Monday, October 29, 2007

More is More: Multiplicity and Generative Art

Douglas Edric Stanley wrote a nice post recently on complexity and gestalts in code and generative graphics. In it he wonders about "all those lovely spindly lines we see populating so many Processing sketches, and how they relate with code stuctures." I've been wondering about the same thing for a while, and Stanley's post has prodded me to chase up a few of these ideas.

Stanley makes some astute observations about the aesthetic economics of generative art; the fact that it costs almost exactly the same, for the programmer, to draw one, a hundred or a million lines. Stanley pursues the machinic-perceptual implications - how simple code structures contribute to the formation of gestalts; but he only hints at what seems like a more interesting question, of how these generative aesthetics relate to their cultural environment: "all of these questions of abstraction and gestalt are in fact questions about our relationship to complexity and the role algorithmic machines (will inevitably) play in negotating our increasing complexity malaise."

I actually don't think complexity is the right concept here. For me complexity refers to causal relations that are networked, looped and intermeshed (as in "complex systems"). These "lovely spindly lines", and Stanley's gestalt-clouds, show us multiplicity but not (necessarily) complexity. Simple, linear processes are just as good at creating multiplicity. There's certainly a relationship here - complex systems often produce multiplicitous forms and structures; and causal complexities embedded in "real" datasets seem to be a reliable source of rich multiplicities - but complexity and multiplicity aren't the same thing. For the moment I want to focus on the aesthetics of multiplicity.

Multiplicity is the uber-motif of current digital generative art - especially the scene around Processing. Look through the Flickr Processing pool and try to find an image that isn't some kind of swarm, cloud, cluster, bunch, array or aggregate (this one is by illogico). The fact that it's easy to do is a partial and not-very-interesting explanation; to go one step further, it's easy and it feels good. Multiplicity offers a certain kind of aesthetic pleasure. There's probably a neuro-aesthetics of multiplicity, if you're into that, which would show how and where it feels good. Ramachandran and Hirstein have suggested that perceptual "binding" - our tendency to join perceptual elements into coherent wholes - is wired into our limbic system, because it's an ecologically useful thing to do. Finding coherence in complex perceptual fields just feels good. The perceptual fields in generative art are almost always playing at the edges of coherence, buzzing between swarm and gestalt - just the "sweet spot" that Ramachandran and Hirstein propose for art in general.

I don't find this explanation very satisfying either, because it doesn't seem to tell us anything much about the processes involved - it's a "just because," and a fairly deterministic one. Another way in is to think formally about the varieties of multiplicity in generative art. I rediscovered Jared Tarbell's wonderful Invader Fractal (below) in the Reas/Fry Processing book recently. It shows a kind of multiplicity that's the same but different to the "spindly lines" aesthetic. Each invader is the product of a simple algorithm; the whole mass is a visualisation of a space of potential - a sample (but not an exhaustive display) of the space of all-possible-25-pixel- invaders. Multiplicity here is a way to get a perceptual grasp on something quite abstract - that space of possibility. We get a visual "feel" for that space, but also a sense of its vastness, a sense of what lies beyond the visualisation. John F. Simon's Every Icon points in the same direction; towards the vastness of even a highly constrained space of possibility (32x32 1-bit pixels).

Perhaps current aesthetics of multiplicity are actually doing something similar. The technical differences are fairly minor; basically a switch in spatial organisation from array to overlay; a compression of instances into a single picture plane. The shortest (and my personal favourite) path to multiplicity in Processing is aggregation: turn off background() and let the sketch redraw. Reduce the opacity of the drawing for an accumulating visualisation of the space of possibility that your sketch is traversing. Multiplicity here isn't an effect or aesthetic for its own sake; it's intrinsically linked to one of the defining qualities of generative systems - their creation of large but distinctive spaces of potential. Multiplicity is again a way to literally sense that space; but also, since it almost never exhausts or saturates that space, it points to an open, ongoing multiplicity; it actualises a subset of a virtual multiplicity, and shows us (as in Every Icon) how traversing that space is only a question of specifics and contingencies. Multiplicity says "and so on"; an actual gesture towards the virtual.

Multiplicity refers to the specific space of potential in any single system, by actualising a subset of points within it; but it also metonymically refers to an even wider space of potential, which is the one that all computational generative art - and in fact all digital culture - traverses. Because of course any system can be tweaked and changed, no chunk of code is immutable or absolute, the machines of the Processing pool are ever-changing things that collectively sample the space of all possible (generative) computation. Just as it refers directly to the space of potential of its own (local) system, generative multiplicity alludes to the unthinkable space-of-spaces that contains that system - a space the system gradually traverses with every change in its code.

This, for me, explains the aesthetic and cultural charge that multiplicity carries. It's a gesture towards an abstract, unthinkable figure; an aesthetics of the virtual, in the Bergson / Deleuze sense of the word. What's more this particular form of virtuality, or possibility - the one accessable through code and computation - is at the core of digital culture and our contemporary situation. Generative multiplicity is, quite literally, a visualisation of that figure.


duncan said...

Great topic, thoroughly enjoyed this post.

Your idea here, that multiplicity in generative art gives us a perceptual grasp on a vast virtual space of possibility delineated by the code, is right on, I think. For me, some of the most interesting code generated art is compelling because it is able to create the aura of a world that extends beyond the simple visual shapes that constitute it. Indeed, the art is the creation of a virtual space, which, in its abstractness, can only be indicated by the relationship or behavior of the multiple parts that make up its concrete visual manifestation.

That successful generative art (or at least some of it) would satisfy this requirement makes sense, I think, as the creative process is in actuality a sculpting of code, i.e. a manipulation of the abstraction that constitutes the conditions under which the actual visual objects of the peice, in their multiplicity, appear. Thus the artist works with the rules that delineate the boundaries of (as you put it) a space of possibility - just as the possibilies of our "real" world is delineated by the abstract rules of physics. In working with code, the artist is effectively playing God in a virtual world rather than simply being the creator of a multitude of shapes.

I'm not sure that much or any of work succeeds at this, but I think you might be interested: .


Claudio Midolo said...

Ciao Mitchell!

I've just referenced this post on the research I'm currently taking on for my Major Studio: Interface class.

you can find it here:

keep up with the great work!

Christo Allegra said...

I think there is an interesting discussion here about singularity and multiplicity and there schizophrenic relationship with one another. The infinite is suggested in these threads of computational thinking and the permutation possibility that they offer. But there is something a little unsatisfyingly utopic in that vision though, even in the act of suggesting that they are describing or imagining possibility.

One doesn't have to be a Kristevian theorist to recognize that the failure of language to describe is built into the construction of meaning and the satisfying hubris of being able to make your machine behave in such a way to perform beyond ones ability, sketching a million lines a second for example, also hints at the hypertrophy that is inherent to the act. As the framerate slows, boundaries emerge and the possibility defines itself as bounded by limitations. God has nothing to do with it.

See Galloway vs. Debord;col1

I think that the lossy aspects of the effort, the inevitable failure of the systems and a definition of its constraints is still necessary if we are to discuss this work within the context of a postmodern dialect and remove generative art from the information politics defined by the cold war theorists that so much of it was influenced by.