This the first of a series of posts by Regan Chasek, an undergraduate researcher with the Walt Whitman Archive at the University of Nebraska-Lincoln. Chasek has received a UCARE research award to create digital visualizations of the data from the Walt Whitman’s Poetry Reprints, 1838-1892 project. Her posts here will trace the issues–ethical, methodological, and interpretive–associated with data visualization for humanities resources.
The Ethical Implications of Visualization
By Regan Chasek
During my time as a student at UNL, I’ve come to notice a few distinct “genres” of scholarly essay. Johanna Drucker’s “Humanities Approaches to Graphical Display” falls into the category I call, “what does it all mean, man?” I label the category this way to invoke the image of an intoxicated person who is questioning the very fabric of reality. While my hypothetical stoner may be asking questions like, “why are grapes round?,” Drucker asks the question of how we can use visualization tools in a way that aligns with the spirit of the humanities.
Before reading this essay, I hadn’t thought about how using visualizations for humanities might be different from the usual statistical use. However, Drucker explains that humanistic knowledge production is based on interpretation, not mere observation. The usual bar graph or pie chart will present its data as fact, which is not conducive to the humanistic mode of knowledge creation. For visualization to be of use to the humanities, we must first figure out a way to display “capta,” or information that is “taken” in a potentially biased way, unlike hard data. Using the word “capta” acknowledges the fact that the information presented cannot possibly constitute the entire truth – somebody had to gather the data, somebody had to create the categories, somebody had to make criteria for each category, and so on. Essentially, Drucker urges the reader to think about how to present information in a way that is interpretive instead of being a mere observation.
On a purely abstract level, this was hard for me to imagine, but Drucker’s examples helped to clarify her ideas. Her most interesting example, to me, was the bar graph of men and women in separate countries. At first, it seemed clear-cut – men and women were in their own, rigid categories, and the countries were their own distinct categories. However, Drucker complicates this by asking the reader to consider the presence of nonbinary individuals, or of people who cross the border regularly, or undocumented immigrants that are not considered citizens. Those who did not fit into neat categories were represented by a dot or bar extending into the space between bars, transgressing the preconfigured boundaries of the graph.
Britt Rusert’s essay “Visualizing the Slave Trade” is a specific application of the kind of critical approach Drucker promotes. Rusert analyzes current methods of creating graphical representations of the North American slave trade, particularly a project titled “The Atlantic Slave Trade in Two Minutes.” While critiquing this project, Rusert directly mentions Drucker’s warning that data visualization comes with “baggage” from its origin as a tool of the natural and social sciences. Such visualizations encourage “representational simplicity,” which is, of course, not always conducive to representing things like the complex horror of the slave trade. These observations by Rusert made Drucker’s argument clear to me.
Because my UCARE project is one centered around data visualization, these essays can be helpful while considering the most ethical way to present the data. Even though a project visualizing Whitman’s poetry reprints may not have the same, obvious implications of a project visualizing the slave trade, it’s still wise to consider the implications that every graph, chart, or map will have.