Composition and improvisation are two terms and two notions historically developed in the context of the occidental script culture. Their application to cultures of oral tradition, the main field of investigation of ethnomusicology, can be helpful for the understanding of music- making in illiterate societies, but this application makes it necessary to adapt the two terms in a special way. The scope of this text is to show the needs and possibilities of such an adaptation and the helpfulness of “composition” and “improvisation” in ethnomusicological perspectives.

     1. “Composition” and “improvisation” in the context of the occidental script culture

        In the context of the occidental script culture, “composition” and “improvisation” are generally considered as forming an antinomy. It is a common opinion that a piece of music has to be notated, written down, to be called a “composition”. It seems that only in written form, music can acquire the premeditated complexity we normally attribute to a composition and the stability indis-pens-able to permit exact reproduction. And these two aspects: premedit-ated complexity of structure and  possibility of exact reproduction, are supposed to be among the main criteria for a composition. On the other hand, it is also a common opinion that “im-pro-v-isation” is in a large measure the contrary of “composition”: non-written, spontaneous in-stead of premeditated, thus generally of lesser complexity of structure, and not reproducible. In the course of western music history - to mention it but briefly - we can observe that at certain moments the relations of composition and improvisation change. Up to the baroque era, improvisation normally forms an integrant part of the execution of a composition. Bach and Händel - and after them even Mozart, Beethoven and Mendelssohn - were great and famous improvisers, but already Bach was one of the first composers who begun to write down exactly what he imagined, even in the ornamentations. Thus in the 19th century, the main aspect of the player’s and singer’s personal input, of their personal expression in the execution of a composition, was no longer improvisation but “interpretation”: a term and a notion applied to music not before the middle of the 19th century. In the western music of the 20th and 21st  centuries however, it is often difficult to distinguish interpretation from improv-isation.
        In the perspective of the traditions and conventions of western art music, there is a strong tendency to consider all non-written music of illiterate societies as “improvised” music. And the absence of scripture and notation seems to implicate the other characteristics of im-provised music as mentioned above: lack of stability and complex structure, impossibility of exact reproduction. This is without any doubt erroneous. Even if we adopt a more positive and less pejorative definition of improvisation, e.g. as offered today by the representatives of the “Free Improvisation”, the idea that non-written music in illiterate societies is automatically im-provised music is still wrong. That is to say: when we adopt John Baily’s definition: “Im-pro-visation is the intention to create unique musical utterances in the act of performance”(1), we stress too exclusively the momentaneous and unrepeatable character of the musical event and overlook the stability of tradition and transmission that is behind and marks also this event.
        In fact, non-written music in illiterate societies is to a high degree composed music, “composed” in a very special sense indeed, but provided with complex structure and stability and able to be reproduced. Helpful in this respect were for the ethnomusicologist what specialists of medieval music found out: that before and at the beginning of a musical script culture there must have been a long tradition of non-written composition, and that especially in the domain of instrumental music, oral tradition and transmission lasted very long.(2)  To give a few examples: On the one hand, concerning vocal music, the text can be written and the singer applies fixed and trained formulas in relation to the syntax of the text. On the other hand, concerning instrum-ental music, there are special models of tone progression and cadence and formulas in con-nection with special instruments and theit tuning. In both cases we find a framework of for-mulas and conventions assuring a high degree of stability to the musical utterance.

     2. Aspects of composition and improvisation in oral tradition

        It is exactly in this “medieval perspective” that we can adopt the term “composition” to music in oral tradition, and this is valuable in the do-main of european folk music as well as in the domain of non-european music. Here too, we find a lot of formulas and conventions that allow high stability and reproduction, and here too, the possibilities and characteristics of the different instruments are especially impor-tant for this purpose. And in the field of vocal music, we have to do with special compositional stabil-ity in all the cultures where the spoken language is a so called “tone language” with defined signific-ations of registers and intervals that have to be respected also in songs. And of course we find in all music cultures pieces of music with a definite title, thus a clear identity, and often linked with the name of the person who sung or played it for the first time, thus with the name of its composer and the period or even the special moment of its creation. The german ethno-mus-ico-logist Artur Simon is certainly right when he points out that most music in oral tradi-tion is composed music.(3)
        But it must be underlined that composition in music cultures of oral transmission is nevertheless not the same as composition in western art music tradition. The main differences are the following:
    (1) In the western tradition, a composition is generally considered as a more or less ab-so-lute work of art. This has its implications such as originality, authent-icity and fidelity in reproduction. Since all these aspects are in connection with the original score and thus with written transmission, it is easy to understand that in cultures with oral transmis-sion (where this reference is absent), the notions of “fidelity to the original” and “exact reproduction” can not be the same. Reproduction in oral transmission meens recollection. But recollection does never lead to a one-to-one copy of an original. Rosenfeld points out: “Perception, recog-nition, memory are not separate processes, as the computational view implies, but an integral proce-dure. … There are no symbols in the brain; there are patterns of activity that acquire different meanings in different contexts.” And relating to Bartletts book on Remembering (Cambridge 1964) he adds: “Remembering is not the re-excitation of innumerable fixed, life-less and frag-mentary traces. It is an imaginative reconstruction, or construction, built out of the relation of our attitude towards a whole active mass of organized past reactions or experi-ence, and to a little outstanding detail which commonly appears in image or in language form. It is thus hard-ly ever really exact, even in the most rudimentary cases of rote recapitulation, and it is not at all important that it should be so.”(4) This characterization of memory allows us to under-stand why in oral transmission a piece of music is never two times exactly the same, even if the musicians tell us that it is the same. They are not wrong, they just do not have our under-standing of authenticity. For them, fidelity and re-creation are not in opposition.
    (2) Another aspect of the non-existence of a score is that the very moment of reproduc-tion or re-creation of a pre-existing composition is of great importance. The execution of a song must be “true” not only or even less in respect to a previous version than to the circum-stances and needs of the moment(5).  But this “functional” aspect of a musical event in illiterate societies has often been overstressed. Too often the meaning was that the social function of an artistic event is the only aspect to take into consideration The consequence was that the other side: the still existing (relative) stability of the “text”, the “composition”, was completely overlooked.(6)  The same phenomenon occured concerning oral literature and there it has been convincingly criticized by Ruth Finnegan.(7)  The function plays a role not instead of the auto-nomous aesthetic value but in addition to it.
        It is exactly under this aspect that both categories, composition and improvisation, are helpful to understand what is going on in a musical event. Improvisation, in this case, is not the whole but a component. It englobes all the elements that occur in addition to the basic “text” adapting it to the circumstances, to the moment and its needs. Improvisation can thus be, following John Baily’s definition, a result of “the intention to create unique musical utterances in the act of performance”, but improvisation can also, so to say, simply occur out of the momentaneous mood of the singer and player or following the reactions of the public.
        To illustrate different combinations of composition and improvisation in orally transmitted music, we give two examples,  both from the Southern Sahara:
    The first example is the type of Hawsa praisesong we frequently heard on market places in the evening, when a group of musicians was there. Such an ensemble normally consists of three musi-ci-ans: a singer playing a fiddle (goge), a woman singing in response to the goge-player and shaking a rattle (caki), and a speaker (“griot”) who plays the role of the animator. The song is ordered (and payed to the musicians) by a person present on the market place for himself or for another person, may be his friend or his (present or absent) girl-friend. The song is ordered for immediate execution, but “to have his own song”, even if it is executed but once, is a durable mark of distinction for both, the person who ordered the song and the praised person. The music event is a mixture of pre-existing textual and musical formulas (praise to Allah, proverbs) on the one hand, thus rather a “composition”, but on the other hand enriched with many special allusions to the commissioning and to the praised person, to an extent that the improvisational elements prevail.
        The second example is the type of epic song we used to hear during evening meetings in a nomadic Twareg tent. Here the ensemble may consist of a woman playing the twareg fiddle anzad, a singer reciting the poem, other women with hand clapping (and perhaps a drum tasawat) and may be another singer adding a response or a burden at certain moments of the song. The literary text, evoking a historical event or telling the story of a journey, has a metric structure, rimes and assonances and is thus a real pre-existing “composition”. And so is also the melody with its fixed repeated formulas. But there can be musical and textual deviations, new ornamentations, special repetitions and even additions taken from another song. The realization comes out of the moment, but the compositional elements prevail.(8)

  1. Cf. Lichtenhahn 1994, p. 123.
  2. Cf. Arlt 1983, Haas 2002, Treitler 1981 and 1991.
  3. Cf. Simon 1984 and 1996.
  4. Rosenfeld 1989, p. 135 f and 197.
  5. About the role of the actual performance in illiterate societies in general cf. Goody 1991.
  6. Cf. Lichtenhahn 2002.
  7. Cf. Finnegan 1970, p. 20 ff.
  8. For details cf. Lichtenhahn 2004.

Bibliography

Arlt, Wulf (1983): “Instrumentalmusik im Mittelalter: Fragen der Rekonstruktion einer schrift-losen Praxis”, in: Basler Jahrbuch für historische Musikpraxis, vol. 7, S. 32-64.

Castelli Gattinara, Gian Carlo (1992): I Tuareg attraverso la loro poesia orale, Roma, Consiglio nazionale delle Ricerche.

Ferand, Ernest T. (1958): Art. „Komposition“, in: Die Musik in Geschichte und Gegenwart (MGG1), vol. 7, Kassel etc., Bärenreiter.

Finnegan, Ruth (1970): Oral Literature in Africa, Oxford, Clarendon Press.

Frisius, Rudolf (1996): Art. „Improvisation, I. Zur Terminologie“, in: MGG2, Sachteil vol. 4, Kassel etc., Bärenreiter.

Goody, Jack and Ian Watt (1991): „Die kulturelle Tradition in nicht-literalen Gesellschaften“, in: Jack Goody, Ian Watt und Kathleen Gough, Entstehung und Folgen der Schriftkultur, Frankfurt a.M., Suhrkamp.

Haas, Max (2002): „Anschauliches Denken - Gedachte Anschaulichkeit. Einige Überlegun-gen zur ‚Musik’ im Mittel-alter“, in: Max Haas, Wolfgang Marx und Fritz Reckow (eds.), Anschauungs- und Denkformen in der Musik, Bern etc., Peter Lang.

Lichtenhahn, Ernst (1994): „Musikalisches Handeln im Spannungsfeld von Norm, Funk-tion und Spontaneität: Bemerkun-gen zur ‚Improvisation’ in aussereuropäischer Musik“, in: Walter Fähndrich (ed.), Improvisation II, Winter-thur, Amadeus, p. 117-125.

---   (2002): „Text ? a helpful ethnomusicological category“, in: Dimitrije Buzarovski (ed.), Contemporary Trends in Musicology and Ethnomusicology, Skopje-Macedonia, Institute for Research and Archiving of Music IRAM, p. 81-89.

---   (2004): „Musizieren in schriftlosen Kulturen“, in: Walter Fähndrich (ed.), Improvisation V, Winter-thur, Amadeus.

Rosenfeld, Israel (1989): The Invention of Memory. A New View of the Brain, with a New Foreword by Oliver Sacks, New York, Basic Books.

Sachs, Klaus-Jürgen (1996): Art. „Komposition“, in: Die Musik in Geschichte und Gegenwart (MGG2), Sachteil vol. 5, Kassel etc., Bärenreiter.

Simon, Artur (1984): „Zur Improvisation in Musikkulturen mit mündlicher Tradition“, in:
Musica 38, p. 37-45.

---   (1996): Art. „Improvisation, VIII. Musikethnologie“, in: MGG2, Sachteil vol. 4, Kassel etc., Bärenreiter.

Treitler, Leo (et al.) (1981): „Transmission and Form in Oral Traditions“, in: Daniel Heartz & Bonnie Wade (eds.): International Musicological Society. Report of the Twelfth Congress Berkeley 1977, Kassel etc., Bärenreiter, p. 139 -211.

Treitler, Leo (1991): „Komponieren unter den Bedingungen mündlicher Überlieferung“, in: Hartmut Möller and Rudolf Stephan (eds.): Die Musik des Mittelalters, Laaber (Neues Handbuch der Musikwissenschaft vol. 2).