Reassessing Backlogs

Backlogs have become a hot-button issue in the archival community in the last few years for good reason. To overcome our backlog problems, we must adopt flexible methods of arranging, describing, and preserving collections and change our ideas about processing and backlog management. We should strive to get a reasonable amount of work done on all our collections so that we no longer think of them as in a backlog.

Numerous surveys have found that the backlog problem is widespread. A 1998 survey by the Association of Research Libraries (ARL) reported that a mean of nearly one-third of the holdings of repositories remained unprocessed. A more recent survey of 100 archival repositories conducted as part of Mark Greene and Dennis Meissner's now seminal study on backlogs shows the problem to be particularly widespread. They report that 60 percent of repositories had more than one-third of their holdings unprocessed, and backlogs are a major problem at 59 percent of repositories.

Greene and Meissner's research and recommendations have sparked a new dialog and energy in the archival community by suggesting different methods for processing that move away from appraising, arranging, describing, and preserving documents on the item level toward a top-down approach that focuses on collection, series, and container-level work, which can be performed much more efficiently. Their suggested methods are designed to increase processing rates in order to make archival materials available to users as quickly as possible.

We need to move in these directions because the pressures causing our backlogs are increasing. Technological developments are forcing our collections to become larger and more complicated, involving many more record formats than ever before. As professionals, we are now being asked to do more than we ever have: staff resources are stretched ever thinner, and budgets rarely increase accordingly. In order to meet these challenges, we must work smarter as we modify our traditional sense of craft by adapting our processing methods to ensure that we process each of our collections, at least minimally, make at least collection-level descriptions of all holdings available online, and use appraisal techniques to determine which collections deserve more detailed treatment.

When I worked in the Manuscripts and Archives Department of the Yale Library, we put these principles to work by developing what came to be called “minimum standards” for processing, which included folder-level arrangement, description, and preservation techniques. We had initially planned to attack most of the collections in the backlog using these methods; however, circumstances changed our conception of the backlog even further.

Because many of the collections in the minimum standards backlog would not meet the more stringent appraisal guidelines we use today, we felt comfortable leaving them in their current state of rudimentary physical and intellectual control, which included collection-level MARC records and inventories.

Also, all collections are stored under optimal preservation conditions. Even though some of our preliminary inventories leave much to be desired, these collections are available to users. We knew that we could always return to provide a greater level of processing, but we will base these decisions primarily and almost exclusively on the perceived research value and use of the collection.

Perhaps more important, we achieved less than we initially intended because all of the staff members assigned to work on processing the minimum standards backlog (and the professional backlog, for that matter) have been pulled away from processing by other, more pressing duties. Chief among these were our efforts to convert the legacy finding aids for our nearly 2400 collections into encoded archival description (EAD), so that we can make them available to the research community via the web. These finding aids include both our fully processed collections and our preliminary inventories. While we would ideally like to do more work on certain collections in the backlog, the basic level of control we have over the collections allows us to prioritize online access to our holdings over backlog processing.

Aside from adopting a more sensible processing approach for the majority of our collections, we also developed other processing mechanisms designed to make incoming materials available as soon as possible. We adapted our accessioning procedures to perform preservation triage, basic arrangement (if necessary), and more robust description.

We have also experimented, with some success, in getting donors to do more extensive work on their own collections, if they are willing to put in the time. We have been able to use donors to rehouse collections, provide simple arrangement, and create inventories. On these occasions, our work has been reduced to editing, quality control, box labeling and barcoding for storage in our shelving facility, and incorporating the inventory files into our finding aid system. Colleagues who work primarily with university archives holdings have had even more success with these methods, as they now require university offices to rebox materials and create folder-level inventories for all accessions as a condition of the archives' agreeing to take the records and provide services.

There are downsides to these approaches. We consciously put a heavier burden of discovery on the researcher, who must now plow through more materials to find documents. This impact is outweighed by the desire to expose the greatest number of collections to our users. Perhaps a greater concern is that public services staff need to retrieve more boxes owing to less granular description. This problem has a silver lining, though: we can track the most heavily used collections and use data to make sensible decisions about which collections to process in a more detailed manner.

Max Evans, director at the National Historical Publications and Records Commission, once said we should call this type of work extensible processing, rather than minimal processing, to get at the idea that we can always return to a collection to do more work to it.

Another trade-off we run is the risk of losing diamonds in the rough, valuable parts of collections that go unidentified. Similarly, without full identification, it becomes more difficult to detect theft. And, finally, collections processed only minimally provide a greater challenge to digitize. It's one of the ironies of our time that as the mass of information we confront requires us to use less detailed means of description, effective metadata for digitization projects often requires greater specificity.

These issues all represent acceptable trade-offs, however. While I do not have any hard evidence, I can report that in my years as head of arrangement and description in manuscripts and archives, where I also worked with readers on the reference desk, I never heard a researcher lodge a complaint about our preliminary inventories. I did witness countless instances of readers finding useful materials in collections that were not fully processed. Moreover, adopting the minimum standards approach liberated staff. Rather than having the ten- to 15-year-backlog albatross hanging around our neck, we were able to make sensible decisions about how to use our resources wisely and strategically.