World War Words: The Creation of a World War II–Specific Vocabulary for the Oral History Collection at The National WWII Museum

Lindsey Barnes and Kim Guise

Abstract: Many interpretative institutions today focus increasingly on narratives, storytelling, and the personal experiences of historical and everyday figures. Providing access to oral histories through a vocabulary focused on describing these stories and experiences is a unique and effective way to share these narratives. This article is a case study of the development of a controlled vocabulary for the oral history collections at The National WWII Museum.

Keywords: controlled vocabulary, The National WWII Museum, World War II

Storytelling has played a central role at The National WWII Museum since its opening as the D-Day Museum in New Orleans in 2000. The museum’s mission is to tell the story of the American experience in the war that changed the world—why it was fought, how it was won, and what it means today—so that all generations will understand the price of freedom and be inspired by what they learn. In 2003, Congress designated the museum as The National WWII Museum. In honoring the legacy of museum founder Stephen Ambrose, who was dedicated to giving voice to the citizen soldier and bringing to light the everyday experiences of men and women, the institution draws heavily on personal accounts, autobiographical materials, and oral histories to accomplish its mission.
The museum has expanded from one building to several and will quadruple in size by the summer of 2016. The museum hosts a growing collection of artifacts, images, documents, and more than seven thousand oral histories. In addition to gallery exhibits, we present traveling exhibitions, host conferences, lead educational tours, and produce documentaries and publications, both scholarly and popular. We are dedicated not only to preserving the artifacts, archives, and oral histories entrusted to us by donors but also to finding new ways to bring this material to life. The museum’s goal is to engage the public on a variety of levels and to create avenues to reach students, researchers, family members, general World War II enthusiasts, and casual visitors—attracting new audiences while engaging old friends.
Of the museum’s seven thousand oral histories, approximately one-third are video, one-third are cassette tape, and one-third are paper transcripts. The museum’s oral histories represent a wide array of World War II experiences with a heavy emphasis on combat stories collected for use in future exhibitions. Excerpts from interviews are used in many museum exhibitions and programs, but the majority of the interviews have not been seen or heard by the public.
The National WWII Museum embarked on an ambitious project in 2009, funded by the Institute of Museum and Library Services (IMLS), to explore innovative ways to navigate, describe, and present oral histories. Our partners on this National Leadership Grant, “Bringing Oral Histories to Life: Unlocking the Power of the Spoken Word,” were The Randforce Associates and National History Day (NHD). The project’s goal was to provide deep and meaningful access to our oral history collection for researchers and students of all ages, beginning with online access to 150 of the museum’s video oral histories.
Despite being a museum with a worldwide scope, we found that a general vocabulary to describe the oral histories in this project did not exist. The vocabularies in common use were not capable of the level of description our institution desired, since we wanted the public and our own staff to be able to search collections by navigating through personal stories. Instead of accessing interviews by searching through transcripts of oral histories, we hoped to provide access via small segments of the entire videos, each indexed using a vocabulary created specifically for this collection. The three overarching goals of the grant were to: (1) create a framework to allow museums and libraries to develop a meaningful vocabulary for oral history collections that enables detailed querying and exploring; (2) develop a model for users to access oral history collections in a flexible and interactive Web 2.0–based environment; (3) improve museum and library engagement with constituents.
Creating the vocabulary was only one aspect of the project. Below are the steps the museum completed to ensure the success of the entire grant:

1. Prioritized 150 oral histories for publication online by developing a matrix to account for diversity and range of stories.
2. Selected software in which to segment, annotate, and index. We chose an open-source solution, Annotator’s Workbench, originally created for use in ethnomusicology.
3. Created a collection-specific experiential vocabulary to aptly describe World War II narratives.
4. Developed standards for segmentation and annotation, as well as a handbook with guidelines for theory and workflow.
5. Segmented, annotated, and indexed 150 oral histories (about 300 hours of video). The museum chose to segment, annotate versus transcribe, and index all video segments.
6. Held focus groups with NHD teachers and students to assess the vocabulary and website.
7. Designed the website’s interface, developed the back-end database, developed video presentations, and created content delivery guidelines.
8. Produced thousands of teacher sourcebooks with our partner, NHD, that provide K–12 lesson plans based on the oral histories and website.
9. Addressed problems that arose while meeting the benchmarks above.

The project website (http://ww2online.org) provides users with the ability to tag, comment, and create their own collections. We hope that users will develop a greater intellectual and emotional connection to the museum’s oral histories through the opportunity to interact with selected video segments. User-applied comments and tags will further classify the videos in a personal way, adding a level of vernacular description to create a flexible and interactive Web 2.0–based environment.

A World War II Vocabulary Solution

As large, centrally themed collections grow within an institution, standard controlled vocabularies, such as the Library of Congress Subject Headings (LCSH) or the Thesaurus of Graphic Materials (TGM), can miss the more specific access points an institution wishes to provide. These and other standard vocabularies are imperative to provide access to large collections and across multiple institutions. Yet these terms often do not drill down to the specificity required to describe recurring events or experiences key to one’s institution. Supplementing standard vocabularies with local terms is now customary at most institutions, but the museum strove to move beyond standard access tools and create a vocabulary specific to the stories of people who experienced World War II. Our grant consultants, The Randforce Associates, helped us to create this new vocabulary, and our grant advisory board helped us to define its goals.
The challenge the museum faced stemmed from providing access into experiences. Creating a vocabulary to describe experience-based oral histories presented a different set of challenges than a vocabulary created to describe items or artifacts. Accurate description through the application of standard vocabulary terms was lacking, because often an interviewee is not discussing an item as if it were being identified in a photograph. The vocabulary needed to be more attuned to the particularities of narrative rather than static subjects. This meant including action words and verbs, such as experiencing, interacting, and overseeing, in the vocabulary terms. The experience, rather than the subject, becomes the search term. These terms provide the researcher with a better sense of what they can access within the museum’s collections of stories.

The Process of Creating a Vocabulary and Recommendations

Although the museum created a vocabulary for access to segmented oral histories, the process described below can apply to any institution interested in creating and applying a local vocabulary to any type of video. These recommendations can also be applied with the same efficiency to nonsegmented oral history collections or to videos that are not oral histories, such as ethnomusicology field recordings. It is our intent that this process helps all institutions interested in indexing video, regardless of the subject matter.
When creating a unique vocabulary, institutions should be fully aware of their audience. During the first steps of vocabulary creation, the team should remember its user base and how those users wish to access the collection. Our users include museum staff, K–12 students, museum members, World War II researchers, and artifact donors’ families. Audiences will most likely range from novices to experts, and therefore the vocabulary terms must be understandable across many levels of knowledge and searching expertise.
The chart (fig. 1) illustrates the process the museum took to create a controlled vocabulary specific to our oral history collection. The vocabulary team was led by the museum archivist, who worked closely with a curator and an oral historian and in cooperation with the entire oral history department. The museum’s oral historians conducted the majority of the interviews and were most familiar with the collection; therefore, their participation was key to understanding the breadth and depth of material.

Fig. 1. Steps in creating a controlled vocabulary

Last week we presented Part I of this article and now here is Part II:

1. Review of existing vocabularies

The vocabulary team, under the guidance of our grant consultants, The Randforce Associates, began by reviewing other locally created vocabularies, both World War II– and non-World War II–related and especially those of other IMLS grant projects. The museum focused most heavily on a coding framework developed by Arkansas Educational Television Network (AETN) for a project entitled “In Their Words” (http://www.intheirwords.org/). Although our team rejected many of the terms used in this project, we agreed that it was crucial to use terms that described experiences rather than objects. For example, the team concentrated on, and later adopted, a version of the AETN term Interactions with Friends/Family. Narrowing our focus on this type of description was instrumental in developing the theory behind the museum’s vocabulary.

2. Review and annotation of sample videos from the museum’s collection

The vocabulary team then individually watched a sample of videos from all service branches and theaters of the war, including the home front, to review a majority of the museum’s collecting themes. Each member noted common themes and shared experiences that repeated across interviews. At this time, the team began to practice annotation in order to explore themes and similarities between the resulting entries. The annotations provided textual evidence for use in later discussions.

3. Key experiences and themes are identified from all videos

The vocabulary team then reviewed their individual annotations together, noting where multiple team members identified similar subjects. The team worked closely with the oral historians during this process to identify the frequency of certain themes. As the oral historians are well versed in the content of the collection, they could discern whether an experience (such as Post-Traumatic Stress Disorder) would be useful as a vocabulary term or whether it did not have a place in the museum’s specific vocabulary.
A good principle to introduce at this step is that of defining the concepts of “too narrow” and “too broad”—in other words, finding out which terms are too specific and too general for the project’s purposes and eliminating those terms from the vocabulary choices. One of the ways in which our team began to understand the distinction between narrow and broad terms was realizing that some experiences might be useful at the interview level, but those same experiences may not be so useful at the segment level. For example, a term such as Pacific War would be too broad, while the naval rating Storekeeper 2nd Class would be too narrow. The team needed to settle on terms that would have relevance to many segments within the collection but not the vast majority.

4. Vocabulary terms are listed based on these key experiences and themes

Once a basic framework of themes was established, the team concentrated on the term itself. The team worked to create a search term that would describe an experience rather than a static subject, often including verbs, such as the term being wounded. All terms were then brought to the table for review. This process was repeated many times during the initial phases of the project, until a cache of terms, based on the chronology of typical service experiences and other prominent themes of the war, developed. With each meeting, additional terms were added or deemed unnecessary based on the consensus of the group.

5. Vocabulary terms are placed in hierarchical structure

The core vocabulary team then took the created list of terms and began to place them in a hierarchical order. As it helped to have a physical manifestation at this stage, the team first cut the terms out of a large sheet of paper and simply began arranging them on a table, putting groups of terms together under broader terms and, in the process, identifying “orphans” (terms with no parent terms in the hierarchy, such as weather) and terms that seemed too analogous to each other (such as correspondence and mail). This exercise was extremely helpful in identifying the structure that our vocabulary had naturally assumed.
At this point, the team arranged the hierarchy of terms together and created the “vocabulary map,” which allowed the team to visually explain the hierarchical structure via a skeletal structure. This created a way for anyone unfamiliar with cataloging and controlled vocabularies to physically and visually identify with the new system they would be using. The map was also extremely helpful for the staff applying the vocabulary terms to the segments because it gave them a way to familiarize themselves with the structure by “reading the map.”

6. The videos are again reviewed and vocabulary terms are applied

After the vocabulary included nearly one hundred terms, it was deemed complete. The team then again individually reviewed video segments, while annotating and applying the new vocabulary terms to them. There were revisions and additions, but the team felt this was a good beginning framework—complex enough to aptly describe our collections but small enough to introduce to novices and to expand upon.

7. Vocabulary terms are evaluated through discussion after application

When the majority of the team used the same term in the same segment, the term was considered a success. All terms applied were then compared through a group discussion, and the final terms were validated based on application within the intended environment. When the term was used improperly or confused with another term, we revised that term to reflect its specific theme. When there was a common theme identified by the group that no term was suitable for, we created another term but made sure it would be pertinent in more than just one oral history.

8. First vocabulary set is created

At this point, the first vocabulary set was created and became our working set for the next few months of more intensive review, during which we introduced a wider range of oral histories. This wider range tested the limits of the initial set and was an opportune time to discuss the small subsets of variant experiences within our broader oral history collection (such as Holocaust survivors and Axis combatants). The vocabulary will remain in perpetual review as the museum adds to its collections and identifies new themes, which is the ninth and final step in initially creating the controlled vocabulary.
When terms were finalized, the team created a list of definitions for each so that every term’s meaning could not be manipulated by judgment or used inappropriately. This vocabulary glossary also helped the team to define each term and stipulate how each would be used to describe different experiences. For example, the term food/mess should be applied to all personal experiences centered on eating or cooking. This demonstrates how all of the terms chosen are defined by actions, even if they do not contain a verb. Such a glossary is also invaluable for training new indexers as they begin to work on the project.
Next, the results of focus groups with NHD students and teachers were analyzed to compare museum-developed vocabulary with NHD-supplied vocabulary of the same oral history clips. The participants were high-performing middle and high school students who opted to attend the sessions during NHD events. In both sessions, we asked the students to list terms describing a series of interview clips. The aim was to compare our professional vocabulary with user-supplied terms and evaluate the success of the coverage of our vocabulary. In analyzing the results, we looked for several categories of applied terms. The data produced showed percentages of terms that fell into the following categories: (1) errors; (2) too broad/interview-level cataloging; (3) supported by the museum-developed vocabulary; (4) too narrow/covered in annotation; (5) valuable for creating new terms. Errors occurred when a student applied a term describing someone as Navy when they were in the Army. Terms such as World War II and others applicable to all interviews were deemed too broad. Very specific or technical terms such as M2 Browning.50 caliber machine gun were deemed too narrow, and other themes not previously considered by the vocabulary team were considered potential new terms. Terms closely related to the museum’s vocabulary were considered successful matches.

Next Steps and the Future of the Project

With the controlled vocabulary in place, the project is now looking to its future and addressing a variety of issues: the imperatives of collecting versus managing, the use of emotional and subjective terms, enabling cross-section searching, and the role of the advisory board.
Collecting versus Managing: The museum has been heavily focused on the urgency of collecting, due to the rapid rate at which World War II veterans are dying. Because of this somber fact, interviews were often conducted, deposited, and then left for future processing and cataloging. A shift in approach has now emerged, broadening the focus from simply collecting interviews to managing and providing access to them. Such a shift is always difficult, but transitioning to a new and innovative approach to access was an even more complex, albeit rewarding, process.
Emotional/Subjective Terms: Decisions had to be made regarding the level of specificity of the description. One such decision resulted in a focus on the objective, leaving out terms related to emotions because of the challenges of subjectivity. It was partially through the focus groups and also through our own work processes that we discovered that viewers/listeners/readers, including our own staff, gravitate toward emotional passages and want to search these videos by emotional, subjective themes, like fear and bravery. During the vocabulary’s development, the team chose not to include emotional terms, such as being afraid or feeling hopeless, because the team felt that professionally applied terms should be based solely in fact. The project’s website will monitor subjective/emotional tags applied by users for possible inclusion in future projects.
Cross-Collection Searching: This grant laid the groundwork for the digitization of and access to the rest of our collections. An important take-away from this project was an institutional solution to enable cross-collection searching based on this vocabulary dedicated to the stories that the museum endeavors to share. All of the museum’s collections are accumulated to support its mission statement of telling the stories of the American experience during World War II; therefore, many of our artifact and archival collections can also be described with this vocabulary. Through a new collections management system, the vocabulary will be applied to all collections, enabling powerful cross-collection searching.
The vocabulary is now an essential way to catalog and identify the museum’s major themes. For example, the term experiencing enemy fire will be applied to an oral history segment of a serviceman describing an enemy attack, but it will also be applied to a helmet with a bullet hole, an after-action report describing enemy fire tactics, or even Axis footage of weapons being fired. At The National WWII Museum, the vocabulary will work well when used with other high-level cataloging areas, such as theater of war and branch of service/civilian, for complex browsing opportunities through the experiences of different people all over the world. However, the wider use of the vocabulary throughout the institution will call for an even greater review of its application and strict control of its scope and purpose.
Advisory Board: As the museum forges its way in creating a new system, new questions arise, and we are constantly addressing them and learning as we go. Our partners, consultants, and the advisory board created to inform our project’s path and outcome were especially helpful in informing every aspect of the project. Prior to this year, we have had relatively little contact with professionals in the oral history field, but this project has sparked discussions with new mentors and partners, and we have discovered many opportunities for collaboration. The museum hopes to continue these relationships long after the grant is finished.
The National WWII Museum looks to become a leader in identifying national trends centering on the use of local, World War II–related controlled vocabularies. We also hope to open a discussion that will connect this community through this case study. As many museums turn toward exhibitions based on narratives, storytelling, and personal experiences, the ability to catalog collections to support these ideas becomes imperative, and a specific vocabulary becomes one of the best core access tools.

Lindsey Barnes and Kim Guise

Lindsey Barnes is the Senior Archivist and Digital Projects Manager at The National WWII Museum. She earned her master’s in library and information science from Louisiana State University. E-mail: Lindsey.Barnes@nationalww2museum.org.

Kim Guise is a Curator and Content Specialist at The National WWII Museum. She earned her master’s in library and information science from Louisiana State University. E-mail: Kimberly.Guise@nationalww2museum.org.