Abstract
We describe our development of a rule-based interactive music performance system (named LOERIC) aimed at performing Irish Traditional Dance Music (ITM) live with a human musician. We followed a practice-led and human-centred approach, drawing on expert knowledge elicited from a three-day workshop at the Irish World Academy of Music and Dance at the University of Limerick, during which thirteen domain experts in ITM performed with our system. We triangulate observations of their interactions, interviews with them afterwards, and our own reflections as system "wranglers" to provide a picture of how LOERIC was gradually configured to perform in stylistically appropriate ways through hands-on practice. We generalise three wider contributions from these findings: the importance of supporting tune-, instrument-, regional- and personal-styles and how we achieved this; how LOERIC's improvisations around traditional tunes walked a line between successfully following the rules of ITM in order to "prove its chops" while being able to break them in order to offer something new; and how its relatively simple engineering enabled real-time interaction with the musicians, though with some limitations.
Keywords
human-AI interaction, music performance modelling, practice-led research, Irish traditional music
How to Cite
Amerotti, M., Benford, S., Sturm, B., Tennent, P., Martinez Avila, J. & Keegan, N., (2026) “Reflecting on an Interactive Performance System for Irish Traditional Dance Music”, Journal of Creative Music Systems 10(1). doi: https://doi.org/10.5920/jcms.1771
50
Views
26
Downloads