The languages spoken in British Central Africa belong to the great Bantu family, which, as is now known, occupies (with a few exceptions) the whole continent of Africa south of a line drawn from the Gulf of Cameroons to the mouth of the Tana River on the east coast. Those spoken within the Protectorate are Nyanja, Yao, the Lomwe dialect of Makua, Tonga, Tumbuka, Nkonde, and a Zulu dialect spoken by the Angoni clans. In Northern Rhodesia we may mention Bisa, Bemba, Luba, and Lunda as the principal languages.

All the Bantu languages are as closely related together as English, Dutch, German, and the Scandinavian dialects. There are several points about them which are extremely interesting to the comparative philologist. They have no grammatical gender—the same pronoun is used for a man and a woman; and, accordingly, most natives who learn English come to grief on this point, like Winwood Reade’s interpreter who asked: ‘What you say when him son be girl?’ On the other hand, nouns are divided into eight or ten classes, each with its own plural inflection, and adjectives and pronouns agreeing with it as they agree with each of the three genders in Latin. This agreement extends also to the verbs.

The Bantu languages further differ from those with which most of us are familiar, in that their inflections are indicated, not by suffixes, but by prefixes—a fact which first meets us in the various and perplexing forms assumed by the names of tribes and countries. Thus Myao is ‘a Yao’ (man or woman), Wayao is ‘Yaos,’ and Chiyao the Yao language. Each noun-class has its own prefix (sometimes much atrophied or even dropped altogether) for singular and for plural, and though these prefixes vary greatly in the different languages, they are always recognisable as having come from the same original, just as we know that the English oak, the German Eiche, the Dutch eik, and the Danish eeg are all derived from one primitive form. The inflectional prefixes of adjectives and verbs are derived from the noun-prefixes, though not always identical with them in form; and the pronouns are modifications of the prefix. In fact, broadly speaking, the prefix may be called a pronoun, and the group of languages under consideration are sometimes called the prefix-pronominal languages.

The careful reader may think that a somewhat Hibernian assertion has been made above—viz. that the prefix is recognisable even where it has been dropped; but this is in fact the case: the pronoun, which must be inserted before the verb, always shows what the lost prefix of the noun has been. Thus we have in Nyanja the word njoka, ‘a snake’; it has no prefix as it stands, but when used in a sentence we find it takes the pronoun i: njoka i luma, ‘the snake bites.’ Now in Zulu, which has kept its prefixes better than Nyanja, we find that ‘snake’ is inyoka.

This principle of agreement, by which all the words governed by the noun repeat its prefix in some form or another at their beginning, is called the alliterative concord, and may be illustrated by the following sentences:—

Nyanja

The pronoun for the class to which mtengo (anciently umtengo) belongs is u, which is quite clearly seen before the verbs ‘to be’ and ‘to fall.’ In the possessive pronoun and the adjective, it is a little disguised, because it becomes w before another vowel (watu = u + atu).

The plural of this is:

Mitengo yatu i-ri yetari, i-dza gwa.

Here the pronoun is i, or, before a vowel, y. R and L are interchangeable; the verb ‘to be’ is usually li after a and u, ri after i.

Another class is thus exemplified:

In the plural:

Ziko zanga zabwino zidzasweka.

Yao

Here the prefix is lu.

(‘This basket is mine, it is not lost; yours is lost.’)

In the plural: Mbeta asi sili syangu, nginisijasika, syenu ’sila sijasiche. The plural prefix corresponding to lu is izim (isim), or izin, and p following this prefix changes to b; hence (isi)mbeta is the plural of lupeta.

There are very few adjectives in the Bantu languages; their place may be supplied by a noun preceded by the possessive particle corresponding to the word ‘of’: thus chiko chabwino is literally ‘a gourd of goodness.’ There are a good many verbs which can be used where we should use adjectives, as ku ipa, ‘to be bad’ (ku is the sign of the infinitive); ku uma, ‘to be dry,’ etc.

Verbs can express, by means of changes in the stem, a number of modifications in their meaning which we have to convey by separate words. These modifications are usually called ‘forms,’ but are really extensions of the principle of ‘voice.’ We have to be content with two voices—the active and passive, with traces of a middle; Hebrew has seven; some Bantu languages have as many as nine or ten; while, counting the secondary and tertiary derivatives, and the compounds, the late W. H. Bentley reckoned out over three hundred forms of one verb, all actually in use, in the language of the Lower Congo.

The aspirate exists neither in Yao nor Nyanja, and when heard in English words is often turned into S; thus the name Hetherwick becomes Salawichi. But the people west of the Shiré use it in words and names borrowed from the Zulus, and seem to find no difficulty with it. L and R are interchangeable, as already stated, or rather it would be more correct to say that the sound intended is really distinct from both and heard by some Europeans as l, by others as r. There are no very difficult sounds, except perhaps ng (pronounced as in ‘sing’) when it comes at the beginning of a word. There are no clicks in any language used in the region under consideration, except the Zulu spoken by some of the Angoni, and in this they tend to disappear. The accent is almost invariably on the penultimate.

The Bantu languages have, of course, no written literature—for we can hardly count the translations, etc., produced by missionaries and their pupils, or even the two or three native newspapers appearing in Cape Colony and Natal. But like most primitive tongues, they are rich in traditional tales, songs, proverbs, etc. Of the folk-stories we shall give some examples in the next chapter. Here are some specimens of Nyanja proverbs:—

‘If you are patient, you will see the eyes of the snail.’

‘Speed in walking in sand is even.’ (‘Bei Nacht sind alle Katzen grau.’)

‘You taste things chopped with an axe, but meat cut up with a knife you don’t get a taste of.’ (The sound of the axe directs passers-by to the place where the food is being prepared—perhaps inside the reed-fence of the kraal—when, of course, they must be asked to partake; had a knife been used, they would have heard nothing, and gone on.)

‘If your neighbour’s beard takes fire, quench it for him’—i.e. you may need a similar service some day.

‘When a man or a reed dies, there grows up another.’ (‘Il n’y a pas d’homme nécessaire.’)

‘Sleep has no favourite.’

‘Lingering met with liers in wait.’

Riddles, as already mentioned, are very popular. They are usually of the simple kind which describes some well-known object in more or less veiled and allusive language, something after the style of

but much more crudely expressed.

‘I built my house without a door’ is one which has the same answer as the above—viz. ‘an egg.’

Others are:—

‘Spin string that we may cross the river.’—A spider.

‘The people are round about, their chief is in the centre.’—A fire, and the people sitting round it.

‘I saw a chief walking along the road with flour on his head.’—Grey hair.

‘Such an one built his house with one post only.’—A mushroom.

‘A large bird covering its young with its wings.’—A house—referring to the roof with its broad eaves.

‘My child cried on the road.’—A hammer.

‘The sick man walks, but does not want to run, but when he sees this, he runs against his will.’—A steep hill (which forces people to run in descending it).

At Likoma they have a set form for riddle-contests, as thus: A. begins, ‘A riddle!’ The rest reply in chorus, ‘Let it come!’ A. ‘I have built my house on the cliff!’ All guess; if their guesses are wrong, A. repeats his riddle. If they still cannot guess right, they say, ‘We pay up oxen.’ A. ‘How many?’ They give a number. If A. is satisfied, he will now explain his riddle—‘the ear’ being the answer to the one given above. If any one guesses right, all clap their hands, and another player asks a fresh riddle.

Another popular amusement might be described as a ‘debate.’ Boys and grown men both delight in it, though with the former it is sometimes the prelude to a fight. One of a party sitting round the fire, or wakeful in the dormitory, will say, ‘Tieni, ti chita mákani’—‘Come, let us have a discussion,’—and will start it, perhaps, by asking whether a hippopotamus can climb a tree. The arguments for and against the proposition are then advanced with the greatest eagerness, till the point is settled, amid volleys of laughter, or the company tired out.

I have never, in Nyanja, come across any of the curious itagu (‘catch-word compositions’) which the Yaos delight in, and which are recited by two or more speakers. The following specimen of a duologue is given by Mr. Duff Macdonald:—

It will be seen that the second speaker repeats the word given by the first (or something like it), and adds another to it, while the first in like manner catches up his last word, or part of it, sometimes giving it a different sense. It is almost impossible to translate this sort of thing, but the following composition on the same lines may serve to show how it is done.

Here, of course, there is no pretence of connection, but the itagu are really connected stories. The language of these itagu is very difficult; either because they are very old, or because words are purposely distorted.

Songs are numerous, and continually improvised afresh as wanted, though many old traditional ones are current, some of which are embodied in tales, and sung in chorus by the audience when the narrator comes to them. Natives nearly always sing when engaged in concerted work, such as paddling, hauling a heavy log, carrying a hammock, etc. They sometimes sing in unison, but not unfrequently in parts. Very often one sings the recitative, another answers, and others add the chorus. There is no metre, properly so called, in the songs, but there is a sort of rhythm, and they usually go very well to chants. Both Yao and Nyanja are exceedingly melodious languages, and it is possible, though not easy, to write rhymed verses in them, especially in trochaic metres, which violate no rule of accent or construction. Many, if not most, however, of the European tunes which have been adapted to native words in mission hymn-books are hopelessly unsuitable, and the result, as regards the accentuation of the words, is sometimes nothing short of grotesque.

No systematic study has yet been made of the native melodies by means of phonographic records; a few of the Nyanja and Chikunda songs have been written down, more or less tentatively by ear, and a good many Chinamwanga tunes have been noted down by Mrs. Dewar, of the Livingstonia Mission. These last, which are all associated with stories, come from a district outside the bounds of the Protectorate, about half-way between Nyasa and Tanganyika.

In general the character of all Bantu music is much the same; the singing has a curious, monotonous, droning effect, which, however, is not without its charm, when heard amid the proper surroundings. It is sometimes said that all the melodies are in the minor key; but this is a mistake. M. Junod, who has made a very careful study of the music of the Baronga, says that the effect which gives rise to this impression is produced by the songs beginning on a high note and descending; and this turns out, on examination, to be the case with many of those collected by Mrs. Dewar, though the height of the opening note is often only comparative. As a specimen, I give the melody (as written down by Mrs. Pringle of Yair) of the famous canoe-song Sina mama.

The meaning of the words (collated from two printed sources and my own notes) is: ‘I have no mother, I have no father; I have no mother, Mary, I have no father; I have no mother, to be nursed by her; I have no mother—thou art my mother, O Mary!’ This song is often heard on the Shiré; but, containing as it does, a faint echo of Romanist teaching, probably originated in one of the Portuguese settlements on the Zambezi.

Another Shiré boat-song is Wachenjera kale, which, when I heard it, I took for a very à propos improvisation, having, I suppose, utterly forgotten the following passage from Livingstone’s Zambezi Expedition, which I must have read, but which struck me as quite new when I came across it a few months later. ‘In general they [the men of Mazaro or Vicenti on the Lower Zambezi] are trained canoe-men, and man many of the canoes plying to Senna and Tette; their pay is small, and, not trusting the traders, they must always have it before they start.... It is possible they may be good-humouredly giving their reason for insisting on being invariably paid in advance in the words of their favourite canoe-song, “Uachingere [uniformity in spelling African words is, even now, not much more than a pious aspiration], uachingere kale,” “You cheated me of old,” or “Thou art slippery, slippery truly.”’ I prefer the former rendering; and, moreover, my men repeated the Wachenjera thrice. There seemed to be no more of the song.

A very pretty corn-pounding song heard at Blantyre is as follows:—

It is not easy to get a satisfactory translation of this, though the words, on the face of them, are not very difficult. ‘Gu! gu! (the sound of the pestle descending into the wooden mortar)—I am going to pound corn; father! mother! Zandia! gu! (I take Zandia to be a proper name). You, child, why are you crying?—They are clever (or, they cheated me) at your mother’s—to cover me up on the fire that I might be burnt—Zandia!’ The ‘child’ addressed is perhaps the corn in the mortar, which cries out and complains of being crushed (‘burnt’), or it may be meant of some maize-cobs put down to roast while the pounding is going on, which may be heard popping and crackling.

Several songs I have taken down are full of allusions to local chiefs and events of which I did not succeed in getting the explanation; they show, at all events, how passing incidents are commemorated and kept in mind. One speaks of ‘Mandala, who ran away from the flag’ (mbendera—the Portuguese bandeira), and ‘Gomani (i.e. Chekusi), who died (or, no doubt, “was kilt entirely”—he being still alive at the time of recitation) in the dambo.’ This may refer to one of the many wars between Chekusi and Chifisi, or Bazale.

Some of the songs are difficult to understand, as, even if not very old, they abound in unfamiliar words and constructions, and also in local allusions, which need explanation to outsiders. One I have written down seems to be about Chekusi’s marriage, and brings in the names of several chiefs. Another says that ‘I have seen Domwe’ (a mountain in Angoniland)—‘Ntaja is dead,—we are ravaged this year.’ Another obscure effusion, after stating that something or other is at Matewere’s (a son of the famous Mponda), goes on to say that ‘I refuse (him or them) the oxhide shield,’ or, maybe, the oxhide to make a shield.

These are ‘Angoni’ songs, and recognised as such on the other side of the river, though the language is not Zulu, but ordinary Nyanja. The original text of the last named is this:—

The rhythm of the songs is rather indefinite; it resembles that of some sailors’ chanties—e.g. the well-known Rio Grande. They often consist of only a few words, repeated ad infinitum, with a refrain of meaningless syllables, sometimes mere open vowel sounds—as: e, e, e, e, o, o, o, o—wo ya yo ho, etc. In canoe-songs and the like, time is marked by the beat of the paddles, the rise and fall of the women’s pestles, and so on; at a dance, it is given by the drums. Some soloist usually leads off with an improvised line, which is either taken up and sung in chorus, or a response to it is so sung, and the principal performer continues till he has exhausted his idea. If the song ‘catches on,’ it is remembered and repeated, and passes into the common stock. Some dances have their recognised songs, as ‘Kanonomera e! e!’ at the Angoni women’s kunju dance, and ‘Leka ululuza mwana hiye! (Stop winnowing, child!)—e! e! e! e!—o! o! o! o!’ at the chamba dance.

Singing, music, and dancing, or other rhythmic action, are very much mixed up together, as is always the case in the elementary stages of those arts; and a combination of all three is practised by the itinerant poet known as the ‘dancing-man.’ Of his instrument, the chimwenyumwenyu, Mr. Barnes says that ‘performers on it are rare and are most welcome guests in any village.’ It is a primitive kind of fiddle, with one string and a gourd resonator, played with a bow, which, when made, has its string passed over the string of the instrument, and so can never be taken off. The man in the illustration, however, appears to be playing on the limba, which has six strings strung on a piece of wood across the mouth of a large gourd, and is played with both thumbs. The gourd is hung round with bits of metal or of shells, to jingle and rattle when it is shaken. The ‘dancing-man’ teaches the children the chorus of his songs, and then, ‘carries on a dialogue of song with his audience, with the excitement and rhythm of an inspired improvisatore.’ Another kind of limba is that shown in the illustration, which was obtained from some Atonga:—a shallow wooden trough with a handle at one end, and pierced at top and bottom with six holes, through which a cord is strung backwards and forwards, and tightened up by winding round the handle. Like the other kind of limba, it is played with the thumbs. But the word limba is of wide application; it (or its plural malimba, marimba) sometimes denotes the xylophone or ‘Kafir piano’ (Ronga timbila), while natives use it for a harmonium, organ, or piano.

Other stringed instruments are the pango, resembling the dancing-man’s limba, but played with a stick or plectrum instead of the thumbs; the mngoli, the body of which is made like a small drum—it has one string with a bridge, and is played with a bow; the kalirangwe, with one string and a gourd resonator, played either with the fingers or a bit of grass; and the very primitive one (mtangala) represented in the illustration, which is played by women only, and is simply a piece of reed, slightly bent, with a string fastened at one end and wound on the other, so that it can be tightened up at pleasure. One end of this is held in the mouth and the string twanged with the finger, producing a very slight but not unpleasant sound, which, as Bishop Colenso remarked of a somewhat similar instrument in Natal, ‘gratifies the performer and annoys nobody else.’

The sansi has a set of iron keys fixed on a wooden sounding-box, and played with the thumbs; it has a piece of metal fixed on the front of the box, to which are attached small discs cut from the shells of the great Achatina snail, so as to clash when shaken, like the bells on a tambourine. A very similar instrument has the keys made of bamboo.

A flute (chitoliro) is made out of a piece of bamboo about a foot long, cut off immediately above and below the joints, so that it is closed at both ends, and from three to six holes bored in the side, some of which are closed with the fingers while playing. Bvalani, the boy with the coronet, used to play on this flute a pretty, though somewhat monotonous, little tune, consisting apparently of three or four notes, repeated over and over again; but neither I nor any other mzungu has yet succeeded in getting a sound out of the one in my possession. It has one hole at one end, to blow into, and three at the other.

Whistles, made out of a small goat’s or antelope’s horn, and used for calling dogs and perhaps for signalling to each other on the road, are worn round the neck by Angoni and Chipetas; and Pan-pipes are made of reeds. Trumpets are made of gourds, sometimes fixed with wax on a long reed; the same word, lipenga, is used for a horn employed in the same way, or for a European key-bugle, or (in hymns) for the Tuba mirum spargens sonum of the Last Day. Some of the people near the south end of Tanganyika have huge trumpets cut out of a large tusk of ivory, like those used on the Upper Congo and elsewhere. One such is figured in Sir H. H. Johnston’s book, p. 465.

The instrument above referred to as the ‘Kafir piano’ is, in a modified form, very popular throughout the Shiré Highlands and on the Lake, and may often be seen in the village bwalo. The Delagoa Bay timbila is portable (see the figure in M. Junod’s Chants et Contes des Baronga, p. 27), with the wooden keys fixed on a flat frame—elsewhere, the frame is curved into the arc of a circle, so that the performer can easily reach all the keys when the instrument is slung round his neck. In both cases, resonators, made of gourds, or the hard shells of the matondo fruit, are attached to the keys. I once saw a very elaborately made and beautifully finished specimen which had come either from Delagoa Bay or Inhambane, and had polished iron keys padded with leather; but this was a sophisticated timbila, scarcely the genuine article. The Nyanja form of it, variously called magologodo, mangondongondo, mangolongondo, and mangolongodingo, usually has to be played in situ, or, if removed, must be carried away in pieces. Two logs of soft wood (banana-stems are the best), perhaps a yard long, are laid on the ground a certain distance apart, and on these are arranged six, or sometimes seven, cross-bars (the Ronga ‘piano’ has ten) cut from the wood of certain trees, and carefully trimmed to shape. Sometimes they are merely laid on the logs, sometimes there are short pegs to keep them in place. The keys on the Ronga instrument are carefully tuned, and each one is cut away underneath in such a way as to make it give a different sound; but some of those I saw at Blantyre seemed to be merely rough bits of wood which fulfilled no condition beyond that of making a noise when struck. It is played by striking the keys with two sticks; the performer holds one in each hand and squats on his heels in front of it: sometimes there are two players, who face each other; the first leads, and the second is said to ‘make a harmony with the one who is playing.’

But the drum is perhaps the commonest and most characteristic instrument, and the one which has been brought to the greatest perfection. There are many different kinds, from the little kandimbe, a mere toy for children, four or five inches across, and tapped with the fingers, to the great mpanje and kunta, five feet or more in length, or the mgulugulu war-drum, which is beaten with sticks to call the people together. None of them have two heads; the body is made of a single piece of wood hollowed out, and the head of goat-skin, or perhaps oxhide; some small drums (more like tambourines) are covered with snake or lizard skin. The sound of the large ones can be heard five or six miles away. Some are beaten with sticks, some with the hand—either with the fist (as the big mpangula, which is supported on a forked piece of wood), or the open palm, or the fingers. Some of the smaller drums are held against the chest and beaten with the open hands, which gives a peculiar, soft, booming sound; one kind is held under the arm; another is laid lengthwise on the ground, and the drummer sits astride it. Still another has legs like a small round stool, and is beaten with two sticks as it stands on the ground. The mfinta drum (large, but not the largest kind) calls the people together when the mabisalila is investigating a case of witchcraft; it is also used in a dance where the performers carry hoes and strike them together. There is a wonderful variety in the notes; ‘the smaller drums are made to answer the big ones, the rapid and slower beats blending in the most perfect time.... There are skilled drummers who go to the dances like a piper at a Scotch wedding’ (Scott).

Drums are tuned when necessary by leaving them in front of a fire, or burning some grass inside them to dry the skin and draw it tighter. The skin is fastened on by small wooden pegs, and has a piece of rubber fastened to the middle of its underside.

Besides the drums, most dances require an additional sound-producing agency in the shape of rattles. These are worn on the arms and legs of the dancers, or shaken in their hands. The commonest kind are made of a hard-shelled fruit called tseche, about two inches or less in diameter; it is allowed to dry till the seeds shake about inside it, and then four or five are strung on a stick, and several of these sticks attached round the ankles of the dancers. Women never wear these maseche at their chamba dance, above referred to; but men always do at their corresponding one, called chitoto.

The subject of dances is a large one, celebrating, as they do, every important event in life, from birth to death, besides ordinary merrymakings which have no particular motive beyond cheerfulness and sociability. In place of attempting to enumerate all the varieties, which would be wearisome and convey no particular impression, I shall content myself with extracting one or two descriptions from my notes. ‘Passing through Mlomba’s village (near Blantyre) found a grand masewero[26] going on. The dancing man was performing, but not singing—calico turban on his head, leather belt under his arms, with a great bunch of long feathers stuck into it in front, some falling down over his waistcloth, others reaching to his shoulders, a wild-cat skin hanging down his back, and dance-rattles on his legs. This dance is called the tseche. There were also six drummers: one sat on the ground and beat his drum (the kind with legs like a stool) with two sticks; the rest held theirs against their chests and beat them with both hands, the drum being supported by a piece of twine passing under it and looped over both wrists. They were well-made, muscular fellows, and danced pretty hard while drumming: this, it seems, is called the nkonde. Two younger boys came forward at intervals and danced pas seuls, and at the end a collection was taken up, chiefly in fowls.’ Sometimes the beads contributed by a gratified audience are put into a hole in the gourd of the chimwenyumwenyu.

I remember the drums going all night long for the chamba dance at Ntumbi (which, by the bye, in spite of the name, has nothing to do with the smoking of the pernicious Indian hemp), and the ball was still in full swing between 7 and 8 A.M., when some of our boys and girls requested permission to go down before school hours and ‘see the Angoni playing.’

Another dance which I witnessed at a Yao village near Blantyre, I am not sure whether to class as a diversion or a ritual solemnity. I think it was the latter, but not (as I was at one time inclined to suppose) the chimbandi, or ‘great unyago,’ which precedes the birth of a woman’s first child (see Macdonald, Africana, i. 128), unless the latter has been considerably modified. In the first place, my friend Chewilaga, who appeared to play the principal part, had a baby about six weeks old; in the second, so far from only women being present, there were three men and a boy working the drums, and one man among a few casual spectators who gathered from outside; and there were other points of difference. Eight or ten women (two of them quite young girls) took part in the dance, led by Chewilaga; they were all freshly anointed, almost dripping with oil, and had on their best calicoes, and (apparently) all their beads, and wore rattles on one leg only. The drummers sat in a row on a form made out of a split log: the three men held their drums against their chests and beat them with their hands; the boy had a four-legged standing drum, which he beat with two sticks. The women—one with a baby tied to her back—stood in front of the band in a semicircle, ‘marking time,’ then formed in couples and ‘set to partners,’ then marched round, in Indian file, then bent forward from the hips, and all danced together in a kind of jigging step; then formed in semicircle again, and so da capo. The song (sung by the dancers) consisted of a few words only, which I failed to catch.

Some of the dances for amusement are confined to one sex; in others, both take part. In one, partners are chosen and led out into the middle; in another, the man who beats the big drum leaves it at intervals and dances alone in the centre of the ring, while every one claps hands to fill up the gap.

The war-dance of the Angoni—executed, perhaps, by hundreds of men leaping into the air at once and beating their shields—is very striking; the Yaos and Anyanja also have one, though the latter are not a particularly warlike race. ‘One in the war-dance,’ says a native account, ‘comes and stretches his leg, stamping down his foot, di! and his gun, di! before his chief, saying, “Chief, we are here, none can come to kill you, for we are not dead yet.”’

The zinyao dances have been already touched on in connection with the mysteries, and the mourning dances in the chapter on funeral ceremonies. The Rev. D. C. Scott thus describes the latter, and at the same time successfully conveys the impression produced by all: ‘The heavy, deep di! di! of the great bass drum, with silence succeeding, broken by the responsive wail and clapping of hands, then with the rapid call of the small garanzi drum, and again with the deep hollow bass, and the never-ceasing circling of the dance, produces a weird sensation only possible in Africa.’[27]