Literary Lab MLA Wrap-Up #3

The next respondent was Matthew Jockers of the University of Nebraska and faculty fellow at the Center for Digital Research in the Humanities. You can read Matt’s full post here. I’ll just summarize some of the main points.

Matt’s role in bringing quantitative approaches to literary study are by now legendary. As he tells the story it began when he assigned 1,200 novels for a class at Stanford. It’s a really funny point if you put it in context: we have these internal debates all the time about making our syllabi — should I go with 12 novels this year (roughly one a week) or the more realistic six or maybe I’ll compromise at 8…When you multiply that by 100 (or now 1000) it’s really a different conversation altogether.

Matt’s most moving point was the fact that 13 of 14 students in the class continued to work on their projects after the semester was over. This was Stanford. They’re as grade hungry as anyone. And yet they continued working when there were no more grades to give. That is a powerful testimony for this kind of work to appeal to our students’ best instincts.

The other aspects Matt prioritized in his remarks were the basic rules of the road for the lab once it got started (lab actually meant a dark room with old furniture):

  • it had to be quantitatively driven;
  • it had to be collaborative;
  • it had to be experimental so that failure was a key part of the process;
  • and the outcome was more narratively driven than argument based — a lab report on process over product.

One of the things I really like about Matt’s contribution to this whole issue is how unapologetically literary it is. Instead of thinking about the lab as a humanities catch-all it can also be a way of profiling disciplinary excellence. Quantitative reading is some of the most intensely language-based I’ve ever engaged in. It allows me to tap into that side of my brain that loves words. I’m not saying this in a retro, let’s throw out the theory and the cultural studies kind of a way. But I am appreciative that it foregrounds what is for me one of the core aspects of our discipline and shows how it can be of value in a broader epistemic shift taking place in the world.

Literary study can think of itself as technologically informed, out in front, and still provide a critical mirror to ourselves and our society, much like many of the writers we study and admire.



The Literary Lab MLA Wrap #2

The next participant in our literary lab roundtable was Laura Mandell, Director of the Initiative for Digital Humanities, Media, and Culture at Texas A&M University as well as editor of the Poetess Archive, on online scholarly edition and database of women poets, 1750-1900, director of 18thConnect, and director of ARC, the Advanced Research Consortium overseeing NINES.

Laura focused on two larger features that humanities laboratories make possible: the first was an emphasis on visualization alongside reading. When we begin reading at a new scale visualization becomes an important aspect of how we make that information accessible and meaningful. My colleagues in computer science tend to see visualization as secondary to quantitative argumentation, and while I see their point I think for us in the humanities aesthetic features are an important part of argumentation. How something looks tells us something about what it might mean to us as individuals or communities. This is also why rhetoric is so important to the way we communicate with each other — we don’t write like the sciences because the writing is part of the point.

Laura’s other big concern echoed Jeffrey Schnapp’s remarks about the pedagogical value of labs, but took it in a different direction. For Laura, the value of the lab is that it can serve as a place of play. As teachers we facilitate their learning as opposed to adjudicate it (thumbs up/down). In distinction to Jeffrey’s interest in the way the lab connects us to the world — and all the financial pressures it brings to bear on our research and learning — Laura emphasized the lab as a closed space, as an outer space, one particularly marked by the idea of consensus-building. Instead of being a space of individuation, where we learn to speak for ourselves, labs are spaces where we learn to build consensual understandings of complex problems.

I really love the idea of the lab as a consensual space. This strikes me as one of its most significant departures from the historical mission of the humanities with its emphases on distinction, elevation, and individualization (your work or the great work). Next up, Matthew Jockers.

The Literary Lab – MLA Wrap-Up #1

Over the next few blog posts I’ll be summarizing some of the excellent insights offered by the panelists who participated in our roundtable, “Theories and Practices of the Literary Lab.” The aim of the roundtable was to address the implications of the increasing number of labs popping up in the humanities, especially now that we are in the process of creating on at McGill. As I mentioned in my opening remarks, how will the lab change the types of research questions we undertake as well as types of knowledge dissemination — whether it be forms of publication or aspects of knowledge transfer related to student supervision and teaching? What kind of teaching and learning spaces are labs? (The full panel abstract can be accessed here.)

Jeffrey Schnapp, now director of the METALab at Harvard, opened the panel with a genealogy of his time at Stanford and the failed attempt to found the Stanford Humanities Lab. He recounted the quip made by his provost in response to the idea: “a humanities lab? isn’t that a paradox?”

Schnapp went on to describe what he saw as the values of the lab for the humanities, that is, the extent to which it isn’t a paradox to think about literature and laboratories in the same breath. These included:

1. its project-based nature. laboratory environments are categorized by an orientation towards contingent goals – “projects” rather than “canons.” They’re more problem-based as opposed to preservation or duration based. But Schnapp also emphasized the idea of fabrication, the way labs are often informed by making things in response to problems. They’re more design oriented than much humanities research.

2. knowledge process versus print product. the laboratory environment is importantly a middle-space, one that emphasizes continuous inquiry and one not always well-served by the discrete nature of print publication. I think this is an important point, but I would disagree only insofar as it is an either/or issue. I think other types of communication might better represent laboratory-thinking, but I think print forms, whether in the form of the article or the monograph, are extremely important as discrete critical reflections on process. Just building, or just experimenting isn’t enough in the sciences – you have to talk about what you do and why, and then place those observations within a larger research context. It may be that print is not the best at imbedding this knowledge socially (wither the footnote?!), but the forms of article and monograph are to me essential in the larger process of knowledge production.

3. entrepreneurial. the laboratory is important to Schnapp because of the way it is more, not less, sensitive to outside pressures. it performs an important bridging function to social relevance and puts pressure on our research questions. for many in the humanities this is very, very bad. independence and irrelevance are our calling cards. but I think Schnapp makes an important wake-up call to have us think about practices that are more connected to the world of funding that we inherently depend on. The university press is another example of the way our ideas have always been under pressure from some market pressures — not as directly as journalism, but still present. As Schnapp put it, the lab is an important bridge, between the academy and the world but also between disciplines. It allows different disciplines to convene in the same place.

He finished with the question of why the literature lab? Why preserve the privileged notion of “literature” even as we move towards a more science- or university-based notion of the lab. I think it is a good question, but one of the things I’m most intrigued about is the way literary study becomes newly relevant as things like text-mining become increasingly useful to a range of disciplines. The literature lab can be a way of invigorating the relevance of this particular disciplinary knowledge — rather than making the cultural studies turn (we study all things and are thus relevant to all things) it profiles a changed understanding of disciplinarity as a new tool of relevance.


Rapid Analysis of 30 Years of Apple Ads

I was catching up on some tech news sites yesterday when I saw a link to an older compilation of some 30 years of Apple print ads. I thought it might be interesting to do some rapid analysis of the text in the ads, in the spirit of my last post on doing quick-and-dirty sentiment analysis of DH Interviews (the main point of these mini-experiment is to describe a set of steps that allow us to engage with the materials fairly quickly in order to determine if more elaborate and time-consuming work might be worthwhile).

One of the big challenges with this mini-experiment is of course that the ads are in graphical form and we’re wanting to do some text analysis – so at some point we’ll need to perform Optical Character Recognition (OCR) to see how well we can extract text from the images. But let’s start by acquiring the images. Continue reading

Sentiment Analysis of DH Interviews

This is part of a new series that I hope to develop on writing step-by-step mini experiments in humanities-oriented reading and analysis of digital texts. These are meant as relatively quick-and-dirty, somewhat pedagogically oriented, erring on the side of playful, little exercises.

Tweet The inaugural edition was prompted by a tweet by Shawn Graham mentioning a series of interviews of digital humanities scholars as part of a digital history course offered by Leslie Madsen-Brooks at Boise State(this is a great idea for an assignment, though I’m not sure how well it would scale if everyone started assigning it:). Not only did I find the interviews fascinating to read, but it struck me as a good, relatively small corpus, to work with for a mini-experiment. I wasn’t the only person to think that, Mike Widner wrote a great post on Graphing Topics in DH Interviews.

As I was reading the interviews I started wondering if automated sentiment analysis could be used to determine how positive or negative each interviewee was in their reported responses. I don’t know about other digital humanists, but I find that people tend to assume at first that I’m an uncritical promoter of all things digital and computational. A cheerleader. Unfailingly positive. I think folks tend to discover fairly quickly that I’m more skeptical and nuanced than they’d assume. In any case, I wondered what a quick-and-dirty experiment in sentiment analysis might reveal about this set of interviews. Continue reading

The Metabolism of Data

At our meeting today we wrestled with the problem of data. In so many cases we often seem to want to run ahead to the really hard problems of what to do with our data. How best to interpret a big set of texts? But we often overlook in the process the equally hard problem of creating data sets. As we experienced in excruciating ways today, pulling together a data set requires huge amounts of time and answering a range of questions with always imperfect answers. Frankly, nothing has reminded me more of being a translator than being a data compiler (I once described being a translator to a friend as learning to become really comfortable with a persistent sense of failure). Continue reading

Archive versus Data

A really nice site called Radio Time Machine that allows you to hear Billboard 100 songs since the 1940s. Its clearly a response to the growing interest in this question of cultural change. Instead of working through the problem analytically with data, it lets users listen for themselves. It’s a brilliant idea and really nice design. Easy to use and a nice gateway for thinking about these questions in more depth.