Posts ⁂ The Postmodern Terminal >_
A Jesuit priest, an IBM contract, and millions of words of Aquinas: how digital humanities became less a set of tools than an argument about interpretation.
In the 1940s a Jesuit priest named Roberto Busa talked IBM into helping him build a concordance of Thomas Aquinas. The Index Thomisticus took decades and ran on punch cards. Busa was doing text mining before anyone had the phrase, and he was doing it because he had a theological question about how Aquinas used a handful of key terms across millions of words. I will argue that digital humanities is a theoretical disposition rather than a toolkit. I will show it by tracing the field from Busa's punch cards through the quarrel over distant reading to the critical turn that asks whose archive gets digitized in the first place.
Franco Moretti put the problem bluntly in Graphs, Maps, Trees and in the work of the Stanford Literary Lab: there are too many texts to read. The canon that literary scholars actually discuss is a sliver of what has been published, and the rest stays invisible. His proposal (audacious, and people are still cross about it) was to stop reading books one at a time and start analyzing large corpora computationally, looking at publication trends, maps of narrative geography, trees of genre. He called it distant reading. Matthew Jockers extended the method in Macroanalysis, surfacing patterns in nineteenth-century fiction that no reader could see book by book.
The objections are good ones. N. Katherine Hayles argues that we need machine reading, hyper reading, and traditional close reading together; the methods are complementary, and distant reading thins out badly when nothing anchors it in a text someone has actually sat with. Johanna Drucker presses harder. There is no such thing as raw data. Every encoding, every database schema, every visualization involves choices about what counts and who is represented. Data in digital form is already interpreted. Numbers do not speak for themselves, which means the humanist's job does not end where the computation begins; it starts there.
By the 2010s the field had turned on itself in the best way. Miriam Posner's essay "What's Next: The Radical, Unrealized Potential of Digital Humanities" insists that race and power are not merely topics for analysis; they are structures built into the tools. Databases have data models; data models make assumptions; those assumptions encode worldviews. Catherine D'Ignazio and Lauren Klein bring feminist theory into the same room in Data Feminism, and their questions are the practical ones: who counts, who is counted, who benefits, who bears the cost.
Christians, of all people, should feel this. We are heirs to centuries of manuscript copying and commentary, and heirs also to archives that erased. Colonial mission records catch Indigenous peoples as objects of conversion. Enslavers documented the religious lives of the people they enslaved far more thoroughly than those people were permitted to document themselves. Women's theological voices were kept off the official record. Projects like Slave Voyages and the Mukurtu platform (built so Indigenous communities govern their own digital archives) do something close to prophetic work: they refuse to let the archive be the only story. Indigenous data sovereignty follows from that refusal, since a community has the right to govern data about itself.
You do not need to write code. If you work with documents, Voyant and Python's NLTK will get you started on themes and language change. If you work with relationships, Gephi and Palladio map them. If you work with archives, Omeka and Mukurtu were built for you, and thinking carefully about how to digitize and describe material is itself scholarship. If you work with survey or program data, a well-made visualization is a form of argument.
One unglamorous rule holds the whole thing together. If you have one copy, you have none; if you have two copies, you have one. Budget for preservation and for file formats, or your prophetic archive dies in a dead drive. Alan Liu warns that in our excitement about big data and large language models we can lose the humanistic commitment to meaning, context, and slow thinking. He is asking the field to hold onto its soul, and that request, writ large, is the whole discipline in one line.
The Postmodern Terminal >_