Winner of a 2012 “Digging Into Data Challenge” grant, the Electronic Locator of Vertical Interval Successions project at McGill’s Schulich School of Music has only been active since January 2012, yet it has already made waves in the musical and “big data” worlds. Lead by Dr. Julie Cumming, the ELVIS research project includes six members McGill, plus others from MIT, Yale and the University of Aberdeen and has successfully built a database of over 5000 searchable scores in symbolic notation, with a team of nine students handling data collection and software development. With over 600 years of polyphonic Western music being considered, the project represents a significant step forward in data analysis for musicologists and music theorists.
While the searching of music is not new, ELVIS’ focus on the analysis of contrapuntal patterns within the scores is a first. The project takes its inspiration from the field of linguistics and the concept of mining large amounts of text to show choice and frequency of words and groups of words. ELVIS carries this though into the realm of interval patterns in musical scores, by tracking and analyzing both vertical intervals (two notes in different voices sounding at the same time) and the melodic motions connecting them.
To facilitate data analysis the McGill team has developed the VIS tool (Vertical Interval Successions) using as its base a programme called music21, developed by scholars at MIT. The VIS tool tabulates the occurrence of contrapuntal structures of different lenghts throughout single or multiple scores. The usage of these contrapuntal structures can then be graphed and compared to show trends and changes in musical style over time. As a result, the ELVIS project has created both a repository for digital material and an extremely powerful research tool.