If there's such a thing as "computing for humanists" is there also such a thing as "humanities for computer scientists?" On implementing interdisciplinarity in the Digital Humanities
This is a just a brief initial thought piece on a question I’ve been asking colleagues about, from whom I’ve not heard the answer I want.
The Digital Humanities is an interdisciplinary field that involves the intersection of computation and the humanities. That means, amongst other things, that neither computation nor humanities is primary to the discipline but both must be present in some way or another. In this way, the Digital Humanities is different from, say, the “History of Science” (History is primary) or “Cognitive approaches to cultural understanding” (Cognitive science is primary).
In actual practice, for most of its life as DH and in its earlier form, Humanities Computing, DH has been mostly the domain of humanists. The people have been located in Humanities departments and the projects have in many ways been developments from and extensions of humanities research. So while Digital Humanities, in terms of content, is more “digital + humanities” instead of the “Humanities of the Digital” (except maybe in special cases like Critical Coding), it has been for a large part in institutional terms a space in which humanists do computing.
An important practical implication of this institutional construction is that we assume that the weaker part of most practitioners’ skill sets involves technology rather than humanities. Thus, if you attend a school like DHSI, you can take courses on programming and project management designed to improve the technological chops of postcolonial theorists; but you can’t take courses on Spivak designed to improve the PoCo skills of Computer Scientists. Likewise, we have a pretty good idea of what a course on “computing for the humanities” ought to look like, but, I suspect, a much poorer idea—or, as I’m discovering, perhaps no idea at all—what a “humanities for the computing” ought to look like.
I think this is increasingly going to be a problem in our field as it becomes more-and-more prominent and attracts more and more interest from disciplines not originally involved in its development. We already have lots of projects involving Geographers, GLAM professionals, New Media specialists, and so on. It is increasingly not uncommon to find people who are interested in DH in traditional comp sci departments and software engineering. What training and skills development in humanities research and skills do practitioners and students in those fields need in order to engage properly with our interdisciplinary subject? We’ve already learned that it is possible to teach those with a humanities background “enough” programming and computer skills to function well in our world; what is the equivalent for those who come from a technological background and need to learn “enough” humanities research and exposition skills to function the same way?
I’ve been asking colleagues about this for a while and what has surprised me is the extent to which people have said, in essence, that it is not possible to do this. I.e. that it is not possible to construct, teach, and especially set work and expectations for a course that aims to teach non-Humanists enough humanities research skills and knowledge to get by.
The basic question I’ve used has been the following: if I had an incoming graduate student for a Masters in DH who had a background in computer visualisation, say, or database programming who wanted to become a Digital Humanities researcher, how should my expectations for that person differ from those of somebody with a background in the Humanities in a course on, say, Post Colonial theory or the Nineteenth Century Novel? Does a graduate student with a technological background entering under such circumstances need to perform at the same level as a student with a more traditional humanities background? And if not, how to we handicap for the disciplinary difference in grading?
I’ve asked this question of people in English, Linguistics, Philosophy of Science, and so on—people who are for the most part used to working interdisciplinarily. And the surprising answer I’ve heard from every single person I’ve asked is that we can’t handicap such a student: i.e. there is no such thing as “good enough for a non-humanist” graduate level work in the Humanities. In fact, most people I’ve asked have gone further: not only is there no such thing as “good enough for a non-humanist” graduate level work, the standard we need to use in assessing such a student is actually “is this good enough for a graduate student in that same humanities domain?” I.e. the standard to use in assessing a graduate student with a background in computer science who is taking a humanities course is whether they meet the standard we’d expect of a student in the same course who had a background in the original discipline. So for a computer engineer to pass a graduate course in Wordsworth, they need to show they are as good as a more traditional grad student trained in literary studies.
There is a startling lack of reciprocity involved in this (imagine if the computer scientists who teach “programming for humanities students” starting insisting our students perform to the same level of sophistication as computer science MSc students). But this all or nothing approach also seems to me to say something bad about either us (Humanists) as scholars and teachers or our domain (Humanities) as a research field. Is it really simply not possible to acquire (and get a grade for) a “working knowledge” level of awareness in the Humanities? Is it really unreasonable to allow somebody who can do DH because they have solid technical skills to be maybe even a little bit clunky or poor at their ability to formulate humanities arguments? Or are we all performance and no content?
Sometimes, I think some of us wonder if we actually do have skills and knowledge that others might want to possess. I once went to a colleague who is a textual critic to ask that person to teach a course on textual criticism to an excellent student from computer science. After I went through the student in question’s skills, the colleague asked “what am I going to be able to teach a student like that?” As if “textual criticism” wasn’t enough of an answer.
But we really do have skills and methods that others can use. I was once on an M.Sc. thesis committee in Computer Science for a student who was doing computational text summation. One day when we were looking at the results of a test run, we discovered that while most worked well, a couple of pieces did not. My Comp. Sci. colleagues suggested that we run another 10,000 texts through the machine to confirm the error rate. But when I looked at the actual examples that had thrown the problems, I discovered that they involved a particular kind of news story, such as you get describing hurricanes or military battles, in which the story begins with a narrative but ends with a list of the damaged, wounded, and dead: our machine was being thrown off by the change-up in form that was coming towards the end of what was, in essence, a specific genre of newspaper article. My colleagues from Computer Science were really surprised: they hadn’t realised that you could define genres like that or then use them to refine your example pool—or perhaps hadn’t thought through the extent to which the different broad sets of examples we were using “Newspaper articles,” “Physics theses,” etc., were only broad categories under which many different distinguishable sub-genres existed.
Presumably, in DH, we want students who are able to apply “humanities” ways of thinking to problems in a similar broad, rule-of-thumb way—or understand other members of an interdisciplinary team when they are applying such methods to common problems we are all working on. What is surprising to me is that I’m not sure we have a method for actually teaching such skills to anybody who doesn’t want to become a traditional researcher in the traditional humanities. I don’t know myself what standards we do use in assessing work by such students. But there has to be something different than our current all or nothing approach.
