Just beyond the hustle of Butler’s busy entrance lies the Studio@Butler, managed by the Center for Teaching and Learning and the Digital Humanities Center. On Fridays from 3 to 5 p.m., when most students begin their weekends, a group of students and faculty members meet to experiment. With their laptops open and screens crowded with code, they look like they belong in Silicon Valley. But they study the humanities, not computer science.
This is the xpMethod Open Lab. Organized by Columbia’s Group for Experimental Methods in the Humanities, this lab consists of thinkers who describe themselves as “a group of literary scholars, sociologists of knowledge, and information designers interested in experimenting with distant reading, lateral reading, cultural analytics, digital philology, micro- and macro-analysis, and textual trans-mediation.” This onerous language sounds academic, far from the sleek language of tech startups.
Indeed, even the term “digital humanities” is not one that all scholars are necessarily familiar with. Nor is it a field that can be easily defined, as it is constantly evolving. Roughly, digital humanities is a field of study at the crossroads of the traditional humanities and computer science. A digital humanist applies computational methods and digital tools to fields within the humanities, such as literature or history.
The GEM meets informally; scholars drop by the open lab to work on an individual or collaborative project, or to check out each other’s work. Phillip Polefrone, a fourth-year Ph.D. student in English literature, sits at a table working on a framework for semi-automated literary mapping, a program that detects and tags locations appearing in literature. A few tables over, Dennis Tenen, an assistant professor in the English and comparative literature department and co-founder of GEM, chats with a graduate student who has come to learn more about digital humanities. On most Fridays, co-founder and English lecturer Grant Wythoff is also mixed in among the group, his screen humming as the members converse.
As an English major, I initially felt insecure when I first started talking to Wythoff, unfamiliar with the jargon he used: “Python,” “digital mapping,” and “critical methods”; they all made me think I was in way over my head.
For Wythoff, who also serves as Columbia’s digital methods and public humanities fellow at the Heyman Center for the Humanities, one doesn’t need an extensive computer science background to study the digital humanities. “Anyone who either studies in the humanities or is a humanities scholar is already a digital humanist, because we use computers to do the work we do everyday,” he says. His office, filled with neatly stacked books, suggests both the creative scatter of a literature scholar and the methodized nature of a coder.
While any Columbia student or faculty member is encouraged to drop in the open lab, and there are digital humanities projects that anyone can get involved in, there is a small core group of graduate students and a few professors. Maybe more Columbia students don’t join because phrases like “lateral reading” and “digital philology” sound so impenetrable, even mechanical, to students whose interests lie in studying human nature through the inexplicable warmth of beat-up books that have passed through the hands of countless scholars before them. To these students, the term “digital humanities” may sound oxymoronic in itself.
So how is it, then, that thinkers like the ones found at the xpMethod Open Lab, whose number fluctuates from 10 to a couple of dozen on any given Friday, reconcile the modern, automated drive of computers with their warm passion for the humanities?
When Wythoff took an interest in digital humanities as a graduate student at Princeton, where he earned his Ph.D. in English literature in 2013, he had no background in computer science. One of his advisers handed him a massive three-ring binder filled with photocopies of texts documenting the history of poetics and asked him if he could turn it into a digital database. It would be his first introduction to interdisciplinary work, and it prompted him to change the direction of his studies to pursue this new field.
Columbia’s lab itself exhibits this interdisciplinary mindset, as scholars from a variety of different academic backgrounds are encouraged to collaborate. “The idea is that projects can organically emerge out of this loose setting,” Wythoff says. “The thing that makes it experimental is that nobody knows, when they are going into that space, what questions are going to be asked, what objects are going to be studied or looked at, what kind of arguments are going to be made, what kind of outcomes we will have.”
Jonathan Reeve, a second-year graduate student in the English and comparative literature department who studies computational literary analysis, might as well be the poster child for computer science and the humanities’ interconnected potential. Reeve, in his heavy beard and winter scarf, looks like a traditional English professor, but his passion for computer science shines through as he runs me through his website.
Like Wythoff, Reeve got into digital humanities through English. As a student working on his master’s degree in the humanities at New York University, Reeve was introduced to digital humanities through literary archives and a web development course, after which he audited a programing class (which he proceeded to drop halfway through, having landed a programming job). Reeve found that literary analysis was easier using computer programs, and since then, he has worked with numerous projects to analyze works.
“Quantitative methods and literature are really a match made in heaven, in a sense,” Reeve says. He goes on to explain to me that society often insists on separating the two: In high school, for example, you are either a math person or a humanities person. But Reeve believes that an English major with skills in computer science, for example, is “the ideal candidate” for digital humanities.
This “ideal candidate” is in higher demand in the job market with each passing day. Reeve explains that skills in computer science or digital methods can create job opportunities for students interested in the humanities who are tempted to abandon their passions in order to pursue fields that are hiring more people, like computer science. He cites natural language processing, a subfield of computer science, as a large emerging field. Tech companies like Google need natural language processors to understand what the words that are being typed into their search engines mean, and people with both literary analysis skills and programming skills are in high demand.
Applying the digital to the humanities releases a world of new academic potential, in addition to giving students a leg up in the job market. For example, any student at Columbia knows exactly what “close reading” means after a year of Literature Humanities, but what about “distant reading”? When you distant read, you look at a text more broadly, with dozens or even hundreds of other texts alongside it. Computational analysis makes this possible, as a computer is able to “read” thousands of books before we could even make it through one page of Paradise Lost.
As Reeve explains, distant reading complements close reading. "We use computers not to come to conclusions based on the data we are getting from the analysis necessarily, but to allow us to find places in the text or ways of close reading the text that we wouldn’t ordinarily be able to," he says.
Beyond computational analysis, digital humanities also allows for new ways in which readers can interact with a text. Wythoff’s most recent book, The Perversity of Things: Hugo Gernsback on Media, Tinkering, and Scientifiction, is an example of the collaborative nature of digital humanities. As well as being printed in paper, the book is being published as a Manifold Edition, a digital publication that addresses the idea that behind every scholarly book lies an entire archive of materials—primary documents, videos, audio files—that can’t fit into the text. With the Manifold Edition, this archive can be linked up to the text, and readers who create an account can interact with it, highlighting and marking up texts for the readership at large to view.
Still, computers as readers have their limits, and sometimes these limits reveal something about a text that we may never have considered otherwise. Reeve runs me through one of his projects, called “Probabilistic Detection of Character Voices in Fiction,” as an example. One of the novels that Reeve extracted texts from for this project was Samuel Richardson's epistolary novel Clarissa, or, the History of a Young Lady. The program was meant to identify different characters by their voices. “It got to 60 or 70 percent of the way there,” Reeve says, implying that the statistical analysis was about 30 or 40 percent off. But even though the program was not completely right, Reeve explains that it elucidated other things about the novel: When the program confused one character’s speech for another’s, it demonstrated that the characters were quite alike. This could give way to closer analysis—close reading, as opposed to digitized distant reading—by a human reader.
“Computers are really, really bad at reading text,” Reeve says, explaining that human readers, specifically those trained in literary analysis, will always be necessary. “We are never going to dispense with close reading, or the traditional methods of the humanities, but having computational analysis at our disposal is a very powerful tool that can teach us a few more things as well.”
Scholars like Reeve prove that computer science and the humanities are not as antithetical to each other as we once thought, and that the term “digital humanities” does not have to be an oxymoron after all.
More and more, literature is becoming intertwined with technology. Indeed, Wythoff believes that, at least at the graduate level, humanities students should be armed with digital methods skills. He’s teaching a class called “Introduction to Digital Media” this spring semester, and Columbia also offers the undergraduate courses “Computing in Context” and “Critical Computing in the Humanities.” The Digital Humanities Center also offers support for scholars in the humanities and history, and its website lists workshops that are run by various departments on campus meant to instruct students on digital methods. But Wythoff would still like to see more widespread study of the digital humanities at Columbia.
“I think it will be more important going forward trying to find a model for standardizing what kinds of digital methods should be talked about, and the kinds of political awareness that humanities scholars should have when they begin doing their graduate work,” he says.
Have fun leafing through our seventh issue, and subscribe to our weekly newsletter, As We See It!