[Thoughts delivered at the "What Is Data in Literary Studies?" roundtable at the Modern Language Association Convention in Chicago, IL on 10 January 2014. Roundtable organized by James English.]

Addressing the question, "what is data in literary studies," offers the chance to enlarge our interpretational procedures to include new methods and materials. But also to apply existing methods of analysis to new materials and questions. Quantitative approaches to archives and texts developed by digital humanists have offered one such expansion. These approaches often treat literature as a data mine. In response, I propose that literature is a heuristic for managing and conceptualizing data. In one strain of my current research on how environmental media make claims to knowledge through their form, I locate this heuristic function in climate change fiction and visual culture.

Novels such as Barbara Kingsolver's Flight Behavior, Nathaniel Rich's Odds Against Tomorrow, and Kim Stanley Robinson's Science in the Capital trilogy comprise the knowledge infrastructure of climate crisis. Now, to say that literature is knowledge producing does not mean that it is all-knowing. Just as climate modelers shelve the Enlightenment ideal of certain knowledge,[1] climate novelists incorporate data with a robust helping of uncertainty. But as part of a knowledge infrastructure, narratives of climate change variously organize and manage data as part of their aesthetic. Devices for information management, they inspire the following methodological question:

How and why might we define the epistemic as a kind of aesthetic?

Kingsolver's 2012 book highlights a version of this problem. The protagonist, Dellarobia, is a stay-at-home mom and lifelong resident of an economically-failing town in Appalachia. About to commit adultery, she stumbles on a wondrous, even miraculous, ecological event: a massive population of monarch butterflies draped like a throbbing, flaming curtain across the forest. Common to other regions, the phenomenon is out of place in Feathertown, Tennessee, but is all the more astounding for that. When an ecologist and his team descend to study the errant migration, Dellarobia learns that the butterflies are a dreadful splendor. In a scene narrating procedures of data collection, she considers that, "if these butterflies were refugees of a horrible misfortune, there could be no beauty in them" (143).

Yet Flight Behavior is a testament to the compatibility of dismaying data and aesthetic beauty. It features dramas of ignorance and attention, of observation and heedlessness, that integrate scientific data and aesthetic experience. The visiting lepidopterist attempts to separate these domains and pronounces, "'Our job is only to describe what exists'" (148). But the novel closes the sharp divide between scientific and aesthetic description. It shows that knowledge only arises through strategies of seeing-as, specifically focalization and simile.

This and other climate novels spark relays between the epistemic and the aesthetic that, for me, invite two inquiries:

First, what are the "epistemic virtues" that fiction espouses, adapts, or challenges? I borrow this concept from Lorraine Daston and Peter Galison's study, Objectivity. They define "epistemic virtues" as "norms that are internalized and enforced by appeal to ethical values, as well as to pragmatic efficacy in securing knowledge."[2]

Broadening the concept of epistemic virtues to the literary field, then, leads to a second inquiry: in what ways do aesthetic features promote knowledge values? These values might be accuracy, humility, subjectivity, or transparency.

These are large questions that these brief notes can't adequately tackle. Instead, I want to highlight a key narrative device for aesthetically disseminating data and establishing epistemic virtues. That device, I argue, is description.

Description is at once a key site for data dissemination in a text and a site where things are made to matter. It calls on readers to attend to the things of the world and, increasingly, to the numerical data that reference those entities. In Cynthia Wall's account, it is a form of verbal visualization that, emerging from a process of selection, "makes something visible, sets it forth, extracts it from its surroundings, and jabs a finger meaningfully at it."[3] Description makes things matter then both by manifesting materialities and by giving them affective and epistemic significance.

In essence, description is a narrative apparatus of information and attention control. Re-examining this instrument of attention—especially in contemporary works that bridge science and art—we might theorize data management within literary studies using extant tools of narrative and formal analysis.