Andrew David King
Public Digital Humanities Capstone student, Summer 2020
My Public Digital Humanities Certificate capstone project focuses on one prong of a two-pronged, ongoing endeavor in self-education pertaining to the application of DH methods in textual and literary criticism. My academic background is in philosophy, literary studies, and creative writing; having had plenty of exposure to close-reading and discursive writing, for years I’ve found tantalizing—and intimidating, to be frank—the opportunities for the use of quantitative tools in criticism. Standing between me and the application of these tools to my own research is a theoretical and practical gap. While the question of whether to start with theory or practice first invokes an unnecessary dichotomy (one could begin with both, bit-by-bit), I’ve decided to use my capstone course, already a month or so underway, to focus on the former. I’d like to develop a better conceptual grasp on what it is that literary scholars who use DH tools think they’re doing with them, and of the argumentative landscape—as evidenced by prominent debates in the DH-literary studies sub-field, and by my own lights—ensconcing DH tools and methods in the broader field of literary studies.
The application of DH tools in literary studies contexts—I’ll use the annoying but less cumbersome acronym DHLS to refer to this sub-field from here on out—has found vocal supporters as well as detractors. While acknowledging the platitude that, yes, the emergence of the global information economy and the era of “big data” must somehow change the way literary critics think about the nature of information and text, what’s less clear to me is what changes, if any, these historical shifts necessitate in the nature of literary-critical practice. Does the fact that I can now digitally search a larger portion of published text than ever before in human history automatically falsify or render suspect those critical narratives whose authors didn’t have this ability? Is all pre-“big data” literary criticism outdated and ineffective, a collection of floppy disks dwarfed by the power of solid-state drives, cloud-hosted databases, and eagle-eyed algorithms? Perhaps no one is actually making these straw-man claims. But given the excited rhetoric—understandable in itself—around DHLS, I might be forgiven for asking. “For whatever reasons, be they practical or theoretical, humanists have tended to resist or avoid computational approaches to the study of literature,” notes Matthew Jockers near the beginning of his 2013 volume Macroanalysis: Digital Methods and Literary History. “Though not ‘everything’ has been digitized, we have reached a tipping point, an event horizon where enough text and literature have been encoded to both allow and, indeed, force us to ask an entirely new set of questions about literature and the literary record” (4).
The questions I’d like to answer for myself all relate in some way to Jockers’s use of the word “force” in the above quotation. To what extent are the methods and results of literary critics necessarily changed, or even obliterated, by the emergence of digital methods? How much of what is recognizable at the close of 20th-century literary criticism—marked by a rejection of the preoccupations of the New Critics for the social, the historical, and the liberatory capacities of textual analysis—can remain the base, the foundation, for what a literary scholar and critic does? One recurring temptation is to think that the more data we gather, the more our questions will “answer themselves”; we’ll asymptotically approach the x-axis of certainty the more information we accrue. It’s true that empirical questions demand empirical answers, and it’s thrilling to watch Goliaths felled by the stones of newly-uncovered facts. (One of my more memorable brushes with DHLS involved witnessing just such a takedown—a presentation by a professor whose careful survey of a corpus painted a far more complex picture than an influential critic’s speculative argument about it did.) But insofar as literary analysis isn’t just an empirical game, how much can DH tools and methods aid it? I anticipate that DH tools can help make easier even some of the most basic forms of bean-counting integral to the small-sample-size literary analyses that we all know and love. What might an analysis of trends in Walt Whitman’s diction over the course of his poetic output suggest—and how might those trends, if trends could be found, interact with qualitative arguments grounded in close reading about specific parts of that output?
Although my own research questions are not, as of now, focused on larger social, cultural, and historical contexts in which certain texts were produced and situated, one can imagine the results of DHLS work modifying dusty dicta about what the “most significant” works of a genre, author, period, etc., were or are. Ian Watt’s The Rise of the Novel comes up for scrutiny by Jockers in Macroanalysis for dealing only with Defoe, Richardson, and Fielding. Acknowledging the study as “magnificent,” Jockers articulates qualms about it that seem largely empirical in nature (Macroanalysis, 7). If part of Watt’s thesis has to do with the flourishing of literary realism in the nineteenth century, and how it may or may not have been foretold in the works of these three authors, then “flourishing,” Jockers says, “certainly seems to be the sort of thing that could, and ought, to be measured”—an empirical quantity, in other words (Macroanalysis, 8). But if this is so, Watt’s monograph falls short, for he had “no such yardstick against which to take a measurement. He had only a few hundred texts that he had read” (Macroanalysis, 8). But can there be an empirically-definable concept of “flourishing”? What about “culturally significant”? It would be interesting to uncover that “the most realist novel of the nineteenth century,” as defined by a set of formal criteria instances of which an algorithm could recover from as vast a corpus of nineteenth-century texts as one could access, was a more or less unknown work. But would such a discovery demand that literary history be rewritten? Would it show that critics like Watt had been misguided, or simply wrong in their judgments about what works were “representative”? Would it necessarily divert the attention of cultural historians away from works whose patterns of cultural influence—regardless of how “representative” of a formally-defined genre or style an algorithm proves them to be—had long been established?
Other questions I’d like to explore include how the results of large-scale DHLS analyses are then presented and interpreted—in particular with respect to digital representations of results—and whether these interpretations don’t in some ways reproduce the “problem” of close reading and small sample sizes that the studies that birthed them were intended to go beyond. At some point, Borges tells us, a sufficiently detailed map becomes the territory itself; what’s the right level of granularity, then, for a given DHLS project, one that’s able to both make use of large amounts of data as well as face head-on the questions it can’t settle by further data-gathering? And for my own sake, for my own research interests, I plan to read and think more about what DHLS can offer formalist, close-reading-oriented analyses of literature, especially poetry. Many of these questions involve getting clearer on which parts, if any, of our work as literary critics and scholars are empirical and which are non-empirical or interpretive. Is there a point at which a thesis about, say, the uses to which Wallace Stevens put the regular stanza becomes falsifiable? Or are there as many ways to look at a blackbird (sorry!) as one can imagine? If the latter were the case, the interpretive world of close reading and formal analysis would have to be thought of as governed, in some sense, by the laws of some non-empirical realm. But, barring the tendency of poets to discuss inspiration in quasi-divine terms, if we dismiss the possibility of the mystical—thinking, maybe, with Stevens, “What is divinity if it can come / Only in silent shadows and in dreams?”—we might then ask: what is the epistemic relationship between specific, material parts of literary works and the broader, more abstract, more philosophical arguments we make about poems, books, and corpora?
Into August, I’ll continue reading my way, like a mole underground, through these questions, pushing past roots and rocks. As for Watt with his pitiable “few hundred texts,” I—not a computer, at least not yet—should be so lucky.