South Africa , Republic of.

(ii) Main musical features.

1. Nguni music.

Nguni is the name applied collectively to the Zulu, Swazi and Xhosa peoples of south-eastern Africa, the largest indigenous group in the country. Their languages and cultures are closely related, and their traditional music is more vocal than instrumental, polyphonic dance-songs being particularly important. About 7,000,000 Nguni live in the Republic of South Africa and in Swaziland; offshoots, who emigrated early in the 19th century, are the Ndebele of Zimbabwe and the Ngoni of Malawi and Zambia (see Malawi, §1).

(i) General background.

(iii) Musical instruments.

(iv) Tonality.

(v) Music and society.

(vi) Western influence and modern developments.

(i) General background.

People resembling the Xhosa who live in the Eastern Cape province and Transkei were encountered there in the 16th century by shipwrecked Portuguese seamen. Early in the 19th century, the military leader Shaka amalgamated various Nguni clans to form the powerful Zulu nation in Natal and kwaZulu. At about the same time, the Swazi nation became established in Swaziland. Nguni languages belong within the Bantu language family, but they show certain features adopted from the neighbouring Khoikhoi (or Hottentots, now almost extinct), most notably in their use of three ‘click’ consonants, written as ‘c’, ‘q’ and ‘x’. As with other Bantu languages, speech-tones influence the shape of vocal melody. A characteristic of the Nguni that is rare elsewhere in Africa (but present in Chinese and German) is the pitch-lowering effect of voiced consonants, which in song often produces rising on-glides.

As with other southern Bantu peoples, the traditional economy of the Nguni is composite; it comprises cattle-rearing, the monopoly of men and boys, and agriculture, which is women’s work. Men used also to do a certain amount of hunting. Since the early 19th century, with the advent of missionaries and settlers, the Nguni have increasingly come under Western influence. Indigenous culture survives only sporadically in some of the remoter rural areas.

(ii) Main musical features.

Strange as it may seem for an African people, the Nguni have no history of drums or percussion ensembles as a basis for their communal dancing. Dancers always sang their own dance music, and although ankle rattles and hand-clapping were sometimes added, the basis of their collective music-making was the unaccompanied dance-song. War-shields were sometimes used percussively by warriors in earlier days, and oxhides were beaten at Xhosa boys’ initiation ceremonies. Drums were not, however, entirely unknown. Medicine men sometimes used them, and a type of friction drum was employed at girls’ coming-of-age ceremonies among the Zulu. Improvised drums and wooden clappers are now used in certain neo-traditional art forms, such as modern Zulu ingoma dancing. Essentially, however, it is clear that in the past the Nguni have specialized in developing vocal polyphony rather than instrumental ensembles or rhythmic complexity.

A striking feature of traditional Nguni choral dance-songs is the principle of non-simultaneous entry of voice parts, and the intricacy of their polyphonic interaction. There are always at least two voice parts with different starting-points; their phrases frequently overlap, but there is usually no common cadence point where the parts achieve a combined resolution. Instead, each voice returns to its starting-point as in a round (though the parts are not identical), and the process is continually repeated. Variations commonly occur in the leading voice part, while the chorus maintains a constant ostinato. A very simple two-part illustration (without variations in the leading voice part) is provided in ex.1; this shows a work-song sung by a trench-digging team in Smith Street, Durban, in 1964. The alignment of the interacting parts is fundamental in such music. This concept is felt so deeply that an individual singer, if asked to demonstrate a traditional choral song, will not merely render a single voice part but will always attempt to present the essentials of at least two parts, the leader and the chorus, by jumping from one to the other whenever a new phrase entry is due.

(iii) Musical instruments.

The most comprehensive survey of traditional Nguni instruments was conducted by Percival Kirby in 1934. Although instruments played little or no part in the traditional communal music of the Nguni, they were certainly used in individual music-making. Flutes and musical bows of several different types were formerly very common. Gourd-resonated musical bows were used for self-accompaniment in solo singing, where the instrument assumed the role of a chorus by supplying an ostinato against which the singer improvised an offset leading part. Many surviving choral songs are said to have been composed in this way. Individual music-making was conceived chiefly as a form of self-expression, not as entertainment for an audience. This concept still persists in towns among rural migrants who play guitars, concertinas or harmonicas for personal enjoyment while walking in the street. To some extent, traditional Nguni musical principles and stylistic features are effectively expressed through such non-African instruments.

Although not strictly musical in function, side-blown animal horns were used by men, mainly for signalling. The Zulu also occasionally used an end-blown bamboo trumpet with oxhorn bell, yielding two or more notes. Several types of small whistle were used, mainly in hunting and in doctoring. Men and boys played flutes, which were mainly associated with cattle-herding. The Zulu umtshingo, a long, obliquely held flute without finger-holes, was sounded by shaping the tongue to serve as an air channel. The 4th to 12th partials of the harmonic series were produced through overblowing and by alternately stopping and unstopping the end with a finger. The making and playing of these flutes, and of the smaller igemfe, was formerly forbidden among the Zulu until the time of the annual umkhosi, the festival of the first fruits. The igemfe was used for duet playing, two flutes being tuned about a semitone apart.

Several types of mouth-resonated musical bow were used for solo playing, but these are now rarely found. They include the Xhosa inkinge, and the Zulu isithontolo shown in fig.2. The stave is held against the mouth and the string plucked with a finger or plectrum or, in the case of one type of instrument, sounded by means of a friction-stick. To produce a melody, different harmonics, usually the 3rd to 6th partials from two or three fundamentals, are selectively resonated by varying the shape of the mouth as in playing the jew’s harp. The commercial jew’s harp has in fact become popular as a substitute for the bow. Another variety of mouth bow, the ugwala or unkwindi, was a ‘stringed wind’ instrument (apparently derived from the Gora of the Khoikhoi), sounded by blowing on a piece of quill that connected the string to the stave.

In earlier days, the classical instrument for self-accompaniment in solo singing was a gourd-resonated bow, the Zulu ugubhu, Swazi ligubhu or Xhosa uhadi (fig.3a). This is a large musical bow, about 1·5 metres long, with a gourd-resonator attached near the lower end, and a single undivided string struck with a piece of thatching grass. The instrument is held vertically in front of the player, so that the circular hole in the gourd faces his left breast or shoulder and can be moved closer or farther away for the selective resonation of harmonics, usually 2nd to 5th partials. Besides the fundamental note yielded by the open string, a second note is obtained by pinching the string near its lower end between the left thumbnail and forefinger, as shown in fig.3a, the remaining three fingers gripping the stave. The interval between the open and stopped notes produced by Xhosa players is usually roughly a whole tone; the outstanding Zulu musician, Princess Constance Magogo kaDinuzulu, uses a semitone varying from 90 to 150 cents on different occasions; both sizes of interval have been noted among Swazi players. Selectively resonated harmonics from the two fundamentals, though relatively faint, are used melodically as a vocal accompaniment. The resultant hexatonic scales obtained from whole-tone and semitone stopping are shown in ex.2; though the open-string fundamental is shown as C, the tuning is often as much as a 5th lower, and the entire series is transposed accordingly.

A second type of gourd bow, the Zulu umakhweyana and the Swazi makhweyane, reputedly borrowed from the Tsonga people of Mozambique early in the 19th century, largely displaced the Zulu ugubhu and the Swazi ligubhu but was not adopted by the Xhosa. This instrument, shown in fig.3b, differs from the earlier type in that the gourd-resonator is slightly smaller and mounted near the centre of the stave instead of at the bottom. In addition, the string is tied back by a wire loop or brace attached to the resonator, so that two open notes are obtainable, one from each segment of the string. These notes are tuned anything from a whole tone to a minor 3rd apart, and a third fundamental, usually a semitone higher, can be produced by stopping with a knuckle the lower segment of the string below the restraining loop. This stopped note has a duller sound, however, and is not always used. Selectively resonated harmonics are used melodically in the same way as on unbraced gourd bows. The notes available from the braced gourd bow, when the two segments of the string are tuned a whole tone apart, are shown in ex.3. Some players may transpose the entire series as much as a minor 3rd higher.

(iv) Tonality.

There is considerable diversity in the scale systems used by different Nguni peoples and also within those used by single language communities. Broadly speaking, perfect 4ths and 5ths appear to be important structural intervals. A few ancient Zulu dance-songs have only three notes, in the descending sequence C–G–F, with the octaves of one or more of these notes. But some apparently older Zulu songs use pentatonic and hexatonic modes containing two semitone intervals. In these, the tonality and chord structure appear to be based on two contrasting triads with roots roughly a semitone apart; this is the same interval that occurs between the roots produced on the ugubhu bow. Both in ugubhu bow songs and in many Zulu and Swazi choral songs, descending hexatonic modes with the notes A–G–F–E–D–C are to be found, based on the contrasting triads C–E–G and D–F–A. The two triads, based on the roots C and D, are used in a contrasting manner. In many songs, the middle note of one triad is omitted, resulting in a pentatonic mode with two semitone intervals. The Swazi most commonly omit the E, and the Zulu often omit the F, while sometimes rendering the E as E. The resultant A–E–D structure then resembles the ancient Zulu C–G–F scale.

The Xhosa, and also Zulu-speakers in southern Natal, most frequently use whole-tone root progressions as typified in the C and D roots of the uhadi bow. Descending hexatonic modes comprising notes from the C and D triads are very common, as in A–G–F–E–D–C–(A); the F may be omitted, resulting in the common pentatonic. However, in the latter case, root progressions between C and D still function, and Nguni pentatonism therefore differs from the purely melodic use of the pentatonic found in many other parts of the world. Major and minor 3rds and 6ths quite often occur as chordal intervals in Nguni polyphonic songs, in addition to the more common 4ths, 5ths and octaves (or unisons), though the former are more transitory than the perfect intervals. But no functional hierarchy of discords and concords seems to operate consistently. Owing to offset phrasing between the voices, there is usually no collective resolution or cadence; instead, the artistic intention is possibly to maintain an ever-changing balance between the constituents, through chordal contrast as well as by other means.

(v) Music and society.

Where traditional ways of life remain relatively intact among the Nguni, music plays an important role for the individual and the community. There are songs for different age-groups, related to different activities and occasions. Many songs are directly functional, either regulating physical actions, as in dancing or a collective task, or being educative by regulating behaviour; they may express group ideas or popular or personal opinion, they may be critical of authority (permissible in song), or they may serve as an essential constituent of a ceremony or social event.

Outstanding composers or performers are admired, but there are no professional musicians. Nearest to being a professional in earlier times was the imbongi, the court praise-poet who recited the praises of the king or chief and his ancestors at important functions. Though izibongo, Zulu praise-poetry, calls for a style of delivery that has melodic features, it is regarded as an art form in its own right and does not fall within the category of vocal music. The traditional Zulu word for singing is ukuhlabelela, from the verb hlabelela, ‘to sing’. However, this term does not exactly match the Western concept of singing; besides excluding izibongo, which to Western ears often resembles a form of praise-song, it includes a form of ‘choral recitation’. This occurs, for example, in several versions of the Zulu recreational isigekle dance. An exaggerated ‘sing-song’ rise and fall of pitch, without exact musical notes, is used, but there is a regular metre, and this seems to be a more important criterion for defining ukuhlabelela than the melodic use of fixed pitch values. Vocal phrasing in Nguni songs often flouts a regular metre, rather than expressing it directly, and word-stresses frequently do not coincide with the physical downbeat of the dance-step or other movements. Consequently, it can be entirely misleading to analyse songs without taking accompanying physical movements into account.

In traditional Nguni society, choral dance music provides the essential basis of orderly social interaction at important ceremonies. Rhythm is always given physical expression through simultaneous actions by the singers themselves, in the form of dance-steps, gestures or the wielding of real or symbolic weapons, implements or regalia. These actions are normally considered inseparable from the music; music and movement are blended to produce an ultimate form of expression involving the complete human being interacting with others of his group. The performance may also be felt to be inseparable from the context of a particular ceremony, and the ceremony to be essentially a part of some sacred or seasonal event, like the impressive royal annual incwala ceremony of Swaziland. Certain incwala dance-songs are forbidden at any other time.

The corresponding Zulu royal ceremony of the first fruits, the umkhosi or ukweshwama, has been re-activated, and British and Zulu military history is in the process of being thoroughly revised. The Swazi incwala ceremony appears to be a mass dramatization of national solidarity under their priest-king, the Ingwenyama or Lion, although the participants are grouped separately according to lineage, regiment, age and sex. An onlooker gains the impression of a vast ‘real-life’ opera or dramatic pageant, for which no audience is intended. The solemn dance-songs are essentially a performer’s art form, a means of collective expression, with national and religious motivation. Their full appreciation requires not a passive audience but direct experience that can be gained only through active participation.

In contrast to the overall solidarity demonstrated in national ceremonies, there is often a strong element of rivalry at smaller social gatherings, parties and weddings, reflecting group differentiation on the basis of locality, family, age or sex. The central feature of a wedding, held at the bridegroom’s home, is an elaborate programme of dances. As if expressing artistically the essential two-family contractual basis of marriage, the bride’s party and that of the groom dance in turn, quite separately, each group seeking to outdo the other and to assert themselves as distinctive and worthy of social recognition.

(vi) Western influence and modern developments.

Through European contact during the past century and a half, many Western musical elements and ideas have been adopted by the Nguni. Traditional instruments are almost extinct, surviving only in some of the remoter rural areas. Traditional Nguni folk music survives only where social life retains a traditional basis. For the past century or more, missionaries and teachers have greatly influenced musical taste. A Zulu hymnbook with European tunes was printed in 1862. The first Xhosa songbook, Amaculo aseLovedale (‘Songs of Lovedale’), appeared in 1884. The tonic sol-fa system was widely taught, and traditional music was increasingly displaced by Western choral music, sacred and secular. Educated Africans also began composing pieces for four-part choir with vernacular words. Outstanding Xhosa pioneers in this field were the Rev. John Knox Bokwe (1855–1922) and Benjamin John Peter Tyamzashe (b 1890); and among the Zulu, Reuben Tholakele Caluza (b 1895) and Alfred Assegai Kumalo (1879–1966). A well-known Xhosa hymn, Nkosi Sikelel’ iAfrika (‘God bless Africa’), composed in 1897 by the Rev. Enoch Sontonga, is now the national anthem of South Africa; its tune was also adopted for the national anthems of Tanzania and Zambia.

Several Xhosa composers have excelled in musical comedy. Todd Thozamile Matshikiza (1921–68) composed the music for two stage productions, King Kong (libretto by H. Bloom) and Mkhumbane (libretto by Alan Paton). After a successful tour in South Africa, King Kong was staged in London in 1960–61. Since that time, the production of African musico-dramatic presentations (mostly with English words) has been growing steadily. A notable artist in this field is Gibson Mtutuzeli Kente (b 1932), who combines the roles of playwright, composer, director and producer. His first four musical plays were Lifa, Manana the Jazz Prophet, Sikalo and Zwi.

Nguni experimentation with Western musical forms began in the late 1600s with the first settlement of Europeans in the Cape. By the 1860s, the mining and manufacturing industries in South Africa had begun to draw heavily on a system of migrant labour, absorbing many thousands of men into emerging urban centres. Muff Andersson (p.38) suggests that by 1914 gramophones and records were sold from ox-carts by the MacKay brothers, agents of the London-based company His Master’s Voice, along with European instruments such as the guitar, concertina, violin and piano. Hybrid musical styles began to emerge in urban shebeens (illegal drinking houses), migrant worker hostels and on the streets, exhibiting a creative blend of traditional forms and newly adopted instrumentations and styles. The steady movement of people to and from the urban centres led to the dissolution of distinct rural and urban social, economic and cultural characteristics. Musical performances thus reflect networks of production and reproduction spanning town and countryside (Coplan, 372).

Although there is a diversity of Swazi and Zhosa performing practices, researchers and the music industry have documented and recorded Zulu music more than that of any other South African cultural group. Images of Zulu have circulated throughout the world since the British encountered Zulu armies in battle in the 1880s, and this may account for the historical predominance of Zulu-related research and entertainment interest.

Although the guitar was most likely introduced to the KwaZulu Natal region as a trade item by Portuguese or Arab sailors as early as the 16th century (Kaye, 351), it became associated with migrancy and an urban identity in the early 1900s. Zulu guitarists localized their instruments by returning them, transposing onto them a rapid plucking style, cyclical form and contrasting whole-tone progressions of traditional umakhweyana gourd bow performing practice (Rycroft, 1977). The guitar came to be associated with the wandering solo performer known as a Maskanda, a word derived from the Afrikaans word musikant which means ‘musician’. Maskanda also refers to a particular musical structure. The first maskanda musician to record commercially in the 1960s was John Bhengu, a highly skilled Zulu guitarist renowned for his ukupika (plucking) style (Allingham, p.385). Since solo acoustic guitar playing was considered to have limited audience appeal, Bhengu switched from acoustic to electric guitar and recorded with full band and backing vocals. He was marketed under the name Phuzushukela (‘sugar drinker’), and his upbeat dance style set the tone for many electric maskanda bands. Although maskanda continues to be associated with a rural Zulu identity, it has also become highly commercial, with radio and television programmes devoted to the genre, nationwide competitions, and the recent development of women professional performers.

Isicathamiya, a male a cappella musical genre developed by Zulu migrant workers at the turn of the 20th century, along with the elaborate network of weekly competitions that helps to define the genre, provides a space within which black South Africans have been able to reflect and act on their fractured world (Erlmann, p.10).

Zulu ingoma dances, once associated with ‘tribal’ dance competitions on the Witwatersrand gold mines, and linked to the political history and ideology of the apartheid government, are now fostered in the schools, and performed at meetings of independent trade unions and important state functions. Ingoma dance troupes perform at weekends in competitions organized at the hostels and mines, and in staged musicals such as the revived IpNtombi. The post-1994 democratic South Africa has inspired the celebration of ethnic identities and cultural roots and has led to the re-emergence of traditional performing practices, values and beliefs, providing the basis for ethnically based social programmes and political parties (Meintjes, p.9).

2. Sotho/Tswana music.

A primary difference between Sotho and Tswana music and that of Nguni peoples of South Africa and Swaziland is the inclusion of elements of praise in vocal music. It has been suggested that the quicker rhythms and tempos of the Sotho and Tswana language groups, when compared to Nguni types (i.e. siSwati and Zulu), contributes to the ease in performance of rapid, recitative-like praise-songs that recount historical events or extol the lives of famous individuals or families. Such praise-singing is sometimes accompanied by choral singing in the background, but more often it interrupts the choral singing (Huskisson, 1982, 374).

(i) Sotho music.

Contemporary Sotho music reflects the continuing modernization and development taking place in southern Africa. The co-extensive nature of old and new in musical performance testifies to the legitimacy of studying both traditional and newer musical aspects of Sotho expressive culture.

Sotho vocal music is essentially pentatonic and performed in call-and-response form. The response performed by a group usually remains static, while the call of an individual is flexible; leaders are often selected for the ability to manipulate words. Older styles of vocal performance included Sotho men performing mohobelo songs and mokorotlo songs, which include praising. Men often perform such songs, which are associated with regiments in khaki pants and white shoes, forming a ‘long L-shaped line, stamping their feet in periodic rhythmic emphasis, at the same time bringing the knobkerry sticks, held aloft in the dance, down to the level of the body’ (Huskisson, 375).

Praise-singing is also an aspect of Sotho women's songs, such as the mokgibo knee dance-song. Women kneel on the ground in a semicircle during these call-and-response songs and perform an elaborate dance with movements in which the knees are raised and lowered. Ululation and hand-clapping accompany the singing. Another type of song is the male diphotha, a step-dance performed in gumboots that are struck together and slapped with the hands in a synchronized rhythm to a series of ‘step’ movements. The diphotha is typically accompanied by a concertina.

Sotho girls perform call-and-response lialolo and metjekong dance-songs that accompany simple movements; many of the girls stand and clap, while others dance. Clapping in polyrhythms is a typical feature of the mokokopelo dance-song performance style of Sotho women. Sotho boys perform lengae initiation songs in a style characteristic of Sotho men; the songs rely on the production of a deep bass melodic line (Huskisson, 376).

(ii) Tswana music.

The Tswana occupy the region of the eastern Transvaal in South Africa and Botswana. Tswana vocal music (dipina) is classified according to its function within a specific social institution.

Music of Tswana boys' initiation rituals (moama) is one such category. According to Johnston, in bowera circumcision schools, initiates spend a considerable amount of time memorizing songs under the supervision of a musician specialist (nake) (Johnston, 890). Initiates also undergo ‘hazing’ during dances such as the secho whipping dance. Completion of the circumcision is marked by a ceremonial procession (thalalagae) before the Tswana boys return to their village.

Young Tswana girls participate in a boyale puberty school, where music and dance also play an integral role. One prominent dance is radikgaratlane, in which a women is masked and disguised as a god, wearing clay horns, a symbol of virility. Upon the completion of the all-night thojane, a ceremonial dance, each young girl is received as ready for marriage and greeted upon return to her family with special megolokwane songs of return.

Music is also featured at the traditional wedding feast, where the setapa dance is performed by guests. In addition, exorcists frequently require their audiences to sing and clap during curing ceremonies. Burial ceremonies (magoga) also involve music, as do go rapelela metsi rain rituals.

Children's vocal music includes pinapalo counting rhymes, tlhaletso nursery rhymes, tshameko ya pina singing games and tsirimanya jingles. Adult work-songs and beer-drinking songs also contribute to the category of Tswana vocal music.

Tswana communal vocal music is often related to the annual agricultural cycle; the post-harvest season usually involves more musical activity, such as the performances of songs and dances during beer drinking. These seasons are outlined by Johnston as follows:Letlhafula (autumn): a time for hoeing songs and work-party songs
Mariga (winter): a time for children's fireside story-songs
Dikgakologo (spring): a time when women and children chase birds from crops by singing lustily in the fields
Selemo (summer): a time for beer-brewing, beer-songs and beer-dances (Johnston, 891)

Tswana songs may be referred to as ‘folklore of great importance’ (mainane a segologolo), or they may be composed by a known composer (motlhami). They may be sung in two or more parts or in unison; in typical call-and-response singing, the response might involve multi-part singing. When singing in two or more parts occurs, the upper part is referred to as sgalodimo and the lower part as segalo. Tswana communal vocal music is primarily pentatonic, with clearly divisive symmetrical metres, relying primarily on the octave, 5th and 4th for harmony. Vocal music is closely associated with the rhythmic movement of the body.

Perhaps the best-known form of Tswana instrumental music is the reed-pipe dance (kubina dithlaka). According to Ballantine, the Tswana do not think of a melody first when they play. Rather, they begin with one or two rhythmic schemes, which are played on the reed-pipes in polyrhythms, ‘but their interrelation and points of coincidence are such that, apart from occasional exceptions, the canons of “harmonic” acceptability are not offended’ (Ballantine, p.55).

The Tswana moropa drum, Lepapata antelope horn and mathlo leg-rattles often accompany dancing, as do a variety of whistles named for their method of construction.

There are four types of musical bows used by the Tswana for accompanimental purposes according to Johnston: lengope, segwana, setinkane and nokukwane (Johnston, pp.891–2). The lengope is a mouth-resonated bow constructed from curved cane and strung with nylon fishing cord. The segwana is larger, with a calabash functioning as resonator. It is struck with a stick and has a cord divided in such a way that two tones a minor 3rd apart are produced. The opening of the calabash may be pressed against the performer's chest to adjust the tone quality. The setinkane is similar to the segwana, however without a resonator. The nokukwane uses a crudely fashioned bow with an arc more pronounced than those of the other bows. It is also struck with a stick and is resonate with a blown up and dried skin milk container (lukuku).

(iv) Modern developments.

3. Venda music.

The Venda, who have lived in and around the Soutpansberg mountains in the Northern province of South Africa for many centuries, have a culture that distinguishes them from other Bantu-speaking people in South Africa and a language that is classed on its own, though it has some affinities with Sotho and Karanga. The Venda were originally shifting cultivators and hunters but later adopted a more settled economy; they also took to keeping cattle as well as goats. They used to live in large villages that were often on mountain slopes and difficult to reach, and every village was administered by a chief or headman and his council. In the first part of the 20th century, people tended to move away from the villages of their rulers and live in homesteads scattered over the hills and mountains, but in many areas they are now being regrouped into villages.

In the past, the music of the Venda was a part of the oral tradition and emerged only in response to the demands of corporate activity. The evidence of an intensive study of musical activity in the Sibasa district between 1956 and 1958 suggests that all Venda children are able to sing and dance as well and as creatively as they speak their language and that subsequent developments of their musical interests and aptitudes are a consequence more of sociological than of psychological or biological factors.

(i) Musical concepts.

(ii) Music and society.

(iii) Musical structure.

(i) Musical concepts.

The term ‘nyimbo dza Vhavenda’ (‘songs of the Venda-speaking people’) includes all tunes that are sung or played on instruments, as well as patterns of words that are recited to a regular metre. It is rhythm, therefore, that distinguishes u imba (singing), from u amba (talking), from u renda (reciting praises) and from u anetshela (narrating). But although it may have no rhythm and is sometimes called u tavha mukosi (‘raising the alarm with a long, loud yell’), a single note blown on a stopped pipe or horn comes into the Venda category of music because the performer ‘makes the instrument cry’. Musical instruments are thus known as zwilidzo (things that are made to cry).

A soloist ‘plants’ (-sima) his song, and the chorus ‘thunders in response’ (-bvumela). A maluselo (dance-leader) shows the step (-sumbedza mulenzhe: ‘shows a leg’), and others ‘pour it out’ (-shela mulenzhe) after him. Great importance is attached to teamwork in dancing, and the verb u tshina (to dance) generally refers to communal dancing, in which all follow the same steps, as distinct from u gaya (to dance a solo). Other more individual styles are u tanga (to dance in a stately fashion), as old women and important people do on special occasions; u pembela (to dance excitedly) especially at the end of an initiation school or the installation of a chief; u thaga (to dance ndayo) at the vhusha girls’ initiation school; and u dabela (to dance independently of and often in the opposite direction to members of an initiation school), as a sign that one has graduated. Most Venda communal dances are basically circular and counter-clockwise: the dancers ‘go round’ (mona) and make ‘a cattle kraal’ (danga).

Singers can indicate the metric patterns of songs by clapping their hands, and they can sing either the solo or the chorus part alone and know exactly where to come in. They do not isolate patterns, nor do they seem to appreciate that there are repetitions of a pattern. People refer to the correct melody or rhythm of a song as kuimbele (the way in which it is sung) or kulidzele (the way in which it is played). Mistakes in performance are recognized, though critics rarely state precisely what is wrong; they know that it does not sound right and are able to correct the mistake by demonstration and argument. Although there is a distinction between ‘hurrying’ (-tavhanya) or ‘delaying’ (-lenga), the tempo during a performance, the tempos of the tshigombela and tshikona dances are not classified respectively as fast or slow: they are ‘different’ and ‘go in opposite directions’ (-fhambana). Time signatures and note values are not recognized, though the word -kokodza (to drag, pull) describes a note that is prolonged, especially at the end of a song.

Because music is conceived as repetitions of basic patterns, there can be no concept of rests in performance, since a rest would immediately destroy the special world of time that music is meant to create. Thus in ex.4, a children’s song, the metric beat does not fall on the syllables -du- tsha and nga-, which are stressed in performance. If people clap to the song, they clap on the syllables Tshi-, -la, -si and -di, so that there is not a rest on the fourth beat, but a total pattern of four beats. Venda music is not founded on melody or on metre, but on a rhythmical stirring of the whole body, of which singing and metre are extensions. When a rest is heard between two drumbeats, it must be understood that for the player it is not a rest; each drumbeat is part of a total body movement in which the hand or a stick strikes the drumskin.

The words -tuku (small, young) and -hulwane (important, senior) are generally used to refer respectively to tones that are high and low in pitch. The word -hulu (big, visibly large in size) is more often used to describe the number of performers and the corresponding loudness of the sound, probably because intensity of tone is not recognized in musical terms: a performer either plays or sings with confidence, and hence with uniform loudness, or indifferently because of shyness, laziness or ignorance of the music. Thus, assuming that the performers are doing well, loud music is at the same time ‘big’, and soft music ‘small’, because of the numbers of people performing it. The sound of the female and male voices are sometimes distinguished by calling the former -sekene (thin) and the latter -denya (thick); pitch within the female and male ranges is further subdivided into high, which ‘closes the throat’, and low, which ‘snores’.

Quality of tone and phrasing, which is invariably legato, is not specifically taken into account: people either ‘play well’ and ‘sing well’, or they do not. Great vigour and energy, precision and virtuosity, are expected of the good performer: a person may sing so well that he ‘nearly bursts his diaphragm’ or dance so that he ‘digs a hole in the ground’ or ‘licks the clouds’, or leaps so high that ‘three people can crawl underneath him’. People like to see and listen to a dynamic, almost destructive performance, when hand-rattles are ‘shaken so that they nearly break’. Quite often a drumskin is torn and a ritual postponed for some hours while it is replaced, or until another drum has been borrowed; leg-rattles disintegrate during a dance, the leather supports of xylophone keys break, and people grow hoarse and lose their voices. Such accidents during a performance do not upset people as they are usually evidence of good, vigorous playing and the intense excitement that goes with it.

The Venda have no word for ‘scale’. They have the word ‘mutavha’, which is used for a complete set of divining dice, metal amulets or stopped pipes, and also for a row of keys on a xylophone or mbila (lamellophone). Thus a mutavha may include more than one octave of a heptatonic or a pentatonic scale, since sets of stopped pipes and lamellophones may be tuned to either of these scales. The Venda recognize the interval of the octave and the fact that heptatonic and pentatonic sets sound different, but they do not express the difference in musical terms, although they name each pipe or key in a mutavha. Traditional Venda melodies have anything from two to seven different tones, but the Venda classify their music on the basis of its social function, which may indirectly affect its structure and especially its rhythmic pattern.

(ii) Music and society.

No fewer than 16 different styles of music are distinguished with different rhythms and combinations of singers and instruments; within these styles, there are further subdivisions with many different songs within each subdivision. There are scores of beer-songs, more than 70 initiation songs of one type and 30 of another; new words are always being added to existing songs and entirely new songs are often composed.

The performance of most communal music is regulated by the rules of the social institutions that it accompanies, but solo instrumentalists can perform at any time of the year without special permission. Some who play the xylophone or lamellophone may accompany singers at a beer-party; others become zwilombe, semi-professional musicians (sing. tshilombe), and from time to time compose new songs or variations of old ones, accompanying themselves on a lamellophone or a musical bow. They are expected to amuse their audiences and are admired for their wit, their mastery of technique and handling of words, and for their ability to clown as well as to protest effectively against any injustice that may need attention.

Both the frequency and conditions of performance of Venda communal music depend to a great extent on the cycle of seasons and the existence of an economic surplus. During the period of planting and weeding, for instance, only important ritual music and work-songs are performed regularly. Towards the end of the weeding season, when the first green maize cobs are appearing, girls begin to practise for their dance, the tshigombela (fig.4), which they would find difficult to dance in the mud of the rainy season, even if they were not required at that time to help with the weeding of the crops, the collection of food and other domestic duties. Circumcision schools are held during the winter, and possession dances and boys’ communal dances take place chiefly during the period of rest between harvest and planting. Communal music is never performed without some kind of reward, either to the performers or to the organizers, so that in a lean year none but the more important items is played. If the countryside resounds with music, especially at night when it is cool, it is a sign of good times. Venda communal music is not a substitute for happiness but an expression of it.

Communal dances also introduce young people to patterns of tribal authority: the music is sponsored by rulers, and one ruler sends his dance-teams on expeditions to other rulers, either to confirm his relationship with them or, if he is a chief and they are headmen, to exact tribute. The mabepha (musical expeditions) consolidate both the lineage ties of rulers, who are separated spatially because of their responsibility for district government, and the neighbourhood ties of clansfolk living in different districts, and hence the bonds between these people and their district headmen. The music of the boys’ and girls’ circumcision schools advertises the power of the doctors who sponsor them, and possession dances enhance the prestige and influence of the families who belong to the different cult groups. Within the traditional music system, ambitious men are able to attract a following and further their interests by means of the music that is performed under their auspices.

Music is therefore an audible and visible sign of social and political groupings in Venda society and the music that a man can command or forbid is a measure of his status. When a ruler holds a domba initiation, all other music in his district is banned, except for his own tshikona (the national dance), beer-songs and personal instrumental music. But nobody is compelled to perform music or to observe these bans, and indeed many Venda Christians ignore them altogether. Diagrams of the relationship between the performance of different styles of Venda music and the passage of the seasons and their political roles are given in Blacking (1973).

Music is an indispensable part of most Venda social institutions, but its transmission depends on their continuity. The Venda assume that every person is capable of musical performance, unless he is deaf; and even then, he ought to be able to dance. In fact people with physical disabilities, such as hunchbacks, seem to excel in music and dancing. Venda dancing consists almost exclusively of rhythmic movement of the lower limbs. When the upper limbs are moved, it is invariably for hand-clapping, drumming or playing a musical instrument; they are also used in dancing, sometimes very vigorously, but chiefly to maintain good balance while the legs are moved.

Dancing is an integral part of Venda communal music. With his body, man creates a special world of time, distinct from the time cycles of natural seasons and cultural events. Just as rhythmical bodily responses to the sounds of music are regarded as the first signs of a child’s interest in music, so participation in communal dancing is generally recognized as the first stage in acquiring musical skills. Small girls copy the dance movements before they participate in the tshigombela and sing the choruses of the songs; they master the dance steps before they attempt to lead a song. Girls usually play the different dance rhythms on the alto drums before they try the straightforward beat of the tenor drum because, although it may not seem so, it is more difficult to maintain a steady beat than to play complex rhythms.

In Venda society, musicians are made according to their birth. Exceptional musical ability is expected of people who are born into certain families or social groups in which musical performance is essential for maintaining their group solidarity. Just as musical performance is the chief factor that justifies the continued existence of an orchestra as a social group, so a Venda possession cult or an initiation school would disintegrate if there were no music.

However, only a few of those who are born into the right group actually emerge as exceptional musicians, and what sets them apart is not so much their ability to do what others cannot do, but that they do it better because they have devoted more time and energy to it. In applauding the mastery of exceptional musicians, the Venda applaud human effort. In being able to recognize mastery in the musical medium, listeners reveal that their general musical competence is no less than that of the musicians whom they applaud. The development of musical ability is therefore a part of every Venda’s experience of growing up, and because the sequence of learning is socially and culturally regulated, music is not necessarily learnt in the order of its musical complexity. Some young people’s music may be technically more difficult than adult music; children often learn pentatonic and hexatonic songs before tritonic and tetratonic songs, simply because these songs are more popular or socially more appropriate.

Most Venda children are competent musicians: they sing and dance to traditional melodies, and many can play at least one musical instrument. But they have no formal musical training. They learn music by imitating the performances of adults and other children. If they do not realize when they are making a mistake, they are soon corrected by more experienced musicians. This does not mean that two performances of the same song must be identical, but that Venda music is based on principles that are acquired partly by learning and partly by assimilation, and that people distinguish between what is or is not specifically Venda about a performance, and are able to create new music according to the same principles. Venda women do not relearn the music of their domba initiation dance (fig.5) when they come together every four or five years to assist the novices: they relive a social situation, and the domba music emerges when the experience is shared under certain conditions of individuality in community. Though the music may sound similar to an outside observer on two successive initiations, it is in fact new to the performers because of the new social situation. Every performance of Venda communal music therefore demands re-creation of a special social situation as much as a repetition of learnt skills.

Venda music is performed in a variety of political contexts and often for specific political purposes. It is also political in the sense that it may involve people in a powerful shared experience within the framework of their cultural experience, and thereby make them more aware of themselves and of their responsibilities towards each other. When two iambic rhythms are combined in canon as in ex.5, the players are not merely using a call-and-response model to produce a surface rhythm that could easily be produced by one performer; each player has his own main beat, so he is expressing musically certain concepts of individuality in community, and of temporal and spatial balance, that are found in other areas of Venda culture. The same principle of sharing in the creation of music is applied to many Venda styles, especially the music of the tshikona shown in ex.6, where each player produces one note of the total pattern.

(iii) Musical structure.

Differences in the styles of Venda music are much influenced by their social functions. The most important communal rituals are called generically by the same word that is used for the large, pot-shaped bass drum, the ngoma. This instrument is found only in the headquarters of chiefs and headmen and, together with the thungwa (the tenor drum) and the murumba (the alto drum), should be played by women (see fig.5).

Sets of drums are kept with sets of 20 or more nanga (stopped pipes), which are tuned to a heptatonic scale ranging over three octaves; these are used to play the national dance, the tshikona. Each player blows one pipe, so that the different notes are combined in a special pattern. The tshikona is performed in the khoro, the public meeting-place, on all important occasions such as the installation of a new ruler, the commemoration of a ruler’s death and the thevhula sacrificial rites at the graves of a ruler’s ancestors. It is performed on Sundays in the urban areas by Venda who have organized themselves into dance teams, with managers, musical directors and other officials.

The full set of drums is also used for the music of the pre-marital domba initiation, in which the girls sing a transformation of what the men play on the pipe ensemble, as shown in ex.7. The music of the tshikona incorporates some of the basic tonal and harmonic principles of Venda music and exerts an influence on other styles that are comparable with its social importance. There are ‘transcriptions’ of the tshikona for most solo instruments, and several children’s songs and songs of the girls’vhusha and tshikanda initiation schools are transformations of, or derived from, the tshikona pattern. The two girls’ schools are less important than the domba, and so their music is accompanied only by the tenor and alto drums and performed in the tshivhambo, the council hut, or the muta, the courtyard of one of the women in charge. Children’s songs are accompanied only by hand-claps or appropriate rhythmic actions.

Full sets of drums may be borrowed from rulers by the organizers of ngoma dza midzimu, dances of spirit possession (fig.6), but they are played in a different way. The doctors who run the masungwi girls’ initiation schools, which were introduced into Venda towards the end of the 19th century, are allowed to own tenor and alto drums. The tenor and alto drums of rulers are borrowed by girls for rehearsing and performing their tshigombela dance and for accompanying the youths’ pipe dances, the tshikanganga, the visa and the givha. These boys’ and girls’ dances are referred to as mitambo (games); the sets of stopped pipes are tuned to a pentatonic scale and made from the ordinary river reed rather than the sacred bamboo that is used for the heptatonic tshikona sets. It is significant that the ‘secular’ pentatonic pipe music was adopted by the Venda in comparatively recent times and that it is never played on a selection of the pipes used for the ancient, heptatonic music (Blacking, Cape Town, 1971).

The sacred music of the mission churches is mostly Western in origin. The Salvation Army uses a European-type bass drum, traditional drums being taboo because of their symbolic associations with ‘pagan’ ritual. Congregations sing hymns in three or four parts, but they often prefer to harmonize in a way that is a novel blend of European and traditional Venda styles. Syncretic styles are a prominent feature of the music of the separatist Christian churches that are run by the Venda themselves. A European-type bass drum, the tshigubu, is used, and the rhythms played are sometimes the same as those of traditional music.

Modern school music is mostly choral and ranges from Western part-songs that are learnt from tonic sol-fa scores for singing competitions, to modern compositions by Venda and other African composers that are sung for pleasure as well as for competitions. The style of this modern Venda music is similar to that of other black composers in South Africa and does not yet betray any special affinity with the unique styles of Venda traditional music. In the late 1950s the guitar and the penny whistle were often used for performing urban jive music.

The circular form of Venda communal dances is well suited to the environment of hills and mountains. The metrical pattern of work-songs is regular and depends on the nature of the work (pounding maize, hoeing fields, clearing weeds etc.), but that of beer-songs may be irregular and emphasized by hand-claps or by the steps of the solo dancers. The irregular metre of many of the ndayo exercises of initiation is played with a stick on the tenor drum and the intervening quavers are filled in on the alto drums, which are played with the hands (ex.8). It is tempting to relate this technique to the vocal call-and-response pattern that characterizes much Venda music; but the alto drums do not complement the tenor’s unfinished pattern as does the response to a call, so much as embellish a pattern that is already adequate, as do the embellishments of a response.

Although the chorus of a song can be sung without a soloist to lead, the solo call cannot be sung without the chorus response. The words and melody of the response are complete in themselves, whether or not they overlap the call, and they may be repeated without change, though it is customary to add complementary parts that agree with the underlying harmonic pattern. Moreover if, as is usual, the tonality shifts regularly in the call-and-response sections of the melody, the tonal centre is established in the response, and so a song without a response is a song without a tonal centre. The importance of tonality and of tonal centres is recognized by the Venda, who call the chief note of every ‘scale’ the phala. The pipe pitched one tone above the phala is called the thakhula (the lifter) because it ‘lifts’ the melody back to its tonal centre. Just as the predominant direction of Venda melodies is descending, so the ‘leading note’ leads down a tone.

In the tshikona, each note of the melody has two companion pitches, which are, in order of harmonic importance, a 5th and a 4th below it (together with their octaves at the 4th and 5th above it). This principle of harmonic equivalence applies throughout the Venda musical tradition, so that the two alternative melodies of the children’s song in ex.9 are regarded as identical. The same principle of harmonic equivalence applies both in improvising additional parts to a choral response (passing notes are also allowed), and in selecting alternative notes when new words of the solo call bring a change in speech-tone patterns that must be reflected in the melody. A soloist is expected both to lead a song confidently and to provide new words for almost every repetition of the basic pattern. Some of these phrases are standard for a particular song or for several songs, and others may be topical improvisations. When women pound maize at night, they may produce a running commentary on local events. There are certain formal rules for adjusting a basic melodic pattern to changes in the speech-tone patterns of the Venda language, and these are learnt at an early age (Blacking, 1967). The more people there are singing a chorus, the more they are expected to ‘fill out’ the basic melody with additional parts. The concern for free-ranging musical development is reflected, for example, in the development of the tshigombela songs, where vocables such as ee and ahee are substituted for words, so that melodies are not subject to the restrictions of speech-tone patterns.

Apart from the kind of social and musical rules cited, the basic structure of every Venda song and the possible course of its development are further modified by each performing situation. Obvious factors are the age and skill of performers and the amount of rehearsal time that they have had. But most important is the fact that the overall form, the number and extent of rhythmic, melodic, harmonic and contrapuntal variations, all the differences between one complete performance of a piece of music and another, are a consequence of the social interaction of performers. The number of performers and audience present, who they are and how they interact with each other, what happens during a performance, who arrives and who leaves, all these and many other social events affect the development of the basic musical pattern. The vitality of Venda traditional music depends largely on the fact that its models are flexible and reflect the organization and values of Venda society as much as certain specifically musical rules. A preliminary set of rules of Venda music has been drawn up (Blacking, African Studies, 1969, and Blacking, 1970). The rules that apply to vocal and communal music apply also to most instrumental music. Even if a physical relationship between the left and right thumbs and the layout of keys on a lamellophone may suggest a ‘walking song’, the music as in ex.10 is conceived within the Venda tonal and harmonic system.

Venda musical instruments have been described in detail by Kirby (1934). Apart from the drums and sets of stopped pipes already mentioned, the most common instruments are the mbila dza madeza, a heptatonic lamellophone with about 27 keys; the mbila tshipai, a pentatonic or hexatonic lamellophone with 11 to 18 keys; various types of signal horns and whistles used by herdboys; side-blown antelope horns used to summon people to the ruler’s place or to announce important events; a three-hole transverse flute and ocarinas that are often played in duet by boys; a number of types of musical bow played by boys and girls and, in the case of two types of bow, by semi-professional musicians; and a large 21-note xylophone, which is played by two people but is now rarely heard.

(iv) Modern developments.

The four decades preceding the coming to power of the African National Congress in 1994 were marked by conflict between supporters of the traditional political order and supporters of democracy. Venda communal music played a role in this conflict by helping to shape new power relations.

Rulers promoted traditional dances that cultivated political loyalty. Their efforts were channelled through government structures such as the Department of Education, which initiated a national dance competition. Refusal to participate in traditional dances elicited fines and accusations of political sabotage. Although coercion played a role, the involvement of adult women in dances formerly not performed by them was a factor of new social responsibilities. Many women took charge of the home economy while men became migrant labourers. Thus, women became a readily available political resource; their new musical roles accelerated change in their social status. This is evident in the changing performing practice of tshikona, a dance known for being a symbol of male social status. Most male residents of the village of Muswodi Tshisimani, for example, had left home in search of work, and so the village became famous for entering a female tshikona team in the national dance competition during the early 1980s. The public shock turned to amazement and eventual acceptance. Most women, however, became involved in tshigombela, a former girls’ dance. Tshigombela songs, which originated around independence (1979), promoted national unity under chiefly rule. The nationalistic content of tshigombela songs reflected a degree of spontaneous reaction to colonialism. For several years after independence, many people supported chiefly rule through tshigombela dancing because they believed that their material conditions would improve.

A change in political consciousness marked communal musical performances from the mid-1980s onwards. This change coincided with the repression of the official political opposition (the Venda Independence People’s Party) and the establishment of a one-party state, which failed to address the economic and political aspirations of large numbers of poor people. Tshigombela songs increasingly challenged government legality. Feelings against political crimes were particularly strong. Mimes of these crimes were performed in tshigombela dances, and many beer-songs expressed anti-government protest.

The Venda government was overthrown in a bloodless military coup during 1990. Communal dancing subsequently decreased in a number of areas and ceased in others. New tshigombela songs emerged with revolutionary topics that were popular with youth cultural clubs affiliated with the African National Congress. Many people now regard this music as less important since it does not advance their socio-economic goals. They prefer to join bands or choirs. Older musicians who perform traditional music find decreasing social acceptance, and most traditional instruments have disappeared. The assimilation of Venda musicians into the global musical culture is evident in the increasing number of English songs, the adaptation of traditional music to a commercial idiom and the emulation of pop stars (see §III below). Fewer people are becoming musicians and more are now consumers of music. Music is not regarded as a reliable career, and formal music education in school is virtually non-existent.

4. Tsonga music.

The Tsonga ethnic group occupies the north-eastern part of what was once called the Transvaal. Tsonga are found in the Transvaal from northern Swaziland and Zululand to the Limpopo river. They also inhabit south-eastern Zimbabwe. In Maputo, the Tsonga occupy the area north of kwaZulu up to the banks of the Zambezi.

The Tsonga have not escaped cross-cultural influences at least since historical and anthropological recordings of groups and nations were made by Western scholars. One scholar who has produced an exhaustive cultural study of the Tsonga is Rev. H.A. Junod. His anthropological study of the Tsonga, The Life of a South African Tribe (London, 1927), attempts to record all aspects of the Tsonga-Shangaan group. Junod included a chapter on Tsonga music before it was subjected to external cultural influences. Tsonga music had two major cultural influences: (a) a Nguni influence, when the Zulus subjugated the Tsonga of Mozambique (c1840), and (b) an influence of Western music culture after missionaries embarked on converting the Tsonga-Shangaan ethnic group to Christianity.

(i) Musical instruments.

The Tsonga do not have many elaborate musical instruments. Traditional music instruments include: comana, rhonge, xipendani, xitende, xizambi, xitiringo, ndhweva and timbila.

The comana is a small drum made from a hollowed tree trunk or large calabash covered with an animal skin. The skin is first soaked in water and then stretched over one side of the trunk or calabash and secured by wooden pegs. With the advent of Western culture, frames of iron sheets were introduced. Large washing basins are now also used as comana frames. The comana drum is used to accompany dancing during the exorcism ritual.

The rhonge is an ocarina made from a dry sala fruit. One large hole and two smaller holes are cut into the dried fruit. Air is blown through the larger hole with the mouth placed in such a way as to produce a sweet whistling sound. The quality of the sound is controlled by blocking and/or opening the smaller openings with the fingers. The rhonge is played mostly by shepherds. A xipendani is a flattened wooden bow approximately half a metre in length. The middle section of the bow is left in its original round stick size. A string or thin wire is tied on each end and pulled to form a bow. The string is divided into two sections by another string or hook-wire pulled across the bow. It is played by plucking the string with a long thorn or needle. Players place their mouths on one side of the flat bow as a resonator. By changing the configuration of the mouth, different pitches are produced. The Xipendani is primarily played by women.

The xitende, like the xipendani, is a bow-like instrument. The differences are that the xitende bow is approximately a metre to a metre and a half long and the wooden stick is not flattened but left round. Both ends of the stick are tapered. Animal sinew or a soft thin wire is stretched from one end of the stick to the other forming a bow. A half-calabash is secured at the centre of the instrument by a wire or string. The half-calabash serves as a resonator. Different sound qualities are obtained by alternately pressing the calabash against the player's breast and drawing the calabash away from the breast. The instrument is played by tapping on the wire with a small stick approximately 30 cm long. The back of the 3rd and 4th fingers touch the string and produce different pitches and sounds. Tin bottle caps (traditionally sea shells) are attached to the bow to add a rattling sound. The xitende is played by a man who uses it to accompany his singing.

The xizambi is a bow made of a stick, 1·5 cm thick and half a metre or shorter in length. It is bowed with a palm leaf 1 cm wide. The middle of the bow is marked by small regular grooves. It is played with a stick to which is attached two or three dry, hollowed sala fruit into which a few bean seeds or stones are placed. It is played by securing one extreme end of the instrument with one hand, while the other end is placed against the mouth with the mulala palm leaf running across the open mouth. The instrument is played with the right hand moving the stick along the grooves while the seeds in the sala fruit provide a rhythmic accompaniment. Different sounds are obtained by changing the shape of the mouth. The xizambi is played by both men and women.

The xitiringo is made from a piece of bamboo approximately 2·5–3·5 cm in diameter with one small hole on one side and three to four small holes on the other side. Both ends of the xitiringo are stopped. The xitiringo is played by blowing air into the one small hole while the fingers of either the left or right hand are placed on the small holes at the one end. Different pitches and sounds are obtained by lifting the fingers from the holes. The xitiringo is primarily played by shepherds and young adult men.

The ndhweva is a hollowed reed or open-end bone of a goat. It is played by pressing the reed or bone against the tongue and blowing hard, thus producing sound when placed at a certain angle. It is played by shepherds. The timbila is basically a Chopi instrument which the Tsonga have adopted. It is similar to the xylophone and is played with two rubber mallets. Male adults play this instrument accompanied by singing and dancing.

The xigubu (drum) is a modern innovation. The xigubu frame is made from a 44-gallon oil drum covered by two soaked ox skins that are then dried in place. The two skins are secured with wet strips of skin approximately 2 cm wide. The drums accompany the xifasi and xincayincayi group dances.

(ii) Vocal music.

Traditional Tsonga vocal music is responsorial. The leader, called musumi, i.e. ‘the starter of a song’, begins by singing the first note or a few notes of a song. This ‘call’ is referred to as ku suma. The remaining singers respond to the call of the musumi by singing a song's choral response. The response is ku hlavelela in Tsonga and the choral group is called vahlaveleri.

Tsonga is a bitonal language with high and low tones, and Tsonga vocal music is to a large extent influenced by speech tones and rhythms. However, the effect of such tones and rhythms is limited by certain musical requirements such as the melismatic vocalization of non-lexical syllables. Another characteristic of Tsonga music is the typical descending melodic cadence at the ends of musical phrases. Other characteristics include the elision of vowels, the contraction or prolongation of final vowels and the contraction or the prolongation of final syllables. Tsonga melodies have descending intervallic contours comprising a 4th, 5th or even 6th from the initial peak to a low-pitched note.

Most Tsonga songs are polyphonic. The musumi sings lead melodies, while the vahlaveleri sing choruses. In the majority of songs, the musumi's solo part carries the lexical part of the lyrics, thus conveying a song's message. The lyrics of the response sections are generally non-lexical, onomatopoeic syllables.

Children's songs are interwoven with games and are generally accompanied by hand-clapping, dancing and chanting. Lullabies are sung mainly by mothers while grinding corn or stamping meal with babies on their backs. The mothers' rhythmic movements rock the babies to sleep while singing.

Songs to exorcise spirits (mancomana) are always accompanied by ncomana drums. The singing is energetic and highly rhythmic, and the ncomana drums beaten by two to four women keep a strict rhythm. When the spirits are ‘out’, the musumi (‘patient’) dances vigorously while the vahlaveleri sing. Singing usually starts in the late evening, lasting until the early hours of the morning.

Initiation songs for girls (tikhomba) are also generally accompanied by the ncomana drum. The singing and the dancing are not as vigorous, however, when compared to the songs used for exorcism. The drumming patterns synchronize with the stamping of the feet by the initiates who number from two to five. There is usually a large group of women vahlaveleri; males are forbidden. Men and boys are allowed only on the last day of the initiation.

Initiation songs for boys (ngoma or murhundu) are sung at an initiation school in a forest around the fire in the evenings. The presence of women is forbidden during such events. The lyrics are generally unintelligible, usually in a foreign language, such as the Ndau language. The songs are generally rhythmically slow such as war dance songs (muchongolo).

Songs for social occasions include those for work parties (matsima, sing. tsima), marriage (nkhuvo) and supplication to the ancestors (mphahlo). These occasions include beer-drinking and are accompanied by singing, hand-clapping and dancing. The singing on such occasions is lively and the dancing is accompanied by graceful body movements.

Songs for dance groups (xifasi and xincayincayi) are usually performed by young men and women. The dancing is taken in turns; women perform in graceful formations, while men engage in wild antics. Dancing is accompanied by singing, hand-clapping, whistles and shaking of the waist by women who wear colourful uniforms, while drums maintain the rhythm.

War dance songs are usually sung in the Nguni language (Zulu), due to the 19th-century Zulu subjugation of the Tsonga people. These songs have a slow tempo punctuated by foot-stamping and by the hitting of ox-hide shields with spears.

Folktale songs (tinsimu ta mintsheketo) punctuate the rich Tsonga folktales. These songs are also performed in the call-and-response style and function as a means of communication among the dramatis personae in folktales. In Tsonga folktales, animals interact with humans on equal terms.

Themes of Tsonga songs other than in folktales mainly comment on social life. Songs often comment on social relationships, for example on marital problems, problems of co-wives, infidelity, abuse of power, social injustice and witchcraft.

(iii) Modern developments.

Tsonga music displays the influence of Western culture, interweaving traditional elements into modern Tsonga music. The call-and-response and polyphonic characteristics remain, as do the descending contours of intervals of a 4th or 5th. Instruments used in contemporary performing ensembles consist of electric guitars and drums, which serve as accompaniment, providing lively rhythms.