Thursday, September 21, 2006

Data, Code & Performance

A few more thoughts and questions following on from the previous post, and responses to it. If data art isn't necessarily concerned with the (apparent) meaning of its datasets, or their empirical basis, then what is it concerned with? Perhaps one answer has something to do with performance. Whatever else it does, this work performs a process that is meaningful in itself. Whatever else it says, it also says, "watch what I do with this data." It displays a data literacy, an ability to acquire, munge, filter, process, map and render. Since it's primarily operating as art, rather than functional visualisation / sonification, it also demonstrates a process of translating or mediating between those domains. This isn't a criticism (necessarily), just trying to think through a few basics, and taking on those points from toxi and infosthetics re. the tension between art and visualisation here. If data art is partly self-referential performance, then what kind of cultural values exist / are constructed around that? Manovich refers to "data-subjectivity" - are data artists exploring / peforming this "super-modern" state of being?

I'm sure there's a connection here somewhere with literal acts of data-performance. I saw some live coding performances at the Medi(t)ations conference in Adelaide (blogged earlier). Brisbane duo aa-cell(Andrew Sorensen and Andrew Brown) played a great set - two laptops, both running Sorensen's own Impromptu environment, with screens projected to show the accumulating code. Here too there was a kind of mediation between computational and cultural domains - a performance of (largely obscure) code structures that generated a sonic structure dense with musical references. It was partly the pulse of a synth kick drum (hand coded, of course) but I came away thinking of Kraftwerk - laptop live coding as the new "man machine."

Live coding has a transparency that a lot of data art lacks - the code structure is gradually constructed, giving an (expert) observer some chance of following the formal, generative structure. Most data art conceals its mapping and munging, offering only an artefact and a promise that yes, this is "the data." Live coding's transparency is itself pretty opaque, though. At least one audience member at the Adelaide performance had no idea that the displayed text bore any relation to the sound. Live coding looks like great fun for the performers (like most improv), but what about the audience? Is data-subjectivity a prerequisite?