So Stephen Marche has written an article on the digital humanities that has caused a bit of a stir in the Twittersphere. As one might expect from Marche’s pugnacious style, this article comes to bury the digital humanities, not to praise them. It’s a curious approach, one that seems to revel in some elements of the digital revolution (most notably EEBO and Google Books) while rejecting others outright, a tension explored at length by Holger Syme in a characteristically excellent piece on his blog. I won’t really get into that here (as there’s nothing I can add to what Holger’s said); instead, I want to look more deeply into the specifics of the digital techniques Marche recoils at.
Marche dislikes the way that the digital humanities has turned literature into data, and robbed it of its context. And this makes sense, to a point, as context is an important part of humanistic inquiry. Likewise, Marche feels uneasy when confronted with what he calls the fascism of algorithms: “Algorithms are inherently fascistic, because they give the comforting illusion of an alterity to human affairs.” These two criticisms bear out, but they are also criticisms that DH has explored, and will likely continue to explore for years to come. As I note in my review of Stephen Ramsay’s excellent book Reading Machines, Ramsay notes that a humanistic use of algorithms will always require a return to the original text to see whether the results of our algorithms stay true to the text, and to avoid pinning our arguments to something that may be a distorted reflection of the text (something he gently criticises Franco Moretti for not paying enough attention to). Ramsay refers to these algorithmic transformations as “deformances,” a term that harkens back to the work of Jerome McGann and Lisa Samuels, who make an argument for deforming literary works so as to better understand how they function. And this is the same technique digital humanists use: we deform texts to exaggerate features and see how they function when taken to the extreme, and then we return to the original to explore in finer detail anything that the deformances may have brought to light. This is, at its heart, the very same type of work that Dominican monks undertook in 1270 when they developed the first biblical concordance, which catalogued all the words in the Bible as well as some major themes and motifs, allowing them to see relationships between different parts of the text that would otherwise be obfuscated by the massive size of the Bible. It’s all about new ways of seeing a text.
The fascism of algorithms has also been explored by a number of digital humanists, and to great effect. David Golumbia’s book The Cultural Logic of Computation explores the way that computational ways of thinking have begun to impose themselves on the non-computational. He looks, for example, at the way that Chomsky’s approach to linguistics has naturalised the use of computers to process language, and notes the ways that algorithms lack the dexterity to properly describe language without constant tweaking and reworking. We can also turn to someone like Lawrence Lessig, who has commented that with changes in copyright law, code has become law, and that perfectly legal uses of ebooks, for example, can be made illegal by the overriding enshrinement of (what is known in Canada as) technological protection measures. Katherine Hayles, an early advocate of digital humanities, has also noted the pervasiveness of computationalism, the subjugation of analogue processes to the digital, which leads to rigid ways of thinking that exclude possibilities that do not fit into quantised, defined, preselected options. So Marche is right in some of his objections to the large-scale effects of digitisation and computationalism, but he seems to be unaware that these are issues that digital humanists have long been keeping their eye one, and that we will continue to explore well into the future.
Marche also seems irritated by the idea that digital humanists are mistaking data for literature, and losing the aesthetic and poetic value of these texts in the process. This is a somewhat strange argument in my mind, as literature and data need not be in so strict opposition; indeed, data can have its own aesthetics. Moby-Dick begins not with the famous line “Call me Ishmael,” but rather with a collection of literary quotations about whales. These pages of quotations plucked from their literary context work to establish the unknowability of the whale that the novel explores in all its detailed descriptions of whales and whaling. Nick Cave’s novel And the Ass Saw the Angel contains a number of lists and charts, all of which map the decline of Ukulore Valley and its residents. Milorad Pavic’s The Dictionary of the Khazars draws much of its strength not from Pavic’s excellent writing, but from the way Pavic divides it into three dictionaries—the Judaic, Christian, and Islamic dictionaries—and juxtaposes these three reference books to find a story that takes place between all three. Ultimately, data can’t be distinguished from literature so cleanly as Marche suggests; literature is as much interpretation as anything. Just as art survived Duchamp’s readymades, literature will survive data.
I must say that I’m disappointed by Marche’s assessment of the digital humanities not just because it’s a rather poor critique of something I hold dear, but rather because Marche himself has made a valuable contribution to digital literature, and seems to have forgotten everything he learned when writing that. Stephen Marche’s interactive novel Lucy Hardin’s Missing Period is a compelling work that follows the titular Lucy’s life shortly after waking up after a one-night stand, on the anniversary of her father’s death. As we read this work, we are given a number of options that shape the way the story plays out. With each choice we make, new possibilities open before us while others recede into the distance. I was TAing in a class last year that Marche came to visit, and while he was there, he shared a number of insights about the making of Lucy Hardin’s Missing Period. The one that has stuck with me the most since then was his argument that interactive works like Lucy Hardin are actually better suited to realist narratives, as they better represent all the different paths our lives can take than a traditional, monolinial work that allows the reader only one possibility. By presenting readers with options that vanish with each choice, Marche felt that he was better representing the reality of our day-to-day existence. This Marche is much more thoughtful, more considered than the one who wrote this rather spotty article for the Los Angeles Review of Books. Admittedly, it’s not a Marche who has embraced data the way that other authors may have, but it is certainly a Marche who sees that even in digital form we are able to create “mushy” truths (as he says). We can create spaces where multiple outcomes hang in the air beside each other, and collapse as we make choices. This incompleteness, an incompleteness Marche feels is lacking in our work with data, is completely native to digital realms. While the Marche of the article decries the failings of the digital, offering critique but not much else, the Marche of Lucy Hardin is very much in the business of pushing boundaries and trying things out, of finding new ways of seeing the world and embracing the unique elements the digital has to offer. I do hope we see more of this latter Marche..