“In some ways, when people hear big data, that causes a lot of anxiety. How do we handle it all? How do we navigate it? How do we make sense of it? There’s a sense that it’s just enough data to be overwhelming.” Researcher and designer Yannis Loukissas posed these questions to assembled Dumbarton Oaks staff and fellows as he delivered a lecture, “The Life and Death of Metadata,” on December 11. Currently an assistant professor of Digital Media in the School of Literature, Media, and Communication at Georgia Tech, Loukissas formerly served as a principal in the metaLAB at Harvard’s Berkman Center for Internet and Society. There, he spent two years working on “The Life and Death of Metadata,” for which he analyzed the records of 70,000 trees, shrubs, and plants accessed by the Arnold Arboretum—Harvard’s “living museum,” located in the Jamaica Plain neighborhood of Boston—over a 140-year period.
The project, Loukissas explained, has two main features: a vast infographic displaying accessions by the Arboretum, each month in each year from 1872 to 2012; and an essay that integrates the displayed data with interviews Loukissas conducted with Arboretum staff and scholars. By uniting data with more traditional historical methods, Loukissas built a case study for understanding the role that metadata can play in illuminating history, both discrete events that become evident when comparing data as well as larger trends revealed by changing patterns of data collection and dissemination.
Walking his audience through the project, Loukissas was able to point out the ways in which visual representations of data could aid historical understanding. Showing a radial version of the data with dates of the year displayed in a circle, Loukissas pointed to blank spaces on the graph that indicated when the Arboretum had made no accessions. “Over here, you can see an interesting gap towards the end of December. That’s Christmas. Here there’s a blank ring that goes all the way around in the ’40s. That’s World War II. . . . You have this sense of the ways in which this data set is rooted not just in a scientific institution but in a broader culture,” he explained.
At the same time, he suggested, the inclusion of the interviews with Arboretum staff is aimed at providing context, as data without context can be unreadable to those who were not present when it was produced. “I hope that projects like this can reveal the ways in which data are partial and sort of unstable,” Loukissas offered. Data are also challenging media, he continued, because of their ability to constantly be remade and altered: “We’re working in a palimpsest form,” he explained.
Ultimately, in the search for creating “humanist perspectives on the technological world,” as Loukissas calls projects like his own, these characteristics of data demonstrate how they are cultural artifacts that come about in a particular place and time. For Loukissas, it is this search for context that drives his work: “We’re only beginning to uncover how we see data from this digital perspective, how we might think of them as entities that are rooted in a place, in a culture, in a community,” he concluded.
Read more about Loukissas’s project, “The Life and Death of Metadata,” here.