The African matrix in jazz harmonic practices.

In 1998, a flyer was circulated announcing the appearance of the Garland Encyclopedia of World Music. On that leaflet, the publishers subjected readers to a test with five questions under this heading: "How Well Do You Know World Music?" The first of these questions in statement form was "American jazz borrows its harmonic structure from European classical music." Readers were supposed to mark whether the statement was true or false. The answer given on the back of the flyer for those still in doubt was "True." Thereby was endorsed one of the most tenacious stereotypes about jazz, the all-embracing notion that harmony in jazz and other African-American music was "European" in origin, while the rhythm was "African."

Unfortunately, the facts are much more complex. Before the topic called "jazz theory" became part of the curriculum in jazz schools across the United States, harmonic practices in jazz were not always so Western or "European." Chord symbols such as [Gm.sup.9], [G.sup.09], [E.sup.o7], [E.sup.(7[flat]5)], and so on, with their implicit reference to Western music theory, had served jazz musicians as a useful notational set, just as the Roman alphabet is useful for writing English, French, Latin, and Kiswahili. These symbols are coins with a hidden face. Jazz chords and progressions have functioned like the system of the orixa in Brazil. To a Catholic, the orixa can be explained as a set of Catholic saints, but a Yoruba from Nigeria will recognize all of them as transcendental beings in the Yoruba religion, and Afro-Brazilians in the Candomble religious meetings will think both ways.

Melville J. Herskovits (1941), to whom we owe much insight into the processes of culture contact, called such phenomena syncretism. Herskovits's terminology, embracing selection, retention, survival, reinterpretation, syncretism, and cultural focus, is still very useful (see Evans 1990, 1999), although occasionally with some necessary conceptual modification. Syncretism, for example, should not be understood as a blend or merger of different cognitive systems; rather, it is, at least originally, an attempt at a parallel, "bilingual" presentation of ideas that one can read in either of the two codes (Kubik 1991, 174-176).

From a standpoint in Western music theory claiming universal applicability, it is often difficult to comprehend that jazz musicians have always converted the tonal-harmonic resources provided by the Western instruments that they played to suit their own concepts, strongly rooted in blues tonality. From my viewpoint, as someone who has spent a lifetime in African cultures and recorded some twenty-six thousand items of African music since 1959 in eighteen countries, jazz harmony at its structural and aesthetic level is based predominantly on African matrices, although it must be added that individual jazz performers, ensembles, and composers vary in the degree to which their harmonic practices and understandings are more African- or more European-derived. It may vary even from one work to another or one performer to another.

I am not the only one who has gathered such a glimpse of the hidden side of the coin. Actually, it was Percival R. Kirby who in his article "A Study of Negro Harmony" (1930) first detected a structural principle of African provenance in the harmonic patterns of Negro spirituals. Later, in 1951, A. M. Jones wrote on "blue notes and hot rhythm"; and there was the eminent Richard A Waterman (1952, 209), who pointed to common evolutionary roots of harmony in European and African music:

   Harmony ... appears in aboriginal music nowhere but in the western
   one-third of the Old World, where it is common in European folk
   music and African tribal music.... [T]here exists a broad intrusive
   belt of Arabic and Arabic-influenced music which stretches across
   the middle of the western area, along both shores of the
   Mediterranean. Since the times of ancient history this alien
   musical outcropping has masked the fact of the previous existence
   of a continuous harmony-using bloc of cultures established earlier
   in the area.

Although Waterman ignored Polynesia, and his view of a former historical unity of European and African harmony is probably an overstatement since the "intrusive" nonharmonic sector existed across North Africa and Mediterranean Europe before the Arabs and Islam arrived (e.g., among Egyptians, Phoenicians, Romans, and Berbers), it is important that Waterman, like Kirby, recognized the precolonial existence of harmony in the music of sub-Saharan Africa.

More recently, Thomas Brothers (1994, 490) has drawn attention to melodic anticipation of harmonic schemes in Louis Armstrong's "Big Butter and Egg Man," recorded in 1926. He writes that "Armstrong's F chord comes a measure early, and as a result the phrase structure of the solo collides with the harmonic rhythm of the accompaniment." Brothers, who carried out fieldwork in Ghana, points to the fact that such behavior serves "an African conception of syntax that involves two levels, a fundamental and a supplemental, with the supplemental moving in and out of agreement with the fundamental." To some, this may sound esoteric in its formulation, but in fact, insights like Brothers' open jazz to cognitive investigation. At one level of inquiry, it may be sufficient to identify the social processes that promote or inhibit innovation, or to study the sounds and their sequences that musician-composers produce, but at another level, we would like to learn about the musical concepts and nonverbal thought patterns that fuel and steer those results.

It seems that in jazz history from the 1920s to the 1950s, different sets of African traits became prominent in succession. Heterophony and responsorial, functional polyphony were dominant in early New Orleans jazz, as in the testimony by Bunk Johnson and musicians such as clarinet player George Lewis, who had remained local (that is, not emigrating to the North during the early 1920s); homophonic multipart structures set the tone of big-band jazz during the swing era of the 1930s, while equitonal melodic principles, clustered chords often based on remote partials, and what has been called the pitch area concept in the blues (Evans 1982, 24) staged a breakthrough in bebop of the 1940s.

In a sense, jazz history was also like a series of volcanic eruptions. Whenever a new jazz style appeared, it brought to the surface something that had been dormant in America for some time, which was then recognized as something with which we have been familiar from somewhere in Africa. In my wife's village in Malawi, I remember that in 1967 there were rap artists; they pronounced their complaints while walking, for everyone to hear. Some were indeed the "public enemy" of the established political hierarchy. We did not call it rap, of course; we simply called it declamatory oral literature.

The fascinating thing about jazz history is that when something new appears, as bebop did most radically in Minton's Playhouse in Harlem in 1941, it often seems to re-create concepts, traits, and aesthetic principles central to some African cultures somewhere on the map. In America, however, they erupt in a new disguise, as if remolded or reconfigured with a different type of clay. Thus, a time journey through jazz and blues history is often like an excursion through different African cultures. In one style of jazz or blues, one gets more traits from the west-central Sudanic savanna and Sahel zone, in another from the Congo and the Guinea coast, and in a third from the Zambezi valley and northern Mozambique. From a viewpoint inside Africa, it is difficult to escape the feeling of deja vu that periodically arises.

The present article is a preliminary exploration of this dimension of jazz history, excluding rhythm and other details that are easy to analyze, while concentrating on the question of tonal-harmonic structures and patterns. The subject is further narrowed by the fact that I deal only with African, and no other, matrices in jazz, but that does not express any deliberate neglect of the other face on the orixa coin.

Recorded African Traditions of Multipart Singing

The age-old presence of specific multipart styles and compositional principles, both vocal and instrumental, in the music of sub-Saharan Africa is no longer much of a dispute. There are numerous works dealing with the subject. In Africa, there are large areas in which only singing in unison and octaves is practiced, such as among pastoralists: the Cushitic speakers of northeast Africa, the Fulbe cattle herders in West Africa, and the Hereto pastoralists in Namibia. However, unison and octave styles do not exclude the development of polyphony; the evidence is the instrumental polyphony on xylophones and harps that developed in the precolonial kingdom of Buganda in East Africa's interlacustrine region, a very complex serial music, with no simultaneous sounds other than octaves allowed (see Kubik 1994). Octave styles in Africa contrast with large pockets of harmonic traditions, for example, along the Guinea coast, from the ancient city of Benin, the Ijesha- and Ekiti-Yoruba--speaking peoples, to the Ivory Coast and southern Ghana, where harmonic composition techniques among the Fanti and Asante were first described by Thomas Edward Bowdich in 1819. Another large "harmonic patch" covers areas of Congo-Brazzaville, Congo-Kinshasa, southwestern Angola, eastern Angola into Zambia, expanding into the Zambezi valley and other parts of southern and eastern Africa (Kubik 1996).

Although we cannot time travel for verification, there is good reason to assume that Africa was one of the planet's cradles of polyphony. Bushmen hunter-gatherers must have discovered the harmonic properties of stretched strings as soon as they had discovered the use of the bow and arrow; and this knowledge has continued until today, with the !Kung', for example, using no fewer than four different kinds of harmonics-producing techniques on hunting bows (Kubik 1994, 217-224). Equally, the pygmies of central Africa developed polyphony probably thousands of years ago--although, in contrast to the savanna-inhabiting bushmen, not from instrumental techniques but from the formants of speech and vocal techniques such as yodel--long before their forests were invaded by Bantu-language speakers from eastern Nigeria and the Cameroons between circa 1000 and 400 B.C. Some of these invaders then adopted the pygmy-style polyphony, while others retained their own multipart experiences.

In 1946, Andre Didier and Gilbert Rouget of the Musee de l'Homme in Paris recorded fascinating multipart singing in the then French Congo: heptatonic harmony in lush triadic chord clusters. I reconfirmed the existence of these styles among the Bongili and Bakota in 1964 during my own travels (see transcription outline in Kubik 1998, 664). In the 1950s, Rouget documented an equiheptatonic singing style among the Baule of the Ivory Coast, with its characteristic neutral thirds. (1) And in 1965 in eastern Angola, I recorded homophonic three- and four-part singing in initiation ceremonies and masked performances among the -Mbwele, -Nkhangala, -Lucazi, and -Cokwe. This is an area displaying a non-modal style that involves chord progressions that fluctuate between major, minor, and neutral triads. An impressive example is a song in the women's secret tuwema (flames) society. (2)

I recorded in 1964 shifting of chord clusters by a semitone in the sya stories among the [TEXT NOT REPRODUCIBLE IN ASCII] settled in the southwestern forests of the Central African Republic. In that culture area, there is a strict parallelism of individual voices. The principle of chord formation in this style could be described as stacking of thirds on top of each other. In principle, this process is unlimited. Sometimes one gets only two voices in parallel major thirds. In another performance of the same tune, one gets three voices, forming major triads. In one version of the song "Atende," which I recorded at Bigene village in 1964, a fourth singer joined with yet another third on top of the triad. Thereby, the progression turned most naturally into parallel movement between two major sevenths, a semitone apart: [Cmaj.sup.7] to [D[flat]maj.sup.7] and back (see Ex. 1).

[ILLUSTRATION 1 OMITTED]

All of these harmonic traditions used to flourish in places that were relatively isolated. There were no influences from church songs or mass-media music. Probably they were centuries-old local developments, and some of them had restricted distribution areas.

In light of the precolonial presence of distinct harmonic practices in many parts of Africa from where people were deported to New World destinations, it would be strange if nothing of the underlying concepts had survived in African-American cultures. Surely, New World cultures were a bedrock of innovation, so things became modified. Blues and jazz were developed as new genres in America; and yet there was resilience, and there were also some strange contemporaneous parallels across the Atlantic divide. It is little known, for example, that about the same time that jazz began to emerge in New Orleans, there occurred analogous developments in Africa, also based on European brass-band music. One was konkomba music on the Gold Coast. We only have relatively recent recordings, such as those of Brookman Mensah and his Sankro Brass Band (see Fig. 1) made by Mose Yotamu in 1978, which were analyzed by Alfons M. Dauer (see Yotamu 1979; Dauer 1985). The analogies between New Orleans' brass-band jazz and Brookman Mensah's music are stunning. Both are heterophonic and set in functional, responsorial polyphony, revealing a common denominator underlying the innovative steps of the late nineteenth century that took place in both America and West Africa.

[FIGURE 1 OMITTED]

In the next section, I discuss in detail some of the more prominent African tonal-harmonic concepts and multipart structuring techniques that appear to form an ever-present background glow to the universe of various jazz styles.

The Perpetuation of Structural and Conceptual Elements from African Music in Jazz: Five Principles

The remote cultural genealogy of various practices and understandings in jazz links with African cultures of the past mainly at the structural and cognitive levels makes research on the subject so difficult. We are not dealing here with phenomena that can be easily measured and verified but with what is hidden, discernible in some forms of behavior that otherwise remain unexplained and in the nonverbal dimension beneath the rationalizations one may get from musicians in interviews. In the area of the visual African arts, the book edited by Mary N. Nooter (1993) is a good introduction to the visual language of secrecy that parallels, in my opinion, an auditory language of secrecy perpetuated in jazz. My colleague Kayombo ka Chinyeka in Zambia has written a book in Lucazi, an eastern Angolan language, on the hidden dimensions of African cultural expressions; he calls his book "Konkha vya vanda" (Search for the Hidden) (Chinyeka [1989]), a perfect descriptive title of what we, the researchers, are trying to do.

There are five major principles of organization and conceptualization of tonal-harmonic elements in African music that seem to have continued and been creatively applied in jazz. These organizational and cognitional principles include

1. the span process,

2. the experience of partials-derived systems,

3. blues tonality,

4. the concept of flexible pitch areas, and

5. equiheptatonic concepts.

I would not claim that this list is complete. Thomas Brothers' (1994) discovery of a conception of syntax involving a fundamental and a supplemental level, for example, could be listed as a sixth principle. Africa is a vast continent, and the connecting lines between African-American and African traditions crisscross the continent like caravan trade routes (see also Brown 1992). Often it is impossible to disentangle the picture historically because we are compelled to rely predominantly on twentieth-century data, projecting into the past. And yet, I feel it is worthwhile to descend to the basement of the building where foundations become visible that may be totally obscured by the noises and the jargon spoken on the first and second floors.

The Span Process

A span is defined as "a stretch between two limits" according to the Longman Dictionary of Contemporary English (1990, 1010). Xylophone players in Africa often transfer visual spans over the keyboard; that is, holding a beater in each hand, they strike two keys on the xylophone that are separated by one key and then shift this bichord in parallel hand movement to lower or higher keys. (3) One would expect such movement to result in harmonic parallelism. It often does, but not necessarily so, depending on the tuning of the instrument and the layout (scale or otherwise) of the notes.

If African xylophones were tuned to the Western diatonic scale--they never were before factories stepped in to produce the commercialized marimba--the result would be chains of third, alternating between major and minor. Lionel Hampton produced them occasionally on his vibes. If, on the other hand, the tuning of the xylophone is pentatonic, such as (from top to bottom) c-a-g-e-d-c, the result of the series of bichords described above would be parallel fourths that are interrupted in one place by a major third. Example 2 shows why that is so.

[ILLUSTRATION 2 OMITTED]

During my earliest work in Africa from October 1959 to October 1960, I encountered vocal harmonic patterns of this kind in southwestern Tanganyika, among the -Pangwa, -Kisi, -Ngoni, and others. (4) At first, I had no explanation for the isolated major third that always appeared in a certain place, until a structural explanation occurred to me. I had chanced upon the "span process" or "skipping process" (as I began to call it) as a generating principle. It turned out to be valid interculturally across Africa, determining the character of many forms of African homophonic multipart singing (instrumental layouts included).

Let me first explain the term homophonic as I use it with reference to African music (Kubik 1996, 1999a). It distinguishes two categories:

1. Multipart organization with analogous movement of individual voices (i.e., homophonic multipart organization)

2. Multipart organization with interlocking movement of individual voices (i.e., polyphonic multipart organization)

In category 1 (homophonic), two or more parts are sung by individual singers in an analogous manner. All parts are rhythmically identical; they start and end at the same point. If the performance is vocal, participants sing the same text. The individual voices usually move together in intervals that are perceived as consonant (i.e., "agreeable"). Motion can be strictly parallel, in which case the participants' voices are always separated by the same interval. Much homophonic singing in Africa, however, is not strictly parallel. There are styles--for example, in eastern Angola and northwestern Zambia--in which all kinds of movement--parallel, oblique, and contrary motion--are considered acceptable. Homophonic multipart singing is associated in Africa with the call-and-response (leader/chorus) form.

In category 2 (polyphonic), a group of people perform together, singing different text phrases (or syllables) in an interlocking style, without any common starting and ending points. The phrases can have different lengths, and they interlock in various ways. At certain points, some of the voices meet, producing simultaneous sounds. This category of multipart organization is not usually linked to the call-and-response (leader/chorus) form. Apart from the polyphony of the bushmen and pygmies, impressive polyphonic styles can be found, for example, in Shona music of Zimbabwe, in Sena music of the Zambezi valley, and in many other places in Africa.

Returning to the example from southwestern Tanganyika, a second singer finds his or her vocal line easily by skipping one note of the scale that they share. The second singer, who may be anyone in the audience or within the performance group, duplicates the pitch line of the first at a different level, most often either two steps down or two steps up the scale. In a pentatonic system such as outlined above (and there are, of course, several types of pentatonicism in Africa), the combination of the two voices will yield characteristic bichord progressions, as in the song "Mphezi mphezi" (see Ex. 3). (5)

[ILLUSTRATION 3 OMITTED]

Proceeding from the pitch line of the first singer, each "harmonizing" note of the second singer is obtained by skipping one note of their common scale. Like the xylophone player who transfers a visual span, so these singers transfer an auditory span that is not defined by "intervals" (e.g., thirds versus fourths) but by an identity in directional movement of the vocal lines combined. This means that what for a Western observer would seem to be an important distinction between fourths as contrasted with thirds is conceptualized in this culture area as an identical linear relationship.

In other African culture areas, there are different relationships of sound, as the span process is applied to a variety of tonal systems. When I realized that it was an interethnic, interculturally valid procedure for the formation of simultaneous sounds, I began to test its applicability across Africa. Soon, it became clear that the span process was a constant. The most important variable on which the sonic result depended was the structure of the underlying tonal system.

What will happen, for example, if one pitch in the simple anhemitonic pentatonic scale that I have just described is altered by a semitone? We do not have to carry out an experiment because such an alteration already exists and has been documented independently in Cameroon and Burkina Faso.

Ever since Paul Oliver's pathbreaking research on the blues and its remote historical links to West African savanna cultures, the so-called west-central Sudanic belt has emerged as a key region for the likely provenance of many of the blues' most salient traits, including the use of bourdon, the strongly centralized tonality, the blue notes, the often declamatory, melismatic singing style, heterophony, and more (Oliver 1970; Kubik 1999a). In December 2000, Hungarian musicologist Andreas Szabo, on one of his numerous visits to Burkina Faso, worked with Mamadou, a xylophone player in the Diabate family settled at Torosso, forty kilometers west of Bobo Dioulasso. Mamadou's xylophone (bala), gourd-resonated and on a stand, was tuned to a "bluesy"-sounding pentatonic scale. He played with several younger musicians, Sadama, Sibiri, and Moussa from the same family, all belonging to the Sembla people, a Mande-speaking group. The auditory impression of a blues-like scale arose from the fact that the second degree (from the bottom) of this pentatonic tuning was raised by approximately a semi-tone compared with the more common pentatonic tunings elsewhere. The resultant bichords were obtained by the player through the visual span process described above. I have seen Szabo's video document. It was recorded in a village that showed no traces of any "returning" blues influence through Ali Farka Toure or other such artists. Obviously, the tuning was part of an established xylophone tradition (see Ex. 4). Although I would not propose that the bala tradition in this village has remained unchanged since the early 1800s, it also happens sometimes that certain traits show surprising stability. For this reason, it is useful simply to pay attention without imposing premature conclusions. Mamadou, the xylophone virtuoso, then created his bichords by shifting his beaters, held parallel, across the keyboard with one "empty" key between.

[ILLUSTRATION 4 OMITTED]

A quarter century earlier, in another key African area, the central Cameroonian savanna, I had recorded a fabulous blues-like grinding song by a Tikar woman. (6) On the same trip in 1964, I also recorded a mourning song by another Tikar woman in the local Chief Ngambe's entourage, the scale pattern of which strangely conformed with that of Mamadou's bala in Burkina Faso. (7) She was accompanied by a man playing a polyidiochord stick zither (mbo loya). The five strings divided into ten sections by the central notched bridge were tuned (relatively, from top to bottom) e[flat]-c-a-g-e-e[flat]-c-a-g-e[flat]. Within this pentatonic scheme, an interference pattern emerged between the middle e in the performer's right-hand playing area and the e[flat] in that of his left hand. It was what accounted for the music's blues-like character. The wailing quality of the female voice and her melodic patterns were other traits. She complained about her husband, who had deserted her. Ngambe village in Cameroon is about two thousand kilometers from the Diabate family's domicile in Burkina Faso. I think that the incidence of an identical tuning pattern found at such a geographic distance is historically significant and suggests a certain antiquity of this pattern. The span process is, of course, absent here (in contrast to the music of the Diabate family in Burkina Faso) because the Tikar singing tradition is based on unisons, octaves, and some heterophony--not on harmony-oriented combinations.

In harmony-employing African traditions (see the rough distribution map in Jones [1959]), the span process often establishes a functional relationship between harmonic style and tonal system. In these cases, it is possible (as I did on many occasions) to reconstruct the "family relationship" between the constituent tones of a tonal system from the type of chords used in the music, and--vice versa--to predict what chords will be used in a musical culture, if one has understood the intricacies of its tonal system. By chords, of course, I mean not only bichords (two voices combined) but any kind of clustering.

In eastern Angola, for example, the underlying heptatonic system invites the application of the span process in two directions simultaneously, up and down the scale, proceeding from a middle voice. Thereby it is possible to explain some of the spectacular progressions in triadic chord clusters that are so characteristic of eastern Angola and that can be heard, often all night, sounding from a mukanda boys' inititation camp. (8)

The harmonic system of the -Mbwela, -Nkhangala, -Cokwe, -Lucazi, and -Lwena (-Luvale) in eastern Angola emphasizes stepwise progression. But singers want to preserve a euphonic identity between the emerging triads across the range of the scale, so intonational adjustments are constantly made. The result is flexibility of several constituent pitches of this scale; they can be modified according to harmonic context. The span process, however, operates independently of this fact. Take the upper c (degree 8) as a point of departure, it can be represented as in Figure 2.

[FIGURE 2 OMITTED]

It is important to understand that (1) the span process here goes in any direction and can be triplicated even to form seventh chords, (2) individual voices tend to move within the margin of a fourth, rarely exceeding the range of a fifth (Kubik 1994, 194), and (3) the notes written as c/c#,f/f#, and g/g# in Example 5 are each to be understood as a single toneme. In the musical language of the Ngangela- and Cokwe-speaking peoples of eastern Angola, the pitch values represented here as c and c#, for example, are not necessarily conceptualized as different pitch units but rather as two alternative intonations of the same toneme, according to context. Therefore, applying the span process from e downward, the note to go with e can be c or c# or even a value somewhere between. The tendency is to intone the "resting" chords, that is, the last bichords in Example 5, as neutral thirds. The songs employ sequences of lush triads, occasionally even seventh chords, and end on bichords that to a jazz ear may sound "altered."

[ILLUSTRATION 5 OMITTED]

Does the span process as a matrix for chord formation survive in African-American music, and in jazz particularly? Most certainly it does, and the evidence is not difficult to procure, although it needs an approach to harmonic analysis somewhat different from the familiar one. In African-American music, the span process survives in homophonic multipart styles, for example, in some spirituals, as Kirby (1930) clearly noticed, vocal quartets, and arrangements of the horn/reed sections in swing jazz, as well as in rhythm and blues/soul music "horn" sections (alto and tenor sax, trumpet and sax, sometimes with added trombone or baritone sax). Where classical European rules of harmony seem to be "broken," such as by parallel fourths or fifths, the explanation often is the underlying span process.

Many of Count Basie's arrangements present a case in point. The famous theme of "One o'Clock Jump," for example, is set for five saxophones (two alto, two tenor, and one baritone) predominantly in harmonic parallelism, itself a feature of African harmonic techniques. The reed and brass sections are in a responsorial relationship, as chorus (reeds) and leader (brass). One would call this four-bar theme a cycle in the African tradition. The theme for reeds is then repeated three times in an almost-identical melodic shape through a twelve-bar blues form. In each repetition, it is "illuminated" by different accompanying chords (on piano and guitar)--in the first line, tonic chords; in the second line (first half), subdominant; and in the third line (first half) dominant. The theme in its vertical dimension is a harmonic progression in added sixths/minor seventh chords. But because it is--so to speak--interpreted from three harmonic standpoints consecutively, the voice lines of the five saxophones are slightly adapted in their repetitions to fit into tonic, subdominant, and dominant harmonies (see Ex. 6).

[ILLUSTRATION 6 OMITTED]

The means by which this is accomplished by Basie follows structural and aesthetic principles that are very different from Western music. It might also reveal something about the history of the blues form. If it is possible to repeat a compact harmonic set of voice lines in quasi-identical shape through the three lines of the twelve-bar blues form, and it sounds fine, then this also brings to light something about the deep structure of the blues, namely, that there is an underlying idea of theme development with no progression at all through "degrees," let alone to a dominant chord--something still treated as somewhat alien in the rural blues. At some time, voice and instrument parts of certain songs within the blues genre would have unfolded over a bourdon-like basis and short rift-like accompanying and responsorial passages. This is not to say that such a tradition should be considered as "older." It is even very likely that through contact with nineteenth-century church hymns and popular songs in the European-American tradition, African-American musicians first embraced the Western common chords enthusiastically, as it has also happened in the development of guitar styles in Africa. But gradually, the African concepts crept in, like Freud's "return of the repressed," and younger musicians ventured increasingly into nonconventional, non-Western expressive techniques. At that stage, a music called blues became known and was then recorded. And ever since, no opportunity has been lost to "disguise" the Western progressions, as is also the case in Count Basie's masterful arrangements. Bebop musicians continued the idea and elaborated on the swing jazz models of presenting a blues theme in three repetitions with little melodic changes. See for example, Dizzy Gillespie with Frank Paparelli on the piano in "Blue n' Boogie" on Dizzy Gillespie and His Sextet and Orchestra or Parker's fantastic "Cool Blues" on Jazz End Blues.

At first hearing, one might not detect any unifying structural principle of African provenance in Count Basie's "One o'Clock Jump," although the auditory impression is regularly described by some of my musical acquaintances in southeast Africa as stunningly African. This concerns in particular the compact set "as if the saxophones were singing makwaya" (in a choir) and the static quality of the added-sixth (respectively, minor seventh) chords that involve the performers in intonating parallel seconds and fourths in several places. At first, this compact sound seems to defy analysis, until one realizes that some of the semitone modifications, for example, e[flat] versus e, might express adjustments similar to those I have described in vocal music of eastern Angola (see Fig. 2).

The remaining work is then almost routine. We now arrange the theme's notes as composed by Basie in the form of a scale and compare them with the slight modifications of the chords formed by the five saxophones, thereby discovering which notes can be understood as single tonemes. If we then apply the span process to this condensed scale, we discover that it explains the chords. I was able to detect that in this scale, e/e[flat] and a/a[flat] each form a single toneme (see Ex. 7). These tones are interchangeable thirds over c and f, respectively.

[ILLUSTRATION 7 OMITTED]

Thus, it can be observed, for example, that the first chord in measure 1 (see Ex. 6), g-e-c-a-g (from top to bottom) derives from the span 5-3-1-6 in Example 7. The lower g, as an octave transposition of the higher one, can be ignored (as a duplication). When this chord is repeated in measure 5, the e is lowered to e[flat], which makes no difference structurally because both e and e[flat] are the same toneme and therefore interchangeable within the span process. Also when checking the spans in Example 7, one must be aware that it is structurally a four-voice arrangement; one note in the voice-line cluster of the five saxophones is always duplicated. It then forms the interval of a second with its neighbor.

All of the chords in Basie's arrangement follow the span process, except one--the penultimate chord in the theme, in measures 3, 7, and 11: g-e-b-a[flat]. Its deviant structure probably derives from its function as a substitution chord of some "dominant" quality; it can be split into the components g plus e and b plus a[flat], both impressing upon the ear their minor-third quality. The auditory trick is that, thereby, g and b are prevented from fully unfolding their conventional dominant function.

Example 8 shows the ten different chords in Basie's theme and its repetitions. To make them comparable, I have rewritten them uniformly, in seventh-chord positions by octave-transposing one or two notes. Succeeding chords in the row (Ex. 8, from left to right) are often obtained through minimal alteration: lowering one pitch by a semitone.

[ILLUSTRATION 8 OMITTED]

The African retentions in Count Basie's arrangement of "One o'Clock Jump" can be summarized as follows:

1. Homophonic treatment of voices, predominantly in parallelism.

2. Creation of a lush compact harmony in homogeneous sound clusters reminiscent of eastern Angolan harmonic singing as well as that of some places on the Guinea coast (Alberts and Alberts 1950). It is not by chance that Ghanaian bandleader E. T. Mensah embarked on a similar aesthetic in the 1950s (see, for example, "O hentse mi lo").

3. Use of a modified diatonic scale with two flexible tonemes, e/e[flat] and a/a[flat], interchangeable in chord formation.

4. Triple application of the span process to the scale, resulting in seventh chords. However, in their distribution to the individual saxophone players, the composer has introduced appropriate octave displacements and duplications that can veil the underlying structure.

The Experience of Partials-Derived Systems

African tonal systems can be divided into three broad families: (1) those developed from the extrapolation of the formants of human speech, (2) those developed from the recognition of partials of the natural harmonic series on tools (instruments), and (3) those developed from projection of the idea of equidistance upon auditory materials. Each category contains an impressive array of possibilities that have led to different tonal systems, that is, tetra-, penta-, hexa-, or heptatonic systems with specific interval relationships (see Kubik 1985, 1994, 1996).

We perceive harmonics in the sounds generated by natural forces such as wind, the wheels of a train, a broken water pipe or someone's vacuum cleaner. These are human experiences that are independent of culture-specific music enculturation, but they can form an ever-present resource for experimentation, especially among children, that begins with a desire to manipulate those sounds, alternating their timbre and reinforcing some of the partials selectively. In speech, it occurs naturally, as Benjamin V. Boone (1994) has shown with reference to blues-like speech in Jelly Roll Morton and Alan Lomax generated by the use of higher partials. Someone brushing his or her teeth while altering the size of the mouth's cavity has rediscovered the natural harmonic series by the method of using the mouth's cavity as a resonator for sound modification and amplification. It is like playing a friction mouth-bow, a chordophone that is common in Angola and Mozambique. (9)

When playing a wind instrument with an inserted mouthpiece, such as a saxophone, a clarinet, or the metal flute that was used in kwela swing jazz of South Africa, the mouth's cavity also functions as a variable resonator. Well-known saxophone players, including Lester Young, Charlie Parker, John Coltrane, and others, explored their physical abilities to modify timbre. In jazz, timbre shaping is one of the most important areas of individual expression. John Coltrane's timbre alterations and manipulations of overtones on the saxophone (e.g., in "The Promise" recorded at Birdland on October 8, 1963) can be seen as the rebirth on American soil of reed techniques inherited from alghaita oboe music in the west-central Sudan. It is worthwhile to compare Coltrane's aesthetics and techniques with those of musicians in the court of the Fulbe ruler of Toungo in northeastern Nigeria and also with the solo performance by a Hausa musician and trader, Amadou Meigogue Garoua, with a small algeita recorded at Yoko, Cameroon, in February 1964. Both recordings have been published on the compact disc that accompanies Africa and the Blues (Kubik 1999a), items 19 and 20.

The manipulation of nonharmonic overtones is one area of expression. Another, of course, is the systematic amplification of harmonics. When brushing one's teeth, one can reinforce harmonics up to partial 8 quite easily. The lowest partial in a harmonics column is called the fundamental. In prehistory, humans used toothbrushes from split plants, we assume, but there were also ample opportunities to discover the harmonic series through the use of other tools. Eventually, some techniques, such as reinforcing the harmonics of a stretched string by mouth or an external resonator, became established.

In Africa, such discoveries probably took place in several places simultaneously within a few thousand years. Until recently, the so-called bushmen in Angola and Namibia have been part of an ecologically determined hunter-gatherer economy, with the hunting bow a central tool. Not surprisingly, therefore, it also became a central musical tool. The tonal system of the !Kung', which I studied in Angola in 1965, was developed from experiences with the stretched strings of hunting bows.

For us today, it is no longer important, as it was at the turn of the twentieth century, to speculate about which might have been first, the "hunting bow" or "musical bow." Psychologically, it is evident that someone discovering one usage is likely soon to discover the other one as well. Thus, it can be inferred that bushmen tonality and polyphony were developed thousands of years ago, long before the specific thirteenth- to fourteenth-century developments of polyphonic techniques in Europe. A totally different issue, however, is the subsequent transmission and diffusion of the ideas, always depending on circumstances. Musical bows seem to have had very little development among Native Americans, for example, although the hunting bow was widespread. Thus, while the hunting bow is nearly universal, musical bows are not. The separation occurred at an early stage, and in some parts of the world they were replaced by complex stringed instruments and/or Jew's harps.

No bushmen were deported to the Americas. The slave traders in Angola despised them. But southwestern Angola was a major slave-raiding area in the eighteenth century, and Bantu-language speakers such as the -Nkhumbi, -Handa, -Cipungu, and others, had been living in an economic symbiosis with the !Kung' for hundreds of years and--as is testified by the research of my Angolan colleague Marcelina Gomes--adopted essentials of bushmen musical-bow techniques. They then took their knowledge to New World destinations. To my surprise, when undertaking a systematic trait-by-trait analysis of an African-American musical-bow player, Eli Owens, recorded by David Evans in 1970 and 1973 in Bogalusa, Lousiana (Evans 1994, 333-334), we discovered that the closest African parallels that emerged were to be found in bushmen-related bow techniques of Angola and Namibia.

There are several partials-derived tonal-harmonic systems in use across Africa. What they sound like depends on multiple variables: first, whether the system comes from vocal discoveries (formants of human speech or overtone singing) or from the use of an instrument; second, if its origins are in instrumental techniques, whether it is based on one or more than one fundamental; third, if there is more than one fundamental, what their interval relationships are; and finally, how high up the series of partials the tones are selectively reinforced.

African musical cultures that incorporate a strong awareness of a basic reference tone often operate from a tonal system that has its origins in the selective use of partials over a single fundamental. Of course, not every bourdon-based style can be linked to such a background. By now, several single-fundamental systems have been identified, for example, among the Wagogo of Tanzania (Kubik 1994, 1996) and the !Ko of the Kalahari in Namibia and Botswana (see video document no. 6, tuning of a pluriarc by Dena Pikenien, 1991/Kubik/Malamusi, Namibia). Such a tonal system is easily identified by the fact that it does not provide any possibility for so-called root progressions (see Blacking 1959), that is, movement between tonal steps (degrees) with harmonic implications.

In some other places of Africa, one can encounter a very different approach to harmony and form: the so-called shifting tonality. Two roots, usually a whole tone apart, are needed to establish musical events shifting between two focal points. In Luba and Kanyoka culture in the southern Democratic Republic of Congo (see, for example, Hugh Tracey's [1973] recordings on AMA-TR 15), the hexatonic material was laid out in the form of two contrasting tonal compounds, transcribed as c--e--g shifting to d--f--a and back. Translated into Western music theory, it could be described as shifting between two chords, C major and D minor. Such systems, usually hexatonic, can be found in abundance in Africa, with or without associated homophonic or polyphonic devices. In jazz, shifting tonality is the hidden African matrix in some of the so-called modal jazz propagated at one time by Miles Davis and by John Coltrane.

But most forms of shifting tonality in Africa include chords. They derive from the exploitation of harmonics over two fundamentals. This is the case among the Fan in Gabon, whose type of mouth bow survived into the 1950s in at least one New World tradition, at Palenque, Colombia (List 1983); it is also the case of the Xhosa in South Africa, and the -Nkhumbi/-Handa in southwestern Angola. The tonal system used by these peoples is based on two harmonic columns up to partial 6 over fundamentals that are a whole tone apart. It generates a hexatonic scale. Spectacular use of this kind of scale for multipart singing can be observed in ekwenje initiation songs, which I recorded in southwestern Angola (Humbi en Handa) (see Ex. 9).

[ILLUSTRATION 9 OMITTED]

In the Caribbean, there is a strong memory of shifting harmony associated with chords of this kind. On guitars, in Venezuela and elsewhere, they are usually represented in transposition as

[ILLUSTRATION OMITTED]

resulting in a kind of "dual fundamental" harmony. In a recording titled "Bo Diddley" that was made in Chicago on March 2, 1955, the early rock 'n' roll guitarist and singer Ellas McDaniel, known as Bo Diddley, can be heard with a passage on the electric guitar in a shifting harmony between f and g (relative pitches) with wow effects. To me, this passage sounds as if it has been played on a ben mouth-resonated musical bow in a bwiti cult meeting in Gabon. McDaniel is unlikely to have had any knowledge of such recordings in 1955 because they then rarely existed. I am convinced, however, that his preferences for such shifting triads on his electric guitar was not a random choice. Even if he adopted this practice from Caribbean styles, it represents a rediscovery of musical-bow harmonies out of some need. But what pushed him to use these harmonies when other blues performers of the time did not? It is significant that he was born and raised in a place on the border between the states of Mississippi and Louisiana. His birthplace, McComb, is only about thirty-five miles from Sandy Hook, where Eli Owens was born and raised. Could it be that there was a stronger historical concentration of a Congo/Angola background in this area than elsewhere in the Deep South? These are questions that we need not answer here, but they remind us that there may be more below the surface than we normally suspect. (10)

Studying African tonal systems comparatively, one can discover yet another significant relationship: the number of roots used in a musical system and the range of harmonics exploited is inversely proportional. Tonal systems built over a single fundamental tend to go high up, at least to partial 9. It is obvious why; one cannot form melodies from lower partials over just one fundamental. It would result in disjunct intervals of octaves, fifths, and fourths. For melodic formation, one needs the higher partials with their narrower intervals, at least from 4 to 9. But if a second or more roots are added, then tonal resources increase, eventually to an extent demanding restrictive measures. Over two fundamentals, one usually goes up to the sixth partial, very rarely up to the seventh, attaining a heptatonic system. But that is the limit. Neither in Africa nor in the blues can any "octatonic" scales be found, an idea (occasionally encountered in the blues literature) that comes from compulsive counting of variations of blue notes.

Tonal systems over three fundamentals do occur in Africa, but the most prominent case is European folk tonality over fundamentals c, f, and g, generating the "three common chords." But tonal systems built on four fundamentals, that is, with root progressions between four degrees, are prominent in Africa. In some areas, such as Busoga and Buganda (East Africa), the four-root cycle is not associated with any harmonic device; tonality in a song simply shifts between degrees. In other musical cultures of Africa, four-root cycles are associated with harmonic sound derived from partials. One such case is the Shona tonal system in Zimbabwe. It is based on four roots, each constituting a column of harmonics up to partials 3 and 4. The result is the characteristic Shona bichords in fifths and fourths over the roots c, e, g, and a (see Ex. 10).

[ILLUSTRATION 10 OMITTED]

Modern adaptation of the Shona chords can be found in the music of Thomas Mapfumo, especially in his early recordings such as on the LP Hokoyo. Although the Shona chords are responsible for the characteristic sounds to be heard in Shona music of Zimbabwe, and also in Sena music of the Lower Zambezi valley, the pitch values of the four roots are not rigorously fixed. Sometimes there is considerable margin for variation, especially in the lower Zambezi valley, where xylophone and bangwe zither players long ago introduced an equiheptatonic temperament. That this must have had considerable influence in conditioning auditory perception in children is evident. Certainly, it could have been better studied fifty years ago, when radios were still rare in local houses. And yet we can tentatively assume that growing up in a musical culture in which "perfect" fourths and fifths have been tempered to create an equiheptatonic scale with a standard interval around 171 cents widens the margin of tolerance in pitch recognition. It can easily make semitone progressions acceptable to the ear. And this is why, besides other factors, (11) I find some strange cognitional affinity between Zambezi music and modern jazz forms.

To our great surprise, in May 1967, when Maurice Djenda and I recorded in a village on the Malawi-Mozambique border, we came across a corn-pounding Sena woman, Fainesi, about age twenty, who was singing a worksong to yodel syllables, modulating in semitone steps through a four-bar cycle. The last "chord" of her cycle was then a perfect modulation to a tonal basis a semitone lower than her start (see Ex. 11a). Because this is yodel, with the lower partials implicated in the progressions, one can rewrite Fainesi's tonal material as a progression through four chords to see what is revealed (see Ex. 11b). There is also a thought experiment behind it. In a science-fiction story, Fainesi, with her tonal understanding, would have been committed to a time journey and settled in the United States with a piano teacher. After having learned to play piano, she would have been taken (with her memory still intact) to Minton's Playhouse in 1941. Charlie Parker would have asked her to make a contribution to his music based on one of her songs remembered since childhood. She would have played the progression shown in Figure 11b with her right hand and later asked Thelonius Monk to add a bass line. Which bass notes would he have added? (This riddle is for the jazz reader to solve. It is my challenge to those who have marked "true" on the Garland Encylopedia's leaflet! It remains to be seen how many solutions will result.)

[ILLUSTRATION 11 OMITTED]

Blues Tonality: Higher Partials, Difference Tones, Blue Notes, and the Interference Pattern

Why do certain forms of African music sound "bluesy" while others do not? In the early 1960s, when I began recording all over the continent, Africa was not yet totally connected to Western mass media. There were still traditions that developed and changed responding mainly to internal forces and intra-African streams of influence. Listeners to my early recordings often detect blues-like music, some of which is now published. (12) These listeners are of the most varied background, come from many different countries, and all have their specific ideas about blues. Yet a tendency is discernible. The music they qualify as "bluesy" mostly comes from within specific language families in a region extending from Burkina Faso and northern Togo to northern Nigeria, into Cameroon, and across the continent as far east as the Ingassana Mountains in the Republic of the Sudan. It comes from peoples speaking Mande, Voltaic, or Adamawa-Eastern languages (Kutin, Chamba, Zande); Hausa; "Bantoid" languages within the Benue-Congo family (such as Tikar); and eastern Sudanic languages such as those spoken by the Ingassana and the Berta with their spectacular horn ensembles recorded by Artur Simon. (13) It is no exaggeration to call this large area across the geographical Sudan the swing belt of Africa. It radiates to areas as far south as the Luo and Abaluhya of western Kenya, especially among lyre players with their declamatory, melismatic vocal style. The second "bluesy" area seems to cover northern Mozambique into Malawi--a much smaller territory than the first--especially as concerns log xylophone music, songs accompanied by fiddle or zither, and so on, mostly by performers within Bantu language Zone P, that is Yao, Shirima, Lomwe (Guthrie 1948). A third "bluesy" area emerged after our research in Namibia, in 1991-1993. Audiences in my lectures usually find the pastoral Herero "bluesy," especially the oucina pentatonic women's songs to hand claps on the off beat. (14)

Can one make anything out of such an impressionistic assessment? I concluded that it was best not to ignore it. I decided to find out what exactly these people (with a variety of cultural backgrounds but some familiarity with jazz and blues) had found to be so "bluesy." I discovered that in many cases, the impression was created by just a few traits that appeared in those musical styles in various combinations and configurations: (a) music with an ever-present drone (bourdon), (b) intervals that included minor thirds and semitones, (c) a sorrowful, wailing song style, and (d) ornamental intonation. Songs with a prominent minor seventh in a penta- to hexatonic framework also sometimes received this designation, as did pieces that featured instrumental play with a clash between a major and minor third or with a specific vocal style. Never did they consider as "bluesy" music with much percussion.

One can take a closer look at some of these traditions. My earliest such encounter was in October 1962 in the Mitukwe Mountains of northern Mozambique. I was studying a small Shirima log xylophone with six slats played by two boys out in the fields to chase away birds and monkeys from damaging the crops. Under the impression of those fast sound sequences over a kind of constant drone, it occurred to me that the tuning perhaps followed the natural harmonic series as far as partial 15 (Kubik 1965, 39). Earlier, in 1960, I had recorded in southwestern Tanganyika two small girls of the -Kisi ethnic group whose voice combinations could only be explained if one assumed that they proceeded from a scale that was constructed from partials 6 to 11 over a single fundamental. (15)

Before going into detail, it will help to recapitulate some of the characteristics of the natural harmonic series. It is important to understand that the upper harmonics are audibly different from Western twelve-tone tunings. The "natural" major third is only 386 cents wide, that is, 14 cents narrower than the tempered third of 400 cents. Further up, the "natural" minor seventh is at 969 cents, that is, 31 cents lower than the minor seventh (b[flat]) on a piano. And the eleventh partial stands at 551 cents, just between a tempered f and f#. The thirteenth harmonic at 840 cents is positioned between a and a[flat] if we take the fundamental to be c. Example 12 gives the approximate notational values of the natural harmonic series, up to partial 11, with deviations from Western tuning marked on top of the notes in cents.

[ILLUSTRATION 12 OMITTED]

It is likely that individuals who grow up exclusively in a tradition where the exploitation of upper harmonics is important tend to "hear" any sonic material as approximations to or deviations from that tonal memory ingrained during childhood and adolescence, as much as conservatory-trained musicians will inevitably project the Western scale on non-Western tonal systems, detect flats and sharps where there are none, and hear nonmodal Angolan singing in terms of major and minor.

After my start in 1962 in the Mitukwe Mountains, I had the opportunity, over the years, of studying many more log xylophones, their tunings, and their music in the same cultural region that extends across northern Mozambique into Malawi. There are clues that can help one to discover the ideas behind this music. One is that the two performers, sitting opposite each other, play interlocking patterns from a relative beat standpoint, which allows them to create very fast music--after all, it is supposed to frighten the birds! However, watching the performers, one soon discovers that their songs are based on a central tone that is also produced in the interlocking manner, both players striking the same key with their right-hand beaters in fast alternation, each from his side. The effect created is that of a bourdon. Because the sounds emitted are of short duration, they are sustained by that method. (16)

The second observation one can make is that there are no root progressions; the bourdon is the unique permanent reference tone. And the third observation is that any note on the xylophone can be played together with any of its other notes; it always sounds agreeable. To Western-trained musicians, this observation may seem stunning, but having gotten this far, we have almost cracked the code of the tonal system of mangwilo, mangolongondo, and similar six- to nine-note log xylophones in this region. There is actually only one music system conceivable in which such extravagance could be allowed and in which "all have won," as Alice in Wonderland would say: a scale that is derived from higher partials over a nonobjectified single fundamental; nonobjectified because it is so deep down that it cannot be used as part of a scale. But why can any note on such a xylophone be struck together with any other? Why is there no distinction between consonance and dissonance, a concept familiar from other African culture areas, where a local tutor will inevitably point out to a student which sounds do not go together and which do, relationships often symbolized by terms such as "boy and girl," "father and mother," and so on. (17)

In African music that is based on a bourdon-like basic reference tone or, alternatively, riffs filling the cyclic structure of a song, I found that the underlying tonal system most likely incorporates the exploitation of higher partials over a single fundamental. The area of partials exploited usually includes the seventh and ninth harmonics but may also extend up to the eleventh, more rarely beyond. In these style areas, songs and instrumental parts can be constructed strictly on the unison/octave principle, or they can allow for chord clusters, in which any notes of the tonal scale can be sounded simultaneously. The result is unlimited freedom in harmonic counterpoint: any note of such a scale goes together with any other note. It never sounds disagreeable. For a long time during fieldwork, I tried to figure out why that should be so. The solution that finally emerged was this: neighboring sounds in a partials-derived scale, if struck together with emphasis, reinforce their common fundamental through the phenomenon of the so-called difference tone. If one strikes those two notes together rather hard, one can hear it. Boogie-woogie pianist Jimmy Yancey most certainly made use of the idea of difference tones, says Alexander Opatrny (2002), one of my graduate students, who observed this phenomenon of "sonic illusion" in Yancey's music independently, unaware of southeast African xylophone styles. Mathematically, a difference tone is defined as the vibration number (in Hertz) obtained if one subtracts from each other the Hertz (c.p.s.) values of two notes struck together. It remains to be analyzed (by Opatrny) which notes Yancey preferably uses to create this effect. That should unveil the deepest level of his tonal-harmonic ideas (see The Unissued 1951 Yancey Wire Recordings).

Gradually, I discovered how musicians were tuning their log xylophones "listening to the wood." If such an instrument is played by three people, one striking a time-line pattern, as in Waisoni Msusa's group, fascinating sound clusters emerge. (18) The harmonic sense of these musicians and their acute perception of higher partials put them on a par with jazz musicians of the 1940s who were using increasingly upward extensions of chords. It is clear that in jazz, particularly bebop, some of these extended chords represent clusters of partials, as do the notes struck simultaneously on a mangwilo. What kind of "cultural memory" could be responsible for that? Recent influences can be excluded because the first recordings ever made in northern Mozambique were cut in 1957 (by Portuguese ethnomusicologist Margot Dias). It is clear that African-American musicians in the 1940s who were beginning to use such cluster chords in jazz were following--consciously or unconsciously--a desire to detach themselves from Western diatonicism and try something else. But why did they take a path analogous to that which xylophone players in southeast Africa had been pursuing for so long?

Jazz of the 1940s was intimately rooted in blues tonal ideas. More than a third of Charlie Parker's compositions were in the blues form. (19) Blues was still a living force in the United States during that period, and with it, the blues tonal system. The latter was the subject of much research during the 1990s. With regard to the origin of the so-called blue notes--those elements in the tonal structure of blues and jazz that deviate most prominently from the Western diatonic scale--Benjamin V. Boone (1994, 8) of the University of South Carolina has argued:

   Could it be that blue notes evolved from the interaction inherent
   in processing the English language through a West African
   linguistic template? Could the creators of the blues with a
   heightened awareness of pitch due to their native language
   traditions ... have intuitively utilized pitch relationships
   from the upper harmonics of the overtone series? This series has
   as its seventh, eleventh and nineteenth members none other than
   the blue notes mentioned above. Furthermore, the blue notes are
   generally played out-of-tune when compared to the Western tempered
   scale, but in-tune with the harmonic series. In fact, the pitches
   used in a blues more accurately and more completely represent the
   first nineteen harmonics of the overtone series than do the
   pitches of the Western major scale.

Boone then goes on to discuss the neurology of pitch recognition, citing the work of Diana Deutsch and others. Although much of Boone's findings should be given serious attention, it is not really necessary to call in the remote nineteenth harmonic to explain the lower blue note, usually transcribed as e[flat] in the key of C. It is also somewhat tenuous to see blue notes as phenomena of acculturation, as Boone suggests, apart from the fact that the idea of "a West African linguistic template" is too generalizing, in view of that region's linguistic diversity, to be acceptable.

And yet, Boone is not really off track. Since my early work in the Mitukwe Mountains of Mozambique, I have had no doubt that higher partials can provide an auditory grid in the formation of a tonal system. But much cogitation is recommended, and much experimental research would have to be carried out to ascertain that tunings exist anywhere in the world's musical cultures that would incorporate a tone representing harmonics as high as partial 15.

Concentrating on West African savanna cultures, it was possible to reconfirm that a majority of the traits of the blues can be traced to areas in West Africa that represent a contact zone between an ancient Nigritic culture world developed by millet agriculturalists and an Arabic/Islamic culture world that had become effective from circa 700 A.D. I found that some of the people in remote central Sudanic mountainous areas--for example, the Adamawa plateau on the Nigeria/Cameroon border (the Kutin, Chamba, Zanganyi, and others that I researched in 1963)--had pentatonic systems that derived from the formants of human speech with an awareness of harmonics up to partial 9.

Eventually, I proposed a simple theory of how the blues tonal system was structured in its beginnings and where it clashes with the Western diatonic scale (Kubik 1999a). There is little doubt concerning the nature of the upper blue note (written as b[flat]) that it represents partial 7 over a single fundamental. But already in West Africa, notably among savanna agriculturalists such as the Kutin of northeastern Nigeria, I detected clues that the section from partial 6 to 9 (range of a fifth) can be used in combination with its own transposition a fifth downward to form a scale, as if the same melody were first articulated by a woman, then by a man. Thus, the sequence d-c-b[flat]-g (from high to low) is continued in transposition: g-f-e[flat]-c. However, since the first section is based on fundamental c, the second on f, the tones b[flat] and e[flat] both represent partial 7, over c and f, respectively. Thus, we do not have to go beyond partial 7 to explain the basic blue notes. That is one advantage of my theory. Each section has its own flavor brought about by a mild auditory awareness of its lower partials: 5, 4, and so on. The combination of the two complete columns, however, must create an interference pattern, with the greatest interference in one place of the combination that eventually crystallized into the lower blue note [+ or -] e[flat]. Its oscillating nature results from the fact that partial 7 of the lower scale (over f) at 969 cents interferes with partial 5 of the higher scale, e at 386 cents (over c). It deviates from the latter by a value of 119 cents (386-267), an audible magnitude.

Earlier, I pointed out two instances of interferences between e and e[flat] in tunings recorded among the Tikar of Cameroon (by myself) and in Burkina Faso (by Andreas Szabo), although a complete integration of two partials-based columns, one over fundamental c, one over fundamental f, has not yet been discovered in any rural area of West Africa. I believe that this particular readaptation of the tonal heritage occurred in America, under the impact of the diatonic system. In other words, African Americans of the nineteenth century with a memory of tonal patterns from the West African savanna region reinterpreted Western tonic and subdominant chords as partials-based columns over two fundamentals, c and f. Therefore, they could be easily extended to include tones representing partials 7, 8, and 9, creating the familiar seventh and ninth chords of the blues. The added sixth was also no problem, as one might guess from Mamadou's bala tuning (Ex. 4). The reinterpretation then resulted in a preference for tonic-to-subdominant shifts that are taken comfortably in the blues, whereas the Western dominant chord remains the "odd man out." There was no model in West African savanna tonality for its accommodation in the emerging blues tonal system.

In some blues, there is a third blue note called a flatted fifth. Actually it is more accurately described as being within the pitch area of a fourth that is augmented. This tone, represented as f[sharp] in the key of C, may have its origin in African tonal traditions than incorporate a projection of partial 11 at 551 cents, although I would not generalize (see Kubik 1999a). Ongoing research on the "flatted fifth" in blues focuses on how this note functions melodically and harmonically in a large sample of recordings. What is certain so far is that partials-based tonal ideas from eighteenth-century West African savanna cultures, and possibly southeast Africa from where people were deported to the United States clandestinely during the final stages of the slave trade, have survived and been perpetuated in the United States. They have decisively influenced tonal-harmonic developments in blues and jazz.

The Concept of Flexible Pitch Areas

David Evans (1982, 24) suggested that blues musicians proceed from an awareness of flexible pitch areas. With reference to a theory--now obsolete--that the blue notes were "neutral thirds" just halfway between e and e[flat], b and b[flat], he wrote:

   In fact, "neutral" probably would best represent an area between
   major and minor where notes can be sung, rather than any specific
   point between them. Blues singers often waver at the third or
   seventh or glide from a lower to a slightly higher pitch. The
   lower part of the third and seventh areas tends to serve as a
   leading tone respectively to the tonic and fifth below, the upper
   part as a leading tone to the fifth and tonic above.

Evans' concept is compatible with the interference margin of 119 cents in my model of the origin of the blues tonal system. But Evans' concept can be extended to jazz, and most certainly the pitch area concept is present in several African musical cultures, although its specific origins and motivations can probably only be defined within each tradition.

Drawing upon my observations in Angola, in the Central African Republic, in Nigeria (when studying alo story songs) and elsewhere, I coined the term elastic scales with reference to certain tonal systems in Africa. I had observed in Zande harp music in 1964 that there was considerable variation in the tunings from musician to musician and even from day to day in the same musician. For example, the highest note on a kundi harp, which I transcribed as an e, was sometimes flatted to the extent of approaching an e[flat] (Kubik 1994, 40, 101-105). To someone who has grown up in the jazz tradition, this gives the strong impression of a blues-like intonation, if the vocalist also follows it. Two Azande harp players I recorded as they were playing together, Bernard Guinahui and Francois Razia, even had the top note of their two harps tuned 58 cents apart, that is, more than a quarter tone. The harps were then played in unison, with those notes generating an enormous friction effect; but the musicians and their audience showed no sign of finding the 58-cents discrepancy uncomfortable. Questions I posed elicited the answer that the harps were "in tune"; so I recorded these tunings.

The harp tuning and singing of another of my harpist friends of 1964, Raymon Zoungakpio, also strikes listeners as "bluesy" because of the same flatted e[flat] tuning layout. (20) This impression is reinforced by the singer's delight in a descending harp passage (from a blue third, e[flat]-c, into an open fourth, d-a, resolved downward to c-g) and by his somewhat raspy voice. Any kind of knowledge of blues or jazz, or other African-American music, can be excluded in Raymon's case.

The concept of flexible pitch areas in African musical cultures cannot be interpreted with a single formula. There may be quite different causal links in different historical circumstances that favor or inhibit elasticity. Sometimes conflict between two or more tonal-harmonic traditions provokes reactions of auditory adjustment, and if the tonal ideas in contact are not too different, those attempts often result in ambiguity. Intonation a little higher or lower would then satisfy either system. But it may also be the nature of the language that allows pitch fluctuation. I observed in Yoruba-speaking communities of Nigeria, when I learned alo story songs, that it was possible to render the same song with different intervals in the overall pentatonic scale. In my contribution to the J. H. Kwabena Nketia festschrift (Kubik 1989a, 146-147), I give an example: two renderings of the song "Ero ti nr'Ojeje." Yoruba is an extremely tonal language, and in this musical culture it is therefore most important that the melodic ductus be maintained in representations of the same song so that the words are unambiguous in their meaning.

In other cultures, the use of a tuning temperament can be a factor in relaxing intonational rigidity. It is logical that in societies whose tonal materials are constructed from partials (no matter whether it is over a single or more than one fundamental), as among the Fan in Gabon and the !Kung' bushmen in Angola and Namibia, intonation will tend to follow accurately the harmonics, for example, of a musical bow. But if, in the course of history, someone introduces a system of temperament, as was the case in the Lower Zambezi valley among the -Sena, children's auditory memory will be conditioned by both: the "perfect" fourth and fifth relationships derived from partials and the near-equidistant tunings of instruments such as xylophones. In these cases, a margin of tolerance in pitch perception and reproduction will evolve. It can be relatively wide, as the example of Fainesi demonstrates. In her yodel song (Ex. 11), she arrived at a chromatic modulation; in a culture that had exposed her to equiheptatonic tunings from her childhood on, she reacted in her vocal adjustments with a compromise: while maintaining the key harmonic intervals of the Sena system, fifths and fourths in perfect shape, at approximately 700 and 500 cents, she adapted her tonal progressions through the four steps (degrees) of the system in a direction toward equiheptatonic tunings. But the memory of the perfect intervals being strong, she did not actually reproduce equiheptatonic progressions of 171 cents in her descending movements; instead, she arrived at semitone intervals and produced a perfect modulation.

This example should make clear that the processes of auditory adaptation and conditioning are subject to complex audio-psychological and cultural factors. I may point out here that jazz of the 1940s--besides the social conflicts well covered in the literature--also rose within audio-psychological conflict. Bebop musicians, particularly Dizzy Gillespie, Thelonius Monk, and Charlie Parker, found ingenious solutions to accomodate their personal tonal-harmonic understanding, which was strongly based on blues tonality plus several African-derived structural ideas, with a musical and cultural environment that had significantly changed since the 1930s. In their reactions to and adaptations of the expanded sound repertoire conquered on the instruments they played, they followed a path that was intertwined with, although not necessarily dependent on, contemporaneous development in European art-composed music.

It could have been predicted that by the 1940s, jazz would increasingly use "substitute" chords and alterations on the basis of inherited blues tonal-harmonic principles. Jazz reasserted itself as an African-American tradition, re-creating in America African extensions along structural principles still transmitted clandestinely on American soil, but with materials--Western instruments and their tunings, contemporary harmonic patterns, popular songs, and so on--that were available to be used as clay. Thus, the strategies that made bebop harmonic developments feasible were different in purpose and meaning from the experiments in harmonic, atonal, and serial development made by European and European-American composers in the early twentieth century. Equiheptatonic Concepts

In precolonial Africa, equiheptatonic tunings were developed in several distinct areas. One was in Mozambique among the -Chopi (see H. Tracey 1970) and in the Lower Zambezi valley. A characteristic recording with a remarkable ending on a "modern" dissonant chord can be heard in a performance by the then-famous valimba xylophone group of Suze recorded by us in 1967; see LPs Opeka Nyimbo in the discography. Equiheptatonic vocal intonation was common in the 1920s around Benin City, Nigeria, as is documented by the 78 rpm record "Ihore" (Amuyun, with vocal and instrumental accompaniment). It was present on the Gold Coast (now Ghana) and on the Ivory Coast, as demonstrated on the record Pondo Kakou, by Gilbert Rouget. Another prominent area with nonmodal harmonic singing and a tendency toward neutral-third intonation and equidistance was the east Angolan culture area, as is demonstrated by my recordings of 1965, published on the album Mukanda na Makisi.

In New World cultures, as is also the case in Africa today, factory-manufactured musical instruments tuned to the Western diatonic/chromatic scale have erased equiheptatonic tunings almost everywhere. In Africa, they have survived, as far as I know, in the xylophone styles of Mozambique, among the -Chopi and the -Sena. Already during the 1950s, xylophone tunings in many other parts of Africa, whatever their previous nature was, were changed to the European heptatonic scale, for example, in southern Cameroon, where a new merengue-based xylophone style became popular with the (now historical) Richard Band de Zoe Tele. The younger generation in formerly equiheptatonic areas, as elsewhere in Africa, has become diatonic in its auditory conditioning (particularly in the Congo and in East Africa) or was conditioned by the African-American reflux: jazz in South Africa, calypso on the Guinea coast, then soul across Africa, and more recently rap and blues in some places in Mali and Senegal. This is the third African generation under massive influence by the media. Auditory conditoning, tunings, scales, tonal relationships, intervals, and so on are matters of early acquisition during childhood and adolescence. Once learned, habitual auditory reaction patterns are projected upon any familar and unfamiliar music. This is why Western-conditioned individuals inevitably "hear" equiheptatonic music in terms of major/minor, although if warned, for example, in ethnomusicology classes, students may become aware of these facts, while their auditory reactions remain unchanged.

And yet, despite what has been the most incisive area of acculturation--because tonal systems are like languages, and a first language leaves behind much deeper engravings in human memory than those learned later--it is stunning that equiheptatonic concepts, besides several other African tonal-harmonic ideas, have survived in the Americas. Despite the massive exposure of African Americans in the nineteenth century to folk-music harmonic patterns in ballads, church songs, and other works as well as popular late nineteenth-century harmonic patterns adopted in the ragtime piano from waltzes, quadrilles, marches, and polkas, some other forms of aesthetic experience continued. These included an auditory sensibility to timbre sequences, spatial imagination based on equidistance in decorative art (Vlach 1978), an array of African-derived polyphonic devices, the call-and-response organization, cyclic form, and many other behavioral traits that survived in African-American communities. Why they survived is an issue that touches on enculturation, group psychology, and depth psychology. It is certain, however, that small-scale circumstances, such as intrafamily transmission and exposure to movement and sound experiences contradicting the mainstream culture, provided much of the humus. These innermost "secrets," "privacies" in behavior, were hard to identify; in any case, they were resilient, and as long as they did not threaten the larger society, society decided to ignore their existence. And yet it would happen that some of those "Africanisms"--considered unworthy by their carriers under pressure from the dominating society--would attract the attention of outsiders. In the nineteenth and early twentieth centures, it would often happen that young European Americans, in reaction to encrusted conventions of their own communities, set out to appropriate the "Negro" stuff--a trend that began with black-face minstrelsy and continued to the Original Dixieland Jazz Band to rock 'n' roll and beyond.

But there was also interest based on appreciation and a deep sense of identification with African-American culture. Bix Beiderbecke learned jazz to achieve personal transformation, and in his piano work, more than on the trumpet, he contributed something unknown before, reinterpreting Claude Debussy from a standpoint almost African American. "Africanisms" did not fail also to impress American composers who had grown up in the Western tradition but quickly adapted. Song composers mostly of Jewish ancestry made an undeniable contribution to jazz. They experimented with their perception of blues tonality. George Gershwin and others reacted to the new tonal material, reinterpreting it from their own background of musical upbringing; but thereby they also contributed to its reinforcement. Jazz would be poorer without "Body and Soul," "How High the Moon," "Lover Man," "I Got Rhythm," and other works by various Tin Pan Alley composers.

Where can we detect equitonal, and particularly equiheptatonic, retentions in jazz? First, we have to clarify what they involve. In African traditions, equiheptatonic tunings arose as a temperament aimed at facilitating melodic transposition across the keyboard of a xylophone, for example. In vocal styles, particularly in eastern Angola, it smoothed out step-by-step chordal progressions without the chords losing their identity. If one divides an octave into seven degrees, one has difficult choices and several possible solutions depending on how one has arrived at such a heptatonic scale in the first place. In eastern Angola, among the -Cokwe, -Luvale, -Mbwela, and -Lucazi, the remote origin of their heptatonicism seems to be in the harmonics that can be produced, for example, on the mouth-resonated friction bow kawayawaya, over two fundamentals that are a whole tone apart, by reinforcing the harmonics up to partial 6. This yields hexatonic melodic material, as shown earlier. The experience was probably extended to an awareness of partial 7 that is easily obtained on the nyavikali end-blown horns that are part of the vandumbu secret society's activities, symbolizing the voices of the dead kings. (21)

From the mouth-bow background with its hexatonic result came the triad-oriented harmonic style of the eastern Angolans. But as soon as a seventh note was added on instruments such as the mucapata lamellophone, possibly inspired by the horn's seventh partial (b[flat] = 969 cents), problems of scale and organization of simultaneous sound inevitably arose. The historical solution was temperature, without exclusion of simultaneous sound, as is demonstrated by the art of Saciteta (born 1895), whom I recorded in 1965. (22) Pie-Claude Ngumu (1975-1976) touched on the same problem with regard to the mendzan gourd-resonated xylophones of southern Cameroon. Here, the seventh tone was called esandi (spoil-sport). While the problems can be similar, the solutions can be different. In eastern Angola, the solution was to (1) maintain the lush chords, (2) expand the original shifting harmony by including bass lines that now often proceeded through four steps, so that it would be easier to accommodate a seventh tone, and (3) maintain overall movement by step. The result was that the clustered triads appearing on any of the four steps had to be corrected constantly by intonation to make them sound as identical as possible.

My impression is that in jazz of the 1940s, analogous problems of scale and harmony arose when motives, even entire themes, began to be transposed a degree downward to form composite themes, as occurs in Dizzy Gillespie's "Groovin' High" (on Dizzy Gillespie and His Sextets and Orchestra) and even in Woody Herman's "Early Autumn." A similar situation can occur melodically, by stepwise shifting, mostly downward, of a phrase in narrow intervals, as can be heard most beautifully in Gillespie's solo trumpet backing of Sarah Vaughan's performance of "Lover Man" with his band (May 11, 1945). Of course, like everyone else, jazz musicians of the time were still captives of the Western chromatic tuning on instruments made in the corporate factories. But the way they began to handle the system melodically and transpose phrases a whole tone or a semitone lower, often without altering their interval structure--dismissing Western harmonic rules--demonstrates the resurgence of a kind of approach to handling melodic and harmonic patterns that is familiar to those of us who have worked in African communities with nonmodal, near-equiheptatonic melodic behavior.