On transcribing, deliberate and creative practice, and strategies for self-generating musical language

Nov 18

How to practice?

As an improviser I’m often wrestling with ways to maximise both the productivity and creativity of my practice time. A few recent conversations with colleagues and students have come back to this question, and while there’s no shortage of great (and not so great…) resources around, I thought I’d share one strategy that has made my practice more systematic and deliberate while constantly opening up new ideas.

As both a music teacher and student, I’m interested in methods of practice that a) extract and study existing language from the universe of vocabulary of a particular genre, but also b) facilitate the manipulation of that language to i) maximise the possible practical use cases thereof, while ii) fostering the development of an individual ‘voice’ and iii) maximising the general quality, ‘flow’ and productivity of practice time.

In a jazz context this inevitably comes back in one way or another to transcription - about which there seems to be as many opinions as there are solos to transcribe. Personally I’ve transcribed about a handful of full solos, and while I’ve found that process insightful and beneficial for ear training I haven’t found it anywhere near as useful or satisfying as exploring and exploding tiny themes or excerpts of a solo - composing/improvising my own variations to transcribed material. To get the most out of each theme, I’ve found it’s essential to be systematic - that is, the process must involve in essence:

Isolating a ‘seed’ concept (i.e. melodies, voicings, rhythmic units, timbral effects, etc) in their simplest form;
Incrementally increasing the complexity of the seed, to stay in a cognitive ‘Goldilocks zone’ (not too hard, not too easy), by;
Using a methodical approach to the expansion and permutation of the seed, in order to explore a broad spectrum of the seed’s possible use cases.

This is not to say I’m against the idea of more intuitively improvising with and/or around a piece of transcribed language - that’s the ultimate goal of this strategy, but going there first runs the risk of leaving a large amount of potential extrapolations of a piece of language unexplored.

To give an example of how I approach this process, I’ll take a couple of lines from Stan Getz on the classic album with the Oscar Peterson Trio and walk through a few techniques for expanding on this 4 bars of material (thanks to saxophonist Chris O’Dea for pointing this one out).

So good! First of all, let’s break this line into two halves. For the sake of isolating smaller seeds, I’ve left out some of the chromatic material at the start and end of the 2nd half.

Taking the first half as our first seed, I’d argue that it could be broken down further into two main components: 1) the chromatic line into the Eb and closure of the Db in b.2; 2) the Db and then triad, leading the top note of the line down into the Bb.

Systematic sequencing

But for now let’s treat this as a whole unit. Especially in this kind of context, my first port of call is to transpose a seed like this through the common chord scale. Let’s assume we treat the Eb7 as a lydian-dominant (i.e. parent scale = Bb melodic minor), and sequence the line down through that scale. To do this, we’ll break the seed down further into five precise components:

1) chromatic line into top note of the scale;
2) chromatic enclosure of the next diatonic note of the scale down (C-Db in the original);
3) diatonic 6th down;
4) root position diatonic triad up;
5) next scale tone down.

This gives us 6 additional permutations:

As you probably noticed, things really start to get interesting around the 4th permutation, where the logic of the line dictates that it start on a note ‘outside’ the conventional chord-scale.

What makes this process particularly satisfying is that to me, each of these variation lines implies a somewhat different melodic direction for any line that would come afterwards (I’ve put a couple of possible extensions in cue size):

You could also change the original system slightly to think of the top two notes of the ascending triad as being 1) the chromatic note below and scale tone above, or 2) the chromatic note below and major 2nd above the last note of the phrase - each of which gives some different, funky results, most of which won’t entirely conform to the ‘conventional’ chord scale:

Rhythmic displacement

Next, let’s combine this exercise with some sort of rhythmic variation. First, let’s just shift the line back a beat with each permutation. I like this process, as the flipped polarity of every 2nd phrase (i.e. starting on beat 4 instead of 3, etc) again changes the character of the line and suggests different lines to follow (and to me, is reminiscent of the types of approaches Lennie Tristano pioneered and taught).

‘Compositional’ variations

Again, each of these imply a very different follow up line. While there’s a wealth of practice material in these processes alone, we’re still taking a relatively conventional approach. So let’s stretch the original idea a bit further. What if:

The whole orientation of the line was inverted?
The individual structural components were intervallically expanded or contracted?
The line was fragmented rhythmically?

As a practice method, each of these four variations could then be sequenced through the chord scale to result in further permutations. Which means already, we have about 45 distinct permutations of the original two-bar phrase - even without considering the variations on the follow-up line that each permutation suggests.

To go a step further rhythmically, we could even displace the original line by an 1/8th note instead of a 1/4, altering the line where necessary - which gives a whole other four rhythmic permutations of the original structure:

Or, we could stretch the melodic components even further to get more of an ‘outside’ effect. Here, the original top note of the line (i.e. first note after the triplet) is approached by [semitone, maj 2nd, semitone], with a few following chromatic alterations (i.e. a falling minor 6th instead of a diatonic 6th).

Harmonic context

But we can still go further, by also considering the various harmonic contexts in which this line, or any variations we’ve made thus far may be useful. For example, our original line will work in any of the related modalities of the Bb melodic minor scale.

Finally, we could also make some minor adjustments to fit the line to contexts it wouldn’t otherwise quite work in if we stuck strictly to the original intervallic logic. For example, where the original seed had a basis in the melodic minor scale, the below examples utilise the major scale (Fig. 12 - B major) and the diminished scale (Fig. 13 - G half/whole) respectively. Once again, we could use these lines over any chord associated with these two parent scales.

Some permutations might be less effective or interesting than others, but again, the point is the systematic approach to the exercise. Most importantly, by methodically transposing and transforming a seed to fit a wide variety of contexts, we’re also implicitly practicing the skill of being able to spontaneously adjust and adapt existing vocabulary to find novel ideas while blowing; the point is not to treat all of these lines as ‘licks’ to be regurgitated, but as structural devices, or grammars that contribute to informed and exploratory improvising. Better yet, it suggests ways that material taken from a solo on a jazz standard might be adapted to fit other stylistic contexts (say, non-harmonic or ‘free’ improvisation).

Macro-strategy considerations

On a macro level, one of the obvious challenges with an approach like this is not to be intimidated by the sheer number of possible extrapolations of a single two bar seed (we’re well over a hundred by Fig. 13!) - let alone any other seeds extracted from a solo (we didn’t even get to the second half of Stan Getz’s original line). It’s just not feasible to cover all of them - and again, that’s not the point of the exercise. The point is working on what we’re working on as we do the exercise - the ability to repurpose language in real time for new improvisatory outcomes.

If the expansion of melodic vocabulary in a jazz context is a creative goal (although the same principals could be applied to other idioms), then practically I’d recommend allocating a specified period of a session to working on seed sets in this way - say 10-20 minutes of an hour. At a minimum, I’ve found it’s useful to do at least the transposition of the seed through the mode or harmonic context, changing the difficulty or adding further elements and more complexity depending on one’s existing skill level and the goal of the practice session. For example. my goal might be to get equally comfortable starting lines on all of the 8th notes of a bar of 4/4, in which case I’m more likely to focus the session on that, and potentially also the modal transposition technique simultaneously (practicing the two together adds another layer of complexity and productivity).

Structured improvisation

Once I’ve run the seed through the largely-systematic part of the strategy, I’ll generally then add some more open-endedness to the process by:

Improvising follow up lines to each transposition of the seed (say, over a four bar section of one chord)
Nestling the seed at some point inside a line over four bars, rather than starting each line with the seed
Playing over a tune, freely adapting the seed to different contexts but always keeping it as the primary theme.

Again, in each case I’m careful to set specific constraints on the exercise, at the same time as focussing on tone, time feel, articulations, accuracy, etc. Each of 1, 2 and 3 above must be in time at a tempo at which I can maintain a grasp on all these elements and focus on the precision of the execution and the eloquence of the line.

Ultimately how much time is allocated to this kind of work is dependent again on the priorities and structure of one’s practice - but that’s another post.

Share the love on your social media of choice if you found this useful, and/or leave a comment below with any other ideas about practice and transcription!

🤓🎹🍻