The most interesting point connected with these coast negroids is their language, the Kisawahili. It was anciently called Kingozi, from Ungozi or the region lying about the Dana, or rather Zana, the river known to its Galla accolœ as ‘Maro,’ and ‘Pokomoni’ from the heathen Pokomo who, living near its course, form the southern boundary of Galla-land proper. The dialect is still spoken with the greatest purity about Patta and the other ancient settlements between Lamu and Mombasah. Oral tongues are essentially fluctuating; having no standard, the roots of words soon wither and die, whilst terms, idioms, and expressions once popular speedily fall into oblivion, and are supplanted by neologisms. Thus the origin of words must often be sought by collation with the wilder kindred dialects of the coast tribes; for instance, the root of ‘Mbua’ (rain), which has died out of Kisawahili, still visits in Kinyika—ku buá, to rain. In Zanzibar Island Kisawahili is most corrupted; the vocabulary, varying with every generation, has become a mere conglomerate which combines South African, Arabic, Persian, Hindi, and even Portuguese, an epitome of local history. On the coast it greatly varies, being constantly modified by the migration and mixture of tribes. Like the Malay of the Indian Islands, it has become the Lingua Franca, the Lingoa Geral of commerce from Ra’as Hafun to the Mozambique and throughout Central Intertropical Africa. This Urdu Zaban, or Hindostani of East Africa, is indispensable to the explorer, who disdains mere ‘geography;’ almost every inland tribe has some vagrant man who can speak it. My principle being never to travel where the language is unknown to me, I was careful to study it at once on arriving at Zanzibar; and though sometimes in the interior question and answer had to pass through three and even four media, immense advantage has derived from the modicum of direct intercourse.
The base of Kisawahili is distinctly African; and, totally unlike its limitrophe the Galla, it grammatically ignores the Semitic element. It is now time for writers to unlearn that, ‘all the languages over the face of the earth, however remotely different and however widely spread, appear to be all reducible to the one or the other of three radically distinct tongues’ (Dr Beke, p. 352, Appendix to Jacob’s Flight. London, Longmans, 1865).[[109]] It is only, I believe, the monogenist pure and simple who in these days would assert ‘there exist three linguistic types, as there are three physical types, the black, the yellow, and the white’ (M. de Quatrefages, p. 31, Anthropological Review, No. xxviii.). To the old and obsolete triad of Turanian, Semitic, and Aryan, or Turanian, Semitic, and Iranian, we must now add at least another pair—without noticing the Asianesian—namely, the American or Sentence language, and the prefixitive South African family. These two great tongues, one extending over half a world, the other through half a continent, are, I believe with Lichtenstein and Marsden, unborrowed, indigenous, and marked with all the peculiarities which distinguish their inventors. Both are idioms which seem to indicate nice linguistic perceptions and high intellectual development; history, however, supplies many cases of civilization simplifying and curtailing the complicated tongues of barbarians, thus making language the means, not the end, of instruction.
The limits of the South African family may be roughly laid down as extending from the Equator to the Cape of Good Hope. The Equatorial Gaboon on the Western Coast[[110]] evidently belongs to it; and upon the Congo river I found that whole sentences of Kisawahili were easily made intelligible to the people.[[111]] Though the language is evidently one in point of construction throughout this immense area, isolation and hostilities between tribes have split it into a multitude of dialects. Almost every people, at the distance of 30 to 50 miles, has its peculiar speech, and in these regions it would not be difficult to collect ‘Specimens of a hundred African Languages.’ The older travellers remarked that the Tower of Babel must have been near the Gulf of Guinea; they would have found the same throughout the interior and Eastern Coast.
My experience[[112]] of the tongues spoken to the west of the Zanzibar coast proper is that their amount of difference greatly varies: some average that of the English counties, others of the three great Neo-Latin languages, whilst in some the degree amounts to that between English, German, and Dutch. And generally, I may remark, the East-West extremities of the lingual area are more closely connected than the North-South: the language of Angola, for instance, is more like Kisawahili than the Sichuana. I am at pain to understand why Dr Krapf should have named this linguistic family, Orphno-(dark-brown) Hamitic, Orphno-Cushite, Nilo-Hamitic, and Nilotic,[[113]] when it is far more intimately connected with the Kafir regions, the Congo and the Zambeze rivers, than with Æthiopia or the Nile Valley proper. Mr Cooley’s term ‘Zangian’ or ‘Zingian’ also unduly limits the area to that of a mere sub-family.
The crux grammaticorum of the great South African language is its highly artificial system of principiatives or preformatives.[[114]] In the three recognized lingual types of the old world the work of inflexion, the business of grammar, and the mechanism of speech disclose themselves at the ends of vocables. In this prefixitive tongue the changes of mood, tense, case, and number, are effected at the beginning of words by prepositive modifying particles, which are evidently contractions of significant terms, and whose apparatus supplies the total want of inflexion. This development, arrested in other languages—the Coptic, for instance—here obtains a significance which isolates it from all linguistic society. The practised student at once discovers that he is dealing with a completely new family by the unusual difficulty which unvaried terminations and initial changes present to one accustomed only to the terminal.
The minor characteristics of the Kisawahili are the peculiarities of the negative system in substantives and adjectives, pronouns, adnouns, and verbs; for instance, Asie, he or she who is not, Isie, it which is not. Secondly, are the broad lines of distinction drawn between words denoting the rational and the irrational, and in a minor degree the rational-animate (as man), and the rational-inanimate (as ass). In most cases the rational-animate affixes Wa as a plural sign: the irrational-animate Ma. Umbu, a sister, properly makes Waumbu, sisters: the ignorant, however, and the Islanders often say Maumbu (sisters) like Map’hunda (asses). Thus personality supplies the place of gender, a phenomenon that already dawns in the Persian and in other Indo-European tongues. Next is the artful and intricated system of irregular plurals, and last, not least, the characteristic alliteration, an assonance apparently the debris of many ancient dialects based upon an euphonious concord not always appreciable by us, and therefore not yet subjected by our writers to rule. We understand, for instance, that an alliterative speaker should say Mtu mema (a good man), and Watu wema (good men); but why is the regularity altered to Máháli pángo (my place), p’hunda zango (my donkey), and Mtu wa Rashidi (Rashid’s man), instead of mango, pango, and ma? These distinctions appear far too empirical, arbitrary, and artificial for the wants of primitive speech.
The Kisawahili is an oral tongue—an illiterate language in the sense assigned to the term by Professor Lepsius. The people, like the Somal and the Gallas, never invented a syllabarium. This absence of alphabet is a curious proof of deficient constructiveness in a race that cultivates rude eloquence, and that speaks dialects which express even delicate shades of meaning: it contrasts wonderfully with the Arabs and Hindus, who adapt to each language some form of Phœnician or Dewanagari. The coast races use the modern Arabic alphabet, which, admirable for its proper purpose, represents African sounds imperfectly, as those of Sindi and Turkish, and is condemned to emulate the anomalous orthography or cacography of our English. The character is large, square, and old-fashioned, resembling later Kufic even more than that of Harar, and he must be a first-rate scholar who can read at sight all the letter of a man to his friend. Literature is confined, to a few sheets upon the subject of Báo or Uganga (Raml or geomancy), to proverbs and proverbial sayings, mostly quatrains; to riddles and rabbit tales, which here represent the hare legends of the Namaquas and the spider stories of the Gold Coast; to Mashairi, or songs rhymeless, measureless, and unmusical, and to ‘Utenzi,’ religious poems, and eulogies of the brave.
In Zanzibar Island Arabic is ever making inroads upon the African tongue, and the student who knows the former will soon master the latter. The first short vocabulary, by Mr Salt, was published in 1814, and was presently followed by others, especially the ‘Soahili vocabulary’ of the late Mr Samuel K. Masury, of Zanzibar (Memoirs of the American Academy, Cambridge, May, 1845), and Mr J. Ross Browne’s ‘Specimen of the Sowhelian Language’ (Etchings of a Whaling Cruise. New York, 1846).[[115]] Strange to say, the ‘Mombas Mission’ translated the Gospels into the obscure local Kinyika, when only three chapters of Genesis and a version of the English Prayer Book (Tubingen, 1850-54) were published ‘in the one language, by the instrumentality of which the missionary and the merchant can master in a short time all the dialects spoken from the Line down to the Cape of Good Hope.’ Dr Krapf’s ‘Outline of the Elements of the Kisauaheli Language’ (Tubingen, 1850) requires great alterations and additions, especially in the alliterative and other characteristic parts of the tongue. Messrs Rebmann and Erhardt, who both were capable of writing a scholar-like book, or of perfecting the ‘Outline,’ turned their attention to the languages of the Nyassa, Usumbara, and the Wakwafi. In 1857 M. Guillain published, as an Appendix to his third volume, a short grammar and vocabulary of the ‘langue Souahhéli:’ they are mere bald sketches, and they convey but the scantiest idea of what they attempt to illustrate. A good study of Kisawahili would facilitate the acquisition of the whole sub-family. For my own use I commenced a grammar intended to illustrate the intricate and difficult combinations and the peculiar euphony which here seems to be the first object of speech: unfortunately my transfer to West Africa left it, like my vocabularies, in a state of MS. My friend Mr Trübner has lately advertised a volume called ‘East African folk-lore, Swahili Tales, as told by the natives of Zanzibar,’ with an English translation by Edward Steere, L.L.D., Rector of Little Steeping, Lincolnshire, and Chaplain to Bishop Tozer (London: Bell and Daldy, 1870);[[116]] and Dr Krapf has proposed to publish the Juo ya Herkal (Book of Heraclius), ‘an account of the wars of Mohammed with Askaf, Governor of Syria, to the Greek Emperor Heraclius, in rhyme; a MS. in ancient Ki-Suahili written in Arabic characters.’ Also ‘Juo ja Utenzi, Poems and Mottoes in rhyme,’ the dialect being that formerly spoken in the Islands of Patta and Lamu. Both the ‘linguistic treasures’ were presented to the Oriental Society of Halle. The last publications which I have seen are ‘Specimens of the Swahili Language’ (Zanzibar, 1866); ‘Collections for a Handbook of the Swahili Language, as spoken at Zanzibar,’ by Bishop Tozer and Rev. E. Steere (Zanzibar, 1865), and the Rev. E. Steere’s ‘Collections for a Handbook of the Shambala Language’ (Zanzibar, 1867), the ‘tongue spoken in the country called in our maps Usumbara, which is a mountainous district on the mainland of Africa, lying opposite to the Island of Pemba, and visible in clear weather from the town of Zanzibar.’
Kisawahili is at once rich and poor. It may contain 20,000 words, of which, perhaps, 3000 are generally used, and 10,000 have been published. Copious to cumbrousness in concrete, collective, and ideal words, it abounds in names of sensuous objects; there is a term for every tree, shrub, plant, grass, and bulb, and I have shown that the several ages of the cocoa-nut are differently called. It wants compounds, abstract and metaphysical expressions: these must be borrowed from the Arabic, fitted with terminal and internal vowels, to suit the tongue, and modified according to the organs of the people, harsh and guttural consonants being exchanged for easy cognates. Even the numerals beyond twenty are mere Semitic corruptions. All new ideas, that of servant, for instance, must be expressed by a short description. In the more advanced South African dialects, as in the Mpongwe of the Gaboon, a compound or a derivative would be found to include all requirements. The sound would be soft and harmonious were it not for the double initial consonants, aspirated or not; for the perpetual reduplications (the Arabic Radif),[[117]] a savage and childish contrivance to intensify the word, and for the undue recurrence of the coarse letter K. Possibly the fondness of the people for tautology may have tended to develop their tautophony and euphony. Abounding in vowels and liquids, the language admits of vast volubility of utterance; in anger or excitement the words flow like a torrent, and each dovetails into its neighbour till the whole speech becomes one vocable. Withal, every vowel has its distinct and equal articulation. It wants the short and obscure sound of the English and other European languages (e. g. a liar, her, first, actor, and hurled) called by us the original vowel sound. Like the Chinese and Maori languages, and the other South African tongues, it confounds the so often convertible letters, the L and the R.[[118]] The slaves, the Wasawahili, and the wild natives mostly prefer the former, e. g. Mabeluki for Mabruki, and the Arabs and civilized speakers the latter, although Mr Cooley (Geography of N’yassi, p. 20) asserts the contrary. The metastasis, however, appears to me often arbitrary, occasioning trouble, e. g. when ku ría (to eat) becomes ku lía (to weep). Dr Livingstone, (chap. xxx. First Expedition) complains of Loangoa, Luenya, and Bazizulu being transformed into Arroangoa, Ruanha, and Morusurus, but he also similarly errs when he converts Karagwah into Kalagwe, and when (p. 266) he uses indifferently Maroro and Maloli. The R is often inserted pleonastically, to prevent hiatus, as Ku potéra for Ku potéa, to lose; Ku pakíra for Ku pakía, to pack. Sometimes, again, it is omitted, as U’ongo for Urongo, a lie. In pronouncing it the tongue tip must (be more vibrated than in our language, which loves to slur over the sound. Aspirated consonants are found, as in Sanskrit, especially B’h, P’h, D’h, T’h, K’h, and G’h. Quiescent consonants are rare in the middle of words; thus the Arabic Mismar (a nail) is changed to Misumari, and treble are unknown. There are only five peculiar sounds[[119]] which are generally mispronounced by the Arabs, and these are mostly of little importance. The dialect is easily learned: many foreigners who cannot speak understand, after a short residence, what is spoken to them. It may be said to have no accent, but a sinking or dropping of the voice at the terminal syllable—possibly the case with Latin hexameters and pentameters—seems to place the ictus upon the penultimate, as Wasawahíli for Wasawahĭli.[[120]] Hence when first writing proper nouns I preferred Mtony and Pangany to Mto-ni and Pangani. Similarly the W when placed between a consonant and a vowel is often so slurred over as hardly to be detected. For instance, Bwáná, master, becomes B’áná, and Unyamwezi might be both written Unyam’ezi were it not liable to confuse the reader. There is also a Spanish ñ (Niña), as in Ñika, the bush, and Ñendo, the P. N. of a district, which I express by Ny, e. g. Nyika and Nyendo. Finally, being a lazy language, which well suits the depressing climate, it takes as little trouble to articulate as Italian: hence, even in the first generation, Arabs and Baloch exchange for it their own guttural and laborious tongues, and their offspring will learn nothing else. This is more curious than the children of the Scandinavians abandoning the father-tongue for Norman and Anglo-Scandinavian, vulgarly called Anglo-Saxon. In East Africa adult settlers forget their mother-tongue,
And now of the slave races proper.