Computational Nature Studies

The ABC's new arts show is about landscape painting; in the first episode aired last week, John Wolseley worked with three Bendigo artists at an old mine site. For mainstream arts TV it was actually pretty good, thanks largely to Wolseley's eccentric persona - somewhere between British naturalist and deep ecology shaman. His ethos also struck me, based around an experiential and material immersion in the landscape. At one point he scraped his canvas over some charred saplings, "collaborating" with them to arrive at a scratchy charcoal underlayer of the developing work.

Over the weekend I visited the coast and wondered about generative art as a medium for responding to landscape. I can't grind up local ochres for a nice impasto, so where does that leave me? Alienated, blinking in the sunlight, laptop plein-air? I did a quick experiment, studying the textures of the surrounding casurina forest and coding up a simple drawing machine in Processing. As in drawing, I worked from close observation, but it was synoptic, rather than specific; looking for patterns, tendencies, signatures, rather than tracing details. And to respond the generative artist has to find a procedure or algorithm, rather than an image - a way of making something. That gap between observation and procedure is productive in itself. Observation constantly challenges the representational ability of the algorithm: the lichen that grows on the casurina trunks is a crucial visual element in that texture, but the procedure I wrote was focused on growing trunks, not lichen. One path leads towards more detailed simulation; another could be to treat the model as a sketch or diagram - a looser and perhaps more poetic representation.


Jon McCormack's work has been exploring this territory for years; Eden, Morphogenesis (pictured) and the latest Bloom are all responses to a specific (Australian) landscape. The audio ecosystem of Eden was inspired by a visit to Litchfield National Park in the Northern Territory. Morphogenesis and Bloom come from more detailed studies of native plants, which are modelled then slightly mutated. McCormack's earlier Turbulence was an all-out digital hypernature; these latest works are crypic variants, rather than fantastic monsters. By "detuning" the model, these nature studies interact with their originals to set up a kind of uncanny interference pattern.

Way back in 1994 McCormack argued for the role of the computational medium as a way of knowing about the world: "You might expect that my ideas about the world are introverted around the machine, in fact the opposite is true. The computer has shown me things about the world that I could not have known, understood or seen any other way. I see and appreciate nature in a fundamentally different way than before." (Wild essay) Is this still true of the Flash/Processing generation of artists? The current aesthetic seems to reflect an environment that is predominantly urban, networked and social. Any exceptions come to mind? Where are the new computational nature studies?