Brad Rittenhouse is a PhD candidate at the University of Miami specializing in 19th-century American literature and the digital humanities. His work thinks about literature as data, and looks at the intersection of literary aesthetics and information management techniques. He is also working on a DH project at UM’s Center for Computational Science, where he is developing a methodology for quantitatively identifying instances of informationally “thick” literary passages.
The inaugural Digital Humanities + Data Journalism Symposium recently took place at the University of Miami, from September 29 to October 1st, drawing together a diverse crowd of academics, journalists, and many in between. As conference convener and Knight Chair in Visual Journalism at UM Alberto Cairo informed the audience in his opening remarks, the somewhat speculative event topic was inspired by a February 2012 tweet by Digital Public Library of America Executive Director Dan Cohen, which hopefully conjectured that “digital journalism and digital humanities are kindred spirits, and that more commerce between the two could be mutually beneficial.” Delivering the first keynote of the weekend-long proceedings, Cohen drew equal inspiration from Thucydides and Reese’s Peanut Butter Cups, lightheartedly suggesting the latter as a metaphor for the symbiotic (and potentially delicious?) roles digital humanists and data journalists could play in our increasingly informational world. In referencing Thucydides, he developed one of the salient themes of the weekend, the notion that digital scholars and journalists alike were primarily motivated by the quest to rise up from data to understanding, to construct knowledge from the complicated and overwhelming.
The other speakers of the weekend included prominent scholars and journalists sharing impressive (sometimes to the point of being daunting) works, methodologies, and questions. In terms of numbers, presenters skewed toward the data journalism side, though most had ties to universities and all of the presentations were academically rigorous. I hesitate to make this observation, but with such an unorthodox conference model, bringing together the professional and academic worlds, I feel it is important to give a sense of the proceedings as a whole. While the symposium was premised upon an exploration of the perhaps speculative connections between DH and DJ, the modes of inquiry were perhaps more similar to those associated with academic rather than professional conferences.
Scott Klein of ProPublica, the keynote on the data journalism side, kicked off the first full day with a paper “Nerds Among the Mathphobes.” In it, he explored the tensions between digital and traditional journalists, and the larger epistemological issues faced by data journalists in telling stories through interfaces. While many on the DJ side stressed the curatorial role of digital journalists, creating clear narratives from messy datasets, Klein suggested that the goal of data journalism was to design interfaces that allowed users to generate their own stories from information.
This “dirtier” vision of data journalism, which cedes authorial power to the user, overlapped with a common narrative on the DH side, where, as speaker and historian Ben Schmidt suggested, developers are often given more freedom to create projects which are ugly and not immediately comprehensible. Focused on the interrelations of the digital humanities and data journalism as disciplines, Schmidt provocatively asked whether the simplification work stressed by many data journalists was in fact a worthy goal, suggesting along with Klein that exploration rather than explanation might be a better heuristic in both the digital humanities and data journalism.
Continuing this theme, Lauren Klein of Georgia Tech presented a paper entitled “Speculative Designs (and Serial Killers): Lessons from the Archive Of Data Visualization,” which explored the eighteenth- and nineteenth-century prehistories of visual representations of information. Landing upon the strange and often inscrutable data visualizations of 19th century educator Elizabeth Peabody, Klein suggested that bias, and in particular gender bias, has historically limited the adoption of what she calls “speculative designs” in the visual rendering of data. While the clean designs of progenitors like William Playfair, which Klein also explored, seem presciently familiar to present-day observers, Peabody’s more intricate designs, which stress the meticulous exploration and manipulation of information, run counter to the ethos of explanation so central to much work in data viz.
While most of the aforementioned speakers fit rather firmly into either DH or DJ traditions, others operated on the borders of the two disciplines. Joint presenters Liliana Bounegru and Jonathan Gray, for instance, who both hold multiple appointments across the DH and DJ worlds, suggested arenas and methodologies for collaboration between the two disciplines. Haile Owasu of mashable.com, too, proposed points of contact in a talk that linked natural language processing (NLP) techniques to groundbreaking work in the analysis and production of targeted journalistic content. Other speakers included Jacqui Maher of Condé Nast, Joe Germuska of the Northwestern University Knight Lab, Geoff McGhee of the Bill Lane Center for the American West at Stanford, and Francis Tseng a prominent designer and data developer. Primarily working out of so-called alternative-academic or academic-adjacent sectors, they presented some of the most radical visions for work that is both computational and humanistic, suggesting points of overlap in far-flung fields like simulation and environmental policy.
Attendees also participated in breakout sessions led by Cairo and other digital experts, including Carnegie-Mellon University’s Scott Weingart and the University of Miami’s Paige Morgan and Abraham Parrish. Some of the sessions, like Cairo’s and Weingart’s, provided crash courses in large topics, like visualization and social network analysis, respectively, while the others, led by Morgan and Parrish, focused on more detailed explorations of GIS and OpenRefine.
Per the DH+DJ Symposium website, located at http://dhdj.com.miami.edu/, work has begun on the 2017 iteration of the conference, and details will be forthcoming soon.