So, I am going to jump topics from either of my usuals. Bear with me!
Data visualization is a fantastic, growing field of artistic work where artists grab large data sets and generate meaningful, compelling visuals from them.
However, when I first started reading about “data art” a few years ago, there was (I think) one source that mentioned, what about using the database itself as the artistic medium? What about art that *is* data?
Something in there combined with the politics of the time gave me an idea that I never felt compelled to finish, but which I’m going to share here because I’m concerned that no one else is doing something like this and maybe they could be.
The specific plan was to take the non-combatant casualty data set from IraqBodyCount, and generate a set of (semi)fictional data based on it. The data set would have one entry for every non-combatant death, with
- a statistically generated American name
- a location of death, also statistically plausible (ie. generated based on population density across the country)
- the nature of the death (from the IraqBodyCount data)
- the original location of the death (in Iraq)
- whatever other data is in the IraqBodyCount entry (date, and I forget what else is there)
The goal was to take a step to break past the disconnect we have for the deaths of those who are “different”, who live far away with names we aren’t familiar with in places we don’t know the geography of. To be able to look at the absolutely massive death toll from the war and start to comprehend what it would look like if this many people died right here. It also served to give a name to the otherwise nameless tally in the IBC data set.
After the data set would’ve been generated, it could have been made available for Google Earth mashups, custom visualizations, or even physical memorials left in the documented locations where these people “died”.
Why am I sharing this?
- I’d love to hear of other existing projects already doing this kind of semi-fictional data generation
- If someone was really hooked by this and wanted to make it happen, I might be interested in partnering on it
- It concerns me that I may have an idea here that *hasn’t* been done, that is, the general idea of making a greater artistic statement by fictionalizing or otherwise transforming the raw data into something that more strongly makes it’s point (without, you know, lying to people)
- 140 chars wasn’t enough.
I’ve gone on record before saying that ideas aren’t worth a lot next to an actual implementation. I suppose I do think that if someone went ahead and did this, it’d be great if they mentioned that I inspired them. But since I don’t see myself making this happen, I’d rather the ideas at least cross-pollinate out there and nobody really owes me anything.
Why am I not making this happen?
- I felt uneasy about mapping the deaths onto America when I am not American. I considered mapping them onto Canada instead, since that’s where I am, but it seemed to lose some of the power of identifying the victims with the aggressors.
- The war front where Canada is more strongly involved, Afghanistan, doesn’t have a project like IraqBodyCount generating a source data set – or at least not that I’ve found. And when I looked into this recently I found out that the war in Afghanistan has produced far, far less civilian / non-combatant casualties than the war in Iraq.
- Generating statistically-plausible American names didn’t look too hard, based on census data, but generating plausible locations for the deaths looked tricky. Accessing some kind of heat map of population density and using that to generate latitude/longitude per person could’ve worked, I think, but (due to the other reasons above) I stalled out on the first small roadblock.