It’s the end to another busy but productive week here at Harvard. After recovering from the circadian devastation that a red-eye flight from San Francisco wrought upon me, it was back to the usual rhythm: Persian classes (readings from Ali Akbar Dekhoda), our digital history, and lots of writing. With only slight exaggeration I can say that I’ve completed revisions to the last chapter of the book manuscript (not the last chapter I’m editing and re-working this autumn, sadly …) on the Afghan-Soviet borderscape in the 1970s and 1980s; doing so has been a crash course in hunting down obscure Tajik- and Uzbek-language books on the one hand, and speaking to my anthropologist colleagues at the Harvard Academy for reading recommendations outside of my usual ken. And there’s a rich trove there: works like Jean and John L. Comaroff’s edited volume Law and Disorder in the Postcolony and Adam T. Smith’s The Political Landscape have proven useful in helping me think of how to deal with my own Soviet sources, and to conceptualize the shifts in the Turkestan borderlands during the 1980s: a period when mass administrative purges in the Tajik and Uzbek SSRs corresponded with the installation of the Najib regime in Afghanistan, its ‘National Reconciliation’ campaign, and the Soviet withdrawal. Can we understand all of these events as part of some broader pattern?
Beyond my usual monkish lifestyle of writing and tasking Harvard Inter-Library Loan librarians with hunting down obscure Soviet books, I’m also beginning to make more serious attempts to map the data I’ve been collecting and collating for Professor O’Neill’s Digital History seminar. As I’ve written before, there are fundamentally two databases here: one having to do with the so-called Death Lists – which tell us about the professional and religious landscape of Afghanistan and the timing and geography of PDPA-led mass killing in 1979; the other having to do with emigration from Afghanistan into Pakistan and Iran during the late 1970s and 1980s. Dealing with the former database (the Death Lists) in ARC GIS is proving to be a bit of a challenge: there’s problems in depicting information accurately, much less beautifully, when so many people were killed in the same place (hundreds of deaths in ‘Ghazni‘ get depicted as one dot, rather than a cluster), and there’s the whole chronological component to deal with. I have a couple of ideas here. Most crudely, it might be possible to use a GIF or a Prezi to create an animated version of the chronology of killing in the spring and summer of 1979. But Harvard’s History Design Studio and other outlets for digital humanities in fair Cambridge are also apparently in the process of making it easier for scholars here to use NeatLine, an application developed at Virginia that would be well-suited to displaying what I really want – the chronological movement of peoples across Central Asia. It would help make more clear the ways in which (Afghan and Soviet) Communism ironically ends up destroying the Afghan dream of ‘Pashtunistan‘ – a Pashtun-dominated state in Central Asia carved out of Afghanistan and western Pakistan – by depopulating the core Pashtun territories of Afghanistan and forcing those people into Pakistan (where they constituted, and constitute a humanitarian crisis and provoke anxiety about the ‘Pashtunization’ of parts of Pakistan, most notably Karachi, typically viewed as the ‘property’ of muhajirs or other nationalities.
Still, after an hour or two banging my head against the keyboard in CGIS yesterday, I’ve begun to produce some rough drafts of some of the graphics I might want to use.
I don’t think it’s bad, but there are still some problems – I’m returning to my Edward Tufte. The green dots in the map above each represent 1500 people who had left from provinces in Afghanistan by 1985; the turquoise dots also represent 1500 people who had arrived into provinces in Pakistan by the same date. The overall picture is clear: the migration provoked by the Soviet invasion was was largely one of peoples from provinces like Paktia, Nangarhar, and Qandahar (populous Pashtun provinces) almost entirely into Pakistan’s Northwest Frontier Province, its Federally Administered Tribal Areas, and, to a much lesser extent, Balochistan. Yet there are many quibbles one could raise here. I’ve used a relatively simple terrain and administrative underlay – are there better options? Choosing to depict population as dots is, I think, the right choice here, but then there’s choices of density and visualization, too. Given the size of the map (or the level of the zoom), depicting one dot as one person would fill up all of the provinces in question. What’s the right balance – visually and analytically – between 1 dot : 1 person and 10,000:1? I’m not yet sure how to answer these questions. Perhaps it’s time for more visits to office hours or – more likely – hours in front of ARC. Mapping is time intensive!
But the fact that mapping and, more broadly, other forms of digital humanities, are so demanding of our time raises questions of the sort that we’re invited to consider in this week’s session on professionalization, assessment of DH scholarship, copyright management, and MOOCs. It’s obvious to me from my everyday work patterns that mapping and building (for example) Omeka sites takes way more time than writing (of first drafts, if not polished prose). This makes for a problem. While conditions and Stimmung vary from department to department, history can be a temperamentally and methodologically conservative discipline. It’s not uncommon for colleagues to fret about making a web site for fear that hiring committees would view a digital presence as “unprofessional.” The attitudes of major publishing houses and their editors can be even more entrenched: several publishers I have spoken with view digital projects essentially as a distraction, while one of the most prestigious academic publishers in the United States that I had to deal with for a project as an undergraduate research assistant did not even use e-mail to accept the corrections to a 1,000 page edited collection of documents I was working on, forcing me to babysit a fax machine in the basement of the Princeton History Department one afternoon. With attitudes like this, it’s no surprise that we often lag behind in assessing DH scholarship. If there’s a place where the dumbest stereotypes of the technocratic about the rest of us Luddites are fulfilled, it might be in academia and academic publishing.
Fortunately, many of this week’s readings address how to change this situation. A piece by Todd Presner, a professor of German and Jewish Studies at UCLA, cogently lays out some starting principles for the review of DH material that I think are a good start. Among them are the principle that “the work must be evaluated in the medium in which it was produced and published. If it’s a website, that means viewing it in a browser with the appropriate plug-ins necessary for the site to work. If it’s a virtual simulation model, that may mean going to a laboratory outfitted with the necessary software and projection systems to view the model. Work that is time based — like videos — will often be represented by stills, but reviewers also need to devote attention to clips in order to fully evaluate the work.” This may sound obvious, but it’s essential: I’ve heard of instances where people want to know why I can’t just print out my maps and mail them to a reader. As a young(ish) digital historian, one is often dealing with huge generational and methodological chasms that are hard to leap.
The obsession with “the argument” as a feature of academic history that I brought up last week is another aspect of this. As classmates have suggested, digital history projects are often better at opening up new horizons of questions than definitely answering existing ones. And there’s nothing wrong with that. I’ve often thought that the more falsifiable an argument is in history, the more open-and-shut it can be, yes – but also the less interesting it can get, too. As we know thanks to the work of scholars like Annette Gordon-Reed, there’s now a strong consensus that Thomas Jefferson fathered several if not all of Sally Hemings‘ children. Knowing that that happened is important for how we think about Jefferson and attitudes about race in 18th century America. But arguably more interesting are the broader panoramas that knowledge of that fact raise: the dividing lines of race in American society then, the history of slave-master sexual relations, the history of defining (some would say defaming) Hemings’ as a “slave girl” rather than just a “woman,” and so on. If you think “the argument” of work like Gordon-Reed’s is that Jefferson fathered Hemings’ children, you’re missing the point.
Just as good print manuscripts raise more questions than they answer, good DH scholarship should do the same. We have, however, to be careful not to fall into the trap of claiming that Omeka projects and the like are the same, or worth “half” of a book or article. Notes Presner: “Is a digital research project “equivalent” to a book published by a university press, an edited volume, a research article, or something else? These sorts of questions are often misguided since they are predicated on comparing fundamentally different knowledge artifacts and, perhaps more problematically, consider print publications as the norm and benchmark from which to measure all other work.” At its most ambitious, the digital history movement has to be about questioning these age-old institutional standards – especially so at a time when our books sell pitiful numbers of copies and departments are under attack by administrators.
As other readings discuss, however, we have to be careful about how we put our projects together. A piece by Dan Cohen and Roy Rosenzweig, “Owning the Past,” discusses some of the pitfalls of copyright in digital humanities scholarship. Reading the piece, however, reminded me of a special set of exceptions that ironically makes working on the modern history of the Soviet Union and Persianate Asia especially convenient (if not worth the tradeoff of using squat latrines when working at Tajik archives). The history of international copyright law and the Soviet Union is complicated, but with some notable exceptions (Prokofiev, Gorkii, authors who published their works in member countries of the Berne Convention before it being published in the USSR) most works that were published in the Soviet Union prior to 1973 – when Moscow acceded to the Universal Copyright Convention – can be freely used, up to and including (in theory) copying them and selling them for profit. (Not that we innocent historians would do something … but still.) Confusingly, through a set of post-Soviet arrangements, pre-1973 works can’t be reproduced in other CIS countries, but to my knowledge, these laws have no force outside of the former Soviet world.
The situation is even more unusual with respect to Afghanistan and Iran. Not only is neither country party to the Berne Convention on copyright (the standard international copyright treaty); the United States makes explicit that it has no copyright agreement at all with either country (the same is also true of Iraq and Turkmenistan, making this the largest copyright-free region in the world). Hence, while Cohen and Rosenzweig wring their hands over the numerous restrictions that (implicitly) Americanists face vis-a-vis copyright law – notably, almost all music is incredibly difficult to use, film slightly less so – simply don’t apply to materials originating in Afghanistan or Iran. In theory, this means that it’s much easier for someone to do ambitious DH work with (for example) Iranian New Wave cinema or indeed anything produced in those countries. One would have thought that the US occupation of Afghanistan – a “strategic partner” – would have changed this; maybe putative US-Iran negotiations and diplomatic normalization will change this, too. But these loopholes make it unusually easy to be a digital historian for this part of the world – and give me another reason to tease our Americanist colleagues for being parochial. (Obviously, there could be serious problems if both you and your pirated collection of Afghan or Iranian works happen to be on Iranian or Afghan soil, or the Internet, at the same time, so one should be careful.)
Last but not least, we’ve invited this week to reflect on the MOOC (massively open online course) movement, a topic that I’ve blogged about in the past and that we frequently discussed in a Rhodes-run University Governance Group discussion club at Oxford. As I’ve written there, I think that much of the usually cited talking points regarding MOOCs – celebrity professors teaching courses! students in Nepal! 500,000 students in one course! – are presented utterly divorced from the context of the dramatic changes in the funding landscape of American universities in the last twenty years. As I’ve written in another post stemming from the University Governance Group, the amount of public funding for public universities in most states in America has been in decline since roughly the early 1990s, which – if you’re paying attention – roughly coincides with the rise of the “development” industry in so many universities. Private fund-raising replaces a public commitment to higher education.
MOOCs, I think, can be seen on one level as part of an attempt by universities to cut costs across the board: an attempt to get even larger-sized lectures and stretch resources even thinner, even as annual attendance costs at public universities top $30,000 a year. A recent memorandum on cuts for graduate funding at UC Berkeley escapes me as I write, but MOOC-ification too often represents an administrator’s dream but a teacher’s nightmare. 500-person lecture classes with one graduate student TA (sometimes the case at a large public university like Berkeley) where students – especially those without means – can barely find support if they’re struggling are small fry compared to the spectre of a 100,000-person course with close to zero administrative overhead. Such courses may make sense for introductory quantitative subjects, but as the press on MOOCs has highlighted, many of these courses have incredibly high dropout rates. It’s unclear how one would manage the grading or assessment of 100,000 essays – Harvard PhD student friends who TA are routinely underwater with loads a percentage of that number – and the repeatedly-cited dream of having Michael Sandel teach every philosophy class is ignorant of the way that humanities scholarship actually works. If history is to fulfill its true function in universities as a dialogue between present and past – rather than mere recitation of ‘one damn thing after another’ – it’s crucial that we emphasize a multiplicity of viewpoints in teaching, and focus on training the next generation of scholars – yes, often by them giving mediocre lectures or classes and learning – rather than sitting soporifically before taped AJP Taylor lectures or their modern equivalent. We have to stay abreast of digital changes to our profession, as many of the readings this week underscore, but as far as MOOCs are concerned, technological Micawberism is not an actual argument.