Monday, March 12, 2007

Lisa Jevbratt - Infome Imaging

Lisa Jevbratt has been doing data art for some time now. Her 1:1 (1999) was one of the first data-vis works to gain critical attention in new media art circles. I re-read some of Jevbratt's writing recently, and the artist pointed me to this 2005 paper, which in part sets out the concept of the Infome. Jevbratt seems to be changing direction - towards bio/eco practices - but her work remains significant, especially while data is the new code/black/whatever, for the Processing generation. The Infome idea is particularly interesting, because it creates a distinctive sense of just what data is.

Jevbratt's Infome is a kind of data cosmology - the Infome is an "all-encompassing network environment/organism that consists of all computers and code." Once you get past the biological analogy, the Infome offers a way to treat data as a kind of material that is concrete and self-sufficient, but also shaped by the (social, political, technological) forces outside it. Data is indexical, but not in the empirical sense of measurement or simple correspondence. Instead Jevbratt uses another material (geological) analogy; the Infome is a kind of landscape in which external forces and structures are overlayed and condensed. Another nice twist is that visualisation becomes recursive: "Images can now simultaneously be reality, since they are part of the Infome, and an imprint of that reality, as if the image produced by a potato stamp were also a potato."

Jevbratt's images of the Infome in 1:1 and Infome Imager Lite aspire to this kind of material directness, making a "slice" or "imprint" of the data. She describes the images as "real, objects for interpretation, not interpretations." This desire to present the data "in itself" closely resembles the "pure data" aesthetics of the audiovisual databenders I mentioned in "Hearing Pure Data" (2004). We can make the same critiques of Jevbratt's work - that we can't see the data in itself, only its specific mapping. Jevbratt does take great care to explain the mappings used, and best of all in IIL she encourages the user to experiment with changing mappings and datasets - an artistic precursor of the public data literacy now mentioned in relation to social data-vis services Swivel and ManyEyes. The critique still stands though; it's clearest in the way Jevbratt wraps all these visualisations around the rectangular picture plane - a structure that has no inherent relation to the data, but a significant relation to the art-world context that these works function in.

For Jevbratt these data-impressions allow us to "use our vision to think" - information and pattern arise from a perceptual process, rather than a computational analysis. Like much other data art, Jevbratt resists providing information, in the sense of meaning or message; instead she offers a substrate for information, a field of potential meaning. She writes of seeking "something unexpected," "hints, suggestions, and openings" that lead us into the Infome itself, its immanent, collective dynamics, even its emergent, distributed agency. It's a kind of data mysticism, but also an attempt to sense the real but otherwise imperceptable shapes of digital culture.