A methodology for the study and modeling of choral intonation practices
Johanna Devaney · 2006
This paper proposes a methodology for modeling choral intonation practices built on the intersection of computa- tion and theoretical approaches. The computational ap- proach consists of pitch tracking each part in numerous choral recordings and applying a statistical machine learn- ing approach to correlating and modeling the collected data. The theoretical approach combines theories of sen- sory consonance with theories of tonal tension and attrac- tion to construct a theoretical model of both vertical and horizontal tuning tendencies. The goal of the methodology is to develop a generalized theoretical model that has been informed and tested by empirical data. those of non-fretted string ensembles, presents a unique challenge because the human voice is not locked into a single tuning system or temperament. At any given point in a piece a choir's tuning cannot be consistently related to a single reference point; rather a combination of horizontal and vertical musical factors form the reference point for the tuning. To further complicate matters, the weighting of these factors often differs in different musical contexts. This paper proposes a methodology for modeling such practices through the intersection of a computer-generated statistical learning model and commonly received knowl- edge in the field music theory. The computational compo- nent of the methodology is built on data extracted by track- ing microtonal pitch variations in recorded choral perform- ances. The theoretical component draws the various branches of music theory that address harmonic and voice- leading practices, musical forces and expectation, acoustics and psychoacoustics, and tuning, temperament, and intona- tion. The computational approach will provide the model with an objective basis, while the theoretical approach will provide a contextual depth that empirical analysis cannot produce. The strength of this methodology is that ap- proaches inform one another at key points in the develop- ment of the model.