is the last copy in the state, or because a faculty member suggests that the volume is a key part of the scientific literature in that field.

Removing books from the collection is a complicated process that requires input from a variety of people: statistics and usage reports from our circulation department, broad knowledge of the field provided by subject librarians, and detailed knowledge of the topic provided by faculty experts. Arbitrary guidelines (e.g. "older than 10 years") are not helpful and can lead to the loss of important works. Just as careful thought goes into adding books to a collection, careful thought needs to go into removing books from a collection.

If you weed out the books that won't be useful to your patrons, the useful ones are easier to discover and more likely to be used. Weeding no-longer-useful books is just as important to collection building as acquiring new books.


* And while folks looking at the history of computing may find such guides possibly useful, we have no such classes or programs at my institution.