Abstract:
An account of instrumental music’s emotional expressiveness has come to be considered a benchmark for theories of expressiveness: because of its abstract character, the attribution of expressive qualities to instrumental music is particularly puzzling. In my dissertation, I contend that a more pluralistic account of musical expressiveness than the ones currently available is necessary. Whereas available accounts normally defend a single phenomenological characterisation of musical expressiveness, I argue that a viable account needs to provide a layered characterisation of the phenomenology of expressive perception. A basic layer of core expressive properties is be characterised by a correspondingly thin phenomenology, whereas a more complex and varied phenomenology corresponds to the historically and cross-culturally variable levels of expressiveness. With regard to the latter aspect, I introduce the concept of a heuristic device. I define heuristic devices as a range of diachronically and synchronically variable strategies that use information from an extra-musical domain in order to create or appreciate music with certain properties. Theories that deny emotional arousal a place in accounting for expressiveness need to determine how music-induced emotions are related to the music’s expressive qualities, as well as whether they contribute to its value as an art form. In order to answer these questions, I develop a general criterion of artistic relevance for music-induced emotions, which is missing from the available literature. I examine a special class of descriptions of music, which I term ‘atmospheric’. Atmospheric descriptions of music are widely found in music criticisms. I argue that currently available theories of musical expressiveness and representation would need substantial qualification if they aim to account for them. Finally, I consider evolutionary explanations of musical expressiveness and the role they may have in philosophical accounts of expressiveness. I contend that an evolutionary explanation that aims at grounding the phenomenology of musical expressiveness should meet five challenges. I argue that the most widely held resemblance theories of musical expressiveness may need such a hypothesis, or at least something that serves a similar role.