The quick, sloping cursive handwriting fills the computer screen. It reads, “Spam, Spam, Spam. All I dream about is Spam.”
These words from an anonymous soldier during World War II were captured in a survey conducted by the Army Research Branch of the U.S. War Department. The soldier’s questionnaire response is one of 65,000 included in The American Soldier in World War II, a digital project that recently received a $350,000 implementation grant from the National Endowment for the Humanities.
“What’s great about these sources is they’re uncensored,” said Edward Gitre, the Virginia Tech assistant professor of history who is leading the project. “The soldiers answered the surveys during the war while they were serving — in uniform and in combat.”
Gitre and his students, along with those of Bradley Nichols, a visiting assistant professor in the Department of History, and other participating universities — are transcribing the documents, which are part of the National Archives collections. The goal is to make the surveys available to scholars and the public.
During the war, the Army Research Branch’s social and behavioral science advisors and staff had been looking for a way to create a more efficient and effective fighting force. The survey was a first step.
Gitre now finds that the collection helps humanize the past. When he first started teaching World War II history, he struggled with the immensity of the subject and the challenge of personalizing such a global war and series of events for his students. This project provided an answer.
“I knew the students would appreciate the sources and find them interesting,” he said. “These soldiers were their age, and because these surveys were firsthand and anonymous, the soldiers spoke frankly. It’s not as though the documents are reporting an oral history 40 years after the fact from recreated memories. The gripes the soldiers had are ones most people would understand.”
The National Archives digitized the 44 rolls of microfiche holding the surveys. Now they are available on Zooniverse.org, a crowdsourced social research platform. Each survey goes through three separate transcriptions to ensure accuracy.
The challenge of translating cursive handwriting provides project participants with a deeper interaction with the content.
“This project is an amazing opportunity to experience firsthand the emotion and the humanity of the Greatest Generation,” said Daniel Hammel, who graduated from Virginia Tech this past spring with a degree in political science. “Often we idolize the soldiers and all those who served in the Second World War, so much so that they have a godlike image; they’re knights in shining armor. These documents do not necessarily reflect this vision. Instead, they give us a down-to-earth understanding of what these men were like and how the war affected their emotions, hopes, and dreams.”
When Gitre first found the collection at the National Archives in 2009, he knew he had discovered something special and unexplored. He thought others would find the contents fascinating and wondered how he might help bring these firsthand accounts to light.
“It wasn’t until I was at Virginia Tech that the pieces started to come together,” he said. “I thought this might be a place where I could make something positive happen.”
When Gitre joined the faculty in 2014, two of his new colleagues — Paul Quigley, the James I. Robertson, Jr. Associate Professor of Civil War Studies, and Kurt Luther, an assistant professor of computer science and history — were working on a digital humanities project, Mapping the Fourth of July. With this website as an inspiration, along with advice from LaDale Winling, an associate professor of history, Gitre began the American Soldier in World War II as an experimental project. And it flourished.
Gitre knew the project would take multiple years to complete. Using research funds, he asked the U.S. National Archives and Records Administration to digitize several test rolls of the microfiche. He experimented with having his students transcribe these and beta-tested an open-sourced software transcription plug-in created by the Crowd Intelligence Lab in the Department of Computer Science. Luther, who directs the lab, provided the proof-of-concept testing that helped Gitre secure the National Endowment for Humanities funding.
Now halfway through the transcription process, Gitre has the support of the National Archives and the Social Science Research Council, which have both been crucial from the start. The university is also providing transdisciplinary collaborations.
In addition, Luther and other data scientists at Virginia Tech and the University of Virginia are exploring algorithmic approaches to derive the best data from what will total 195,000 transcriptions. Michael Hughes, a professor in the Department of Sociology, is providing a sociologist’s perspective, while the university’s Advanced Research Computing is hosting the data.
The implementation grant provides funds for the extraction of survey data from out-of-date computer files, completion of the transcription efforts, and creation of a professional open-access website that will be designed by the digital development agency Cast Iron Coding. Together, these efforts will reconstruct an immense portrait of the U.S. Army during World War II.
Although many students are transcribing and reflecting on the surveys, the project requires additional volunteers. Gitre encourages anyone who would like to take part to sign up on Zooniverse.org. The site offers detailed instructions for participants. Gitre also has plans for additional transcribe-a-thons, including one this fall in honor of Veterans Day.
For all those involved, Gitre’s words to his students apply. To volunteer with this endeavor, he told them, is a way participate in a historic effort to preserve the past.
“These are sources mostly unseen since World War II, and we’re putting them out into the community,” he said. “Students like the idea that their role in the project is more than just getting a grade; it’s also public service.”
— Written by Leslie King