Quantitative approaches to literature have a history stretching back to the nineteenth century. But for most of that time, numbers seemed limited to straightforward empirical questions: How many feet in a line of poetry? How many books sold in 1950? Neutral mathematical rules were obviously useless for interpretive questions that required perspectival reflection, including reflection on what Gadamer called the “fore-meanings and prejudices” brought to the process by the interpreter herself (269). Even as quantitative methods grew more prominent in the first decade of this century, their opposition to interpretation seemed unchanged. Franco Moretti, for instance, suggested that numbers would be useful for “the explanation of general structures” rather than “interpretation” (91).
But that opposition has lately broken down. Humanists who use numbers don’t necessarily accept that quantitative methods are neutral or universal. In Data Feminism, for instance, Lauren Klein and Catherine D’Ignazio emphasize that “all science … is undertaken by individuals, each with a particular standpoint.” Humanistic methods of data analysis might need to reflect standpoint theory. This premise is echoed in Ted Underwood’s Distant Horizons, which argues that machine learning is “very good at representing perspectival problems” and “deeply compatible with an antifoundational approach to interpretation” (xiv, 67). Oddest of all, statisticians increasingly seem to agree, warning researchers that “multiple perspectives and context dependence are actually basic conditions of scientific inquiry” (Gelman and Hennig, 969, 993).
So the wall that once separated quantitative analysis from hermeneutics seems to be crumbling on both sides. But this is not to say that scholars share any consensus about what will fill the space. Literary research using numbers may aim at goals ranging from formalism to Marxist critique. Meanwhile, poststructuralists—including many queer theorists—often remain skeptical about the interpretive affordances of numbers, since familiar kinds of quantitative analysis still seem to require stable categories and variables. Johanna Drucker, among others, has argued that a truly humanistic approach to data will require the invention of new methods. Finally, there are still critics who deny that computational methods can be used for literary interpretation at all—a perspective voiced most recently by Nan Z. Da.
We have assembled three speakers with different perspectives on this debate, in order to explore a range of possible positions. Speakers will keep to less than fifteen minutes each, leaving time for Jonathan Arac to respond in a sympathetically skeptical way, framing recent debates in the longer history of critical theory, and glancing forward to their future.
Laura McGrath will start the conversation by arguing that “Sociological Literary Theory Needs Computation.” Katherine Bode has argued influentially that quantitative datasets require better social and historical grounding in order to tell us anything about culture. McGrath will propose a corollary: computational models are necessary for sociological literary theory to understand the processes and procedures of mediation. Relying on a case study from contemporary publishing, she will show that computational models can help us see into the “black box” of complex concepts like influence or symbolic capital that are central to cultural administration and sociological criticism.
In “Difference and Data,” Richard Jean So will address a common misconception among literary scholars: that quantitative analysis relies on fixed categories and is thus inappropriate for the study of literary texts. Critical theory has often aimed in particular to destabilize or deconstruct categories of identity, such as gender and race. So will use a brief case study focused on race in US fiction to show how sophisticated computational models drawn from new developments in machine learning can be used precisely to expose the contingency and instability of socially constructed categories, underlining the compatibility of computation with existing critical theories of race and difference.
In “Beyond Distance, Beyond Theory,” Lauren Klein will explore the intellectual inheritances that link quantitative literary analysis to critical theory, without necessarily seating it in the chair theory used to occupy. Klein begins by positing that critical theory has lost its once “outsized role” in the profession, leaving literary scholars with a valuable set of concepts and techniques for destabilizing cultural categories and for probing the limits of literature as a representational form. Quantitative literary analysis inherits the same task: it should pursue the nuance and complexity in its objects of study while remaining aware of what they cannot disclose. Klein uses quantitative means to surface the visible and invisible labor of Black women newspaper editors in the nineteenth-century United States, but also argues that the fullest sense of their labor can be gained through an approach that deconstructs the concept of labor, and then looks to what additional forms of work remain outside that concept—and outside the text. Reflection on the limits of representation can lead quantitative literary analysis beyond the axis of “distant” and “close."
References
Bode, Katherine. “The Equivalence of ‘Close’ and ‘Distant’ Reading; or, Toward a New Object for Data-Rich Literary History,” Modern Language Quarterly, vol. 78, no. 1, 2016, pp. 77-106.
Da, Nan Z. “The Computational Case against Computational Literary Studies.” Critical Inquiry, vol. 45, 2019, pp. 601-39.
Drucker, Johanna. “Humanistic Theory and Digital Scholarship.” Debates in the Digital Humanities, edited by Matthew K. Gold. University of Minnesota Press, 2012. http://dhdebates.gc.cuny.edu/debates/text/34
Gadamer, Hans-Georg. Truth and Method. Second revised ed. Trans. Joel Weinsheimer and Donald G. Marshall. Continuum, 1999.
Gelman, Andrew and Christian Hennig. “Beyond Subjective and Objective in Statistics,” Journal of the Royal Statistical Society A, vol 180, no. 4, 2017, pp. 967-1033.
Klein, Lauren and Catherine D’Ignazio. Data Feminism. MIT Press Open, 2018. https://bookbook.pubpub.org/data-feminism
Moretti, Franco. Graphs, Maps, Trees: Abstract Models for Literary History. Verso, 2005.
Underwood, Ted. Distant Horizons: Digital Evidence and Literary Change. Chicago: University of Chicago Press, 2019.