We are getting closer to the essential nature of improvisation if we see this account of what it means to play a standard as an expression of the special temporal logic of improvisation. In the last chapter, we saw that the ability to improvise is the result of complex forms of rehearsal and immersive practice (Einübung). I maintained that even though mechanical and routine elements of musical practice do play a role in acquiring an ability to improvise, the path to gaining one’s own musical voice is by no means a straightforward one.

The same point applies to carrying out the ability to improvise in concrete situations: it does not follow from the fact that learning to improvise involves mechanical or routine abilities that improvisation itself is a mechanical or routine practice. Naturally there are cases where improvisation can slide into the routine or mechanical. But improvisations like this are lacking, that is, they are unsuccessful instances of the ability to improvise. In this sense every account of improvisation that equates it with playing certain scales or melodic combinations or rhythmical figures over certain harmonic combinations is simply false. It is true that scales or certain set melodic phrases – so called licks – can indeed play a role in the ability to improvise.

But truly executing on the ability to improvise does not consist in recalling scales or licks. It consists instead in something far more general. It consists in establishing a certain coherence of various musical articulations, a kind of coherence that is not determined in advance through a score or some other blueprint. This mention of coherence should not be taken to mean that improvisation is indebted to some ideal of unity. This should not be taken to mean that unity is an explicit goal of the improviser, and that improvisations cannot be fragmentary, leaping or broken (zerissen). Rather, what this should be taken to mean is that each individual articulation of a musical idea get its meaning within the context of other ones.

Let us take the playing of a standard as an example and imagine that the musicians in a group each consecutively play solos on its form. One soloist might begin with a small melodic idea that matches the rhythm of how the melody of the standard is shaped. He potentially establishes a musical idea by playing this phrase once. In doing so, he might for example maintain the rhythm, but displace it over a progressing harmony. Or he might let it begin in a different time. Or he plays the exact same notes, but this time in a different rhythm. In doing so he is not committed to following the harmonic form of the standard. For example, he could decide to play the melodic idea in a key that does not at all seem to fit it, but that could serve as a way of creating tension or dissolving the normal sense of the standard. This is especially true with respect to the dominants and dominant sept harmonies in various harmonies – and it makes up a crucial means of creating and resolving tension within inside out modal improvising.

With this example in mind, it becomes clear that there is no certain blueprint that an improvisation can fall back on in order to guarantee in advance that it will be coherently structured. In light of this we can say: whatever an improviser does will only acquire its specific meaning in light of what is done later. To put this into a game metaphor: the meaning of every move changes in the light of later moves. Each move gets its meaning in the light of all future moves. Of course, thinking of it in terms of chess moves is problematic in the sense that with improvising there are no fixed rules that establish in advance what is even a possible move and what is not. In improvising, one cannot make an error as one might do in chess by making a pawn move like a bishop. In chess, if one moves a pawn like a bishop, then one is no longer playing chess; but if one transgresses the harmonic structure of a standard, then one has not necessarily stopped playing the standard. This has to do with the temporality that is specific to improvisation, as was illustrated in the opening example of an improvisation: the temporality of improvisation is both retrospective and retroactive in a specific way.

In philosophical terms, one can say that improvisation has a retroactive logic. The beginning move – for example, playing a melodic phrase that takes up the rhythm of a melody but uses different notes – does not establish anything for certain. Paradoxically, it is what the improviser does next that sets down what was done before. It is only with a view to the latter musical articulations that it will become clear what the meaning of the prior musical articulations was. If the improviser repeats the initial musical phrase in a different rhythm, then it will prove to mean something different than if he repeats it with the same rhythm but displaced to a whole octave lower or higher. Jazz performances can gain a very special kind of intensity, not in the least, from the fact that the whole meaning of the improvisation stands at each moment at the disposal of the performers.

What has been said up to now is still not sufficient for the following reason: when composed musical works are played it is also generally the case that the individual melodic and harmonic elements get their meaning through what happens throughout the further development of the melody and harmony.

To put it simply, it also happens with the performance of works that the individual elements that make them up get constituted through contrast to other elements. But there is an essential difference: in an improvisation it gets restlessly renegotiated over the course of a performance what an element is, whereas in giving a rendition of a work, it does not get restlessly renegotiated in a performance what an element is.

In the latter case, it gets established through the way that musicians use their training to work through the musical score. In playing a composed work, a musician can consider in advance how to play a certain passage. In the case of improvisation, it may well also be the case that a musician has already played the standard being performed countless times in the past, but there is less that follows from this than in the case of a rendition of a work.

This is an edited extract from Philosophy Of Jazz by Daniel Martin Feige (Lexington Books).

You can read Andy Hamilton’s review of the book in The Wire 497. The print magazine is sold out in our online shop, but new and current subscribers can read the review and the entire issue online via the digital library.