This project is a practice-based investigation into the creation of instability in composition and instrument design. The research began by exploring concepts of fragility before broadening towards ideas of instability, which I created through different ‘agents’ such as score, instrument, or performer. These agents can be split into two categories: forced instability; and chance instability. When instability is forced, a failure of sound to behave as expected is an unavoidable consequence, for example modifying an instrument so the resulting sound is unpredictable. Chance instability relies on failures of sound which are hidden within the notation, instrument capability, or performance situation. An example of this is the music of Jakob Ullmann, which is so quiet it has the potential to become inaudible (Thurley, 2015, p. 5). In this case, instability is not forced upon the performer, but there is a chance that it may emerge. Through my research, I have found forced instability to be the most consistent outcome for creating music which contains contingencies and instabilities, as it provides a performance situation where instability is an expectation rather than a consequence. This led me to consider ideas of instrument modification, exploring Scott Mclaughlin’s research on ‘material indeterminacy’ as a starting point (2022, p. 153).