Way back in 2005 (about 4o Internet years), I taught our intro to digital history for graduate students, the first of two required courses for our PhD students and these days a requirement for our MA students in Art History. Then, for a variety of reasons having mostly to do with a sojourn in administration, I didn’t teach the course again until this semester.
As you might imagine, when I looked at my syllabus from 2005 for inspiration, all I could do was laugh. I think I ended up using something like 20% of the readings from 2005. Many of the weekly topics were the same, but the flood of writing about digital history/humanities, and the many new tools, especially computational tools, available to us now meant that I had to substantially rethink my approach to the course. We’ve wrapped up for the semester and as I sit down to start grading my students’ final projects next week, I’ll be looking at a whole lot of work that I never could have imagined in 2005.
To be sure, much of what my students were interested in hasn’t changed a lot. Their questions as historians and art historians in 2013 are very similar to the questions they were asking in 2005. The big difference is the tools available to them for answering those questions.
And of those tools and/or approaches to digital historical/art historical analysis that really stand out are text analysis, especially topic modeling with tools like Mallet, and geographic analysis with GIS tools such as ArcGIS. In both cases, the students using these tools have engaged in what we’re now calling “distant reading” of large volumes of information and then using one or both of tools like Mallet and ArcGIS to analyze their data and then display their findings. I don’t know if anyone even used the term “distant reading” in 2005. I know I didn’t.
Tools that would allow the distant reading of large corpora of historical sources certainly existed in 2005, but the bar to entry was so high that no one in our department, and I suspect in any history department, was incorporating these sorts of tools into their teaching. These days, however, the bar to entry has dropped substantially and so a number of my students were able to do some pretty sophisticated work in topic modeling and geographic analysis.
The other thing that makes this sort of work possible, of course, is the mass digitization of sources for the students to then use these tools on. Distant reading depends on these corpora and at least for teaching we can’t ask our students to do the work without easy access to the sources they need. Which is why we need to keep advocating for open access to digitized historical sources.
Finally, few of my students would have been able to do the work they did without the technological support and advice of a couple of the more advanced students in the room who showed up on campus this fall already reasonably adept with these tools. Had they not been available (and so willing) to support their colleagues in the course, the results I am looking at now would have been much less diverse.
For me, the big result is that before I teach the course again (I sure hope I won’t wait another eight years) I need to rethink the overall structure of the syllabus. This time around I took the existing structure and just plugged in new topics and readings. And while that worked well enough, it didn’t work well enough for me. Next time around I’m going to restructure the course more specifically around four or five key topics in digital history/art history and leave it up to the students to fill in around the margins of those topics. That way we’ll have more time to do more sophisticated work on fewer things.
What those things are will be a function of how long a hiatus I take between this semester and the next time I teach the course.