Hah! I tricked you. I don’t intend to define digital humanities here—too much blood has already been spilled over that subject. I’m sure we all remember the terrible digital humanities / humanities computing wars of 2004, now commemorated yearly under a Big Tent in the U.S., Europe, or in 2015, Australia. Most of us still suffer from ACH or
ALLC (edit: I’ve been reminded the more politically correct acronym these days is EADH).
Instead, I’m here to report the findings of an extremely informal survey, with a sample size of 5, inspired by Paige Morgan’s question of what courses an undergraduate interested in digital humanities should take:
Should undergrads w/ humanities majors & interest in #digitalhumanities grad work pursue a second CS major, or a CS minor?
— Paige Morgan (@paigecmorgan) November 20, 2013
The question inspired a long discussion, worth reading through if you’re interested in digital humanities curricula. I suggested, were the undergrad interested in the heavily computational humanities (like Ted Underwood, Ben Schmidt, etc.), they might take linear algebra, statistics for social science, programming 1 & 2, web development, and a social science (like psych) research methods course, along with all their regular humanities courses. Others suggested to remove some and include others, and of course all of these are pipe dreams unless our mystery undergrad is in the six year program.
The discussion got me thinking: how did the digital humanists we know and love get to where they are today? Given that the basic ethos of DH is that if you want to know something, you just have to ask, I decided to ask a few well-respected DHers how someone might go about reaching expertise in their subject matter. This isn’t a question of how to define digital humanities, but of the sorts of things the digital humanists we know and love learned to get where they are today. I asked:
Some of you may have seen this tweet by Paige Morgan this morning, asking about what classes an undergraduate student should take hoping to pursue DH. I’ve emailed you, a random and diverse smattering of highly recognizable names associated with DH, in the hopes of getting a broader answer than we were able to generate through twitter alone.
I know you’re all extremely busy, so please excuse my unsolicited semi-mass email and no worries if you don’t get around to replying.
If you do reply, however, I’d love to get a list of undergraduate courses (traditional humanities or otherwise) that you believe was or would be instrumental to the research you do. My list, for example, would include historical methods, philosophy of science, linear algebra, statistics, programming, and web development. I’ll take the list of lists and write up a short blog post about them, because I believe it would be beneficial for many new students who are interested in pursuing DH in all its guises. I’d also welcome suggestions for other people and “schools of DH” I’m sure to have missed.
And because the people in DH are awesome and forthcoming, I got many replies back. I’ll list them first here, and then attempt some preliminary synthesis below.
The first reply was from Ted Underwood, who was afraid my question skirted a bit too close to defining DH, saying:
No matter how heavily I hedge and qualify my response (“this is just a personal list relevant to the particular kind of research I do …”), people will tend to read lists like this as tacit/covert/latent efforts to DEFINE DH — an enterprise from which I never harvest anything but thorns.
Thankfully he came back to me a bit later, saying he’d worked up the nerve to reply to my survey because he’s “coming to the conclusion that this is a vital question we can’t afford to duck, even if it’s controversial [emphasis added]”. Ted continued:
So here goes, with three provisos:
- I’m talking only about my own field (literary text mining), and not about the larger entity called “DH,” which may be too deeply diverse to fit into a single curriculum.
- A lot of this is not stuff I actually took in the classroom.
- I really don’t have strong opinions about how much of this should be taken as an undergrad, and what can wait for grad school. In practice, no undergrad is going to prepare themselves specifically for literary text mining (at least, I hope not). They should be aiming at some broader target.
But at some point, as preparation for literary text-mining, I’d recommend
- A lot of courses in literary history and critical theory (you probably need a major’s worth of courses in some aspect of literary studies).
- At least one semester of experience programming. Two semesters is better. But existing CS courses may not be the most efficient delivery system. You probably don’t need big-O notation. You do need data structures. You may not need to sweat the fine points of encapsulation. You probably do need to know about version control. I think there’s room for a “Programming for Humanists” course here.
- Maybe one semester of linguistics (I took historical linguistics, but corpus linguistics would also work).
- Statistics — a methods course for social scientists would be great.
- At least one course in data mining / machine learning. This may presuppose more math than one semester of statistics will provide, so
- Your recommendation of linear algebra is probably also a good idea.
I doubt all of that will fit in anyone’s undergrad degree. So in practice, any undergrad with courses in literary history plus a semester or two of programming experience, and perhaps statistics, would be doing very well.
So Underwood’s reply was that focusing too much in undergrad is not necessarily ideal, but were an undergraduate interested in literary text mining, they wouldn’t go astray with literary history, critical theory, a programming for humanists course, linguistics, statistics, data mining, and potentially linear algebra.
While Underwood is pretty well known for his computational literary history, Johanna Drucker is probably most well known in our circles for her work in DH criticism. Her reply was concise and helpful:
Look at http://dh101.humanities.ucla.edu
In the best of all possible worlds, this would be followed by specialized classes in database design, scripting for the humanities, GIS/mapping, virtual worlds design, metadata/classification/culture, XML/markup, and data mining (textual corpora, image data mining, network analysis), and complex systems modeling, as well as upper division courses in disciplines (close/distant reading for literary studies, historical methods and mapping etc.).
The site she points is an online coursebook that provides a broad overview of DH concepts, along with exercises and tutorials, that would make a good basic course on the groundwork of DH. She then lists a familiar list of computer-related and humanities course that might be useful.
The next reply came from Melissa Terras, the director of the DH center (I’m sorry, centre) at UCL. Her response was a bit more general:
My first response is that they must be interested in Humanities research – and make the transition to being taught about Humanities, to doing research in the Humanities, and get the bug for finding out new information about a Humanities topic. It doesn’t matter what the Humanities subject is – but they must understand Humanities research questions, and what it means to undertake new research in the Humanities proper. (Doesn’t matter if their research project has no computing component, it’s about a hunger for new knowledge in this area, rather than digesting prior knowledge).
Like Underwood and Drucker, Terras is stressing that students cannot forget the humanities for the digital.
Then they must become information literate, and IT literate. We have a variety of training courses at our institution, and there is also the “European Driving License in IT” which is basic IT skills. They must get the bug for learning more about computing too. They’ll know after some basic courses whether they are a natural fit to computing.
Without the bug to do research, and the bug to understand more about computing, they are sunk for pursuing DH. These are the two main prerequisites.
Interestingly (but not surprisingly, given general DH trends), Terras frames passion about computing as more important than any particular skill.
Once they get the bug, then taking whatever courses are on offer to them at their institution – either for credit modules, or pure training courses in various IT methods, would stand them in good stead. For example, you are not going to get a degree course in Photoshop, but attending 6 hours of training in that…. plus spreadsheets, plus databases, plus XML, plus web design, would prepare you for pursuing a variety of other courses. Even if the institution doesnt offer taught DH courses, chances are they offer training in IT. They need to get their hands dirty, and to love learning more about computing, and the information environment we inhabit.
Her stress on hyper-focused courses of a few hours each is also interesting, and very much in line with our “workshop and summer school”-focused training mindset in DH.
It’s at that stage I’d be looking for a master’s program in DH, to take the learning of both IT and the humanities to a different level. Your list excludes people who have done “pure” humanities as an undergrad to pursuing DH, and actually, I think DH needs people who are, ya know, obsessed with Byzantine Sculpture in the first instance, but aren’t afraid of learning new aspects of computing without having any undergrad credit courses in it.
I’d also say that there is plenty room for people who do it the other way around – undergrads in comp sci, who then learn and get the bug for humanities research.
Terras continued that taking everything as an undergraduate would equate more to liberal arts or information science than a pure humanities degree:
As with all of these things, it depends on the make up of the individual programs. In my undergrad, I did 6 courses in my final year. If I had taken all of the ones you suggest: (historical methods, philosophy of science, linear algebra, statistics, programming, and web development) then I wouldn’t have been able to take any humanities courses! which would mean I was doing liberal arts, or information science, rather than a pure humanities degree. This will be a problem for many – just sayin’. 🙂
But yes, I think the key thing really is the *interest* and the *passion*. If your institution doesnt allow that type of courses as part of a humanities degree, you haven’t shot yourself in the foot, you just need to learn computing some other way…
Self-teaching is something that I think most people reading this blog can get behind (or commiserate with). I’m glad Terras shifted my focus away from undergraduate courses, and more on how to get a DH education.
John Walsh is known in the DH world for his work on TEI, XML, and other formal data models of humanities media. He replied:
I started undergrad as a fine arts major (graphic design) at Ohio University, before switching to English literary studies. As an art major, I was required during my freshman year to take “Comparative Arts I & II,” in which we studied mostly the formal aspects of literature, visual arts, music, and architecture. Each of the two classes occupied a ten-week “quarter” (fall winter spring summer), rather than a semester. At the time OU had a department of comparative arts, which has since become the School of Interdisciplinary Arts.
In any case, they were fascinating classes, and until you asked the question, I hadn’t really considered those courses in the context of DH, but they were definitely relevant and influential to my own work. I took these courses in the 80s, but I imagine an updated version that took into account digital media and digital representations of non-digital media would be especially useful. The study of the formal aspects of these different art forms and media and shared issues of composition and construction gave me a solid foundation for my own work constructing things to model and represent these formal characteristics and relationships.
Walsh was the first one to single out a specific humanities course as particularly beneficial to the DH agenda. It makes sense: the course appears to have crossed many boundaries, focusing particularly on formal similarities. I’d hazard that this approach is at the heart of many of the more computational and formal areas of digital humanities (but perhaps less so for those areas more aligned with new media or critical theory).
I agree web development should be in the mix somewhere, along with something like Ryan Cordell’s “Text Technologies” that would cover various representations of text/documents and a look at their production, digital and otherwise, as well as tools (text analysis, topic modeling, visualization) for doing interesting things with those texts/documents.
Otherwise, Walsh’s courses aligned with those of Underwood and Drucker.
Matt Jockers‘ expertise, like Underwoods, tends toward computational literary history and criticism. His reply was short and to the point:
The thing I see missing here are courses Linguistics and Machine Learning. Specifically courses in computational linguistics, corpus linguistics, and NLP. The later are sometimes found in the CS depts. and sometimes in linguistics, it depends. Likewise, courses in Machine Learning are sometimes found in Statistics (as at Stanford) and sometimes in CS (as at UNL).
Jockers, like Underwood, mentioned that I was missing linguistics. On the twitter conversation, Heather Froehlich pointed out the same deficiency. He and Underwood also pointed out machine learning, which are particularly useful for the sort of research they both do.
I was initially surprised by how homogeneous the answers were, given the much-touted diversity of the digital humanities. I had asked a few others to get back to me, who for various reasons couldn’t get back to me at the time, situated more closely in the new media, alt-ac, and library camps, but even the similarity among those I asked was a bit surprising. Is it that DH is slowly canonizing around particular axes and methods, or is my selection criteria just woefully biased? I wouldn’t be too surprised if it were the latter.
In the end, it seems (at least according to life-paths of these particular digital humanists), the modern digital humanist should be a passionate generalist, well-versed in their particular field of humanistic inquiry, and decently-versed in a dizzying array of subjects and methods that are tied to computers in some way or another. The path is not necessarily one an undergraduate curriculum is well-suited for, but the self-motivated have many potential sources for education.
I was initially hoping to turn this short survey into a list of potential undergraduate curricula for different DH paths (much like my list of DH syllabi), but it seems we’re either not yet at that stage, or DH is particularly ill-suited for the undergraduate-style curricula. I’m hoping some of you will leave comments on the areas of DH I’ve clearly missed, but from the view thus-far, there seems to be more similarities than differences.