Thanks yet again to Nathan Yau and flowingdata.com for bringing this to our attention. Researchers looking into air pollution at Berkeley’s Department of Environmental Science, Policy, and Management have taken complex data on the composition of chemicals in the air at some locations ranging from sylvan mountain idyll to hellish trunk road tunnel and sought to make the differences easier for the non-expert to spot. Did they use a graph? A video? Nope, they made it into sound. Now this is a very interesting approach, and one that has almost never been taken by scientists, although there are many examples of sound artists and electronic musicians turning data into sound from the “other side”. We are very early on in the evolution of data audibilization and it requires a lot of explanation to the newcomer (i.e. everybody) but it is quite rewarding to persist with listening and trying to work it out.
As a footnote, I found it hard not to think there is a heavy moral content to the choice of data and locations (not to mention the glib Rachel Carson-esque reference to birth defects without the need to cite the evidence), and we would expect to find this reflected in the aesthetic choices for converting data to sound, but actually these are quite value-neutral. And that deserves some praise. Few people who know me from the stats world would know that I used to make some sound art and electroacoustic compositions, so I appreciate how hard it is to present sound that doesn’t play to the gallery with cliches.