Digital Boot Camp for Military Historians

An after action report

HUMANITIES, January/February 2015, Volume 36, Number 1

“My brain is full,” said Jeff McClurken, who was sitting behind me.

I felt the same way. For the past two days, McClurken, a professor at the University of Mary Washington, and I had been part of a workshop on digital methods for military historians. The “boot camp” was designed to give scholars interested in digital humanities a chance to explore the possibilities for their scholarship while learning how to use some common tools. We’d cleaned data, mapped correspondence networks, georectified maps, and created online exhibits. And that was after lunch. In the mornings, we learned how scholars were already using digital tools to expand the boundaries of military and diplomatic history. Being mentally tired was almost a badge of honor after the intensive tutorials and repeatedly saying, “Whoa.”

The workshop, which took place October 10 and 11, 2014, was hosted by Northeastern University. A collaboration between NULab for Texts, Maps, and Networks, NEH’s Office of Digital Humanities, and the Society for Military History, it was funded under NEH’s Standing Together initiative, which seeks to highlight the experiences of veterans.

The workshop was the brainchild of Heather Streets-Salter, chair of Northeastern’s history department, and Abby Mullen, a doctoral student there. “I wanted to give military historians a chance to put their foot in the door with new technologies,” said Mullen, who is writing her dissertation on the Barbary Wars. “Coming from a digital history and military history background, I noticed there wasn’t a lot of overlap between the two disciplines and I thought this would be a helpful way to bridge the gap.”

Social and cultural historians have embraced digital humanities, jumping at the chance to apply analytical tools to large datasets like the census, court cases, and literary texts. Mapping the Republic of Letters has blazed new paths of scholarship on Voltaire and Benjamin Franklin by using network analysis to delve into their vast correspondence networks. Visualizing Emancipation pairs historical records with Geographical Information Systems (GIS) mapping to show how and when slavery began to collapse during and after the Civil War. The Proceedings of the Old Bailey has put 197,745 trials at London’s central criminal court into a searchable database, making it easier to sift through Britain’s legal history and use the records, which offer a trove of information about the lives of average people.

The field of military history, on the other hand, has been slow to embrace digital tools, which is ironic, given how high tech the military has become. It’s also a missed opportunity. Troop orders, muster rolls, draft records, and battle movements are ripe for data mining and mapping. And work that once had to be painstakingly done by hand, like plotting the delivery of orders up and down the line or sketching the flow of a battle, can now be done faster and with more flexibility by computer.

The workshop at Northeastern was designed to help the field catch up. It also reflects a commitment by NEH’s Office of Digital Humanities to provide opportunities for scholars to increase their knowledge of a growing field and learn about new methodologies. That’s how fifty of us, ranging from graduate students to senior scholars to museum professionals, found ourselves fighting for space on power strips to keep our laptops running. In most cases, we were novices rather than master coders, who were looking to see what digital tools could offer our work. I was there both as a reporter for Humanities, and as someone who works on World War II and made the switch to digital for my archival research.

On Friday morning, Robert Warren, a postdoctoral fellow at the Big Data Institute at Dalhousie University in Halifax, showed how he and other scholars from the Muninn Project are extracting data from World War I records, ranging from memos to war memorials. The datasets are being used, among other things, to track relationships between soldiers and regiments to armies in the field—and map their locations as the war progresses, using GIS. Micki Kaufman, a graduate student at CUNY Graduate Center, shared her work on the Quantifying Kissinger project. Kaufman uses visualization tools to parse the 17,500 pages of memorandums documenting in-person and telephone conversations conducted by Kissinger while working for the Nixon administration. The resulting graphs and visualizations depict when concern about Daniel Ellsberg spiked or the ebb and flow of Vietnam policy, while also showing what else was being discussed at the time.

The kind of work that Warren and Kaufman do requires a dataset—and the results of a project are only as good as its data. On Friday afternoon, Jean Bauer, associate director of the Center for Digital Humanities at Princeton University and the lead developer of Project Quincy, showed us how to clean our data. This does not involve Windex. The process entails carefully making changes or additions to a dataset to introduce consistency while still preserving complexity. For example, a person with the name “William Barnes” might be noted in the records as “William,” “Wm.,” “Will,” or even “Bill.” By tagging all of these iterations as “William Barnes,” any subsequent searches or visualizations for “William Barnes” will display all the associated records.

Once we’d tidied up our practice dataset, Bauer, who graciously answered all questions, gave us a tutorial on Gephi, an open source visualization tool. Across the room, correspondence networks began appearing on screens in every shape and color—or not, in my case. I had chosen the wrong option and my network inverted instead of expanded. When I tried to fix it, I made it worse. (Sigh.) Not to be deterred, I started over, and soon colorful lines streaked across my screen.

Saturday started rainy and cold, but we didn’t care, as we were tucked back into classrooms at Northeastern, coffee and tea perched beside our laptops. During the morning session, Alberto Giordano, professor of geography at Texas State University, walked us through Geographies of the Holocaust, which uses historical data to visualize the geographical evolution of the Holocaust. The project has mapped the expansion of the SS concentration-camp system across Europe, traced the arrests of Italian Jews, and plotted changes to the Budapest ghetto. As someone who has worked on the Holocaust, I was transfixed by how the visualizations showed the evolution of Nazi policies, but I was also slightly sick to my stomach knowing that the points on the screen represented someone’s internment or death. Giordano also emphasized precisely this point: You must never lose sight of the human factor behind the data and its representations.

Ed Triplett, a PhD student at the University of Virginia, also shared how he was using digital tools to reconstruct the viewsheds—what you could see from a hill or turret—of fortresses across medieval Spain. Being able to keep an eye on a pass or an approach was important for deciding where to locate a structure or how to arrange artillery. When Triplett put up a slide that showed a camera attached to a kite to illustrate how he took aerial photographs of crumbling fortresses for use in 3-D reconstructions, there was an audible gasp of wonder from the room. (There were also some murmurings about whether or not it was too late to consider a career in architectural history.)

In the afternoon, we once again became students and could attend two of the three tutorials offered. Under Mullen’s guidance, I learned how to georectify a map, using the open source MapWarper. After you upload an image of a historical map, you look for places of commonality on the old map and the modern map—a street corner, the outline of a park, an intersection—and drop pins marking the locations. Once the pins are in place, the software morphs the historical map to the dimensions of the modern map—and layers them together so you can see the differences and commonalities. For historians looking to describe how a city has changed over time or figure out what once stood where, the mashup is invaluable.

For my second session, I attended a tutorial of Neatline, which allows users to create annotated online maps and timelines. David McClure, one of the lead developers of the tool, taught the tutorial. Over the next hour and a half, a flat timeline of Napoleon’s advance into Russia turned into a dynamic webpage with text pop-ups, artwork, and links to further reading.

A third session was offered by Scott Nesbit, assistant professor of digital humanities at University of Georgia, on using ArcGIS to do thematic mapping. Nesbit is also one of the directors of Visualizing Emancipation. Since I missed his session, I asked him why a historian might want to do thematic mapping. “We have a lot of information and sources about events that happened in a particular place and time,” said Nesbit. “Knowing more about those relationships is bound to bear fruit.” How close, for example, were railroad tracks and depots to battlefields in the Civil War? Nesbit, however, cautions that using tools like GIS can be a challenge for a historian. “Our evidence is often fuzzy, and this program demands precision.”

Along with the admonitions to back up our work and take good notes about our methodology, the presenters also emphasized the need to think about the choices we made along the way.

“I was impressed with the philosophical element of some of the discussions and the warnings that the presenters gave. ‘Remember, this isn’t a neutral process and you need to be very clear about what you’re privileging,’” said Daniel Franke, assistant professor of history at the U.S. Military Academy. Indeed, just as a book has an argument, so does a dataset. The scholar makes choices about what to include or not to include, which influences the results.

Collaboration was also a big theme. Warren, Giordano, Bauer, and Nesbit all collaborate with other scholars on their digital humanities projects. But for historians used to thinking of themselves as solo practitioners, the idea of teaming up with someone can be daunting.

“Collaboration isn’t something that is foreign to historians. It’s just not something that is often acknowledged,” said Mullen. “The acknowledgments page talks about the archivist and the person who proofread your manuscript. Those same kind of connections are forged by a digital project. It’s not just the expertise that you bring to the table, but also the expertise that others bring as well and that all works together to become a whole.”

Mullen concedes that venturing into new fields can be taxing, because it requires historians to stretch themselves and learn new methods. But the reward comes with being able to ask new and different questions—and the results they produce. As the workshop wound down, people were already talking about how they could start using what they learned. I even passed on some World War II ship logs to Warren, who was interested to see what he might be able to mine from them.

I also had a chance to ask Bauer if she had any advice for military historians who were thinking about wading into digital humanities but hadn’t made the plunge yet. “Every expert was once a beginner,” she said. “If this is something that you really feel your sources are calling for, if you found projects that are making use of methodologies in ways that would allow you to address these questions, then give it a try. Be brave.”