the trip was "generative," i keep telling people. this means that it was not always fun but that it was always provocative -- of new ideas, ambitions and seeming necessities as well as of excitement and frustration. (don't worry, it was pretty fun too. also incredibly beautiful!)
we set up shop in the apartment that serves also as hrdag's bogota office, six people and seven computers and cords everywhere, and we hacked and hacked and hacked.
when we weren't hacking, we were busy eating new fruits (guanábana, lulo and feijoa were my favorites, although the sweet and crunchy "happy poop fruit" -- not its real name, i promise -- was also yummy) or eating at one of bogota's seemingly endless supply of fine restaurants, or talking with colleagues from a million and one organizations about data, politics, and the politics of data.
there's not enough anonymity in all the interwebs for me to do our many conversations justice, but i think i can safely say that colombia's war is a place in which the empirical, who-did-what-to-whom (and how many times and where) truth is clear to no one. maybe the broad outlines are coming into focus, but that's an exceptionally weak maybe, a maybe that in any case can only speak for a few times and places. there's no excuse for it -- i should know better! -- but in my naivete i was taken aback by the extent to which violence data and the interpretation thereof have become their own front in this conflict. (indeed, information and interpretation seem to be the one place in this conflict with conventional front lines: clearly identifiable, relatively impermeable and bitterly contested.)
"interesting" doesn't begin to do justice to our talks with colleagues outside of HRDAG. it doesn't really begin to do justice to our internal conversations, either. what was initially envisioned as a "let's get the colombia project really moving, hip hip hooray" sort of junket turned into an all-out frenzy (and i mean that in the best way), with data from at least four countries in various stages of disrepair and deduplication. my own work gave rise to a number of wackily intense sleep-deprived mood swings, as it became clear that (a) the data are nowhere near their finally ready state and (b) i can probably begin the matching process anyway. [this is Very Good News, as i'm counting on those descriptives for APSA in september.]
the matching process -- generally the object around which the last week was built -- also gave us plentiful opportunities to have bigger conversations about epistemologies, specifically the differing levels of epistemology involved in building the software, implementing the software, grokking the results that the software can offer, and trusting the grokked stuff. and then of course, the social scientists among us were on our high horses about what might be causally claimed on the basis of our trust in the grokked stuff.
in sum, the week was a single long opportunity to think [much more clearly than social scientists typically seem to] about the sensitivity of our analyses to the processes that generate our variables. qualitatively-oriented folks have been doing this for some time (occasionally self-defeatingly so), but dismantling the awesome and not-so-well-deserved rhetorical power of The Comprehensive Data Set has been a long time coming.
long story short, i'm all fired up and also kind of terrified -- i'll be back in bogota (plus maybe medellin?) next month, by which point i'll need to have mastered all the available information on conflict think tanks and NGO's and advanced my spanish well beyond its current state, which can be summarized as "i understand this, but damned if i can figure out any words to say in response." and i'm also trying to learn some more python? and finish my prospectus? and move? at the moment my ok-o-meter is swinging between RAWK! and OMGWTF!!1!, but i'm confident that i'll quit oscillating and settle somewhere in the middle.