Self-organisation, as manifest, for example, by swarms, flocks, herds and other collectives, is a powerful natural force, capable of generating large and sustained structures. Yet the individuals who participate in these social groups may not even be aware of the structures that they are creating. Almost certainly, these structures emerge through the application of simple, local interactions. Improvised music is an uncertain activity, characterised by a lack of top-down organisation and busy, local activity between improvisers. Emerging structures may only be perceivable at a (temporal) distance. The development of higher-level musical structure arises from interactions at lower levels, and we propose here that the self-organisation of social animals provides a very suggestive analogy. This paper builds a model of interactivity based on stigmergy, the process by which social insects communicate indirectly by environment modification. The improvisational element of our model arises from the dynamics of a particle swarm. A process called interpretation extracts musical parameters from the aural environment, and uses these parameters to place attractors in the environment of the swarm, after which stigmergy can take place. The particle positions are reinterpreted as parameterised audio events. This paper describes this model and two applications, Swarm Music and Swarm Granulator.

This paper introduces the concept of variable-coupled iterated map networks and explores its application to generation of musical textures. Such networks consist of one or more interlinked nodes. Each node consists of an iterated map function with a time-delay factor that schedules successive iterations. The value broadcast by a node can drive the variables and time-delay factor of any other nodes in the network, including itself. Lehmer's Linear Congruence Formula, an iterated map normally used for production of pseudo-random numbers, is explored for its own potential as a pattern generator and is used as the iterated map in the nodes in the examples presented. The capacity of these networks to produce richly gestural behaviours and mid-term modulation of behaviour is demonstrated.

Pattern theory provides a set of principles for constructing generative models of the information contained in natural signals, such as images or sound. Consequently, it also represents a useful language within which to develop generative systems of art. A pattern theory inspired framework and set of algorithms for interactive computer music composition are presented in the form of a self-organising hidden Markov model – a modular, graphical approach to the representation and spontaneous organisation of sound events in time and in parameter space. The result constitutes a system for composing stochastic music which incorporates creative and structural ideas such as uncertainty, variability, hierarchy and complexity, and which bears a strong relationship to realistic models of statistical physics or perceptual systems. The pattern theory approach to composition provides an elegant set of organisational principles for the production of sound by computer. Further, its machine learning underpinnings suggest many interesting applications to emergent tasks concerning the learning and creative modification of musical forms.

This paper focuses on the infrastructure and aesthetic approach used in PIWeCS: a Public Space Interactive Web-based Composition System. The concern was to increase the sense of dialogue between human and machine agency in an interactive work by adapting Paine's (2002) notion of a conversational model of interaction as a ‘complex system’. The machine implementation of PIWeCS is achieved through integrating intelligent agent programming with MAX/MSP. Human input is through a web infrastructure. The conversation is initiated and continued by participants through arrangements and composition based on short performed samples of traditional New Zealand Maori instruments. The system allows the extension of a composition through the electroacoustic manipulation of the source material.

Atmospherics/Weather Works is an interdisciplinary project in the sonification of storms and other meteorological events generated directly from atmospheric data produced by a highly detailed and physically accurate simulation of weather systems. This paper discusses the background, conception and execution of the first stage of the project that has resulted in several performances, stereo recordings, a multi-channel spatialised sound installation, and an interactive sound environment.

Absolute (or perfect) pitch exists in fewer than 1/10,000 of the adult population and many claim that it cannot be taught. On the other hand, research suggests that the mechanisms for absolute pitch exist in us all but access is inhibited during early maturation. We here argue that this inhibition can be switched off by artificially turning off part of the brain, allowing everyone access to absolute pitch. This possibility has profound implications for understanding the strategies adopted by the complex networks of the mind. We describe agent-based modelling techniques to understand the computational rationale for these inhibitory processes and to develop new artificial music recognition and synthesis techniques.

Various algorithmic techniques are available for generating music, many of which come from the field of artificial intelligence, which is rich with potential in this regard. However, the musical appropriateness of these techniques is less clearly understood. In this paper, I will report on a study that aimed to describe the characteristics of two of these techniques, rule-based and genetic algorithms, as they apply to melody generation. The appropriateness of these characteristics in contributing to well-formed melodies was judged by aesthetic criteria. The results indicate that most combinations of rules, mutations and evolutionary selection result in poor or average melodies, but that careful combination of these techniques can generate melodies that are not simply well-formed but in many cases display some elegance and novelty.

The field of electroacoustic improvisation in the Netherlands is largely male dominated. Flutist, improviser and composer Anne La Berge signalled this and other problems related to gender in ‘Kraakgeluiden’, a Dutch venue for improvisation in electroacoustic music. Four possible factors contributing to gender discrimination are proposed: the stereotypical gendered concept of an improviser, differences in communicative styles, the dominance of male networks, and music technology itself. This is followed by a consideration of gendered differences in music improvisation. It is posited that male improvisers often seek to establish a hierarchy among themselves, while women performers tend to project themselves less to the forefront. Women improvisers may traditionally have a different sense of autonomy, in adherence to gender norms. While music education might offer a means to change behaviour patterns, it appears that these gender norms are internalised at a surprisingly young age. While concentrating on electroacoustic improvisation, this essay also includes relevant experiences of other improvisers. Gender discrimination in electroacoustic music improvisation is a complex issue deserving more research.

Microsound is an emerging approach to music composition and analysis which places emphasis on extremely brief time-scales, usually a tenth of a second or less, as well as an integration of this micro-time level with the time-levels of sound gestures, sections, movements and whole pieces. This paper summarises some of the technical issues involved in microsonic analysis/composition, traces a history of microsonic techniques in contemporary music, and examines some of its aesthetic implications in a social context.