Big Music Data, Musicology, and the Study of Recorded Music: Three Case Studies
Stephen Cottrell · The Musical Quarterly · 2018
Music Information Retrieval (MIR)—a term that embraces a range of interdisciplinary but science-based approaches to gain information about music—remains for many musicologists a rather enigmatic part of the music studies field. Some of the reasons for this are considered below. The more recent arrival of Big Music Data (BMD)—very large-scale collections of data pertaining to music in the form of audio recordings, consumer preferences, bibliographic information, etc.—is an emergent topic that, as David Huron argues, contributes to the transformation of musicology from a data-poor to a data-rich field.1 This article considers the relationships between MIR, BMD, and musicology, particularly in relation to the analysis of recorded music. It starts by providing a brief overview of the study of recorded music in the different parts of the music studies field, before giving some insights into ongoing MIR work and how this supplements and augments musicological work elsewhere. The main part of the article looks at the opportunities afforded by software such as that generated as part of the recently established Digital Music Lab, as well as reflecting on both the technical and the conceptual challenges that BMD provide in relation to musicology. Three case studies are detailed, which evidence the type of musicological insights that can be gained from investigating BMD. The article concludes by introducing Franco Moretti’s concept of “distant reading” as a hermeneutic device to explain the possible benefits of BMD for musicologists.