Improvisation in music is a concept that can easily mislead. The popular cliché is to contrast jazz and Indian raga with the detailed, fixed prescriptions of WAM; rather, it’s profitable to subsume improvisation under the whole process of musical creation, considering more and less flexible frameworks for performance.

Bruno Nettl has paid much attention to the subject, co-editing two splendid books:

• Bruno Nettl and Melinda Russell (eds.), In the course of performance: studies in the world of musical improvisation (1998)
• Gabriel Solis and Bruno Nettl (eds.), Musical improvisation: art, education, and society (2009).

In Chapter 4 of his masterly The study of ethnomusicology: thirty-three discussions, “Inspiration and perspiration: creative processes”, Nettl unpacks the issues in typically illuminating fashion. Pondering “What is the nature of musical creation?”, he explores the continuums mediating between composition and performance. Here’s a typical passage showing how he treats all human musickings on a equal footing:

Schubert is said to have composed a song while waiting to be served at a restaurant (presumably by a slow waiter), quickly writing it on the back of a menu; Mozart turned out some of his serenades and sonatas almost overnight; and Theodore Last Star, a Blackfoot singer and medicine man, had visions in each of which, in the space of a brief moment, he learned, from a guardian spirit, a new song. But Brahms labored for years on his first symphony, Beethoven planned and sketched ideas for his Ninth for over two decades, and William Shakespear, an Arapaho elder, said that when he took a motif from one song, something else from another, and a phrase from a third, thus making up a Peyote song, it might take him a good part of an afternoon. The xylophonist of a Chopi orchestra made up music as he went along, but he was constrained by rules articulated by his leader. The North Indian sitarist sits down before his audience and creates a performance of new music on the spot, but he can do this only because for hours every day he practices exercises that he has memorised, as he maintains in his mind a musical vocabulary on which he can draw and a group of rules that tell him, once he has selected a raga, what he must, may, or cannot do. A Kentucky mountaineer about 1910 sang “The Two Sisters” in a tavern, his friends admiring a new twist he put into the refrain. And the overjoyed Bach lover after the cello recital exclaims, “She’s never played it like this before. She makes the suite live like no one else.” In some sense, each of these musicians has created music, but music scholars actually know very little about the way in which such music comes about, especially in its innovative aspect, which is what they most admire. They believe, as Blum (2009) explains in detail, that when music is produced (in any sense of the word), something new is being created.

There is innovation—of different sorts—in the composition of a symphony, in the jazz improvisation on a well-known show tune, in the unique rendition of a Japanese chamber work that has been handed down with little change for generations, or in a rendition of a standard string quartet. Ethnomusicologists in particular must deal with what is new, new in a sense generally understood by them but new also within the specific cognitive framework and understanding of its culture.

But what is “newness”? Speaking cross-culturally, what may be heard as new composition in one culture might be regarded as simple variation in another. Judging the degree of innovation is a tricky business. The Persian improviser who by the standards of European composition gives his audience something different each time he performs may not be, in his own manner of musical thought, doing something really new, but simply “playing a particular mode”. By contrast, the Blackfoot singer who learned a song in a vision may have thought of it as a new song, even if objectively it sounded virtually identical to a song that had been received by one of his friends in another vision. The South Indian musician with a penchant for giving her audience unexpectedly strange vocalizations runs the risk of rendering something outside the realms of propriety and being criticized for not knowing her basic material. The American composer who writes a piece inspired by Hindemith or Stravinsky might be criticized for presenting something belonging to the past and thus not properly innovative.

Nettl goes on to suggest three intersecting continuums:

• From spiritual to intellectual: music is inspired (just listen to Aretha’s Amazing Grace), but

it is also the result of the manipulation and rearrangement of the units of a given vocabulary, of hard work and concentration. The concepts of inspiration, of genius, and of acquiring music directly from supernatural sources are very widespread among human societies, simple and complex. Haydn worked regular hours and depended on some kind of inspiration; when it did not come, he prayed for it, rather like the Native American seeking a vision who is also, in effect, praying for songs. At the other end of the line is the concept of composition as an essentially intellectual activity, in which the musician consciously manipulates the materials, or building blocks, of music […]. […] The listener may be unaware of all the care that went into the preparation of this complex structure. But such an approach is not limited to societies with written notation and music theory texts. Native American composers of Peyote songs may be equally careful, using and abiding by general structural principles that govern the song, musically making clear a number of intricate relationships, deriving new phrases from earlier ones, all within a rather rigidly defined formal framework. Yet it seems unlikely that the typical Native American listener understands the details of the structure.

The two ends of the continuum merge: Mozart’s music sounds to many divinely inspired, and we know it was often composed quickly, yet has incredible consistency and great complexity. The songs of the Yahi of California, sung by Ishi, the last “wild” Indian, each ten seconds long and using only some three or four tones, exhibit considerable sophistication in their internal interrelationships, with a logic not totally unlike that of Mozart. An Iranian musician says that his improvised performance comes “from the heart”, but analysis shows us highly structured and sophisticated patterns unique to the performer. A performer of improvised Indian alapana learns a vast repertory of melodic and rhythmic units that can and must be interrelated in many ways, exhibiting her skill in showing the multitude of combinations she can control, yet many in her culture regard this music as essentially spiritual. Each case confronts us with aspects of both ends of the continuum, obviously in different proportions.

• From improvisation to composition—two versions of the same process. Again, Nettl compares Schubert’s rapid composition of a sonata with the assembly of an Indian improvisation: “the fact that Schubert used paper and pen might actually be considered incidental”. He suggests that even the gigantic labour of composing a symphony is cognate with the technique of a Yahi Indian composer, and that Horowitz’s multiple renditions of Beethoven may be compared with an Arabic musician performing the maqam.
• From precomposition to composition to revision: another process common worldwide, played out over years or minutes.

Nettl then stresses the false dichotomy between the composing of “art” and “folk” musics:

There seems no reason to regard composition in cultures with oral and written traditions as different species. […] In each culture, the musician is “given” something and then has the job of adding something else, but there are many different kinds of “given” and “added”.

Discussing the balance between the two, he concedes that there are many societies in which innovation is restricted, adducing Anglo-American folk song and South Indian kriti.

What is “given” to the creator of music are the building blocks and the rules of what may be done with them; innovation consists of how the options are exercised.

Finally Nettl homes in on “improvisation”—whose definition as “the creation of music in the course of performance” already looks dubious. As always, he gives useful leads to the whole history of research on such topics. He praises Paul Berliner’s Thinking in jazz, the classic analysis of the diverse elements on which jazz performers draw.

As improvisation received more attention, ethnomusicologists also began to see it as complex syndrome of behaviours, and the distinction from traditionally conceived composition began to blur.

Left: Clara Schumann with Robert; right: Dariush Talai.

He notes the role of improvisation in WAM (cf. On “learning the wrong music”), from early performances where “the quality of musicianship is judged by the degree to which the improvised piece sounds as if it were not improvised” to the 19th-century rhapsodies and impromptus, which were “composed and written out but seem to be intended to make the listener think they are improvised, or at least somehow connected with an immediacy and spontaneity of creation”. Organists are perhaps the main perpetuators of the tradition of improvisation in WAM (see here and here).

It’s good to find WAM taking its place where it belongs, within ethnomusicology, or “all the musics of the world”. In the later history of WAM, as increasingly prescriptive notation (a red herring: see also here), and the recording industry, came to limit improvisation, I can’t help feeling that WAM musicians have sacrificed a lot.

Themes in the collection Musical improvisation include Ukrainian funeral laments, jazz, and Persian music—and several authors write on WAM, including Robert Levin, whose renditions of Mozart, notably his improvised cadenzas, are so brilliant. We might now hear much of the romantic piano repertoire as improvisations to which performers gave a fixed form, as Messiaen did later with his Messe de la Pentecôte. An earlier instance might be the astounding harpsichord cadenza of Brandenburg 5.

Indeed, it makes sense to suggest that the kinds of things we call improvisation exhibit such variety—everything from simply adding ornaments to a composition to totally (well, almost totally) “free” improvisation, from oral composition to following precise rules in re-creation—and are practised in so many cultures that improvisation ought to be considered the central form of music creation, with traditional Western-style composition, with pen and ink, as a highly developed subtype.

In his introduction to In the course of performance Nettl also mentions Albert Lord’s 1965 The singer of tales and the study of Gregorian chant. Among topics covered in the volume are Javanese gamelan, African–American girls’ singing games, Italian folk song, and the Preludes of Clara Schumann (for an idea of how she might have improvised, do listen to Hélène Grimaud‘s playing in the Brahms concertos—in particular the slow movements!). See also the Dream songs of aboriginal people in Australia.

All this also suggests ways of understanding the whole range of musicking in China: shawm bands (whose music, misleadingly, often sounds “improvised”), silk-and-bamboo (perhaps akin to Irish music—the heterophony of instrumental ensembles is often akin to the ornamentations and divisions of early WAM); as well as folk song, the qin zither (e.g. here), the parameters of performing Daoist ritual, and so on. And it makes yet another caveat against reification: performance as process.

Reading Nettl is endlessly stimulating.

See also Bernard Lortat-Jacob at 80.