Can a Genetic Algorithm Think Like a Composer?

Aus de_evolutionary_art_org
Wechseln zu: Navigation, Suche


Andrew Gartland Jones: Can a Genetic Algorithm Think Like a Composer? In: Generative Art 2002.



There has now been a substantial body of work utilising Genetic Algorithms (GA) for the purpose of musical composition. A common point of discussion is how far GA’s can simulate not just the musical output of human composers, but also the process of composing itself. This paper begins by discussing the suitability of using a GA for composition, and goes on to describe a generative music system (by the author), that utilises a domain specific, knowledge rich GA. The system acts on a supplied 2-bar musical phrase (up to 4 parts), and evolves musical fragments towards a supplied target. The aim is to provide interim points on the evolutionary path, which represents a ‘new’ musical ideas audibly based on the supplied fragments.

The paper concludes that the system is able to model at least part of the creative process of composition, and is effective at producing musically successful results. (Audio download sources of its output are included to support this conclusion).

The system was used to generate music included in an interactive installation work, exhibited at Brighton Arts Festival 2002, and other applications under developed that use the algorithm are discussed.

Extended Abstract


Used References

1. Hiller, L, & Issacson, L. Musical Composition with a High-Speed Digital Computer, 1958, Journal of the Audio Engineering Society.

2. Miranda, E,R. On the Origins and Evolution of Music in Virtual Worlds, in Creative Evolutionary Systems, ed. Bentley, P, Corne, D, Academic Press (2002).

3. Jacob, B. L. Algorithmic Composition as a Model of Creativity, in Organised Sound, vol. 1, no. 3, pp. 157-165. Cambridge University Press. December 1996.

4. Cohen, H. A self-Defining Game for One Player: On the Nature of Creativity and the Possibility of Creative Computer Programs. Leonardo, Vol. 35, No1, pp. 59-64, (2002).

5. Werner, G.M., & Todd, P.M. (1998) Frankensteinian Methods for Evolutionary Music Composition, Musical Networks: Parallel distributed perception and performance. Cambridge, MA: MIT Press/Bradford Books.

6. Wiggins, G. Papadopoulos, G. Phon-Amnuaisuk, S, Tuson, A. Evolutionary Methods for Musical Composition. Proceedings of the CASYS98 Workshop on Anticipation, Music & Cognition, Li`ege, (1998).


Full Text

intern file

Sonstige Links