The CDH is accepting proposals from faculty and staff for our 2024–25 Collaborative Research Grants. Grantees are matched with CDH staff with relevant expertise—often Research Software Engineers (RSEs)—to work on short-term (approximately six-month) projects that integrate methods, sources, or critique from the humanities with computation, data science, machine learning and AI, and/or design. Those interested in applying now or in the future are invited to an information session on February 7 at noon at the CDH (B-Floor, Firestone Library); a Zoom link is available upon request.
So, what does an RSE do?
Let's keep it simple: “Research software engineers work on research projects using software engineering,” explains CDH RSE Ryan Heuser. Recent CDH RSE collaborations have resulted in the Geography of Taste, a prototype interactive map for exploring data from the Shakespeare and Company Project (an earlier long-term partnership), with Joshua Kotin (English); and simulations using agent-based modeling to explore risk attitudes and decision making, with Lara Buchak (Philosophy). Lead RSE Rebecca Sutton Koeser adds that at the CDH, RSEs “have enough context on humanities research to understand the goals of a project and help faculty collaborators figure out what’s possible, what’s exciting, and figure out together how to get there.” Both Heuser and Koeser earned their Ph.D. in English, whereas our newest RSE, computer scientist Laure Thompson, has worked with everything from medieval manuscripts to science fiction novels.
My project is to figure out what happens when you incorporate non-standard risk-attitudes into game theory. I’ve been interested in this question for years, and tried unsuccessfully to learn the programming skills to do it myself. Rebecca [Koeser] supplied the expertise, and it was also helpful to be able to talk to her about the ideas behind the work. It was a terrific partnership.
–Professor Lara Buchak
How does a collaboration with an RSE work?
Collaboration begins with faculty and CDH staff working together to plan and write a formal document called a charter. “Chartering is very important for scoping projects, clarifying roles and expectations, and planning out major checkpoints or goals,” Koeser says. Determining goals and scope is all the more crucial for short-term collaborations, which often comprise a small portion of larger projects. Once work on the collaboration begins, collaborators typically meet two to four times a month, with communication on Slack between meetings. Explains Koeser: “If we’re developing software, we need collaborators to review what we’re building. It’s always a conversation with a lot of back and forth—the whole point of these collaborative projects is that none of us could do the project on our own, we need the mix of skills and expertise.”
My favorite part of working with faculty is the encounter of minds and abilities that fosters mutual respect and true collaboration.
–CDH RSE Ryan Heuser
How do I know if a project is sufficiently developed to begin work with an RSE?
If the project includes a lot of data, notes Koeser, “the data needs to be sufficiently curated or gathered (and any rights issues or agreements established) in order to support the software engineering or computational work planned.” If you're not to that point yet, we can help you find on-campus resources to get there. Consider scheduling a consultation or reaching out to one of our campus partners.
If the project is not data-intensive, you should have a clear research question in mind. “We need some idea of what software we will build or what problem needs to be solved, even if we don’t have the specifics," she says. Adds CDH Faculty Director and Associate Professor of English Meredith Martin, who directs the CDH’s flagship project, the Princeton Prosody Archive: “The book proposal is a good model—what are the projects or outputs that are similar to or would be in conversation with what you are hoping to build with the CDH? What gap will this fill?”
If your project doesn’t look like something CDH has worked on before, that doesn’t necessarily mean we can’t take it on. Buchak explains that “my project was very different than some other projects advertised on [the CDH] website, and to my knowledge there hasn’t been a lot of collaboration with philosophy before. But it ended up working out really well and . . . [we] both learned a lot.”
And, remember, says Koeser, “when you submit your proposal, that isn’t the end; we’ll work together with your project team and CDH staff with project management, project design, and technical expertise to shape whatever you propose into something ambitious and achievable in our time frame.”
Know what questions you want to answer and the audiences you want to reach. Know as much about your data as possible—its gaps and shortcomings, its potential.
–Associate Professor Joshua Kotin
How can a CDH collaboration advance research?
For Martin, “the best RSE collaborations result in software that has an immediate impact on your work but has the potential to impact work by other scholars in your field or even beyond.” For example, Kotin notes that his work with Heuser on the Geography of Taste “allowed me to both visualize—and more importantly—analyze my data in new ways. The application we built together allows researchers to understand connections between Shakespeare and Company—its members and books—and the geography of Paris.” Moreover, a recent CDH blog post looked at the research impact of the Princeton Geniza Project. Adds Koeser: “Most collaborative projects will have a number of different scholarly outputs, and my colleagues and I are actively thinking about how to make those different outputs more visible and citable.”
RSE collaborations can also change how scholars think about their work more broadly. Martin’s collaborations on the Princeton Prosody Archive (PPA) are crucial to her forthcoming book Poetry’s Data. Without these partnerships, she explains, she “would not have been able to conceive of the way that design makes arguments, or to think about how the search navigation was part of the argumentative structure of the data in the PPA”—ideas she engages with in her book.