Dreaming of Neogeography … and Nightmares About the Spatial History of Atrocities

After a brief but exhilarating trip to Moscow for what turned out to be a fantastic conference on the history of Russian and Soviet photography, your humble narrator finds himself spending this overcast weekend recovering from jet lag and nursing cups of tea whilst taking care of the chores that inevitably pile up when one spends life on airplanes. There’s plenty to be done. After some thinking and conversations with fellow graduate students at Moscow, I think I have a strong idea for how to revise one of the chapters of Developing Powers on northern Afghanistan, and I’ll have the chance to test out a few of the ideas at Harvard’s Russian and East European History workshop in a few weeks’ time. Visitors are constantly arriving in Cambridge, prompting your author to make runs for the favorite tea of the day (chai one week, sleepy chamomile the other). Readings in centuries-old Persian texts pile up, too, making the more modern Afghan texts I use in the manuscript – difficult but at least written decades rather than centuries ago. As always, the seminar I’ve been taking on digital history continues to prompt valuable ideas about how to transform the project beyond just a “mere” monograph made from dead trees.

A screen capture of the Persian-language Afghan “Death Lists,” which provide a comprehensive view into the workings of the PDPA regime from 1978-1979

The theme for this week of the class has been neogeography – how we use GIS tools from ArcGIS and MapInfo to ever-developing tools like NeatLine and PastMapper – and its applications to our own work. It’s a timely topic for discussion. Just as I had submitted an initial proposal to Professor O’Neill for a digital history project having to do with the outflow of refugees from Afghanistan into Pakistan and Iran throughout the 1980s, this past week the Netherlands’ National Prosecutor’s Office published a groundbreaking series of detailed lists of the more than 4,000 people who were murdered by the Communist PDPA (People’s Democratic Party of Afghanistan) regime from 1978-1979. These killings constitute a terror separate from the one the country saw after the Soviet invasion in December 1979; following the overthrow and murder of President Muhammad Daoud Khan in July 1978, an internally fractured but still potent ultra-nationalist Communist Party embarked on a killing spree of land owners, Shi’ites, former members of the Daoud regime, and members of the wide spectrum of old anti-Daoud Left groups. The files – available here in Persian through the Dutch authorities – which contain detailed information on the names, professions, and ambiguously-named “places” of the victims, provide one of the most comprehensive looks yet to the Afghan (and international) public about the timing, scale, and geography of the atrocities.

The New York Times’ piece describes how the files became public:

The chain of events that led to the lists’ discovery began with an asylum request by Amanullah Osman, the head of interrogation for Afghan intelligence in 1978 and 1979, who fled to the Netherlands in 1993. In his asylum interview, according to the prosecutor’s office, Mr. Osman admitted to signing documents concerning people who were to be executed. “That was expected and desired of me,” he said. “If you don’t go along with it, you can never attain such a high position.”

The Dutch denied him asylum but never expelled him, and eventually opened up a war crimes investigation. That led them to a 93-year-old Afghan refugee in Germany who gave them the death lists, which she had gotten from a former United Nations official, Felix Ermacora, who had never released them. Dutch authorities said they were confident of the lists’ authenticity.

The prosecution was dropped in 2012 when Mr. Osman died, and the Dutch decided to release the lists. “The close relatives of the deceased in this case have the right to know the truth about the circumstances of the disappearance and the final fate of their loved ones,” the prosecutor’s office said.

In the context of post-1979 Afghanistan this all makes for emotional – and potentially explosive – material. Many people knew nothing about what happened to their relatives since the day they disappeared in the year-and-a-half rule of Nur Mohammad Taraki and Hafizullah Amin. Naturally, however, there are calls for war crimes trials against the captors and jail officials who carried out these executions. But given the spree of bloodshed which 1978 began in the troubled Central Asian country – one that has not yet ended – others fear that opening a formal domestic or international war crimes tribunal would prove exceptionally complicated and fraught in a country that is, in many places, still effectively at civil war with itself. Making public later files that document 1980s-era atrocities – which already exist thanks to the painstaking work of Afghan and international archaeologists – would potentially implicate large swaths of the country’s political élite in human rights offenses. In short, the discoveries and release of the files constitute just the beginning of what will, one day – hopefully under conditions of peace – be a long public reckoning with more than thirty years of mass killing. All of the problems one sees today in Guatemala or Dhaka surrounding the use of historical materials and war crimes trials – the former a topic exhumed in some detail by fellow Harvard professor Kirsten Weld – are there to explode.

Beyond the images of progress trumpeted by the PDPA (here in a 1978 or 1979 Pashto-language documentary film)thousands were brutally tortured and executed. Might the tools of neogeography help us better penetrate these myths and do this traumatic history justice?

But I digress. For the historian of Central Asia – if one like myself more concerned with development and modernization than mass killing and atrocities – the death lists hold enormous potential, if also a fraught one, for better understanding the past. Can the tools of neogeography potentially help us to better understand – whether as academics or as people making scholarly interventions into a fraught domestic Afghan debate – these dark days of the country’s past? Or do they slide, as early 1990s-era critics of GIS techniques argued, slip too comfortably into an autistic positivism that does little either to understand the past and/or facilitate truth and reconciliation in post-genocide societies?

I’m still putting together a usable spreadsheet from the Dutch documents – their PDFs convert sloppily to Excel documents – and so I don’t yet have a workable .XLS or .CSV database to use as a test platform for many of the tools O’Neill has invited us to test this week, but a tour of several of the suggested options we have on the seminar syllabus (and one that’s not) invites several thoughts on how one might write the history of atrocities in 20th century Central Asia. One that has particularly grabbed my eye since we were introduced to it last week (albeit indirectly) is GeoTrellis, a piece of software produced by Azavea, a Philadelphia-based software firm that produces GIS software primarily for private sector clients. A team of digital historians at the University of Richmond, having secured an NEH Office of Digital Humanities start-up grant, used GeoTrellis towards the end of “visualizing emancipation,” the name of an impressive project that follows the timing of the release of slaves throughout the American South (and North) in the years during and after the American Civil War. The key concept here is that GeoTrellis allows us to track not only the place (inherent almost to the definition of GIS software) but also the timing of events. By seamlessly “playing” one map layer after another in chronological sequence, we get a sense of quantitative change over time that bolsters our (hopefully polished) prose writing.

“Visualizing Emancipation” with GeoTrellis

The application for data like the Afghan Death Lists – which include the date of execution of many victims as well as their “crime” and place of origin – through software like GeoTrellis is potentially enormous. With a source like the Death Lists, of course, it’s crucial to remember that we’re limited by the terms of the executioners, and have to be mindful about reifying them further. Many innocent Afghan Shi’as were murdered for having committed the crime of being a “Khomeini,” while by far the most common charge against ordinary individuals was “Ikhwani” (member of the Muslim Brotherhood), a charge that could be applied with equal impunity towards hardened followers of the brutal commander Gulbuddin Hekmatyar or anyone who stood in the way of the PDPA regime’s megalomaniacal ambitions to create a Pashtun-dominated Central Asian state that would modernize itself leaps and bounds faster than had even the Soviet Union – no matter if it took mountains of Afghan corpses or a direct war with Pakistan to accomplish this goal.

Yet if the historian keeps in mind the criminal nature of the files she has at hand here, there’s a potential wealth of information we could glean about the history of Afghanistan by plugging a database derived from the Death Lists into software like GeoTrellis. We might get a better sense of the timing of the confessional killing directed from Kabul, as well as the extent to which the PDPA regime – dominated at first by ultra-nationalist Pashtuns from eastern Afghanistan who overthrew the (southern) Durrani regime of Daoud – devoured its own versus eliminating opponents from southern tribes first. We’d get a better sense, too, of the limits of PDPA power (how many victims came from Kabul and the environs, as opposed to the hard-to-reach Hazarajat massif or the mountains of Badakhshan in northeastern Afghanistan?). A more detailed analysis of the killing under Nur Mohammad Taraki as opposed to Hafizullah Amin (a fellow PDPA leader who murdered Taraki in September 1979, alienating the PDPA’s patrons in Moscow) would help us better contextualize Soviet decision-making in the autumn of that year and help understand (or discredit) Soviet claims that one of the legitimate reasons it had to invade was the unsustainable civil war and reign of terror that Amin (as opposed to Taraki) had unleashed on the country. In short, there’s lots of motivation there for your humble narrator to keep on updating the spreadsheet to turn it into something that the gears of GIS programs like GeoTrellis might accept.

Other pieces of software that we were invited to examine this week offer their own advantages, albeit ones perhaps not as well suited for the Death Lists (as opposed to other kinds of data). For my original proposal for our Digital History project, I intend(ed?) to use ArcGIS, one of the most widespread GIS suites that does (or so I thought) most of what I need to depict the displacement of refuges from their homes in Afghanistan to the neighboring countries (other than the Soviet Union). Practically speaking, as someone who’s migrating from using MapInfo and hence makes a lot of dumb mistakes in non-MapInfo or non-Google products, one of the advantages of Arc is that there’s plenty of expertise around for it, particularly at Harvard’s Center for Geospatial Analysis. Having smart Harvard undergrads to exploit ask for help is an important and useful resource, one that’s not available (yet?) for less general-use products like Time Map. More seriously, though, one fears that in piggybacking on to using ArcGIS, or R (another popular visualization tool), historians risk simply enslaving themselves to the existing tools used by quantitative political scientists, rather than reflecting critically on the specialized tools they might need for their own work.

Neatline: amazing for micro-history, less so for big-scale change?

For me personally, that’s where tools like Neatline, a “a geotemporal exhibit-builder” that allows historians to build-in visual narratives to the geographical data they’ve accumulated. Scholars have used Neatline both as a general-use mapping tool (one that looks more attractive than most Arc maps, I would add), but also – more creatively – as a tool to embed micro-narrative sources (diaries, travel accounts, and so on) into maps. Add to this the fact that it integrates seamlessly with Omeka, and the temptation to build seamless exhibitions incorporating Neatline maps is strong indeed. However, since the basic thrust (at least as advertised) of Neatline is its ability to track micro-level movement across (usually) historical maps, I wonder how useful it is for analyses of macro change of the kind that I’d hope to tell with my refugee data. If only I knew – if I had the data to tell – where individuals from Afghan villages went to in Pakistan or Iran, it’s easy to imagine how one might draw beautiful maps of the escape from the PDPA into the relative safety of the Northwest Frontier Province, or slums in Tehran. To my knowledge, however – or maybe to my ignorance [perhaps the UNHCR archives in Geneva have something] – such data just isn’t there, at least not behind closed doors in Peshawar, Tehran, or Islamabad. But maybe that’s what the coming weeks are for: scouring the resources of Harvard and other libraries to see if there is something – maybe in Iranian or Pakistani publications – that might help us tell this micro-story. Yet there’s the question, too, of whether I really want to (or can) narrate a story of escape, when perhaps what’s most important (and what the data does help us do) is a story of the depopulation of many parts of Afghanistan. Maybe Arc is just enough?

I’ll have to give it a think. For now, there remains – as always – much writing, much reading, and much thinking to do. Things could be worse.


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s