NOTE ON CLICKS IN THE BANTU LANGUAGES

Author(s): A. Werner

Source: Journal of the Royal African Society, Vol. 2, No. 8 (Jul., 1903), pp. 416-421

Note from moderator: Alice Werner worked as a Blantyre Mission missionary in Nyasaland in 1890s and was more acquainted with the Maseko Ngoni having worked among some Maseko Ngoni as a teacher. 

IT is generally conceded that the clicks which occur in Xosa, in Zulu, and, to a limited extent in Sesuto, have been borrowed from the Hottentots. Accordingly we find a greater number of click-words, though not a greater variety of clicks, used by the Xosas than by the Zulus, the former having been more in contact with the previous occupants of the country than the latter. Dr. McCall Theal (History of South Africa, II., 196) says that the clicks "were introduced by females spared when the hordes to which they belonged were conquered, as is evident, not only from tradition, but from the words in which they occur being chiefly those pertaining to the occupations of women." This, however, scarcely holds good, at least as far as Zulu is concerned, as the following list of words (which might easily be made longer) will show. (It is scarcely necessary to point out that c stands for the dental click, q for the palatal, and x for the lateral).

icala = a debt, lawsuit,

isigcogco = a man's head-ring,

incwadi = a message; hence a letter, book, etc.,

icebo = a device,

isiceme = a line or row of men,

icakide = a weasel,

umcibitsholo = an arrow,

iqawe = a brave man,

iqude = a cock,

iqanda = an egg,

inqola = a waggon,

amahwangqa = whiskers,

uxamu = an iguana,

ixcgu = an old man,

isigaxa = a lump, mass :-

together with the verbs -cela (ask), -cita (destroy), -qina (be strong), -qala (begin), -qeda (finish), -xotska (drive out, as cattle) and a good many others. None of these words can be called specially appropriate to women; but it is possible that they, or some of them, were borrowed from the captives by Zulu women, or in certain cases, perhaps even by men, to replace words which could not be used on account of uku-hlonipa. The Mang'anja mlandu, pretty well corresponding in meaning to icala, is found in several cognate languages of the south-eastern region, but not in Zulu; is it too much to conjecture that it might have been tabooed as the name of a chief Umlandu, and the Hottentot icala (or the word which now appears in this form) adopted in its place ?

On the other hand, the following words may be accounted distinctively feminine in character:-

iqoma = a basket for holding mealies,

-coleka = to be fine (as flour, etc.),

-colisa = to grind fine,

icansi = a rush sleeping-mat (usually made by women),

umcaba = boiled mealies,

-buqa = to harrow in seed.

Isicatulo (a shoe) is probably a recent word, introduced (along with the thing) by missionaries. The derivation, according to Dr. R. J. Colenso (whose assistance, in this and other points, I gratefully acknowledge) is from uku-catula, to walk slowly, toddle, (as small children: this might well be a woman's word), which, in its turn, comes from ukuti catu (much the same meaning). Catu is one of those interjectional adjectives or adverbs (usually introduced, and brought into grammatical connection by the verb "to say "), which are a characteristic feature of the Bantu languages. (It is worth noting that, in Zulu, a large number of these contain clicks.)

It is natural to suppose that the words containing clicks were borrowed from the Hottentot language, rather than that the clicks were imported into words of Bantu origin. Accordingly, the Rev. J. Torrend says that "among the Kafir words which contain clicks, there are few which have equivalents radically identical with them in other Bantu languages. But, in Appleyard's Kafir Grammar (a work which, though published many years ago, and needing some corrections where the outlying languages are concerned, contains a great deal of sound learning), we read:

"It would be wrong to suppose, however, that every word which contains a click sound is of Hottentot derivation. It may well be doubted, indeed, whether any click-words have come from that source except a few nouns, and perhaps a few particles. The fact appears to be that the Kafirs have substituted the Hottentot clicks for other characters, and have thus simply changed the form and sound of their own words. This is borne out by the comparison of a few roots, where both the original and the adopted form of words are still in use. Thus:

namatela and ncamatela = to stick to,

nyamekela and ncamekela = to care for,

tyabatyabaza and cabacabaza = to walk in fear,

isitywetywe and isicwecwe = a flat object,

tyatyamba and qaqamba = to yield pain,

nyotula and ncotula = to pluck out,

tsitsha and tshica = to spit,

twebula and xwebula = to bark trees,

qika and qiqa = to comprehend,

hluma and cuma = to grow,

tyanda and canda = to cleave,

tola and cola = to pick up."

It might be objected that, in the absence of further proof, it would be difficult to tell which of these two alternate forms is the original. The click being presumably the more difficult sound, it would seem more reasonable to suppose that it had been modified than added: the question could only be decided by finding these words in other Bantu languages. The only one of which I could say for certain that this is the case is tola, which exists, with the same meaning, in Mang'anja, and perhaps elsewhere. But further research might reveal analogues to the others.

Before I had seen Father Torrend's book, I had for some time been endeavouring to discover (chiefly in Mang'anja) the analogues to Zulu click-words, and had been struck by the extreme difficulty of finding any. I should be glad to know if any student of African languages can add to the list; but must premise that I should be very sorry to dogmatize on the subject, and do not by any means feel certain (except, perhaps, in the case of cima) that the words compared are really the same.

Cima = to extinguish (a light, etc.). Mang'anja, zima (which properly means "to go out," the causative zimitsa

being equivalent to the usual Zulu meaning) Swahili, zima. Herero, thema, (th as in "that"); Sechwana, tima; Duala, dima; Kongo, jima; Bobangi, (Middle Congo) limwa (" to go out") and limwisa (" to put out "). Herr Meinhof (Lautlehre der Bantuspracken, p. 172) suggests the primitive form "lima, ndima, erl6schen." (With this may be compared Mang'anja mdima, "darkness"). It seems unlikely that such an obviously Bantu word could have been borrowed from a race whose known linguistic influence was only exercised on the southern extremity of the Bantu area.

-ncane = small. Mang'anja -ng'ono. This latter form, so far as I have been able to discover, stands alone. In fact, the words meaning "small " are exceedingly divergent, and it seems impossible to reduce them to a common root. Yao -nandi, Swahili -dogo, Ronga -tongo, Zeguha -dodo, Nyamwezi -guhi (probably identical with the word for "short"); Kongo -akete (this is perhaps connected with Bobangi and Lunyoro -ke), and Luganda -tono, may serve as specimens. Meinhof gives no primitive form for this word, while -kulu, "large," and -kupi, "short," occur in his list of " Ur- Bantu" roots. The "ringing ng" (n ) seems quite a possible sound to be substituted for the nasalized click (nc) by those unable to pronounce the latter. The isolated position of the Mang'anja word suggests the question whether it could have been borrowed-if not from the Hottentot, from some language allied to it. Sir H. H. Johnston has expressed the opinion that the tribes about Mount Mlanje show traces of a pre-Bantu element. Various points in the physique and customs of the Mang'anja, especially of the small, wiry people west of the Shire, commonly, but inaccurately called "Angoni," would seem to point to the same conclusion. If it be true that the Hottentots are of North African origin, and directly or indirectly connected with Egypt, we have, perhaps, here a hint as to the provenance of those industries, spinning, weaving, smith's work, etc., in which the Mang'anja, Mashona, and other more or less subject tribes have the advantage of their Zulu conquerors. The similarity of implements, processes, patterns, etc., to those used by the ancient Egyptians has often been pointed out.

Isicamelo = a wooden pillow, or rather neck-rest (of the kind used in Africa from time immemorial, and found in Egyptian tombs) :-Mang'anja, mtsamiro, from tsamira, to lean upon. Ts is a combination which never occurs in Zulu, though it does in Xosa; it seems a very natural substitute for the dental click. Tsamira and camela would be the " applied" forms of a verb cama or tsama: which of these is original and which derived, comparative philology has yet to determine.

Iciba = a pool, might be Mang'anja dziwe: the sound shifting b = w is of regular occurrence (cf. Z. ukubaba = M. kuwawa) and the d (or "hardened stem ") takes the place of the lost prefix i(li). There is also a Zulu word (with a different class-prefix) isi-ziba.

-cwazimula = to shine, glitter. Mang'anja, nyazimira (from nyazi, adj. or adv.).

-qonqota = to knock; Mang'anja, gogoda or guguda; but they might be independent onomatopceias.

-enqaka = to catch (a ball, etc.), Mang'anja, yakha as in "Ndiye wa ku yakha mpira "-" he is one who catches the ball" (mpira) (= he is one of the players).

Iqanda = an egg; Ronga tanda. Yao, (li)ndanda. I cannot find any further parallels to these forms. Mang'anja, dzira; Swahili, yayi. The roots -yi, -gi, and -ki, seem of frequent occurrence.

In a list of Zulu words as used by Chekusi's Angoni, which I took down at Ntumbi (West Shire District) in 1894, I find the words amaqanda (eggs), isizwembe (a wooden ladle), and licansi (a mat), all written with a k, subsequently corrected to c. This is more probably because my ear failed to discriminate between the clicks, than because my informant (a very intelligent old woman, who had lived for a long time at Chekusi's kraal, but, I think, was not a Zulu by birth) pronounced them alike; but it is possible that, in the course of their northern wanderings, the Angoni have reduced the three sounds to one. M. Edouard Foa (Du Zambeze au Congo Francais, p. 74) says that Mpezeni's people called the head-ring (isigcogco) chijojo, which looks as though they had substituted j (probably the French sound as in jeune is intended) for the soft dental click gc; but it is also possible that M. Foa (or his Mang'anja boys, if he did not get the information direct,) failed, like myself, to catch the click. 

It is perhaps worth noting, by the way, that the Angoni, like the Xosas, drop the first i only, of the prefix ili, which the Zulus contract into i, and say lilanga (ilanga), lizulu, licansi, etc. The prefix umu is shortened into mu rather than um: mufana for umfana; and the a of ama is usually dropped, as in Mang'anja (maqanda, etc.), except in the case of monosyllabic roots :-ama-fu, ama-nzi, etc.

A. WERNER.