The Challenge of Data in Digital Musicology

Most of our work in the humanities is increasingly driven by digital technology. Musicology is no exception and the field is undergoing the same revolution as all disciplines in the humanities. There are at least two key areas in which digital technology is transforming research: access and scale. Technology, and the internet in particular, has radically changed how we can access data, but also how we can make research results accessible to others. Correlatively, the scope of projects can be broadened to a completely new extent. What does this mean for musicology? Scholars in musicology base their work on a wide range of materials. Since most of the music that forms our heritage in Western culture has been preserved in a text-based form, this is by far, the most widely used type of material for musicological studies. Handwritten and printed sources constitute the core data, but historical studies also rely on various types of textual and archival material, be they letter writings, libretti, or inventories of diverse kind. These are essential for understanding the socio-economic context in which the music sources were written or produced and for better understanding of specific aspects, such as performance practice of the time. Performance practice study itself may also be based on sound recordings when focusing on relatively recent history, as it is often the case for studies in ethno-musicology or in folk-songs (Cook, 2010). Obtaining access to the sources has always been a struggle for musicologists. Only a few years ago, studying a particular source meant first locating the relevant sources using printed bibliographies, writing to the holding library, and then waiting for a microfilm to be prepared and sent out. The process could take months and be unpredictably expensive, with no guarantee of success. Such an obstacle seriously reduced the breadth of research musicologists could reasonably envisage, with a consequent inclination toward close-reading approaches on a restricted set of sources. With the coming of the digital world, the situation changed. Many resources are now available online, including the bibliographic finding aids, which makes locating sources significantly easier. Collections are being digitized and made accessible online, which greatly facilitates access to them for musicologists. This is also the case for secondary sources. Some projects are composer-specific, such as the Digital Archive of the Beethoven-Haus, others are repertoire-oriented, such as the digital image archive for medieval manuscripts (DIAMM) or based on a particular library collection, such as the Julliard Manuscript Collection, to cite only three examples. In the archives, digital cameras are often allowed and can be used to capture sources quickly. It is now straightforward for scholars to store thousands of images on their personal computer, in the cloud, or even share them on community websites, although this in its turn raises new copyright concerns.What other issues need to be addressed? Digital access in musicology is still overwhelmingly linked to images. Several important digital musicology research projects, such as the OCVE and the Edirom projects, focusing mostly on philological issues have been very successful in relying extensively on digital image resources (Bradley and Vetch, 2007; Bohl et al., 2011). However, digital musicology projects that address a wide range of other issues, such as music analysis or music searching, require access to the music itself in digital form, are referred to as content-based resources. Musicology has never been behind other disciplines for experimenting with computational approaches in these domains, quite on the