"Or I Could Leave It Hang-innnng..."

"I could go up? Or I could go down! Or I could leave it hang-innnng..."

I thought of this last week while I was recording some lessons about improvising on the spiritual "Motherless Child." For almost as long as I've been playing, I've been thinking about how to practice improvisation. Where's the balancing point between learning specific notes to play, and figuring out what to do with those notes yourself? Learning to improvise exists along a continuum. On one end, you've got the primary-source approach: listen to and transcribe solos, analyze them yourself, and then do various things with that information –  spit it back whole, harvest various individual licks from it, or go deep enough to draw your own conclusions about phrasing, vocabulary and syntax. On the other end, you've got the raw materials approach: we have these chords, we have these notes (organized into scales) that fit over them – what can we make?

I'm a fan of the primary source approach, but if you stop short of truly absorbing and drawing conclusions from what you've learned, only using the solo or arrangement you've captured as a script rather than a resource, you end up with a kind of classical music, albeit a roots-based one, rather than an improvisatory one. I do love me some Vivaldi, but that's not why I pick up my OO-18 to do some fingerpicking. I'm also a fan of the raw materials school, but without a parallel pursuit of primary sources – listening to, transcribing and absorbing the music of foundational artists in a given genre –  it can quickly comes to resemble doing a jigsaw puzzle without looking at the picture on the box. Super imaginative and fun, but it may not come out looking like the thing you were going for.

So when I teach improvisation, I try to provide specific examples, but keep them short. Rather than present a whole solo, I'll say "Here's a lick that takes you from IV to I. See how it resolves to the downbeat? See how the accented notes form a slowly rising chromatic thread through the chord progression?" And then, "Here's one more. Notice how this time we're starting after the downbeat? How we keep coming back to this one common tone on top?" And finally, "So that's the idea. See if you can create a couple more of these, using what we've just looked at as a model. Keep the chord tones of whatever chord you're on in mind, use the same phrasing, but choose a different direction and landing point for your lick."

I've run into more than one songwriting teacher who starts beginners off writing parodies. Everyone knows – or thinks they know – what a country song sounds like, the strategy goes. So you can use that form as a model, and learn a lot about rhyme scheme, phrasing, chord progressions and verse-chorus form in the process of trying to write your own barroom tearjerker. Along the way, you might discover how hard a good honky-tonk song is to write after all, and – bonus – acquire new respect for the Merle Haggards and Buck Owenses of the world.

This is the same idea: yeah, we're trying to learn how to put notes over chords (the raw materials approach), but few things exist in a stylistic vacuum. So if we're trying to solo on the blues, let's look at how some specific blues musicians solo (the primary source approach), and learn what the stylistic containers are for that genre. Along the way, we might acquire a new respect for the humble pentatonic scale, or for people who are "just playing blues licks."

So back to my friend the dialect coach. In this lesson I was recently recording, I found myself taking one specific lick and saying, "Ok, look. The original version of this thing resolves to the root. But check it out! We could resolve it to the 5th, instead. Or we could play the whole thing up an octave. In which case we could also resolve that up on the fifth, back down to the root, or somewhere in between." And suddenly, from one lick, we have four possibilities.

You dig? "I could go up? I could go down! Or I could leave it hang-innnng..."