"The image of rappers and champagne is pretty legendary and dramatic and problematic and funny," says Tahir Hemphill.
Hemphill, an artist and a Harvard University Hiphop Archive Fellow, is talking me through an interactive map he co-created that charts more than 800 mentions to champagne brands in rap songs from the past 35 years. The graphic is a tiny, tangible sample of Hemphill's much more ambitious undertaking: Hip-Hop Word Count, a living, searchable almanac of hip-hop music.
Hip-Hop Word Count has been a pet project of Hemphill's since 2006, when the trained artist and engineer was crafting rap data visualizations by hand. When casual fans started sending in requests for him to analyze particular artists and albums, he figured out a way to automate the process and came up with an early prototype.
To date, the Hip-Hop Word Count database includes more than 50,000 rap songs from 3,000 artists, from 1979 to the present. A built-in search engine lets any user extract highly nuanced information from this ever-growing catalog of rap music.
Take champagne, for example. I can pinpoint the exact year Moët overtook Dom Pérignon in popularity (1993). I can also see a steep dropoff in mentions of Cristal in 2008, two years after rap mogul Jay-Z publicly boycotted the brand.
At one point, conversation turns to one of the ultimate brand battles in hip-hop: Nike or Adidas? To clarify which sneaker reigns supreme in rap, I ask Hemphill if there's a way to examine subdata under broad keywords—for example, could I isolate and study just the mentions to "Jordans" within Nike?
"You can go down the rabbit hole with this stuff," Hemphill says of the database's ability to let you forge connections between seemingly disparate datasets, people, and places.
Hip-Hop Word Count is by no means limited to deep analysis of rap lyrics—it can be applied to any body of text. Much in the way that other rap-meets-tech platform, Rap Genius, has expanded to explicate everything from poetry to legal documents to Bible verses, Hemphill says he can eventually see a potential business in licensing his technology to universities, schools, ad agencies, and marketers. He says he'd also like to open up his APIs to data visualization designers and social scientists.
"Our work is never really done," he says. "It's about collecting and refining knowledge of the living language in hip-hop culture, so there's never going to be a completion date."
[Poster Photo by Joel Arbaje]