The system, which I will call Rhizomes is much more than the 'Performance' or the perceived elements comprising the 'Performance'. I see it as a whole dynamic ecosystem which supports the 'Performance' / art-work. This includes practical and academic research through the whole of this MA, as well as computer software and hardware development. Rhizomes has grown / emerged out of this ecosystem soup, which Curson and I have created (Curson and I could be described as simply vital soup ingredients or as within system active autonomous agents like 'Maxwell's demon' (in a closed thermodynamic work system Kaufmann, 2000) which are an integral part of the self-organisation of the complex system).

Throughout our MA research, Curson has been simultaneously evolving the software and Virtual Models (Curson, 2004) in a feedback process. Curson creates a version of the software. We play with this through improvisation of Pc and Vc and also test the software against the ability to re-create work. Through our play we seek awareness of limitations of the software. Curson also critically interrogates the functionality and user friendliness of each version. For example, he created a cross hair device which allowed the virtual camera (which we move to create the realised Vc) to orientate the axis of Vc flexibly to the live dancer, rather than being on a static and typical horizontal / vertical alignment (a cross-axis). From this development, we now understand that our previous Pc improvisation tended toward a fixed cross-axis on the stage, in part because the Vc relates most easily to a cross-axis.

I also use our creative play to stimulate my ideas for choreographic relationships between Vc and Pc, which I suggest to Curson, for example suggesting that two models be visible and two pieces of music play simultaneously. This means Curson has then to solve a new series of problems to further develop the software. Thus feedback loops exist within this complex living system of Rhizomes.

Curson has been aiming to create a generic software package which is flexible enough to operate over a wide range of choreographic processes, but can recreate movements easily and keep operator choices at any one time down to a functional number. These are seemingly at opposite poles (McKenzie, 1994), and Curson has had to develop lateral solutions, by altering his perspective on the nature of the Vc to (Rhizomes) System relationship. Each new perspective has emerged via lines of flight from a previous plateau of understanding that holds our current practical research, the current software, and my current academic reading. For example, during the creation of FLUX in phase II Curson (2004) started to see that the Vc could be regarded as a significant extra dancer, rather than an animated (albeit in real time) film or video. Reading on Hypertext created the stimulation for a 'line of flight' to gather impetus taking with it our practical research desires of connecting Vc-Pc scenes in a web of possibilities. Thus our then current plateau-position (of having simply a set of Vc-Pc relationships) took flight toward creating Vc bites that could be implemented in any order in real time, in relation to a dancer's similar real-time improvisation choices. Curson developed software with the ability to hold a palette of movement sequences that could be invoked during improvisation at any time, just as would an improvising dancer (ibid. for history of software development up to Flux). When we started researching Complexity Theory for Rhizomes, Curson had the further revelation of creating a whole new software package for our work, which could operate not just like a dancer but like a ‘thinking’ dancer with the ability to:

1. Create relative movement transferable between models, rather than absolute movement of one model
2. Create a modular system that allows processes to be constructed creatively & to flexibly network with the Vc processes that are being used (module packages which can link and control other modules)
3. Create an ability to link and relink these modules in different patterns to allow the Vc process to change easily during a performance (parameters and their boundaries can mutate / transform in real-time easily)
4. Create feedback loops inherent within the computer system (not just relying on Curson as the mediary (or Maxwell Demon) supplying the feedback)
5. Deriving movement from more than one source e.g. joysticks, maths equations, sensors

Curson has facilitated our software to evolve in the emerging directions 'suggested' by both the system and our creative play. Retrospective reflection has shown us how each version of the software package has had inherent properties, which favoured particular Vc movements and relationships with the Pc.. The system is evolving an increasingly complex hierarchy-network structure as described by Kaufman (2000) and Heylighen (1988) with concomitant emergence of new properties. These new emerging properties of the Vc system provide newly defined boundaries or constraints from within which Rhizomes 'work' is channelled (see ‘Work’ and ‘Work’ in Rhizomes). That is, the Pc-Vc relationships evolve and expand into new possibilities as the Vc system evolves, for example into creating 3rd order structures of intercutting between scenes/models yet continuing with the same movement vector (Jemma's Transition Phrase).

So Rhizomes exhibits emergence of new properties as it evolves with self-organisation, as well as being non-linear and composed of part-order part-chaos. It has perhaps quantum consciousness. It has an identity bound up within the intraconnections of its elements.

Rhizomes is a living evolving complex system / an autonomous agent.

3.2 Home Page, 'Beginning', Conclusion, Acknowledgements, References, Rhizomes Performances, Performances as References