Over at Five Thirty-Eight, Jody Avirgan interviews Fotini Christia, an MIT political scientist who was given 10 million cell-phone records by a mobile phone company to do big data research. Christia explains that the motive for the phone company for handing over the records had to do with wanting to further development research in Yemen.
The audio and interview are worth a listen. Also this:
Jody Avirgan: I wonder if you can start by describing the challenge that you were trying to solve by using this data?
Fotini Christia: Yemen is a fascinating place because it is a hotspot and a place of trouble. It has been an issue for the U.S. in terms of terrorism, instability, continued conflict. And though we do have quite a bit of anecdotal evidence [about the country], it tends to be very selective. It’s usually from journalists that can be on the ground in very particular places. So there’s a lot we don’t get to hear about Yemen because it’s so hard to do social scientific or analytic work on the ground. It’s not a place that has rich census data. It’s not a place that has rich household-level data, recent survey or polling data. So people tried to be creative about where else you can get information.
That line about journalists speaks to questions of epistemology and, as someone who is anthropologically trained, its assumption that ‘evidence’ derives from mass data rather than familiarity with ‘particular places’ struck me.
In the interview, Christia describes drone attacks as an “exogenous event,” like an earthquake (her example), but this is a strange characterization for a number of reasons. For one, an exogenous event is one that has no particular relationship to the sociality of a place, but the US itself argues that they bomb particular people because they are doing militant-y things on the ground. Second, at least for FATA (and I would suspect Yemen), people are in fact making judgments about what might make them targetable (i.e. don’t talk to so-and-so; don’t make phone calls here or there or say this or that word; don’t go to this place at this hour) and are trying to avoid it. Moreover, opposition groups have repeatedly killed people they suspect to be informants in the aftermath of a drone attack. Others speculate that drone bombing is sometimes caught up in local rivalries where a person may (mis)inform or allege that his rival is a terrorist. In short, drones are not exogenous except in the theoretical frame that flattens place into blank space.
And finally, I haven’t read the paper, but in the interview at least, the deployment of the category of religion (i.e. Christia says they can see from the phone records how much religion structures life) is simplistic.
And finally, this kind of big data research opens for me all kinds of ethical questions about the researcher’s relationship to her subjects who have no idea that their metadata and information, down to the individual level it sounds like from the interview, has been handed over.