Oral Literature in Africa
Visit the homepage for downloads, additional resources and more...
Browse the Book

9. Lyric

Occasions. Subject Matter. Form. Composition.

In the sense of ‘a short poem that is sung’, lyric is probably the most common form of poetry in subsaharan Africa. It is not always recognized that these songs, in which the musical element is of such obvious importance, are in fact poems. It is true that the verbal aspect sometimes appears less developed than in the lengthy poems that are delivered in spoken or recitative style, like some of the praise poems, hymns, or hunting chants that have already been described. But this should not prevent us from calling them poems. We should remember that classical Greek or Elizabethan lyrics were equally designed to be sung. Indeed, in its original form of a poem in a musical setting, lyric is one of the most important kinds of African oral literature.

So far, with a few exceptions,1 the poetry we have considered has mainly been associated with relatively formal events. The lyric songs discussed here are for more informal occasions. Whereas much other poetry depends on a specialist and even esoteric tradition, these involve popular participation. The verbal content of these songs tends to be short (though the actual performance may be lengthy) and is often ephemeral. There is usually plenty of improvisation. Unlike the general pattern of Western European folk-songs, the individual singer does not tend to stand out in a dominant position as against a passive audience,2 but instead interacts with a chorus. Yet these lyric songs still provide wide scope for individual expression.


Songs appear in an almost unlimited number of contexts. In words that might be applied more widely than to the Ibo of whom he was writing, Osadebay speaks of the ‘wealth of culture and fine feelings which find expression in our music and poetry. We sing when we fight, we sing when we work, we sing when we love, we sing when we hate, we sing when a child is born, we sing when death takes a toll’ (Osadebay 1949: 154).

Rites de passage are very common occasions for singing. There are songs associated with birth, with initiation and puberty, betrothal, marriage, acquiring a new title or status, and funeral and memorial celebrations. The most serious of these songs, in which the verbal element is elaborated at length, cannot be called lyrical. But often such ceremonies are in fact more an occasion for festivity, which includes song, than a solemn ritual with specially designated music, and the gatherings normal at these times are a reason for singing for its own sake.

Weddings, for example, are popular occasions for comment in song—by no means always involving praise of the newly wedded pair:

Serpent que tu es!

Chien que tu es!

Tu fais: oua-oua!(Junod 1897: 46)

sing the bride’s friends about her husband among the Ronga, amid a series of songs cheerfully warning her of the ill-treatment she will without doubt receive at the hands of his parents. In more reflective and personal style is the Ganda song of farewell by the young girl about to be married, with the repetitions typical of this form:

Oh, I am gone,

Oh, I am gone,

Call my father that I may say farewell to him,

Oh, I am gone.

Father has already sold me,

Mother has received a high price for me,

Oh, I am gone. (Sempebwa 1948: 18)3

It is likely that advantage will be taken of this opportunity to sing songs on many other topics.

Many of these songs are for dancing. A particular song type is sometimes inextricably tied up with a particular dance. Thus the Swahili used to have a special gungu song for the ‘pounding figure’ of the dance:

Give me a chair that I may sit down and hold (the guitar)

Let me sing a serenade for my Palm-daughter

Let me sing for my wife

She who takes away my grief and sorrow.

(Knappert 1966: 130; cf. Steere 1906: 47)

This occasion for song is, if anything, increasing, and many examples could be quoted like the Zulu ‘town dancing songs’ quoted by Tracey, where the words are subordinate to the dance:

This is the girl that jilted me,

The wretch of a girl that jilted me.

At Durban, the dance leaders are afraid of us! (Tracey 1948b: 61)4

Zululand, my home, I love you.

Goodbye, Willie I like you too.

We are the boys.(Ibid.: 66)5

The same kind of mood, of recreation and light-hearted enjoyment, is evident in many of the ‘drinking songs’. These too, for all their lightness, may express the thought in true lyric manner, with economy and grace. In a Shona drinking song, the original is only seven words in all:

Keep it dark!

Don’t tell your wife,

For your wife is a log

That is smouldering surely!

Keep it dark!(Tracey 1933, no. 9)

There are sometimes more formalized occasions for the singing of lyrics. One could mention the recent interest in the short balwo lyric among the Somali.6 Special balwo parties became fashionable in the towns. People would recite the lyrics they knew or compose new ones, and the recitations would be interrupted for tea and conversation (Andrzejewski 1967: 11). Popular and occasional bands among the Akan also sometimes perform on specifically recreational occasions (Nketia 1962: 16–17; 1963b: Ch. 6). Again, in many areas the radio nowadays frequently creates opportunities for lyrics to be performed.

All over the continent it is a common pattern for stories to be interrupted from time to time by a song, usually led by the storyteller, while the audience act as his chorus.7 Sometimes these songs amount to quite long poems, and are then often in recitative. Short verses are also very common, sometimes with many nonsense syllables to fill in the rhythm and tune, with repetition over and over again between leader and chorus. One Limba story, about ‘The clever cat’, has a verse of this kind:

The story is about a cat who proposes to initiate the young rat maidens into the bondo (women’s society). They, like all young girls, are eager to enter—but the cat’s one desire is to have a chance of eating them! The cat pretends to act in the usual way of a bondo senior woman. She lines them all up and leads the singing, telling them not to look round. She sings:

When we go,

Let no one look behind oh!

When the cat is free, fo fen.

The chorus of young rats take up the same words:

When we go,

Let no one look behind oh!

When the cat is free, fo fen

in the way young initiates do in real life. The rhythmic and melodic song is repeated in the story perhaps eight or ten times, first by the cat (the narrator), then by the rat initiates (the audience) who have quickly picked up the tune. But while the singing is going on, what the cat is really doing is to quietly pick off the rats one by one as they sing with their backs to her. At last only one is left, still singing the song. Just in time, she looks round, and throws herself out of the way and escapes.(for the full story see Finnegan 1967: 333–4)

A story like this appeals to its audience partly because of the amusing form of words and the parody of the usually very serious initiation ceremony, but perhaps most of all because of the attractive song which, in terms of the time spent repeating it over and over, took up as long as the prose narrative. Simple as the words were in themselves, the audience all joined in enthusiastically, overlapping slightly with the leader’s last note and half dancing as they sang; they would, it seemed, have continued indefinitely had not the leader finally broken into their response to continue his narration.

The same song is sometimes repeated at different points in the story, a kind of signature tune with slight variations on the words to fit the development of the plot. The structure of the story is thus marked by the recurrence of the song in each new episode. Another Limba example can make this plain:

The plot is the intentionally fantastic and humorous one of the hero Sara and his endeavours to kill and eat a guinea-fowl he had caught without sharing it with any of his friends. But the bird is a magical one and the more Sara tries to kill and eat it, going through all the usual preparations and cooking procedures, the more it sings back at him. At last he eats it—but even in his stomach the bird sings and demands to be excreted; and in the final effort, Sara dies.

Each of the many parallel stages of the plot is marked by the same song, with variations to suit the event, the last phrase and response being repeated several times by narrator and audience with the same tune throughout. First, the guinea-fowl is discovered in the snare, and it sings:

Sara is coming to loose me,

Sara is coming to loose me.

Here he found a path, a night passed,

Here he came and put a snare for me,

The guinea-fowl,

The guinea-fowl,

Ko de ba ko naligbe8

What is your name?

What is your name?

(Response) Tambarenke, Tambarenke.

What is your name?

Tambarenke, Tambarenke.

What is your name?

Tambarenke, Tambarenke . . ., etc.

Sara looses the bird from the noose, and brings it home to prepare for eating. Again the bird sings:

Sara is coming to pluck me,

Sara is coming to pluck me.

Here he found a path, a night passed,

Here he came and put a snare for me,

The guinea-fowl,

The guinea-fowl,

Ko de ba ko nagligbe

What is your name?

What is your name?

Tambarenke, Tambarenke.

What is your name?

Tambarenke, Tambarenke . . .

As the story continues, new first lines appear:

Sara is coming to cut me up . . .

Sara is coming to pound me . . .9

Sara is coming to mould me . . .

Sara is coming to put me in (to the pot) . . .

Sara is coming to take me out . . .

Sara is coming to eat me . . .

Sara is going to lie down . . .

And, finally,

Sara is going to excrete me . . .(Full story in Finnegan 1967: 284–90)

The linguistic content of songs in Limba stories, as in some others, is relatively limited, and for the audience their main interest lies in the rhythm and melody and the fact that they can participate in the singing. In some other cases, however, such as some Akan stories, the words are more developed. The following is a variation on a very common theme:

Elephant and Antelope are said to have made very good friends in the forest. Elephant being the stronger and wealthier of the two was able to lay on sumptuous meals every day to which he invited Antelope. One day he expressed the desire to visit Antelope in his house. This embarrassed Antelope for he also wanted to give him a good meal. It occurred to him after failing to get any meat that Mother Antelope was the answer, so he caused her to be killed and used. When Elephant arrived he was greatly surprised by the delicious meal and asked to see Mother Antelope. But Antelope succeeded in putting this off. After the meal however, Elephant again asked for Mother Antelope and Antelope replied in a song as follows:

Elephant, please don’t worry me.

Have you ever seen a poor man

And a wealthy man exchange things equally?

Elephant Akwaa Brenkoto that commands his destiny,

Elephant that plucks the tops of trees on his right,

King of musketry, father and king,

Birefi Akuampon, mighty one to whom all stray goods are sent to be used.

Yes; let us proceed,

Mother Antelope, I have stewed her.

Yes, let us proceed.

Mother Antelope, I have used her to redeem myself. Yes, let us proceed. (Nketia 1958b: 19)


The subjects of the many different songs sung on these various occasions include just about every topic imaginable. There are songs about wives, husbands, marriage, animals, chiefs, this year’s tax, the latest football match, a recent intrigue, the plight of a cripple dependent on his family, an amusing incident, a friend’s treachery or an enemy’s vices, the relationship between variety in the human and the natural world—and so on according to the genre of song involved, the context of performance, and the poetic inspiration of the singer.

It has frequently been remarked that African poems about nature are few and far between, and there is truth in this assertion. Certainly there seems to be little in common between most African lyrics and the romantic interest in ‘Nature’ typical of certain epochs of the English poetic tradition, and lyrics about people, events, and personal experience are more common. But observation of the natural world, especially the animal world is often significant. Take the simple little song about a brook recorded in Malawi in the nineteenth century. The effect is an imitation of the sound of the brook and it is sung ‘softly and soothingly’ in a subdued voice; the main point is to reflect the tune of the water rather than describe in words, though a picture is given of the bank of the little stream (chiko) and the prickly bush that grows by it (likwanya):

1st voice

Likwanya likunyanya ku chiko.




1st voice

Likwanya likunyanya ku chiko

2nd voice

Anya nya-nya-nya-le e.

Then the two voices interchange lines twice, with the final response:

Anyanyale.(Macdonald i, 1882: 49)

Again, we could mention the case of Somali poetry which ‘is imbued with a consciousness of the beauties and cruelties of nature’ (Andrzejewski 1967: 9). For instance, the simple lyric ‘O Distant Lightning! Have you deceived me?’2 gains its emotive tone from the inspiration of rain and its life-giving and beautiful results. Lightning often presages rain, and this symbolizes hope. But sometimes the hope is disappointed and the rain-clouds move away. So here the poet is writing of love, but calls the girl ‘Distant Lightning’, expressing his disappointment in love in terms of natural forces.

Songs associated with birds are very common. Sometimes the song is envisaged as sung by the bird itself, and at least in part is an onomatopoeic representation of the call. We could instance the many lyrics supposed to be sung and exchanged by birds among the Beti of the Cameroons. The ngiai afan (genderme silvatique) sings of the insecurity of life:

Point de sécurité en forêt.

(Mvie e se a fie.

Point de sécurité en forêt.

Mvie a se fie).(Anya-Noa 1965: 129)

The female kolvodo ban nga (magpie) in one of her songs praises the virtues of work:

Va au travail.

(Kel’ esie o.

Va au travail.

Kel’ esie o.

Si tu entends dire:

O wog na:

‘C’est une fille d’homme’

‘Ngon mot’

C’est grâce au travail.


Si tu entends dire:

O wog na:

‘C’est une fille d’homme’

‘Ngon mot’

C’est grâce au travail.


Le pays serait-il généreux,

Nnam akab,

Ne sois pas quémandeur.

Te bo zaq.

Le pays serait-il généreux,

Nnam akab,

Ne sois pas quémandeur.

Te bo zaq.) (Anya-Noa 1965: 124–5)

The Zulu songs attributed to birds attempt to represent something of the nature and appearance of the bird as well as its cry—and cast a sly glance at humanity too. The bird called uthekwane (hammerkop or heron) is pictured strolling gracefully by the waterside, with his fine-looking crest and shapely thighs—symbolizing vanity:

I myself, have often said:—Thekwane! You, with your crest, your leisurely strolling when frequenting the spring, at the time it has been opened up—mark you as a very fine fellow. You have large thighs. (Dunning 1946: 44)

Other Zulu bird songs involve interchange between the hen and the cock, the male in deep bass, the hen higher. The song of the insingizi (hornbill or turkey-buzzard) is really a comment on married life, particularly the last line of the cock’s exhausted rejoinder to his wife’s constant nagging:


Where, where is (the) meat? Where, where is (the) meat?


There’s none, it’s up in the trees above (bis)


Where, where are the worms? (bis)


There are none, there are no worms (bis)


Are there none, are there none over there? (bis)


Oh! get away with you! Where will I get them from? (bis)


Look for them, look for them over there (bis)


There are none, there are none over there (bis)


I am going, I am going, I am going home to my people (bis)


Go, go, you have long since said so (bis).(Dunning 1946: 33)

Most elaborate of all is the song of self-assertion attributed to the iqola, the fiscal shrike. In it the cock utters his proverbial cries of ‘Goshi! Goshi! Dadi! Dadi!’, cries which are supposed to describe the sounds made by the movements of his wings and feet as well as the ejaculations he utters as part of his great display. He is pictured as turning his head to the right, then to the left, surveying himself in self-admiration. His cry really amounts to saying ‘I am the personification of everything that is Majestic and Powerful and my ornaments jingle and rattle in perfect rhythm’. He sings:

Goshi! Goshi! Dadi! Dadi!

Who do I kill (stab)? Who do I kill? Who do I kill?

I kill the relations of these (indicating his victims) outright! outright!

I kill the relations of these outright! outright!

I kill the relations of these outright! outright!

Sanxokwe, Sanxokwe (addressing her Majesty)

I’ll pay your bridewealth (lobola) with a red beast

I’ll pay your bridewealth with a red beast

I’ll pay your bridewealth with a red beast.

When men drink beer, they become intoxicated,

They take up their sticks

And they (the sticks) clashing together sound xakaxaka, xakaxaka, xakaxaka.

I have been across the Umdawane10

Where I ate up the big dance.11

I caught a small bird, I fixed it on the end of a slender twig very early this morning.

I repeated this by catching a Fantail Warbler early this morning

And fixed it on the end of a slender twig.

I drank the blood of a bird early this morning.

I struck its little stomach, it became red with blood at that very moment,

Because I am the King of Birds.

Goshi! Goshi! Dadi! Dadi!

Bayede! Bayede! (Salute me royally) Khuleka! Khuleka! (Make obeisance to me) Nkosi! Nkosi! [Address me as] King!(Dunning 1946: 45–6)12

Finally, in a rather different style, is the brief but pathetic Nyanja song of the unloved night-jar:

Moon, you must shine, shine that I may eat the tadpoles;

I sit on a stone, and my bones all rattle.

If it were not for my big mouth,

The maidens would be crying for me.(Rattray 1907: 164)

Songs about, or attributed to, animals seem to be less common than those associated with birds. But some certainly exist particularly in South and Central Africa. The brief Hottentot song about a baboon gives a vivid little picture of his typical occupation:

There, I’ve got you, I’ve got you, I’ve got you . . .

Crack, crack, what a louse . . .

It bit me, what a louse . . .

Crack, crack, what a louse . . .

It bit me, what a louse . . .(Stopa 1938: 101)

and is cast in the typical form of a sung lyric, with plenty of scope for repetition and, apparently, for chorus responses. Among the South African Bantu the tradition of praising seems still strong, and recent praises (although strictly of a different order from the songs quoted in this chapter) are much more simple and lyrical in concept than the lengthy and grandiose praises of traditional culture. Thus Hurutshe men describe a hare:


Son of the little dark brown one with spots,

Little yellow one, leaper from the stubbles,

Yonder is the son of the little dark brown one

Leaper from the treeless plain

Leaper from the trunks of trees;

It leaps up, and stretches its tail

And it places its ears on its shoulders

Ga-re-ya-gaa-koo!(Merwe 1941: 328–9)14

Among pastoral peoples, songs are often composed and sung in praise of individual beasts. Cattle come to mean far more to their owners than mere economic sustenance, and are accepted as emotional and evocative topics for deeply felt expression. This can be seen in the songs collected by recent investigators from the Nilotic cattle-keeping people, and also from a Dinka song published early in the century. The individual singer typically praises his own bull in an outpouring of personal pride:

My Bull is as white as the silvery fish in the river; as white as the egret on the river bank; as white as new milk.

His bellowing is like the roar of the Turk’s cannon from the great river.

My bull is as dark as the rain-cloud, that comes with the storm.

He is like Summer and Winter; half of him dark as the thundercloud; half of him as white as sunshine.

His hump shines like the morning star.

His forehead is as red as the arum’s [hornbill] wattles. His forehead is like a banner; seen by the people from afar.

He is like the rainbow.

I shall water him at the river, and drive

My enemies from the water with my spear.

Let them water their cattle at the well;

The river for me and my bull.

Drink, O Bull, of the river. Am I not here with

My spear to protect you?(Cummins 1904: 162)

But songs describing animals, or even birds, are apparently far less common than those in which the main interest is human life. In fact this can be seen even in many of the songs ostensibly about birds, for the bite of the comment is often its veiled relevance for human action, character, aspiration, or absurdity. There are lyrics about every facet of human activity. Love and marriage are probably the commonest themes, and the remainder of this section will illustrate some of these songs.

Marriage is a topic that can be treated many different ways. Not only its attractions are indicated in song, but also its difficulties or absurdities. Thus one of the Ganda songs connected with marriage lightly warns young suitors:

When he sees a pretty girl he falls for her,

‘I will go with you, let us go’.

Not knowing that he is going with a girl with a fiery temper.

(Sempebwa 1948: 17)

Among the Shi of the Eastern Congo, again, marital relationships are the most common single subject in songs, many of them concerned with marital and pre-marital strife. One of the popular forms is a song describing a girl’s rejection of her suitor because she thinks him too poor:

‘You want to marry me, but what can you give me?

A nice field?’

‘No, I have only a house’.

‘What? You have nothing but a house? How would we live? Go to Bukavu; there you can earn plenty of money. You can buy food and other things’.

‘No, I won’t go. I don’t know the people there. I have always lived here, and I know the people and want to stay here’.

‘You are a stupid man. You want me to marry me but you have nothing. If you don’t go to Bukavu and earn money to buy me things then I won’t marry you’.(Merriam 1954: 45)

A different point of view is expressed in one of the many Chopi songs on this subject. Here the girl is pictured as sad and solitary without her husband; like so many others he has gone off many hundreds of miles to work in the mines. And yet there is something in common—a comment on a woman’s demand for material possessions:

I am most distressed,

I am most distressed as my man has gone off to work,

And he does not give me clothes to wear,

Not even black cloth.(Tracey 1948a: 46)

The number of love songs recorded is surprising—at least to those brought up to the idea that the concept of personal love is bound to be lacking in African cultures. Even the idea of courtly and romantic love is not always absent. It seems, for instance, to occur to some extent among the Hausa, whose rich tradition of love poetry is now influencing surrounding people.15 Fletcher quotes a simple Hausa song of love, ‘To Dakabo, a maiden’:

Dakabo is tin!

Dakabo is copper!

Dakabo is silver!

Dakabo is gold!

Where greatness is a fortune

The thing desired is (obtained only) with time.

Thy things are my things,

My things are thy things,

Thy mother is my mother,

My mother is thy mother,

Thy father is my father,

My father is thy father!

Be patient, O maid!

Be patient, young maiden!(Fletcher 1912: 65)

The Somali balwo (later called heello) are even more striking examples of romantic and emotional love poetry (see especially Andrzejewski 1967). These are short lyric love poems that have become popular recently and are particularly associated with the new urban generation. The balwo is characterized by extreme brevity—it usually consists of only two lines—and a condensed and cryptic imagery expressed in ‘miniature’ form. It is sung to a distinct tune with syncopated rhythms, but there are relatively few of these tunes and thousands of different poems. There are two, related, themes in these lyrics: first, those addressed to a beloved woman, in hope of marriage; and secondly those to a woman admired from afar off, even one seen only once whom the poet can have little hope of seeing again. This theme of romantic and frustrated love gives rise, it seems, to genuine and deeply felt emotion, expressed in a condensed and symbolic form arising from one central image:

Woman, lovely as lightning at dawn,

Speak to me even once.

I long for you, as one

Whose dhow in summer winds

Is blown adrift and lost,

Longs for land, and finds—

Again the compass tells—

A grey and empty sea.(Laurence 1954: 31)

If I say to myself ‘Conceal your love!’

Who will conceal my tears?

Like a tall tree which, fallen, was set alight,

I am ashes.(Andrzejewski 1967: 13)

My heart is single and cannot be divided,

And it is fastened on a single hope; Oh you who might be the moon.

(Andrzejewski and Lewis 1964: 146; also pp. 49–51)

The romantic love poem is not just confined to the coast. The Nyamwezi of central Tanganyika around Tabora can sing:

My love is soft and tender,

My love Saada comforts me,

My love has a voice like a fine instrument of music.(Tracey 1963: 20)

Not all African love songs, however, are in the romantic, even ecstatic vein perhaps more typical of areas like Hausa country or the East Coast, long influenced by Arabic culture. There are many ways of describing this fertile theme. The Kuanyama Ambo of South West Africa have a series of brief antiphonal love poems used in courtship, with call and response between man and girl. Usually some analogy of a general rather than a personal kind is made between nature and human relationships:

Where one sees birds in their flight, there is water;

Where one hears the sound of women’s laughter, there is a kraal.

A palm stick bow does not like the rainy season (it warps);

A woman fond of a man does not like to be among people.

(Loeb 1950: 847, 848)16

An analogy with nature is also made in a very light-hearted love song by a young Soga in East Africa:

All things in nature love one another.

The lips love the teeth,

The beard loves the chin,

And all the little ants go ‘brrr-r-r-r’ together.(Tracey 1963: 20)

Zulu love poetry often seems to be by women, a feature that has parallels elsewhere in Africa. Dhlomo gives one girl’s song that is both realistic and romantic:

Never shall I fall in love with a suckling.

Joy, joy, O mother, this one sleeps unrealising.

Never shall I fall in love with one who is no ladies’ man.

Joy, joy, O mother, this one sleeps unrealising.

I would like to fall in love with a dashing he-man.

Joy, joy, O mother, this one sleeps unrealising.

Would love him-who-appears-and-causes-heart-aches!

Joy, joy, O mother, this one sleeps unrealising.

Yes, I would like a whirlwind of a man!

Joy, joy, O mother, this one sleeps unrealising. (Dhlomo 1947: 7)

In much more disillusioned vein is another Zulu love song, this time by an older woman living in Durban, where she runs her own small group of singers. The song expresses all her despair and the mundane yet heart-breaking aspects of parting:

I thought you loved me,

Yet I am wasting my time on you.

I thought we would be parted only by death,

But to-day you have disappointed me.

You will never be anything.

You are a disgrace, worthless and unreliable.

Bring my things. I will put them in my pillow.

You take yours and put them under your armpit.

You deceived me.(Tracey 1948b: 41)

Among the Luo of Kenya, too, love songs are sung by women. The final examples of love poetry will be taken from their oigo lyrics, one of the many types of songs in Luo country.17 These are love songs in a slightly different sense from the ones already quoted.

The oigo are songs sung by young girls on their way to visit the young men they are courting. The girls walk to the hut where they are to be entertained by the men, by the light of the full moon. As they go, they sing these songs, individually or in groups, taking it in turns to sing the whole way. ‘There was no formal order of singing; the more musically gifted girls or the more effusive took the leading part according to their mood’ (Owuor 1961: 51). Meanwhile the young men are waiting, straining their ears for the first sounds of the song. When it is heard, one of them announces to the rest, at the top of his voice: ‘The landing has taken place, they have arrived’. The girls come and are welcomed with gifts. And then the evening’s entertainment proceeds, the men playing on reed flutes while the girls sing their oigo songs.

These songs have their own special form. The tunes are simple and rather repetitive with an insistent rhythm. The most striking aspect is the singer’s vocal style. ‘The singer trills in a bird-like voice and conveys an impression of being possessed by the stream of song within her, breathless and helpless. The emotions expressed are often sorrowful and almost hysterical, yet the singer exults in her ability to sing endlessly like a bird’ (Owuor 1961: 52). This distinctive style comes out, even in translation, in the following poem. The characteristic refrain, doree ree yo, is far more repetitive and appealing than can be represented in an English text:

I am possessed,

A bird bursting on high with the ree lament

I am the untiring singer.

Dear bird, let’s sing in rivalry

Our doree ree yo . . .;

It is my wayward self,

Singing in rivalry

The doree ree yo;

I am the untiring singer

That rocks far-off Mombasa

With the aree ree yo;

It is the voice crying the doree

That rocks far-off Nakuru;

I am the compelling Ondoro drum,

The bird bursting with the doree’s plaintive tones;

I am the untiring singer

Choking herself with the doree ree yo.(Owuor 1961: 53)

Sometimes the emphasis of the song is on the sorrow of the singer, or the way she is possessed by the song. At other times we are given a picture of another side of her nature—wilful and unpredictable, her impulsiveness breaking through the ordinary rules of behaviour. This comes out in one song that is arranged round the image of a family setting out, led by the favourite bull who symbolizes their unity. Impulsively, the girl runs ahead to keep up with the animal, in spite of the pain in her chest from her exertion:

Our bull is starting off for Holo,

The Kapiyo clan have fine cattle.

Our bull is starting off for Holo,

The Kapiyo clan have fine cattle.

Then the giggling one said,

Then the playful one said,

(How amusing)

The impulsive ree singer

Is a forest creature lamenting the pain in her chest;

The forest creature lamenting the pain in her chest,

The spirited one lamenting the pain in her chest,

The giggling ree singer

Is a forest creature lamenting the pain in her chest,

The Nyagwe Gune lamenting the pain in her chest,

The impulsive ree singer

Is a forest creature lamenting the pain in her chest.

Our bull is starting off for Holo,

The Kapiyo have fine cattle;

The Kadulo clan is a bull which starts off for Holo,

The Kapiyo have fine cattle.(Owuor 1961: 54)

In these songs, a special picture of girlhood is presented. It is one which does not necessarily correspond in all ways to the reality, but forms a conventional part of this particular form of art:

She lives in a dreamland, though much tempered by the idealised role she longs to fill in the community . . . As with a bird, singing appears to be the natural outpouring of the life force itself. The prestige of clan and family depended not only on the prowess of its young men but also on the zealous way in which its women represented its interests in song and dance. For a group of girls the oigo was a means of announcing their presence and of differentiating themselves from the older married women; for an individual a way of expressing her idiosyncrasies (Owuor 1962: 52).

I’m still complaining,

Crying the ree ree ree,

I’m still complaining,

Ever tearful with the ree ree ree,

I’m still complaining;

The redo-singer’s unceasing complaint,

Scion of young women

Still complaining,

Ever tearful with the ree ree ree,

I’m still complaining.

I am in love with the oigo;

I cry the ree ree ree

Infatuated with the oigo;

The redo-singer’s unceasing complaint

Blasting Amimo’s hearth

With constant complaining;

Ever tearful with the ree ree ree,

I’m still complaining.18


Songs in Africa are very frequently in antiphonal form. That is, there is response of some kind between soloist and chorus, and the song depends on the alternation between the two parts. The role of the soloist (or ‘cantor’) is crucial. It is he who decides on the song, and when it should start and end. Even more important, he can introduce variations on the basic theme of the song in contrast to the part of the chorus, which is more or less fixed. In other cases, the soloist has complete scope to improvise his part of the verse as he chooses (apart perhaps from the very first line). This type of composition results in many impromptu and often ephemeral lyrics.

Within the general antiphonal form, which has often been mentioned as one of the main characteristics of African song, there are several possible variations. This is partly a question of who the performers are. Sometimes, for instance, there is more than one cantor; two or even three may interchange verses with each other as well as with the accompanying chorus. In other special musical types, the singers take turns in leading the singing, or two answer each other’s song. But, as will appear, even in the most basic type (one leader/one chorus) there is scope for variety and elaboration.

One of the simplest forms, at first sight at least, and one that seems to occur widely in Africa, is repetition of two phrases between soloist and chorus. Nketia terms this pattern the ‘call and response’ form and shows how even this type of antiphony can be elaborated in actual performance (Nketia 1962: 28ff).19 At its simplest level, one that occurs, for instance, in children’s games or other action songs, there is merely a repeated interchange between leader and group, the first singing his own phrase (A), the chorus coming in with theirs (B).20 But there are also more complex forms.

Various techniques of elaboration of the basic A-B form may be employed. Variations in text, in melody or both may be introduced in the cantor’s phrase (A) while the balancing responsive phrase (B) sung by the chorus remains the same. Interest may be further enhanced by varying the beginning and ending point of the cantor’s phrase in such a way as to make this part overlap with the chorus response. In addition to these, a little elaboration in the form of a short introduction based on the words of the song may be sung by the cantor or by a member of the chorus who wishes to start a new song before the leading phrase (A) is begun. Examples of this will be found in the music of kple worship of the Ga people. It is also greatly exploited in Adangme klama music (Ibid.: 29).

Songs founded on this type of repetition are basically short, though the actual repetitions may be drawn out almost indefinitely. Further extensions of the basic principle are also common. One might be built up on a kind of sequential pattern so that A and B are repeated at different levels, resulting in a form of A B A1 B1. The complete unit (now of four sections, or even of six, eight, or more) can be repeated several times over. In this type too the cantor is at liberty to introduce slight variations, melodic or textual. The words of the chorus usually remain the same, though in some elaborations they are changed while the cantor’s part stays the same. As can be seen, many other combinations are also possible—like, for instance, the A1 B, A2 B, A3 B pattern of many Limba songs.

All these elaborations of the ‘call and response’ pattern basically involve the balance of sections sung by leader and chorus against each other, and depend essentially on repetition. This raises the problem of how the song is ended. Sometimes the end is abrupt and the leader simply stops; but at other times he joins in the chorus response, often with a prolonged final note. In other songs there is a special closing refrain.

Another type of antiphonal collaboration between leader and chorus is the ‘solo and chorused refrain’ (Nketia 1962: 30–1). In songs of this pattern there is not the same balanced alternation between the two parts. Instead the soloist merely introduces the song. The cantor might sing the entire verse of the song right through once, and this is then repeated by the chorus. An example of this is the simple but effective Ghanaian song:

I sleep long and soundly;

Suddenly the door creaks.

I open my eyes confused,

And find my love standing by.

Mother Adu, I am dying.

Adu, kinsman of Odurowa,

What matters death to me?(Nketia 1963a: 37)21

In the Ewe nyayito dance songs, in which new words are continually being composed, the first cantor sings the whole song through unaccompanied. By singing in a dramatic tone he can encourage people to join the dance (Jones 1959: 75). In other cases, the cantor sings only a short introductory phrase, and the chorus then sings the main song. The form is highly flexible:

When the cantor has sung through, he may sing a short leading phrase before the chorus comes in. This leading phrase may also be added to a cantor’s introduction. Further, the main chorus refrain can be interrupted by a cantor at appropriate points . . . Furthermore a number of cantors may take turns at leading the chorus. Either of them may sing an introductory phrase before the chorus comes in, or they may take turns at leading each new verse. Sometimes cantors singing in twos are encountered. All these show that this form is flexible, and that there is room for building up complex sectional patterns on the basis of the singing roles taken by the participants.(Nketia 1962: 31)

There are other possible variants. There are various combinations of the two main types described, including songs like the well-known Adangme klama, which open with an introductory section by the cantor sung in free rhythm, followed by a section in strict tempo with a solo lead and chorus refrain (or overlapping solo and chorus parts), repeated three or more times; each new stanza can then be treated in much the same way as the song proceeds (Nketia 1958a: 28). Sometimes basically solo songs in declamatory style are supplemented by a chorus or instrumental addition (Nketia 1962: 31). In other songs the antiphony is between two soloists rather than solo and chorus. Thus the Kassena-Nankani of northern Ghana have a special type of song in which a young man who wishes to sing the praises of a girl conventionally asks the assistance of a friend: this results in a kind of duet by the two men accompanied by gourd percussion (Nketia 1962: 27). Alternatively the antiphony may be between two choruses. This is the common pattern, for instance, in the Limba women’s song which accompanies the boys’ gbondokale dance and involves almost endless repetition of only a few phrases. Another example is the Zulu wedding song where, after the leader has stated the theme, it is taken up first by the chorus of women, and then by the men who answer with a contrasting theme, overlapping with the women’s singing (Cope 1959: 35).

It is clear that the antiphonal form provides scope for far more flexibility, rich elaboration, and varied interpretation than is immediately apparent from the bald statement that this is the characteristic structure of African songs. It is also a most suitable form for the purposes to which it is put. It makes possible both the exploitation of an expert and creative leader, and popular participation by all those who wish or are expected to join in. The repetition and lack of demand on the chorus also make it particularly appropriate for dancing. Finally the balanced antiphony both gives the poem a clear structure and adds to its musical attractiveness.

We must not, however, exaggerate the significance of this very common antiphonal type of song and thus overlook the fact that some songs are primarily for soloists only. Thus one of the song types recorded from Zambia, the impango, seems to be designed primarily for solo singing (Jones 1943: 11–12). Men among the Bushmen sing personal and plaintive songs as solos (Lomax 1962: 438–9); and certain types of songs—such as lullabies and sometimes love and herding songs—always tend to be sung by individuals. Such songs can develop the verbal content, unlike the antiphonal songs that normally seem to involve a lot of repetition. It is by no means always clear in the sources how far a song is in fact sung by chorus and leader and how far just by one person, because those taking down texts tend to avoid repetitious phrases and to transcribe the song as if it were sung by one person only. The Akan ‘maiden songs’ are a good example of how one could easily assume that there is only one singer. Nketia in fact, with characteristic precision, explains that these are sung by groups of women, each taking it in turns to lead the verses of the song; in the case cited here the last three lines are sung by the chorus. But in most other sources this explanation would not be added and the words would have suggested a single singer. The song is in honour of a loved one:

He is coming, he is coming,

Treading along on camel blanket in triumph.

Yes, stranger, we are bestirring ourselves.

Agyei the warrior is drunk,

The green mamba with fearful eyes.

Yes, Agyei the warrior,

He is treading along on camel blanket in triumph,

Make way for him.

He is coming, he is coming.

Treading along on sandals (i.e. on men).

Yes, stranger, we are bestirring ourselves.

Adum Agyei is drunk.

The Green Mamba, Afaafa Adu.

Yes, Agyei the warrior,

He is treading along on camel blanket in triumph, Make way for him.

(Nketia 1958b: 20; see also Nketia 1963b: 51)

The musical side of these lyrics, unlike spoken or semi-chanted poetry, is of vital importance. The verbal expression and the melody of the song are interdependent. So much is clear—but beyond this there are many areas of uncertainty. For one thing, the relative weight given to melody and to verbal content seems to vary in different areas and between different genres of song. For instance, the work songs designed to accompany and lighten rhythmic labour lay little stress on the words, and much more on the melody and rhythm, while in love songs the words take on greater interest. Further, there seems to be no firm agreement among musicologists about how far, when discussing African lyrics, one can generalize about such matters as scale, melody structure, rhythm, and harmony;22 few detailed studies have been published for particular areas.23

One point of interest is the question of the exact connection between spoken and sung tone, especially in the highly tonal languages characteristic of parts of Africa. Again, there is some controversy on this score, but it seems clear that there is often a relationship between the tones of speech and the melody, so that the melodic pattern is influenced by linguistic considerations. This is well documented for some West African languages. The relationship seems to be flexible, with the possibility of variation and tone modifications. Nketia sums up the position for the several Ghanaian tone languages:

What the intonation of a song text provides . . . are tone patterns or syllable relationships and not the actual melodic notes that are to be employed. We would not, in traditional Ghanaian music, expect a high tone always to be sung in the upper or middle compass, or a low tone in the middle or lower compass. Within each compass we would only expect the melodic working-out of high, mid and low tone relationships. The verbal intonation would not provide us with the beginning or ending tone, but it may guide the immediate direction of movement from the beginning tone or movement towards the ending tone . . . The tonal relationship between words and melody is not rigid. It is flexible. While it is important to take the intonation curve into account so that the words of a song may be readily recognised, it must be emphasised that the ‘art’ of the song lies in the departures that are made from this guide where appropriate, on purely melodic grounds. Thus the use of ascending interlocking patterns or pendular movement where the intonation shows a descending trend, or the use of rising seconds where intonation is level belongs to the ‘art’ of the song. However, it would be as wrong to assume rigid relationship as it would be to conclude that because such deviations occur, the tones of words are unimportant in the construction of melodies (Nketia 1962: 52).24 Thus, although tone/melody relationships in these languages allow a certain degree of freedom, the link between the two is a complex one, and composition and extemporization demand a high degree of skill.

A further vexed question is that of rhythm. The fundamental importance of rhythm in vocal as in other African music is widely accepted, but there is little agreement as to its exact structure. One helpful distinction is between songs in ‘free’ and those in relatively ‘strict’ rhythm (Ibid. 64). In the former songs (or portions of songs) the singing is not co-ordinated with any bodily rhythmic activity such as work or dancing. The very common songs to strict time, however, have a beat that is articulated with dancing, rhythmic movement, percussion by instruments, or hand-clapping, all of which contribute to the form and attractiveness of the song. These rhythms are worked out in many different ways in various types of song, but one commonly recurring musical feature seems to be the simultaneous use of more than one metre at a time, as a way of heightening the rhythmic tension.25

The accompaniment takes many different forms, depending, among other things, on the geographical area and its resources,26 on the genius of the particular people, and on the different genres within a single culture. It is common, for instance, to find some types of songs regularly without accompaniment, others with just clapping and/or dancing, others again with many different kinds of instrumental accompaniment, conventionally graded according to the song, the singers, or the occasion.

When we come to the verbal style of these poems, it is almost impossible to generalize. As would be expected in poetry, there is a tendency to use a language somewhat different from that of everyday speech. This is particularly evident in the case of sung lyrics, where the melodic line imposes its own requirements, and in tonal languages, where there is the additional complication of the relationship between tune and tone. Connected with the importance attached to the musical aspect in these relatively short, sung lyrics is the frequent occurrence of meaningless words and onomatopoeic sounds which fill in the line, add length to the song as actually performed, and are used especially in chorus responses. Some songs, too, tend to be verbally fragmentary rather than fully developed poems as far as the words are concerned, though the fragments themselves may have a terse poetic interest (e.g. the examples in Nketia 1958b: 15–16). But there are many variations between different types of songs, each with its own style and diction, and, indeed, in contrast to comments on the subject-matter and contexts of songs, there is relatively little published work available.27


How far can these lyrics be said to be truly personal expressions of experience? This raises the difficult question of composition—difficult mainly because so little interest seems to have been shown in this aspect of African poetry. Many commentators, even when they try to take this into account, content themselves with labelling specific songs as ‘traditional’ or ‘improvised’ without considering in what senses these words are used. But even this is better than the other still common approach of apparently explaining away the problem by classifying the lyrics as ‘folk-songs’ which can, it is then assumed, be happily attributed to ‘the folk’, so that the question of composition does not arise at all.

It is certainly clear that some songs retain their popularity for many years. This may happen less to incidental and recreational songs (like most of the lyrics described here) than to songs definitely tied to particular solemn occasions such as initiation or religious ritual. A common pattern—demanding further research—may be for the music to remain basically the same while the words change.28 But even with light-hearted dance songs it does seem that some (words as well as music) remain popular for so long that they might with some justice be termed ‘traditional’. Others, however—and this is much more commonly mentioned in recent sources—are ephemeral only. The Ibo, for instance, are said to create impromptu poems all the time and forget them (Green 1948: 842). Again, among the Kamba most songs are ‘improvised’ (with the exception of circumcision songs, which reappear in the same form again and again), and with dance songs the leader of the singing and dancing must make a new one when the old one is worn out—about every month or so (Lindblom iii, 1934: 40). Similar comments have been made about songs among many African peoples.

Even with a familiar song there is room for variations on words or tune in actual delivery so that each performance in a sense may be a ‘new’ song. It must be remembered that these variations on a basic theme are more likely in societies that do not share our stress on the fixing nature of the written word, the concept of a single ‘correct’ form attributed to a single author. Even such obvious points as the number of repetitions used by a particular leader, the order of the verses, the variations by instruments in an accompanied song, and the varied movements of dancers—all these contribute to the finished work of art as a unique performance of which the verbal text of the song is only one element.

There is one further aspect. The leader of the song adds new verses arising from the basic themes recognized by him and the chorus. Tracey describes this process in Southern Rhodesia. The chorus parts of a song are expected to remain the same, but the soloist (mushauri) introduces the song and is allowed full scope for originality during its performance. If he is not able to compose his new verse swiftly enough to keep his initiative, he either repeats the last verse several times to allow himself time for thought or, if necessary, yodels the tune, and finally sings to his neighbour to replace him in the lead (Tracey 1929: 97). This pattern by which the antiphonal form is exploited through improvisation by the leader and relatively unvaried support by the chorus seems to be very common indeed. Unless there are definite reasons for retaining sanctioned words, it seems generally rather rare for such songs to be repeated exactly from performance to performance—there is always scope for some variation by the leader (this is apparently also sometimes extended to the improvised performance of quite ‘new’ songs in terms of the melody and the form of the words. At least in some cases, choruses are quick to pick up the melody and words, often after having heard them just once or twice from the leader, and to sing them enthusiastically even though they were previously unknown to them).

But one must not be so impressed by the excellences of African improvisation that everything is attributed to spontaneous creation. There is, first, the obvious point that improvisation takes place within certain conventional artistic forms known both to the soloist and also, perhaps equally important, to the chorus. More significantly, certain commentators make it clear that serious and conscious composition also takes place.

One of the more detailed accounts of such composition is given by Tracey in his description of musical composition among the Chopi. The Chopi ngodo is an orchestral dance in nine to eleven movements that certain skilled and known musicians compose anew every two years or so. The stress is on the music and its elaboration. It is worth quoting his description at some length here, for this dependence of the words on the music is by no means unparalleled:

A description of how Katini and Gumukomu set about composing a new orchestral dance will show how musically advanced these men are. Both of them say that the first thing they do is to find appropriate words for their song and compose the verses of the lyric before the music.29 The subject-matter may be gay, sad, or purely documentary. In every case it is highly topical and appropriate to the locality, so much so, in fact, that most of the allusions would be caught only by those in close touch with the villagers and the district.

To return to the composer: when he has decided upon the words of his poem, or, in the case of a long poem, the opening verse, he must now find his melody. Chichopi, in common with other Bantu languages, is a tone language, and the sounds of the words themselves almost suggest a melodic flow of tones. This is developed rhythmically, as Gilbert and Sullivan did in their light operas, in one or other of the well-defined patterns which characterize their national verse, with clever use of repetition and offset phrases. The verses are not always metrically alike, as one would naturally expect of a tone language, but all bear a family relationship to the prototype lines. As often as not, the final verse sung to the coda is a repeat of the statement or first line of the poem. In this they follow a well-recognised trick of the trade which is exploited so frequently in our own popular songs . . .

The verse and the leitmotive now fixed in the composer’s mind, he sits at his instrument [xylophone], over which his hands wander with expert deftness, and picks out the melody . . . After a while, during which his right hand becomes accustomed to the new tune, his left will begin to fill in the harmonies or contra-melody with 1 well-understood sequences, punctuated with rhythmic surprises suggested by the ebb and flow of the words. Now the right hand will wander away from the melody, mapsui, into a variation, kuhambana, and as he sings the words over to himself the contrapuntal accompaniment will begin to form under his hands . . .

They now have the primary melodic line of the poem—the subject or leitmotive—and the secondary melodic accompaniment—the orchestral sentence—which fits the words contrapuntally, with a number of variations and sequences . . .(Tracey 1948a: 2–3; 4–5)

Though the lyrics and their music are topical and relatively ephemeral, they are certainly not totally impromptu; in describing the process of their creation we can more suitably speak of artistic inspiration coupled with studied technique than of ‘improvisation’.

Something of the same process occurs with several song types in Zambia (Jones 1943). Among the Ila and Tonga there is commonly an interest in the personal ownership of songs: individuals are often expected to sing one of their own songs—a young man on the day of his marriage, for instance, a young girl on the day she is allowed to wear adult dress. Among their many types of songs are those called impango. These are sung by women only, at beer drinks or at work, and each woman must have her own personal repertoire of impango songs to sing as solos. One woman stands up at a time and sings her song in a very high and fast style. Meanwhile her intimate friends or her relatives may get up from time to time and interrupt the song with praise and small gifts. Impango composition is known to be difficult, and in every village there are a few women who are especially skilled in this art. What happens when a woman wants to make an impango is that she first thinks out the rather lengthy words—it may be praise of herself, her lover, or her husband—and then calls in some of her women friends to help her. Together they go to a well-known maker of impango songs. After hearing the woman’s ideas, she then, often over a period of several days, composes the complete tune for the whole song. She calls a party of women to practise it each evening after supper, and they continue until the impango is complete and has been mastered by the whole party. The group is then disbanded and the woman who ‘owns’ the song continues to sing it on her own, knowing that if she forgets at any point she can ask one of the practice party to help her. She is now fully mistress of her impango and proud of her accomplishment. Whenever she is invited to a festival she keeps ‘singing it in her heart’ until it is finally time for her to stand up and sing it in public (Jones 1943: 11–12).

The composition of another type of song, the inyimbo, is a simpler matter. The same sort of procedure is followed, but as these songs are shorter and simpler, the process is quicker. There are three main forms of this type of song, and the correct one must be used. The typical occasion of performance is for people to gather and sit down, and then start clapping or beating with sticks. A man or woman then stands up and dances; and as the owner of the song sings it right through, people pick it up and then sing it through themselves several times, followed by the owner again, then back to the group. There are also other types of song: the mapobolo song is characterized by brief words and a short tune that a woman first composes herself (working out at least the words or the tune), her friends then helping her to complete it before the actual performance in antiphonal form; while the zitengulo or women’s mourning songs are composed completely by the individual, with no help from others; she starts to sing little by little and gradually adds the words and melody until the song is complete (Ibid.: 13–15).

There are, then, many different forms of song among the Ila and Tonga, and each has its own recognized mode of composition. What is striking is the emphasis on the care involved in composition and on the idea of personal ownership.

Song composition in non-literate cultures almost necessarily involves co-operation, particularly where there is an accompaniment by chorus, instruments, or dancing, and where, as so often in African lyrics, there is an emphasis both on performance and on participation by the audience. But that there can also be a purely personal element of the greatest significance in moulding the song is clear from the Chopi and Zambian examples.30 How far this personal contribution is recognized by the people themselves seems to vary; even within one group certain songs may be regarded as the property of named individuals, while others are not.31 But it is quite possible that further investigation of a topic that has hitherto been ignored will show that many other African peoples besides those mentioned engage not only in the art of improvisation but also in a process of long-considered and reflective individual creation.

1 In particular the work song which could have been treated under the present heading.

2 See Lomax 1962 on the general contrasts in this respect between Africa, Europe, the Orient etc.

3 For other examples of marriage songs see e.g. Beaton 1935 (Bari); Leslau 1947 (Harari), Dufays 1909 (Ruanda).

4 i.e. a boast by the (Johannesburg) dancers that no one can dance better than they—their reputation has even reached Durban!

5 Nearly all collections of poems include some dance songs. See also Stappers 1954; Emsheimer 1937 (not seen); Beaton 1940, 1938; v. Funke 1920–1; Clark 1965; Traoré 1942; Littmann 1926; Vansina 1955; Nketia 1957.

6 On which see below.

7 On the function of songs in stories see Ch. 13: 385. Also Belinga 1965: 55ff.

8 Apparently nonsense words.

9 Meat is often pounded in a mortar, then moulded into balls.

10 A fabulous river.

11 i.e. won all the prizes.

12 The line division is not quite clear in the text and I may have interpreted it incorrectly in places.

13 The shout given when the hare jumps up from its lair.

14 Animal songs also occur in Central Africa (Lamba) where they bear some resemblance to Southern Bantu prises (Doke 1934: 365).

15 See Mayssal 1965: 81, on Hausa influence on the Cameroons Fulani.

16 Loeb’s general interpretation, however, is highly doubtful.

17 I write in the present though in fact these songs are now a thing of the past. The description is taken from Owuor 1961.

18 Owuor 1962: 53. Other references on love songs include Tracey 1963: 19–20 (examples and general discussion); Knappert 1967a (Swahili); G. Schürle and A. Klingenheben, ‘Afrilcanische Liebeslieder’ (Duala and Zaramo), ZKS 3, 1913–14; Chadwicks iii, 1940: 668ff. (Tuareg); E. Von Funke, ‘Einige Tanz-und Liebeslieder der Haussa’, ZES 11, 1920–1; Tescaroli 1961, Ch. 4 (Sudan); E. Cerulli, ‘Poesie di guerra e di amore dei Galla’, Arch, antrop. e etnol. 5, 1942 (reference in IAI Bibl. (A) by R. Jones, North-East Africa, 1959: 33)’; D. Earthy, ‘A Chopi Love-song’, Africa 4, 1931. For other discussions or examples of ‘lyrics’ see J. Vansina, ‘La chanson lyrique chez les Kuba’, jeune Afrique 27, 1958; T. Tsala, ‘Minlan mi mved (chants lyriques)’, Recherches et etudes camerounaises 2, 1960 (Beti); L. Longmore, ‘Music and Song among the Bantu People in Urban Areas on the Witwatersrand’, Afr. music Soc. Newsletter 1. 6, 1953; and references in following sections. For written Swahili forms see Knappert 1966: 128f., 136.

19 I draw heavily on Nketia’s analysis here: though he is working primarily on Ghanian music, his analysis has a wider application; see also Rycroft 1967.

20 The children’s singing games quoted in Ch. 11 include some examples of this basic form.

21 The emotion of love is, as often in Akan poetry likened to that of suffering and death.

22 For recent general discussions, see Merriam 1962, 1965; Jones 1959, Ch. 9; Rouget 1961; Tracey 1964; Adande 1952; Nketia 1964.

23 Though see Tracey 1948a (Chopi); Jones 1943, 1949 (Zambia), 1959 (Ewe); Nketia 1962 (Ghana); Brandel 1961 (Central Africa); also further references in Merriam 1965 and Gaskin 1965.

24 Similar points have been made by other writers: e.g. on Ewe ‘tone and tune’ see Jones i, 1959, Ch. 10; Igbo, Green 1948: 841, Wescott 1962; Yoruba, King 1961: 38ff.; Bantu generally, Westphal 1948; West Africa, Schneider 1961; Chopi, Tracey 1948a: 4–5; Ngala, Carrington 1966/7 (reference in Africa 38, 1968: 110); also general discussion in Wangler 1963.

25 On rhythm, see Merriam 1965: 455–6 (and further references given there).

26 See Tracey 1954a: 8–9, on the effect of the environment on the choice of instruments.

27 One exception is the consideration of the Somali balwo in Andrzejewski 1967.

28 In Tanganyika, for example, the poet seldom composes the time, but is free in his choice of the text; thus it is rare for a tune to be associated with one text only (Koritschoner 1937: 51). I noticed a similar pattern with certain types of songs, particularly dance songs—in Limba; see also Helser 1930: 65 (Bura), Andrezejewski 1967 (Somali).

29 This is not necessarily the most common method of procedure. Contrast, for instance, Ngoni composition where ‘it is always a single inspiration which leads the composer to find the right words and the right music’ (Read 1937: 3).

30 Cf. also the Luo nyatiti songs mentioned in Ch. 4: 90ff., and the Somali poets who spend hours or days composing their works (Andrzejewski and Lewis 1964: 45).

31 E.g. Hurutsche (Merwe 1941: 307). For some other discussions of the process of composition and attitudes to it see Babalola 1966: 46ff. (Yoruba) de Dampierre 1963: 21ff. (Nzakara); Read 1937: 3 (Ngoni); Nettl 1954b, 1956: 12–19 (general).