Collaborating with the Behaving Machine: Simple Adaptive Systems for Generative and Interactive Music


Situated at the intersection of interactive computer music and generative art, this thesis is inspired by research in Artificial Life and Autonomous Robotics and applies some of the principles and methods of these fields in a practical music context. As such the project points toward a paradigm for computer music research and performance which complements current mainstream approaches and develops upon existing creative applications of Artificial Life research.
Many artists have adopted engineering techniques from the field of Artificial Life research as they seem to support a richer interactive experience with computers than is often achieved in digital interactive art. Moreover, the low level aspects of life which the research programmes aim to model are often evident in these artistic appropriations in the form of bizarre and abstract but curiously familiar digital forms that somehow, despite their silicon make-up, appear to accord with biological convention.
There has been some similar research in the musical domain, but this has focused on a very small selection of specific models and techniques which have been predominantly applied as compositional tools rather than for use in live generative music. This thesis considers the advantages of the Alife approach for contemporary computer musicians and offers specific examples of simple adaptive systems as components for both compositional and performance tools.
These models have been implemented in a range of generative and interactive works which are described here. These include generative sound installations, interactive installations and a performance system for collaborative man-machine improvisation. Public response at exhibitions and concerts suggests that the approach taken here holds much promise.

Submitted Oct 2006. University of Sussex. CCNR/ Creative Systems Lab



Full thesis [ pdf ] (40.6MB) [ zip ]
Or individual chapters:
Ch - [ pdf ] Frontmatter (contents etc)
Ch 1. [ pdf ] Introduction
Ch 2. [ pdf ] All Computers are Interactive, but Some are More Interactive Than Others.
Ch 3. [ pdf ] Computer Automated Algorithmic Composition in Research and Practice
Ch 4. [ pdf ] Mimesis, Alife Art and Music
Ch 5. [ pdf ] Studies in Simple Adaptive Dynamical Systems
Ch 6. [ pdf ] Composing Generative Systems
Ch 7. [ pdf ] Ashby's Grandmother's Footsteps (an Interactive Installation)
Ch 8. [ pdf ] Self-Karaoke Machines: Collaborative Man-Machine Improvisations
Ch 9. [ pdf ] Recapitulation and Future Work
Refs [ pdf ] References
Appendices [ pdf ] Appendices


Max patches [ .zip ] (UB externals with examples)


Audio examples discussed in the text:
early homeostat experiments [ .zip ]
cellular automata tests [ .zip ]
neural oscillator experiments [ .zip ]
installation and performance examples [ .zip ]