Improvisation

The usual early start this morning and, with coffee in hand, climbed up on the old stool and took some J.S.Bach models for improvisation. One is below, as I've been messing about with Baroque dance suites recently, noting the different characteristics of Allemande, Gigue, Courante etc. Actually with Gigues there's more variety in meter than I thought (see also below) and that's the best thing about referring to a model (original piece): they depart in significant ways from my 'ideas' concerning them. Sometimes, it could be a simple question of meter - I imagine Gigues to be always in quick 6/8, for example:

whereas they can be in a variety of tempos, these sometimes dictated by the complexity of the figuration and ornamentation:

In the book of Partitas I'm now looking at, there are two more types, all somehow instantly recognisable as Gigues:

One, a kind of fast, spinning motion:

The other, also in 12/8, but with a more hefty feel:

All these different rhythmic motifs serve as useful templates for gigue improvisation, and while it's not always comfortable to use a model - the complexity and seeming perfection of J.S.Bach's writing can be a little overwhelming! - I find the variety of ideas pushes me in good directions.

Sometimes, a variety of approaches to analysis of models is helpful and insightful. For example, I was recently very taken with an ouverture of J.S.Bach (from the 1st Orchestral Suite, BWV1066), possibly due to the excellent performance of Trevor Pinnock et al.(!).. This ouverture is in the French style, but the first section manages to evade expectations while somehow satisfying those expectations at the same time. My first step was to notate the bass line with figures.

What's interesting to me is the variety of rhythmic motif and melodic character in the line. Originally my impression of french ouverture was to be kind of locked in to a dotted rhythm, with everything changing at the same time: a type of metrical straitjacket from which there was no escape or relief! This bass line on the contrary, moves when the melody rests; it alternates flowing semiquavers with dotted rhythms, but the dotted rhythms are unpredictable.. some move through the bar, some move at the beginning and others wait until the end. 

All these observations serve as instructions for improvisation. In my mind they become written like this:

In the L.H.

1. Alternate (mix up) flowing semiquavers with dotted rhythms.
2. Keep the bass stationary until the melody rests, then move.
3. Vary the timing of dotted rhythms: put some on the first two beats of the bar; others on the last two beats.

There are many other observations/instructions to try. Possibly these ones won't work: following one of these instructions might make the music sound fussy or unnatural; possibly an instruction works better as a sub-direction to a different or new instruction (for example, No.2 might work better  in conjunction with a melodic instruction towards the R.H. This would then serve as the refinement of an accompaniment, rather than as a main generative instruction); possibly too, an instruction might work very well for a phrase or two and then needs to be dropped, in order to allow more variety or to let the music develop naturally.

Clearly this is a zone of experimentation. I follow my curiosity, my appetite to uncover a generative principle or two, the decisions of the composer which produced the things that I like. The techniques gained are ephemeral, yet lasting.

Another analysis of this section proved more insightful, and concerned the harmonic rhythm. This is in a sense more subjective and difficult to analise (the rhythmic dominance of the bass and treble lines make it difficult to correctly perceive the actual moments when the harmony genuinely shifts) and it's possible that a musician would disagree with the following interpretation which I offer. What I found interesting is the independence of harmonic motion from the more predictable formulas of the french ouverture style, as outlined in the bass and treble. I think it is this motion that caused the effect that I originally found so compelling while listening to the ouverture.

Looking now at the points where the harmony changes, it seems to me that more variety is almost impossible.. So this became another instruction from this analysis, which, while it is not an easy one, is I think quite valuable for effective improvisation.

1. Look to vary the harmonic motion, even within - perhaps particularly within - predictable rhythmic structures.