Special Collections and Scholarly Communications

Donald Waters' thoughts on The Changing Role of Special Collections in Scholarly Communications come from the broad perspective of a funding organization (The Andrew W. Mellon Foundation), which allows him an informed view of the world of special collections without the potential bias of someone deeply invested in a specific project. Though written five years ago, this piece is still relevant.

Regarding Waters' opinion on the relative commonness or distinctiveness of collections, I think it is important to distinguish between current material and archival material.  As scholarly publishing is ever more thoroughly packaged and sold in huge lots (subject or title collections sold by publishers or "complete" aggregations from EBSCO, ProQuest, and others), and purchases consolidate at the level of the library consortium, homogenization of basic research holdings is an inevitable outcome.  On the other hand, as libraries are increasingly pressed to make space for more social or student-centered needs within their existing footprints, legacy print collections are indeed likely to become more distinctive, a boutique selection tailored to specific campus needs.  This will probably be most noticeable at smaller, non-PHD-granting institutions where it is difficult to make a value case for what one of my bosses called "the book museum" model. These smaller collections will rely more heavily on sharing arrangements with other institutions, with expanded metadata requirements to ensure discoverability and support collection management at individual campuses (e.g. which library holds, or will hold, the last print copy of a title, allowing others to discard the local copy).

I think Waters makes a terrific suggestion that special collections librarians work with the scholars at their institution to prioritize which collections are digitized and processed.  Given the huge backlogs and extraordinarily varied nature of most collections, accommodating current faculty needs seems like an excellent way to direct the workflow, justify the expenditure to campus stakeholders, and build important bridges between the library and faculty researchers.

Another great idea in theory is having researchers assign metadata to the materials they are working on. I think this would work well in the examples cited, where the researcher is working on a fellowship.  It's a great way to "pay forward" some of the grant money, and leverages the deep subject knowledge of the researcher.  However, I think the system would not scale well to the larger world of all digital libraries and special collections.  Faculty balancing research, teaching, and administrative duties are unlikely to be interested in "helping" the library, or have the spare time to commit even if they were; thus the metadata tasks would either go undone or be relegated to graduate students who might or might not have the expertise or commitment necessary to do the job well.

I had the pleasure last semester of interviewing Dr. Sylvia Huot, a scholar involved in the early development of one of the Mellon-funded projects mentioned in the article, The Roman de la Rose Digital Library. In regard to their original plans for full transcription and tagging, she told me that despite small volunteer transcription projects going on at various universities involved in the project, the leaders eventually realized how impractical it was to attempt transcription of the entire library and were forced to abandon the effort.  "No one with the skills to do a proper transcription job was willing or able to volunteer, and it would have required an enormous influx of funding to hire even skilled graduate students to complete the work."

There is little doubt of the value of excellent metadata, accurate transcriptions, open availability, and cross-institution linking and discovery for special collections.  Usage begets citations begets more usage, which pleases stakeholders and funding organizations and leads to more digitization projects in a virtuous cycle. However, there must be a better way than simply shifting the burden from one constituency to another.  There must be either greater efficiency, perhaps through automation or better tools, or greater funding, if this valuable goal is to be reached.