Teaching rule-based algorithmic composition: the PWGL library cluster rules

This session reports on an approach to research - informed learning (research - based learning, according to Jenkins et al. (2007)) in the field of Music Technology. In the unit Algorithmic Composition, students learn how to create computer programs that assist the music composition process (using an easy to learn visual programming system). They then use their programs to compose music with them. Our students typically want to compose in a mainstream musical idiom, e.g., virtually all students aim for tonal music, and most often they want a clear rhythmic structure. Constraint programming is a proven approach to successfully model complex mus ic theories like harmony. I recently developed a software library that greatly simplifies the constraint - based modelling of tonal and metric music. More specifically, this new library (called Cluster Rules) provides a collection of predefined musical rules (constraints) for the new music constraint system Cluster Engine by örjan Sandred (University of Manitoba, Canada). The collection includes various rules on rhythm, melody, harmony and counterpoint. These predefined rules offer a low floor (students easil y get started), but also allow for a high ceiling (highly complex music theories can be modelled freely, by defining further rules for Cluster Engine from scratch). This session will demonstrate the new software, motivate its design, discuss how students u sed this software to generate musical material for their compositions, and it will report on challenges met in that process.