Even on the basis of our brief acquaintance this semester, it is safe to say that critical work involving quantitative methods often seems, on the whole, to be more about what it hasn’t taught us. Like Moretti—who was forced to problematize and even revoke large swathes of his argument in the final pages of his essay—Underwood is constantly qualifying his findings, which in any case point in no clear direction, having failed to support any of his initial ideas. Indeed, if there is any rhetorical gesture that marks this genre of criticism, one is tempted to say, it is surely the cautious backpedaling of uncertain conclusions.
Of course, that their method may be more productive of ambiguity than certainty, I am sure distant reading’s practitioners would be the first to agree. Not for nothing does Underwood feel compelled to close his essay with a section titled “What have we learned?”—a tacit admission that the answer has heretofore been far from obvious. But the larger question that remains for me, at the close of Underwood’s article, is what kind of knowledge such methods are actually intended to produce. What would learning look like, on a good day? To put it another way: if we are adopting a critical practice that involves, as Underwood puts it, “bracketing the quest to identify underlying factors that really cause and explain the phenomenon being studied” (p. 2 in the downloaded PDF version), then to what degree are we really sacrificing understanding? And what kinds of new understanding are we looking to gain, that could compensate for our inability to explain or account for genre?
In this respect, it is striking how often our received notions of genre find affirmation here (for example, that Poe created the blueprint for all subsequent detective fiction), while our more theoretically sophisticated and nuanced understandings (genres as “fluid and tenuous constructions”) are troubled and unsettled (4, 5). The picture of genre that Underwood’s research paints is mainly new, it seems, insofar as it is so old: its novelty, from our current vantage, is precisely its outdatedness. In these pages, we are returned to a vision of genre as stable, unchanging, and intrinsically identifiable: detective fiction began with Poe, and remained fixed and coherent ever after. Ironically, this is the exact opposite of Underwood’s intention, having set out with a fluid view of genre based in “blurry family resemblances” and “the shifting practices of particular historical actors” (2).
It might behoove us to ask how Underwood’s quantitative methods have produced a result so antithetical to his own theoretical premises. In the meantime, the further irony is that even this state of straightforward coherence remains essentially opaque and inexplicable in his model. This is because, for Underwood, the linguistic consistency that marks the long-term coherence of the detective genre does not define it (“our goal is no longer to define a genre” –2): i.e., the genre’s coherence does not consist in its linguistic consistency (Underwood makes clear that genre is not a linguistic phenomenon), “words just happen to be convenient predictive clues”—predictive of a coherence, in other words, that is comprised of something else (2). What that something else is, the model has no way to tell us.
Underwood concludes his essay by opining “that quantitative methods can give literary scholars descriptive resources” (8); his basic move, throughout, has been to trade definition for description: instead of trying to define genre, we’ll just describe how it behaves. I am happy to meet him on those terms, which seem like a worthwhile trade-off to make: but I remain unsure of what description he has actually provided. All he can do, it seems, is tell us that there is indeed a thing called genre out there (but only sometimes—see the case of the Gothic): its content and features remain, if anything, more mysterious than they were before.