Critical Spatial Scholarship

There was a brief moment as I read Matthew Wilkens' contribution to Debates in the Digital Humanities where I felt swept along in the dream of a data driven English department. I imagined undecided freshman filing into their undergraduate education. They were confronted with an English department split into two tracks, one traditional the other digital, "data driven". The two tracks shared, of course, a core curriculum. Everyone read a bit of Shakespeare, everyone took at least a basic literary analysis course, and everyone at least heard the name "Foucault" in the same sentence as "sexuality" at some point in their education. Where the tracks differed however was in their commitment to the tradition of close reading or, conversely, to "distant" reading, perhaps even literary mapping (though Moretti's work gives us reason to question the validity of the word "map" to describe his kind of scholarship). The two tracks shared some classes together and spoke a similar scholarly language, but they also had the opportunity to specialize into the particularities of their particular reading commitment. I imagined something similar to a university's multiplicitous engineering offerings: computer, chemical, mechanical, electrical.

My reverie came upon me as I read Wilkens concluding remarks on the necessary displacement of scholarship that must accompany a move by the humanities into broader textual analysis. He says: "If we do that--shift more of our critical capacity to such projects--there will be a couple of important consequences. For one thing, we'll almost certainly become worse close readers." (256) For Wilkens, scholarship is a constant question of opportunity cost. It is this principle of opportunity cost which has kept the canon firmly in place, even after decades of poststructuralist fashionability. We are only so many eyes, Wilkens seems to suggest, and we must consider the cost of directing those eyes elsewhere, away from the two hundred or so sacral texts which anchor the discourse fields of literary scholarship. I will return to this idea in a moment.

Jo Guldi, in her series of blogs on the "spatial turn," seems to suggest a historically traceable trend in myriad fields of scholarship. She likens this "turn" to the linguistic turn with which English departments are far more familiar. Guldi explains that a "turn" is concerned with a retrospective reevaluation. It is the movement of a discourse field in the process of rearticulation according to a newly (re)introduced logic. Thus the spatial turn is the incorporation of spatial logics into seemingly disparate modes of knowledge production. This is not a particularly new phenomenon. Guldi locates the beginnings of the spatial turn in literature before the Civil War. However, in much the same way that digital technologies reinvigorated debates about the history of the book and the canon (even as they effaced the longstanding tradition of exploring both), Guldi sees new tools like GIS as a refreshing force in the application of spatial logic to other fields.

Franco Moretti certainly seems to agree by way of his experimentation with spatial logic in his "Maps" of Graphs Maps Trees. Though he admits that his spatial representations aren't really maps (a map, he says, would find value in a location "as such") they are certainly a spatialization of literary texts. While I like the approach, and I think that his diagrams of various village narratives from the nineteenth century present an otherwise hard-to-see set of information about the genre, it also feels strangely like a kind of close reading as well. The conclusions Moretti draws are, more than most of his claims throughout the text, founded in the question he starts with, and it is never quite clear why he chooses to arrange the texts according to the concentric circle logic that he selects. If this is a kind of shift in textual analysis toward scientific modeling, then it is a very strange shift. The methodology is comprised of thoroughly uninterrogated assumptions about the "best" way to represent the interior spaces of a given series of texts. It is also unclear how this particular methodology could generalize out to other genres of texts, and whether it would even be worth doing.

The diagrams in Wilkens' text seem to me to have much more utility. His net is cast broad enough and his methodology is distant enough from close reading that his conclusions are, I think, genuinely insightful. Sometimes, I am even surprised by what he finds. Most importantly for me, however, is the very specific revisions that his data suggests. This, to me, is the sign of productive scholarly work: surprising conclusions that when properly reckoned, bring about change in the body of knowledge to which it belongs. In the case of Wilkens, the revision is to "American regionalism," a critical historical construct with implications across a whole spectrum of humanities disciplines. I wonder if the difference between Wilkens' work and Moretti's isn't ideological motivation. Moretti's work, while interesting, remains inscrutable to me. It's lively, experimental, and evocative, but only sporadically does it have a political locus. Wilkens, on the other hand, begins with the clearly enunciated assumption that canons are in some way damaging to understanding the human activity we call "literature." This, I feel, is quite different from what Moretti does when his question already contains the answer, as it seems to in "Maps." Wilkens' question has less to do with a hunch about the nature of the texts and more to do with the conviction that these other texts that lie outside the purview of canonization are indeed worth investigation.

Perhaps what the digital humanities have been lacking is neither good ideas nor clever experimentation nor academic rigor. Perhaps it was merely lacking the spine of critique.

Convergent Divergence (Divergent Convergence): Doodling Moretti's Trees

In his thorough and quite civil response to Franco Moretti's Graphs Maps Trees: Abstract Models for Literary History, Cristopher Prendergast returns (ironically?) to a critique of Moretti's circularity. Though this critique appears in a number of different locations throughout his essay, I take its most concrete manifestation to be in Prendergrast's discussion of the reciprocal relationship between convergence and divergence among the branches of Moretti's trees. In Moretti's text, the convergence is the precondition for the divergence, and divergence provides ever more opportunities for convergence as well. Prednergast seems to take this as a "chicken-egg" problem. He says,

These are however strictly reversible propositions: if convergence preupposes divergence, then divergence presupposes convergence. We are back in the chicken-and-egg world, yet again opening onto the infinite regress that leads back to a hypothetical First Cause. It is best to avoid this morass. (57)

Though Tsuda counters many of Prendergast's arguments against Moretti more effectively than I could, I want to take a bit of time to examine this particular hang up, this concern with circularity and convergence as it relates to divergence and vice versa.

I object, first, to this notion of circularity, and instead want to reaffirm along with what I think Moretti would call reciprocality. There is nothing to my eyes in Graphs Maps Trees that suggests a search for whichever came first. Moretti's goal in outlining the give and take of divergence and convergence is conjoined to his attempt to combine the wave with the tree. The point in that combination is the perpetual preexistence of the one over the other. The effect of Moretti's divergent convergence (convergent divergence) is one of entanglement, not linear causation. He simply has the good sense to not append "quantum" as an adjective. Prendergast is sensing in this entanglement a question which isn't there. Moretti has no need to answer which came first, divergence or convergence. Neil Degrasse Tyson once said in response to the chicken-egg question: "The egg, but it wasn't laid by a chicken." In much the same way, Moretti's divergence and convergence morph and change, creating a heritage with as many conjunctions as disjunctions.

In Moretti's discussion I was reminded once again of the similarities between the models that DH theorists generate in an attempt to chart literary history and the decision-making trees that form the backbone of simulated experiences. Moretti's tree of clues is certainly of the same structural family as the reader's progress through a "Choose Your Own Adventure" book. This is evolution written as OHCO. Here I wonder at the possibilities of McGann's deformance and Drucker's speculative computing as they apply to a simulated, computer generated representation of divergence and convergence. How might someone play through literary history, utilizing data from the other 99.5%, the unread, and as a result generate new divergences and convergences that would otherwise be invisible?

Last year I became briefly addicted to a small mobile game called Doodle God developed by JoyBits. It is, in a few important ways, a possible resolution to Prendergast's concerns about circularity as well as possible imaginative platform for playing in Moretti's trees. In Doodle God, the player starts with four basic elements: Earth, Water, Air, and Fire. From these elements, players combine and recombine them in different arrangements, generating new elements along the way. As each combination yields yet more elements, the game creates a tree of equations, tracking the heritage of each new convergence. Here again we encounter the fecundity of a text. Doodle God is a small, simulated example of numerous discourses in DH which have come to a head in Moretti's text. It conjoins the concept of convergence and divergence to Latour's fecundity as well as McGann and Drucker's speculations. The results of these combinations are often quite unpredictable, though in hindsight, always explicable. This seems to me to be the logic of divergent convergence (convergent divergence). It is progressive yet reciprocal. The question I would like to ask is this: How can we doodle literary history?