|
|
Steph Buongiorno, Emory University
Many computational tools used in historical research were originally developed in other disciplines and later adapted for history. As historians increasingly work with larger and more computationally intensive datasets, however, the field requires methods designed specifically for the scale of historical analysis. "Design-based history research"—in which tool development is itself a research goal—is therefore a necessary paradigm for addressing questions posed by large-scale data, and "design-based history pedagogy" is equally important for preparing students to contribute to that work. This paper describes the central challenges involved in developing The Dissentometer and the design-based history pedagogy that emerged through the project. Rather than applying an out-of-the-box tool to a historical question, Dissentometer was developed with a team of undergraduate students to create a method for analyzing how historical narratives converge and diverge across large-scale multilingual corpora. This kind of work, however, presents a central challenge: as the design evolves, knowledge of the system becomes distributed across the team, so that each member holds only partial and unequal knowledge of the project as a whole. This complexity arises from the interdependence of multiple components, design decisions, code scripts, and iterative stages of development. Design therefore becomes a fluid and evolving process of discovery rather than a simple application of existing technologies. This paper shows that such distributed and partial knowledge is a feature, not a bug, of design-based history research and proposes a collaborative team architecture for coordinating decentralized knowledge across diverse team members and bringing complex digital history projects to completion.
No extended abstract or paper available
Presented in Session 111. Text Mining Across Space, Time, and Languages