Uncovering Social History: An Interdepartmental Approach to Scrapbook Digitization

Scrapbooks are a compelling digitization opportunity for those who are willing to undertake the numerous challenges that result from working with these unique items. In 2011, the University Libraries of the University of North Carolina at Greensboro (UNCG) began what would become a three-year project to process and digitize hundreds of scrapbooks held by the Martha Blakeney Hodges Special Collections and University Archives (SCUA). These scrapbooks document the history of the university, as well as social and cultural issues surrounding the period of their creation. Many reflect both an individual and an institutional perspective, often containing materials absent or overlooked in the archival record groups of the university. The collection was selected for digitization to facilitate access, use, and preservation of the scrapbooks while limiting the physical stress on the original materials.

As the scrapbook project team began to plan this endeavor, it quickly became apparent that limited professional guidance was available on practical approaches to processing and digitizing large scrapbook collections. The dearth of information relating to this type of initiative ultimately compelled the project team to write its own rule book, creating innovative approaches to the many challenges presented by archival processing, digitization, online presentation, metadata creation, and preservation of scrapbooks. This article documents the process and discoveries that came out of this project.

Library and archival literature focuses heavily on the numerous challenges inherent to scrapbook collections. Scrapbooks are frequently composed of complex and diverse sets of materials.¹ These materials often present preservation difficulties, causing archivists and other staff to face the consequences of acidic paper, rapid deterioration, problematic adhesives, and effects of poor and damaging storage environments.² The contextual interdependence of objects contained in scrapbooks poses further challenges, especially in cases where organic or acidic items raise the threat of deterioration or damage to nearby materials. Furthermore, the unique nature of scrapbooks requires archivists to treat each one individually,³ which can strain available resources—time, money, personnel—involved in the handling of these collections.⁴ Yet, some of the physical challenges that can cause such headaches for archivists are the very things that make these items so valuable to researchers,⁵ specifically, the authentic format and arrangement, with materials provided in their original order.⁶

Scrapbooks are highly social artifacts—snapshots of life⁷ that have been likened to physical surrogates of today's online social media.⁸ Researchers looking for personalized historical viewpoints may be particularly interested in scrapbook collections.⁹ For example, scrapbooks created by students or student groups can be important in the study of higher education and the student experience.¹⁰ Their research value makes preservation an important topic for personnel working with scrapbooks, a fact reflected in the literature.

The diversity of materials found in scrapbooks means that a one-size-fits-all preservation plan may not be possible for every collection. A recognized first step to stabilizing scrapbooks is the use of appropriate storage methods,¹¹ including controlled temperature and humidity, archival boxes, and flat storage.¹² The interleaving of scrapbook pages with acid-free paper can also help alleviate some deterioration concerns,¹³ but the literature disagrees on the topics of scrapbook disassembly and removal of problematic materials. Scrapbooks may hold organic matter—pressed flowers, cigarette butts, and other such items—and archivists may wish to remove these items to keep them from further affecting nearby items that may be of greater value.¹⁴ Additionally, scrapbooks suffering from worn, stretched, or otherwise weakened bindings may benefit from being disassembled, but doing so may affect their original order and arrangement and diminish their research value, as noted by Lisa Wood.¹⁵ Archivists working with scrapbooks must balance preservation concerns and research needs when making conservation decisions.

Based on potential research value and preservation issues, scrapbooks are often excellent candidates for digitization. Digital surrogates can allow users to interact with the materials without causing further physical damage,¹⁶ alleviating some preservation concerns. But digitization is not without its challenges. Using flatbed scanners is problematic for many scrapbooks, as doing so may damage bindings and/or interior materials.¹⁷ Therefore, overhead or book scanners are recommended for use in scrapbook digitization.¹⁸ Once scrapbooks are digitized, staff must consider how to provide access to them, which creates further challenges. Many scrapbooks contain folded or layered materials as well as items held in envelopes or other enclosures. These materials create difficulties in replicating the traditional experience of turning pages within a scrapbook and physically interacting with its context-based items.¹⁹ Allison B. Zhang noted that the Premium Page Flip component of Flash 8 can be used to digitally mimic the effect of turning pages, and she also pointed out that choosing page-naming conventions is important when it comes to structural metadata for these items.²⁰ Besides these references, discussions of metadata and workflow issues in regard to archival scrapbooks are almost entirely absent from existing literature. This article attempts to address some of these missing pieces.

The University Archives Scrapbook Collection comprises 241 scrapbooks dating from 1906 to 2002 and is a unique resource that documents the academic, social, and cultural history of UNCG. Founded in 1891 as the State Normal and Industrial School, this small public women's college was initially established to train young women to teach when North Carolina's literacy level was at an all-time low. In addition to teaching pedagogy, the school also offered curricula in business and home economics. The scrapbooks reflect the history and institutional memory of the university and, in many cases, provide the most comprehensive source of information about its traditions, events, and social history. Candid photographs, invitations, programs, class songs and poems, dance cards, newspaper clippings, and colorful ephemera illustrate the student experience.

Prior to the start of this project, archivists at UNCG had identified the University Archives Scrapbook Collection as a valuable resource that needed immediate attention. Digitization would provide the opportunity to process the collection and make it accessible to the public in an online format that would simulate the physical items. This would entail assessing the condition of each scrapbook, creating metadata relating to the object as an archival resource, digitizing and displaying the scrapbooks online, and safely rehousing them in a manner that would contribute to their future preservation. Digitization would also serve as a pilot for future projects involving other scrapbook collections held by SCUA.

The first step was to present the scrapbook project to the library's Digital Projects Priorities Team for approval. Once the project was accepted, a multi-departmental team was assembled, which included the digital projects coordinator (Electronic Resources and Information Technology Department, ERIT), the photograph archivist and the special projects archivist (SCUA), and the meta-data cataloger (Cataloging Department). Other library staff and student workers were added to the project team as needed.

Each team member was responsible for his or her area of expertise. The archivists, with their knowledge of university history and research patterns, would arrange and describe the scrapbooks. The digital projects coordinator would oversee the digitization of the scrapbooks and integrate them into the CONTENTdm platform. The metadata cataloger would create descriptive meta-data based on information collected by the archivists.

The project team understood that unique challenges would dictate the processing and scanning of the scrapbooks, but the practical impact of some issues could not be predicted. For example, the archivists recognized that the scrapbooks' poor condition would likely cause problems during the scanning process, but they did not realize the full extent to which the scrapbooks' size and fragility would affect processing and metadata collection. The discovery of detached items, poor adhesives, crumbling paper, organic materials, and folded, multipart documents slowed the project at every stage. Other hurdles included negotiating an interdepartmental workflow that would maximize the efficiency of the project and best match the individual proficiencies of each contributor. Team members also had to refine communication strategies, standardize project metadata, and create a workable digital representation of the scrapbooks. While the project posed many challenges, it also provided an opportunity to devise techniques and approaches that could be incorporated into a streamlined prototype for future collaborative digitization projects.

When digitization began, the team realized the importance of creating an efficient project workflow that would enable all departments to access and edit information. The initial plan was to use an online spreadsheet kept on a shared drive as the project-tracking document. The metadata and processing status would be updated in real time using this spreadsheet, keeping the entire team abreast of every aspect of the project. This process ultimately failed to meet the needs of the archivists, who found that certain physical aspects of the scrapbooks challenged the prescribed workflow plan. The scrapbooks, both commercially produced and handmade, ranged in size from 6 inches by 4¾ inches to 25 inches by 19¾ inches. Managing large and fragile scrapbooks in the same physical space as a laptop computer proved cumbersome, making it impractical to simultaneously process the materials while entering metadata in the online spreadsheet.

Additionally, the spreadsheet did not adequately track the scrapbooks as they progressed through the project workflow. Staff members and student workers from several different departments were responsible for transporting and digitizing the scrapbooks, but their involvement in the process was not always documented as intended. Therefore, the archivists were often unsure of the location and digitization status of the scrapbooks on any given day. For these reasons, a paper form was created to collect basic metadata and track movement of items between departments. These paper forms remained with the scrapbooks as they moved through each department, and copies were also kept in SCUA at all times.

The paper form greatly simplified the logistics of the initial metadata workflow. It successfully incorporated information such as assigned item number, title, creation date, creator, size, number of pages, and physical description. Because the project included a preservation component, the archivists also assessed the condition of each scrapbook, noting tears, discoloration, and any access considerations. This became a good opportunity to review the scrapbooks and document unique materials. The archivists took particular note of candid photographs of building interiors, campus views, and school events while also recording unique school traditions such as the Daisy Chain, Tree Night, and Morning Watch. Prominent names and important campus, community, and national events were also noted. Initially, the archivists were expected to create a list of controlled vocabulary terms as part of the metadata creation, but it quickly became apparent that they had little experience with this aspect of description. They were instead asked to create specific content notes based on standardized subject terms provided by the catalogers. Additionally, the archivists drafted free-form notes from which the catalogers could construct descriptive narratives.

Many of the preservation concerns that the project team faced were consistent with those presented in current literature; poor-quality paper, problematic adhesives, and inadequate storage environments all posed challenges to the processing and digitization of this collection. Preservation was handled on a case-by-case basis, primarily through proper interleaving and storage.

After the digitization and metadata creation process, archivists interleaved pages with acid-free tissue and stored the scrapbooks in archival boxes. Loose photographs were sleeved, and organic materials removed, stabilized, or encapsulated when necessary. As most of their bindings had been taken apart prior to scanning, the scrapbooks were reassembled and tied gently with string. Several scrapbooks were restricted from handling or exhibiting due to their unstable condition.

Particularly problematic were those scrapbooks created between 1970 and 2001 using magnetic page photo albums. These contained mostly photographs, ephemera, and publication clippings, many of which were stuck to the adhesive on the page boards and could not be extracted without extensive damage. Most of the plastic covers had long ceased to fasten to the page. It was ultimately decided that the albums would be scanned “as is” and that the scrapbooks would be boxed without removing the photographs or materials.

Only one scrapbook was in such poor condition that it required conservation. This scrapbook, which chronicled the large May Day event that took place on the campus in 1912, was considered unique and historically significant. It had been extensively damaged by strip adhesive that was a part of the original design of the commercially produced scrapbook. The glue had puckered and created a rippled effect on the pages, and the photographs that comprised most of the scrapbook were likewise affected. Because the photographs were irreplaceable and in good condition apart from the rippling, they were extracted and preserved separately. The photographs were carefully removed from the scrapbook and flattened, then lightly hydrated and placed between sheets of Reemay and blotter paper. Afterward, they were placed under moderate pressure to reduce the amount of cockling. Once this process was completed, the photographs were identified and numbered by original page placement, sleeved, and put in a folder within the scrapbook storage box. The remaining materials were left in the scrapbook, as most were published articles that would have torn and crumbled if extracted from the pages.

Existing digitized scrapbook collections run the gamut from minimalist presentations of one or two scrapbooks to more elaborate displays of scrapbooks as parts of other collections. Interestingly, several of the projects examined appeared to be displaying scans of microfilm copies rather than of the actual scrapbooks, which may reflect both preservation concerns and the degree of difficulty involved in scanning scrapbooks. This option seems best suited to an item like the Pennsylvania Necrology Scrap Book, which is primarily textual in nature as opposed to a more traditional, photo- or ephemera-based scrapbook.²¹

Some digitized collections, such as the Frederick Douglass Scrapbook of the Monroe County (NY) Library System²² or the Briggs Scrapbooks of the Virginia Department of Historical Resources,²³ are really no more than PDF files linked from static HTML pages within the repository's website. They lack detailed meta-data, and no real attempt has been made to display a hierarchical relationship between pages and the embedded content within attached items, such as mul-tipage booklets and folded letters. Because many of these scrapbooks contain primarily newspaper clippings, letters and folded items are less of an issue, and the need for robust description of content formats is decreased.

Somewhat more elaborate collections, such as the Harry Houdini Scrapbook Collection of the Ransom Center at the University of Texas at Austin, are generally presented in CONTENTdm or a comparable content management system and displayed as page images in JPEG format.²⁴ Some archives make an attempt to demonstrate the relationship of fold-outs and multipage attached items, usually through consecutive pages that show the same page with items folded, unfolded, repositioned, and so on. Similar presentation strategies are used for the Historic ET Collection Glenda McKissic Scrapbook (Texas A&M University)²⁵ and the Iowa Women's Archive Founders Collection (University of Iowa).²⁶ The College of Charleston also displays its Scrapbooks and Photo Albums Collection in a comparable fashion using the open-source Fedora platform rather than CONTENTdm.²⁷

Most of these collections feature a small number of scrapbooks and relatively thin descriptive metadata, although the Iowa project does include some genre terms for scrapbook contents. Iowa also uses scans of individual items as a way to display attached content, but does not relate these items to their “source” pages other than by displaying them consecutively. Similarly, the North Carolina Digital Heritage Center (NCDHC) has digitized over two hundred scrapbooks from various repositories within the state.²⁸ NCDHC also uses CONTENTdm and displays page elements as sequential pages rather than as part of a hierarchy. Many of the items do have significant descriptive metadata, however, which individual contributing institutions generally supply.

The physical challenge of scanning fragile and cumbersome scrapbooks on flatbed scanners was considerable; many of the scrapbooks were literally crumbling, making further damage a significant possibility. Many of the smaller and less fragile volumes were scanned on an Epson 10000XL flatbed scanner. The Digital Projects unit was later able to purchase a Bookeye 3R-2 overhead scanner, which staff used to scan larger and more fragile volumes.

The goal for online display of the scrapbooks was to replicate as closely as possible the experience of browsing an actual scrapbook, paying particular attention to the relationship between “master” pages and the items attached to them. This goal required that the project team devise a scanning procedure that would maintain this relationship while minimizing the additional work required to display that relationship online using the CONTENTdm platform.

Structuring the digital files to match the physical items and provide the most helpful browsing experience was a significant concern. The project team set up file naming conventions and folder structures to address this issue, facilitating project organization and contributing significantly to metadata creation. A hierarchical system was employed wherein each master page was first scanned to show how it might appear to a user leafing through the scrapbook; these files were named “Page 001.tif” and so on. Scans of items attached to the master pages were saved in discrete subdirectories given unique names to reflect both the master page and the attached item's content—for example, “Page 002-Program” contained scans of a program attached to page 2 of the scrapbook—so that the file and folder names would constitute a reasonably descriptive title for each scan (see Figure 1). This allowed for automated uploads into CONTENTdm using the “Monograph Compound Object” functionality, which duplicates the file and folder hierarchy and displays these descriptive filenames by default in the DC. Title field for each image (see Figure 2). Thus, a fairly significant metadata element was created immediately upon digitization.

Initially, the team hoped that temporary student assistants would be able to complete much of this work. Unfortunately, the complexity of the file and folder structure, the fragility of the materials, and the numerous “judgment calls” (e.g., scanning of blank pages, handling unusual or large items, etc.) soon made it clear that using short-term student workers was not the best option at this stage of the project. Full-time departmental technicians learned the rather intricate process through repetition and were better able to make independent decisions based on experience. The limited tenure of student workers never allowed them to achieve the necessary level of proficiency, resulting in error-prone work and questions that slowed the digitization process considerably.

In addition to providing access for researchers, it was important to the project team that the metadata for this collection be cross-searchable with other local digital collections housed in CONTENTdm, as well as with external collections and content providers, such as the Digital Public Library of America and OCLC's WorldCat database. The University Libraries implemented OCLC's WorldShare Management Services as its catalog provider in the summer of 2013. Digital project metadata is made available on that platform via the WorldCat Digital Collection Gateway, which utilizes a Dublin Core metadata map. With its wide use locally, nationally, and internationally, Dublin Core was the obvious choice for the descriptive metadata standard for this project. The final metadata scheme relied heavily on Dublin Core fields, supplemented with local fields to provide additional descriptive information to local users.

Many UNCG collections are presented with full-text transcripts and otherwise thin descriptive metadata. In selecting the metadata fields for the scrapbook collection, the project team decided to document the collection with fuller descriptive metadata than the in-house standard. While this collection includes many textual items, it was not a good candidate for scanning with optical character recognition (OCR) due to the variety of visual, tactile, and handwritten materials found alongside text-based items. The lack of text transcripts and the need to provide access to rare and unique research materials compelled the project team to spend more staff time on metadata creation and entry on the scrapbooks than on most other collections.

The metadata workflow for this collection involved personnel in all three cooperating departments. Archivists in SCUA captured initial descriptive meta-data on the paper forms mentioned previously, focusing on material formats and items of particular importance. Scrapbooks were transferred to Digital Projects for scanning, and scanning staff entered basic descriptive metadata on the paper forms, including working title, date, and size information. Personnel in the Cataloging Department entered data into a unified metadata spreadsheet, standardizing the data before sharing it with the archivists, who ensured that scrapbook content was reflected accurately.

The cataloging staff also played an essential role in the choice, development, and application of controlled vocabularies for this collection. Collection-level metadata included Library of Congress Subject Heading (LCSH) terms, applied across-the-board to all items to facilitate item collocation in cross-repository searches. An additional controlled vocabulary, called “content formats,” was assigned at the scrapbook level. This local vocabulary was developed in-house through cooperation between catalogers and archivists and was used to document materials of particular importance within the scrapbooks. After initial discussions with the archivists regarding important materials known to exist in the collection, the metadata cataloger drew up a draft list of terms for this vocabulary, using existing terms from the Art and Architecture Thesaurus and the Thesaurus for Graphic Materials where applicable. For materials with no existing standard term, including items related to local traditions such as “Rat Day,” “Tree Day,” and “Gym Meet,” the metadata cataloger created local terms, in consultation with the archivists. The vocabulary saw additions over the course of the project, as new and different material types were discovered in the collection.

While much of the descriptive data was straightforward to enter and format, the title field presented some unexpected complications. Some scrapbooks had no title at all. Others had titles applied by their original creators, but those titles were not always indicative of the content, for example, “^~G^~” for a student legislature scrapbook or “Lullabye of Birdland” for a student dance scrapbook. Additionally, many scrapbooks could be considered parts within a series, as individual organizations created them on an annual basis. The project team strove to create data that would satisfy the research needs of the patrons and the internal needs of the University Archives, all the while meeting appropriate national standards for description and access. To accomplish this, the team decided to use two title fields. The transcribed title field lists information taken directly from the resource (where applicable), while the standardized title field provides series- and content-related information within brackets: [North Carolina Student Legislature Scrapbook, 1992–1993] and [Student Dance Scrapbooks: Junior-Senior Classes Ball, “Lullaby of Birdland,” 1959]. This method allows easy identification and collocation of collection materials, while still reflecting existing data as item creators applied them.

Scrapbook digitization will always be a challenging undertaking, presenting both anticipated and unexpected challenges. In the case of the UNCG scrapbook digitization project, the problems were well worth the results. The 241 scrapbooks have now been digitized and made available online for use by students, faculty, and other researchers. In addition to enhancing the accessibility and visibility of the collection, the project laid important groundwork for future digitization projects in the University Libraries. The lessons learned in this project could also be beneficial to other institutions with similar collections. The project team recognized four areas of particular importance for application in future projects: interdepartmental collaboration, communication, workflow, and use of temporary support staff.

As an interdepartmental endeavor, it was obvious from the start that collaboration would be integral to the success of the project. But the project team learned that simple collaboration was not sufficient. Each team member brought a unique skill set to the team, and it was critical to match the existing skill sets with the needs of the project. For example, rather than training archivists to enter metadata in unfamiliar online forms and systems, it made more sense to have catalogers take the lead on metadata entry and standardization. This freed the archivists to focus on the materials—their area of expertise. The collaborative structure of the core project team—the photograph archivist, the special projects archivist, the digital projects coordinator, and the metadata cataloger—was essential to the success of the project, as it provided a foundation for the three-year scrapbook digitization effort.

The project also underscored the importance of interdepartmental communication. While the working relationships of the team members were strong before digitization began, intricacies of departmental logistics, unfamiliar and specialized terminology, and a complicated workflow had the potential to cause difficulties throughout the course of the project. After a somewhat slow start, the team members realized that to make the project a success, they needed to communicate regularly—not only to update each other on the status of the project, but also to enact changes to expectations and workflow components. It also became apparent that this communication needed a face-to-face aspect in addition to the online project workflow document. This type of contact served to strengthen the working relationships among the project team members and helped to maintain a smooth workflow.

The workflow itself may have been the most problematic aspect of the project. The expectation that an online workflow document could track the scrapbooks as they moved through the departments and assist with base quality control was not realized. While the online spreadsheet worked well for the digital projects unit and for the catalogers, the archivists found the system awkward and difficult to use and were concerned that this paperless system was not able to adequately track the scrapbooks. The creation of the paper form solved several problems; it allowed the archivists a much more flexible approach for capturing metadata, and it created a document that would follow the scrapbooks as they made their way through the departments. Ultimately, the project incorporated a hybrid paper and electronic system that met the needs of all departments, and this approach will be considered as appropriate for future projects.

One issue the project team did not expect to face related to the use of temporary support staff. Mostly comprised of student workers, this group was pulled into the project as needed to assist with a variety of tasks. The archivists did not use temporary staff, as those students did not have the institutional memory to recognize and extract important information, nor did they know how to process and store the fragile scrapbooks. In the cataloging department, students were able to contribute a great deal to the entry of online metadata—a time-consuming task with this uncharacteristically metadata-heavy collection. These students already had basic metadata training and were able to learn the particular needs of this project relatively quickly. The metadata cataloger provided additional training when needed, as well as quality checks of the students' work. The digital projects unit had initially hoped that its student workers would take on most of the scanning process, but a lack of continuity in scheduling and the complicated nature of the project made their use impractical.

The UNCG scrapbook digitization project proved to be challenging but successful, allowing access to a unique and historically significant collection, which is now available at http://libcdm1.uncg.edu/cdm/search/collection/ui/searchterm/ua111*/mode/all/order/title. As expected, it became a pilot for collaborative dig-itization projects in the UNCG University Libraries, providing the framework for a $200,000 multiyear Library Services Technology Act grant, which included digitization of additional scrapbooks and related materials. Further research is needed to understand how archival institutions are handling born-digital scrapbook materials in their collections, specifically with respect to preservation, online display, and descriptive metadata.