Popular Music Subgroup Update

Project Blog | Project Updates | Nicole Biamonte | November 22, 2021

Popular Music Subgroup Update

The Popular Music subgroup of the Diversity Working Group is currently conducting a pilot study of timbre in popular song and its interactions with musical form and texture. We are encoding and analyzing a corpus of 400 popular songs released 1990-1999, comprising 100 songs each from four disparate genres: pop, hip hop, heavy metal and country. We developed a novel sampling method for song selection that balances genre typicality with considerations of gender and racial diversity as well as chronological representation. This method is designed to counteract systemic biases (such as those in the Rolling Stone corpus, which is heavily weighted toward music from the late 1960s through the 1980s by white, male artists) and better align dataset diversity with real-world diversity. We are drafting an article discussing our corpus, the ethically motivated sampling strategies we used to increase representation of minoritized artists, and what equitable benchmarks for diverse artist representation in popular-music corpora might look like. Future work will expand the corpus to include more genres and decades.

The project is led by Nicole Biamonte and postdocs Lindsey Reymore and Matthew Zeller (all at McGill), working with ACTOR members Leigh VanHandel (UBC) and Ben Duinker (U. Toronto postdoc), nonmembers Nicholas Shea (Arizona State University) and Christopher William White (U. Massachusetts - Amherst), and student members Jade Roth and Jeremy Tatar (McGill).