Thursday, July 3, 2014

Are musical genres real and constraining or are they mixing together polymorphically?

Monica Lee (doctoral candidate at U Chicago) and Dan Silver (sociology, U Toronto) have a fascinating "big data" post on musical genres. They examine data from over a million bands on MySpace to determine what kinds of music go together. They start with a thesis Bruce Springsteen presented in a keynote speech at South-by-Southwest in 2012:
There are so many sub–genres and fashions, two–tone, acid rock, alternative dance, alternative metal, alternative rock, art punk, art rock, avant garde metal, black metal, Christian metal, heavy metal, funk metal, bland metal, medieval metal, indie metal, melodic death metal, melodic black metal, metal core…psychedelic rock, punk rock, hip hop, rap rock, rap metal, Nintendo core [he goes on for quite a while]… Just add neo– and post– to everything I said, and mention them all again. Yeah, and rock & roll
That is, genre doesn't matter any more because distinctions are so fluid. In Springsteen's words:
It’s all just what you’re bringing when the lights go down. It’s your teachers, your influences, your personal history; and at the end of the day, it’s the power and purpose of your music that still matters.
So, Lee and Silver ask:
Do we truly live in a “post-authentic” musical world, where norms and conventions around musical creation have weakened and individual creativity is born to run free (pun!)?
Their answer: Not really:
We found little support for his idea that American popular musicians freely roam across genres; they mostly operate according to what seem to be strong conventions about what genres go together. But not everywhere to the same degree. Indeed, if we conjure a picture of the Austin SxSW crowd, it looks a lot like the regression results: college students, radio stations, and the record industry. While America in general may not conform to the Springsteen Hypothesis, in some contexts it comes closer than others, and Austin is probably one of them. Considered in this way we might take Springsteen’s speech less as a general proposition and more as a specific expression of the expectations he and his audience have about the nature of musical creativity, one which is by no means universally shared. In other words, he was preaching to the choir.
Just how they get there – that paragraph is third from the end of their long and rich post – is fascinating.

They find that all that music falls into 17 clusters (e.g. Jammy, Popular, Dark, Rave, Keeping the Beat Alive). So genres are real. Conventions matter. Second, "less popular bands are on average somewhat more unconventional than popular bands, but the difference is not large." Finally, "it appears that college towns have the most scene-crossing while racially diverse metros anchor the main streams of American popular music."

H/t Tyler Cowen.


  1. In the early days of my graduate work I did an experiment that included having people list their top 10 genres of music. We had to abandon that approach due to too much variance in the levels of categorization people used. Some only used broad categories like pop, rock, techno, rap, etc. Whereas others had extensive lists of sub-genres like pop rock, pop punk, alt rock, rap rock, etc. The categorization structures varied too much to compare. Since then I've become fascinated by how these differences in thought organization and subjective perception shape other thinking and behavior, mainly decision making.

    1. Yes, categorization is a tricky business, and important.