A frequent phenomenon in music is when two metric layers conflict, but neither layer is dominant. While this conflict often involves, in Western music, interactions between "notated" metric layers, indicated by time signatures and bars, and "heard" metric layers, given by cues in the sounding music, such conflict may appear in improvisatory and/or non-written contexts as well, as in many non-Western musics. The primary purpose of this dissertation is to formalize an analytical system general enough to encompass all of these instances of metric conflict, but specific enough to generate precise and interesting analytical statements.;To achieve this balance of generality and precision, this dissertation first constructs a new analytical system based on Lewin's "generalized interval system" (GIS) concept, introducing a new set of metric GISs along with specialized techniques that allow the analyst to organize, compare, and relate the available GIS intervals. These techniques include intervallic decomposition, expansion and contraction of intervals to various levels of pulse, and the formation of equivalence classes of expansion- and contraction-related intervals. The remainder of the dissertation applies this new system to analytical and theoretical studies drawn from nineteenth-century and South Indian Carnatic music that motivate important additions to the overall system, as well as providing significant metric insights into the pieces being studied. In sum, this dissertation provides a precise, unified approach to metric conflict that suggests substantial new possibilities for metric analysis.
展开▼
