Welcome to Jonathan Talks, a place where I actually write down my thoughts on numerous subjects at length instead of throwing their quip versions into the abyss that is social media. (Please feel free to subscribe for free.) Today, I will be talking about the digital humanities and the project from which I hope to build a dissertation for my PhD at New York University. I call it, “The Art of Literary Modeling”, for reasons which I hope will become clear to you as I write.
So we begin
Substack is reminding me that beginnings are difficult. This is almost always true for a digital humanities project. Sometimes you begin with a question; other times you begin with a dataset. There are so many small acts of interpretation and modeling involved between that beginning and the acts of experimentation, analysis, and subsequent interpretation you originally had in mind that those beginning acts become difficult to keep track of and explicitly document.
For example, let’s say I had a question I’d like to answer about Shakespeare’s Othello. How do I begin? I have to ask questions and perform some preparatory acts. Has my question already been asked? If I still want to ask it, what edition of the play should I utilize to answer that question? Once I select that edition, what part(s) of the play do I want to look to in order to find potential answers? Ultimately, the most explicit, tricky, and highly interpretive question that I must ask and answer is how I would like to quantify the digital text so that I can perform some modeling over it. And all of those choices often go undocumented or at most, quietly mentioned as an after thought in favor of the more attractive: the question I want to answer and – if I am lucky – some potential answers to that question.
More to the point, these acts are not something that feel particularly worth mentioning. As one professor (whose views I still value) put it: ‘This is process text.’ That’s true. To report on preparation seems like a banal, perfunctory act with a dismissible result, but for humanists delving into a mathematical world it turns out that it is foundational for our claims. It is part of the essence of what differentiates claims in the digital humanities from claims in the sciences (i.e. natural language processing): a robust, qualitative familiarity with our data. We can find that understanding within the initial promising idea of teaming a computer science-born programmer with a humanities-born researcher. That premise includes the claim that humanities-oriented researchers are more intimately familiar with the cultural objects being digitized and modeled. We humanists are also often interested in the edge cases, the small minutiae or legomena that serve to defy central tendency in statistics and act as a fulcrum for phenomena in a larger context. Recognition of those cases and their potential interpretive function is also part of the promise of humanities research. These are the hooks upon which we can hang the highlights of our findings, making our research outputs captivating and perhaps encouraging for others to delve further into the topics of our work.
Your House; Your Perspective
I liken this process of preparation to what Michael Pollan writes about in A Place of My Own: it’s like building a house. Consider the dataset to be our landscape. We want to build a house on top of a portion of it. And we want to situate our structure in such a way that we can look at other parts of that landscape. First, you must imagine the house. How big is its footprint? What direction will it face? What path will the sun take across your house? What view(s) will you have from its various windows and entryways? We hope to make each determination in such a way that we are best able to answer our question(s).
Sometimes we deal with “the best” materials we can produce or locate at the time, and with a shrug take that material and begin our work. But this does not feel adequate. In light of justifying our claims both to our skeptical humanities audiences and our skeptical science and engineering audiences, and for the sake of important and recent concerns like reproducibility I have sought to describe in a new way the process of beginning a digital humanities project and to bring to bear actual quantitative and qualitative foundations upon which we can justify the acts of preparation, modeling, and interpretation. As you can imagine, there is only so much generalization to be found here. These are a series of research acts driven by perspective and context.
There is so much more for me to share, but for today I will close here with this question for you to personally consider. For a moment, forget about your particular research focus and consider your process. What is your techne? What is your craft? And I don’t mean some generally understood label of it, or the technology it involves. What makes your approach to answering questions different or unique? What is the art by which you model the worlds you would like us all to explore along with you?