In December 2012, the Mudd Library announced that we had received a grant from the National Historical Publications and Records Commission (NHPRC) to digitize the most frequently accessed portions of six highly-used collections documenting United States foreign policy and the origins of the Cold War. We are pleased to announce that as of December 2015, all of the series and subseries selected for digitization— over 350,000 pages of documents— are freely available to view and download from Princeton University’s finding aids site (a complete page breakdown by collection is listed below). Now individuals anywhere in the world can read John Foster Dulles’s first major speech outlining the policy of “massive retaliation,” George Kennan’s unsent letter to Walter Lippmann regarding containment, and a myriad of other one-of-a-kind materials from any computer or device, at any time of day.

The digitization of archival materials is an expansion of the Mudd Library’s ongoing mission to make our holdings accessible to a wider set of users. While the completion of this specific project is an important step forward in its own right, we also knew that this project was going to be part of a bigger picture. From the start, our goal was to use the lessons learned from this project to create sustainable large-scale digitization workflows for future implementation at the Mudd Library, and potentially other archival repositories, as well.

With this goal in mind, we gathered data throughout the project on our productivity and costs. Most of the digitization work was accomplished with the help of an outside vendor, and the vendor’s metrics are based on quotes and monthly reports to Mudd staff. For the in-house scanning of portions of the John Foster Dulles Papers, we experimented with three different methods of digitization: an overhead Zeutschel scanner, a microfilm scanner for materials that had previously been microfilmed, and a networked photocopier with a sheet-feeder for the duplicative photocopies of Dulles’s correspondence. The in-house digitization metrics were collected by asking student scanning technicians to fill out a minimal, time-stamped form at the beginning and end of each set of scans, which Mudd staff then analyzed. Our final results are summarized in the table below. The page totals reflect only the number of images that were eventually uploaded to the findings aids site, not the number of pages that were rescanned. However, we included time spent conducting quality control assessments and rescanning documents in the total amount of time spent per method.

Our final metrics indicate that both vendor-supplied and in-house digitization can be employed successfully on a large scale, but that digitization by an outside vendor significantly decreases the amount time required to complete large-scale projects. We found that the convenience of outsourcing was enhanced by its surprising cost effectiveness, a consequence of the comparatively little Mudd staff oversight that was required for the outsourced portion of the project to proceed smoothly. (The final per-page cost for the vendor scanning takes into account all overhead and indirect costs, not just the direct fees charged by the vendor. These costs are also reflected in the other scanning methods.)

Though our results indicate that vendor-supplied digitization is more efficient for large-scale projects, Mudd staff members believe that metrics based on a greater number of in-house images would provide more insight into the benefits and challenges of conducting digitization internally. Personnel changes, unforeseen description issues with the John Foster Dulles Papers, and, most significantly, problems with our scanning equipment delayed digitization work and made it impractical to digitize all of the selected series from the John Foster Dulles Papers in-house as initially planned. However, we plan to do additional scanning work in the coming months in order to increase our understanding of these alternatives. In particular, we will gather metrics on microfilm and especially flatbed scanning, as we have now resolved a number of technical issues. We do not plan to continue testing digitization with a networked photocopier, as this machine ultimately produced unusable images, and we do not run originals through a sheet-feeder, extremely limiting the possible choices.

Mudd staff members also want to explore the possibility of conducting future digitization projects using other in-house scanning methods. Within the last year, public services staff at the Mudd Library have started using a tablet for the majority of our scanning requests from patrons. We are also interested in examining the use of digital photography for digitization of our materials, including the possibility of uploading photographs taken by patrons directly to the finding aids. The Mudd Library will continue to experiment with various capture methods that balance speed with the requirements for good image capture and report out on them as appropriate.

We wish to thank the NHPRC for their generous support and our student assistants for all of their dedicated work on this project.