The Digital Adjutant in Digital Humanities Context
The Digital Adjutant situates itself within the broader environment of digital humanities insofar as it strives to present another way of using digital tools to convert abstract statistical data into meaningful and manipulable digital visualizations. In this way, the project maintains a similarity to statistical data visualization projects like Data USA, the United States Census Bureau Data Visualization Gallery, and the Civil War Research Database (CWRD).
Data USA defines itself as an attempt to place “public US Government data in your [the user’s] hands” by offering “an open, easy-to-use platform that turns data into knowledge.”1 The platform provides access to an impressive volume of statistical data concerning every locale in the United States through a series of engaging and interactive charts and graphs which are, to a varying degree, manipulable by the user. This manipulation “allows millions of people to conduct their own analyses and create their own stories about America – its people, places, industries, skill sets, and educational institutions.”2 Due to the sheer volume of the data utilized, Data USA cannot realistically use such data to model the entirety of the communities it represents. Because of this, the platform relies on more traditional graphical visualizations that shrink the scale of data down to a manageable size while simultaneously organizing it into a meaningful form accessible at a single glance.
The United States Census Bureau Data Visualization Gallery approaches similar objectives, but with sources that consider the entirety of the nation as a whole. Instead of offering access to community-level data, the USCB gallery provides geospatial visuals in conjunction with more traditional statistical charts and graphs in an effort to make intelligible the mountain of demographic data compiled within decennial Federal censuses. Like DataUSA, USCB states its primary objective as “making data available to the public,” and marks its “Data Visualization Gallery” as “an early part of that effort.”3
Similarly, and perhaps most pertinently, the Civil War Research Database promotes itself as “the definitive online resource for researching the soldiers, regiments, and battles of the American Civil War.” Developed and maintained by Historical Data Systems, Inc., the platform is essentially a digitized online database of “over 4 million soldiers,” boasting “thousands of regimental rosters, and officer profiles” that “will continue to grow as new information is loaded bi-annually.” CWRD provides analysis charts and traditional statistical graphs that enable users to “identify a large-scale trend and then focus down to the regiment or individual soldier,” the developer explains.4 Most interestingly, users deploy the “query tool” to “view a chart analyzing death by disease and can then explore the Regimental Casualty Analysis and the Regimental Assignment charts to research this subject further.” In the end, these charts “can be viewed through a single interface that allows for an immediate visual understanding of complex data.”5
In reality, however, CWRD is less impressive. The sheer scale of data the database contains means that widespread errors are common. The decision to aggregate individual soldiers’ records in order to develop quantitative statistics for analysis of small-unit dynamics ignores the fact that personnel records are often far more fragmentary than unit-level sources like Record Group 393. In fact, CWRD makes no use (or even mention) of organizational sources like company and regimental-level orderly books, morning reports, etc. Instead, the database attempts to produce its own statistics based on fragmentary sources – statistics that often fall far short of those recorded at the time, and offer serious dangers to the researcher. In short, the mass commercialization of access to such data (CWRD is a subscription-based platform) and a lack of a historian’s eye to the approach to carefully compiling and analyzing such data, means that CWRD is a resource that must be used with extreme care. Moreover, of the limited charts and graphs available (none of which – despite promises – can be produced at the regimental level), all are of a traditional line graph style, and there is no attempt to offer visualizations of the database’s statistics in any other format.
Beyond simply utilizing digital methods to transform data into a visual format for researcher and public utilization, The Digital Adjutant also speaks to a broader digital humanities literature concerned with the making of historical arguments through the process of data visualization. Within the Civil War field, Robert K. Nelson’s Digital Scholarship Lab at the University of Richmond remains the veritable field headquarters of digital humanities scholarship. By transforming extant data from manuscript sources into “digital tools and digital media,” DSL’s “Hidden Patterns of the Civil War” suite of projects is able to “uncover and represent patterns that are not easy to find when we look at particular pieces of evidence in isolation and only become evident when we visualize a wealth of evidence in graphs, maps, and models.”6 With projects like Mining the Dispatch, which utilizes text mining and topic modeling methodologies to analyze the vast textual archives of the Richmond Dispatch newspaper across the Civil War years, Voting America: Civil War Elections, which animates a manipulable election result map to illustrate change over time in a number of interactive ways, and Scale and Freedom, which offers maps that have been transformed into 3D visualizations themselves, Nelson and DSL are clearly very much in the business of making historical arguments. Central to these arguments (both those rooted in text and digital mapping) is the process seeking to “uncover patterns in Civil War history that are most evident when we write and think about them while using new tools and techniques.” Such patterns, the DSL lab argues, “enrich our scholarship and suggest stories that still remain to be told.”7
The Digital Adjutant seeks to accomplish much the same objective. By aggregating erstwhile separated company records, and piecing together data at the regimental level that has either been lost to history or was never properly compiled in the first place, the project provides for the recreation and identification of patterns that are at the heart of historical argumentation. For example, the 116th Illinois Volunteer Infantry, which serves as the unit focus of this working prototype, published no regimental history at the close of the war. Only two archived manuscript collections (letters) from the regiment remain extant. The story of the 116th is very much a story that still “remains to be told.” The analytical trajectory of such a story, as well as arguments for how and why such a story progressed as it did, emerges from the data when visualized digitally in a number of formats. By reproducing the visual experience of observing each company on the parade field, offering digital data analysis tools to explore questions of manpower strength, sickness and disease rates, battlefield attrition, and leader-to-follower ratios over time, The Digital Adjutant breathes life back into the musty records of these forgotten military organizations, and opens doors to more comprehensive historical investigations of the same.