Deleuzian simulation models
19/03/2012 Comments Off on Deleuzian simulation models
I picked up a book a while back from a university bookstore solely for its title: A Deleuzian Century? I have started it a few times but not finished it — it’s that sort of book (or I’m that sort of reader). Foucault suggested that this century would be Deleuzian, that Gilles Deleuze, a French philosopher, would be central to thinking in the 21st century.
I made it through both volumes of Deleuze and Guattari’s Capitalism and Schizophrenia and also read Brian Massumi’s excellent A User’s Guide to Capitalism and Schizophrenia. I admit that I don’t ‘get’ Deleuze. I can read the sentences and paragraphs, I can start picking up concepts like bodies-without-organs and rhizomes, but I am always outside the concepts the same way one cooks from an unfamiliar recipe.
Nevertheless, I am beginning to think that Foucault was right.
One Deleuzian concept that I find utopian and naive is the rhizome. A rhizomatic plant sends stems through the ground, putting roots down and leaves up from new nodes. The structure is nonhierarchical. Deleuze contrasts this type of growth with things like trees, which have a center trunk, a hierarchy of limbs and twigs, and an integrated existence (it can be killed as a whole). He uses the rhizome metaphor to describe new ways of becoming that are lateral and invasive, that don’t depend on hierarchy or permission. It is meant as a liberating metaphor.
I have just been looking at cellular automata models, including Schelling’s model, The Game of Life, and Wolfram’s classification of one-dimensional models. They are discussed in a excellent on-line course by Scott Page. It occurred to me that these are mathematical representations of Deleuzian thought. They are presented as flat models, as full depictions of their worlds from which patterns of organization spontaneously emerge.
Key to both the Deleuzian metaphor and cellular automata model is the idea of ’emergent properties’. Patterns and organisation are thought to occur spontaneously as a result of individual elements just doing what they do. For Deleuze, these are new ways of being — I hesitate to say new ‘identities’, although I think that’s what he’s aiming at. For simulation modellers, it is complex order arising from simplicity, such as Wolfram Rule 110.
Simulation model thinking is rhizomatic thinking. Like John Wheeler’s ‘it from bit’, it considers that the binary on/off can be the basis for all organisation and the rest is just emergent properties. The simple rules of a Wolfram model push themselves through blank grids to establish new patterns, which can repeat themselves with relying on the rest of the pattern.
Simulation models are becoming more important in economic and policy analysis. Stats NZ, for example, recently hosted Martin Spielauer from Statistics Canada to talk about simulation modelling. As these models become more accepted, so will the underlying thinking.
We are becoming Deleuzian without even knowing it.