Statistics in Historical Musicology

Gustar, Andrew James (2014). Statistics in Historical Musicology. PhD thesis The Open University.

DOI: https://doi.org/10.21954/ou.ro.0000a37b

Abstract

Statistical techniques are well established in many historical disciplines and are used extensively in music analysis, music perception, and performance studies. However, statisticians have largely ignored the many music catalogues, databases, dictionaries, encyclopedias, lists and other datasets compiled by institutions and individuals over the last few centuries. Such datasets present fascinating historical snapshots of the musical world, and statistical analysis of them can reveal much about the changing characteristics of the population of musical works and their composers, and about the datasets and their compilers. In this thesis, statistical methodologies have been applied to several case studies covering, among other things, music publishing and recording, composers’ migration patterns, nineteenth-century biographical dictionaries, and trends in key and time signatures. These case studies illustrate the insights to be gained from quantitative techniques; the statistical characteristics of the populations of works and composers; the limitations of the predominantly qualitative approach to historical musicology; and some practical and theoretical issues associated with applying statistical techniques to musical datasets. Quantitative methods have much to offer historical musicology, revealing new insights, quantifying and contextualising existing information, providing a measure of the quality of historical sources, revealing the biases inherent in music historiography, and giving a collective voice to the many minor and obscure works and composers that have historically formed the vast majority of musical activity but who have been largely absent from the received history of music.