In honor of International Jazz Day we have a guest posting from one of our longest-tenured music analysts who specializes in the jazz genre. Jeffrey Burr is not only a Pandora music analyst, he is also a guitarist who’s played with with Norah Jones, Dizzy Gillespie, Ari Hoenig and has released several albums.
However you choose to celebrate International Jazz Day, we hope this posting sheds a little light on the genre from the perspective of a Music Genome Project specialist.
In an ancient age lost to the mists of time, the Music Genome Supermen forged the Jazz Genome, the framework that we music analysts use to characterize songs in the Jazz oeuvre. I’d like to talk a bit about what makes jazz special in the context of the music analysis work I do for Pandora.
One of the major performance conventions in most of mainstream jazz is for the players to structure their performance as two statements of the melody (the “Head”), one at the beginning and one at the end, with a longish episode of improvisations in between. For the purposes of our music analysis, we call this form “Head/Solo/Head,” and we have a gene, that we use to characterize how closely a song adheres to this formula. A song might have additional sections, an omission of the head, or other structural trickery, in which case we might score this gene somewhat towards the “through-composed” end of the scale.
So, what’s going on in that long episode of improvisations? Most often, the players are looping the underlying harmony of the song and improvising fresh, exciting melodies over the top of it. To do it well, a player typically draws upon a knowledge of harmonic theory, a familiarity with what older players have already done, and a very personal idea of what sounds cool.
To me, improvisation is the very marrow of what jazz is. Improvisation is the indispensible element of a jazz song, whether it’s just some tasty ad-lib statements behind a vocal feature, or every part of an entire performance. Our section of musical genes dealing with improvisation helps our music-matching algorithm connect similar sorts of improvisers for Pandora playlists. These genes relate to several important dimensions of improvising.
One of the most important dimensions of improvisation we cover is the “Inside-to-Outside” gene, which characterizes whether or not an improviser is staying very close to the underlying harmony of the composition. Some masters, like Miles Davis, can construct interesting lines out of little else besides the basic chord/scales of a song. However, a major thread in the historical development of mainstream jazz improvisation had to do with players implying alternative harmonic progressions, superimposed over the usual harmony of a song. It’s commonly referred to as playing “outside” or just “out,” and a lovely feeling of mystery, and of risk, is generated this way. As with all of the Music Genome Project genes, there are varying degrees of “outness” covering a very wide gamut of possibilities, from “occasional use of chromatic neighbor tones” all the way to “we don’t need no stinkin’ scales.”
We have additional genes related to improvised solos, these include note density, emotional intensity, and clarity of articulation.
By now the jazz cognoscenti have surely abandoned me for simplifying too much truth out of the matter with this post, but I’m hoping that just a bit of background information might help the uninitiated to better appreciate this fine form of music.