BOUGHT WITH THE INCOME FROM THE SAGE ENDOWMENT FUND THE GIFT OF 1891 kB..iAn - - iJilJai Cornell University Library PL 8025.K81 Language-study based on Bantu 3 1924 026 932 867 The original of this book is in the Cornell University Library. There are no known copyright restrictions in the United States on the use of the text. http://www.archive.org/details/cu31924026932867 LANGUAGE-STUDY BASED ON BANTU OK AN INQUIRY INTO THE LAWS OF ROOT-FORMATION, THE ORIGINAL PLURAL, THE SEXUAL DUAL, AND THE PRINCIPLES OF WORD-COMPARISON ; ^Tables JHustrating tbe primitive ipronomtnal SBStem restored in tbe Htrican asantu ffamitg o( Speecb. BY THE EEV. F. W. KOLBE, OF THE LONDON MISSIONARY SOCIETY, FOEMERLY OF THE RHENISH HEEERO mission; AUTHOR OP "an enqlish-herero dictionary." LONDON: TRUBNER & CO., LUDGATE HILL. 1888. [All rights reserved.] V CCRa^EiLN fUNIVLRe:TY LI8RAPV BALLANTYNE, HANSON AND CO. EDINBURGH AND LONDON PREFACE. The author, who since 1853 has been a missionary of the London Missionary Society, and before that time was connected with the Rhenish Herero Mission in Damaraland, is, as far as philology is concerned, a self-taught worker. When sent to Damaraland in 1848, he could not possibly have dreamt of ever writing a treatise on the Principles of Language. But when, in conjunction with his colleagues, the Eev. Dr. C. H. Hahn and the Rev. J. Rath, he studied Herero, he was from the very first fascinated with its marvellous structural regularity and wealth of pronominal forms. Fortunately, he knew from Gesenius's Hebrew Grammar and Heyse's Lehrbuch der Deut- schen Sprache that there are only three primitive vowels (a, i, ii), and was struck to find only these three in the pronominal roots of Herero ; and in grouping the formative prefixes of the noun (or roots of pronouns) according to their consonantal sounds, he conjectured that there must be some difierence of meaning between such forms as oka- {KA), otyi- (^KI), oku- (KV), and that this difference must be caused hy the voivels. And turning to the verb, he received the same impression, though what that difference might be he had not the remotest conception. So plodding on, he was by degrees led to the discovery of the vowel-laws — laws which are by no means confined to Bantu, but are traceable as well in the roots of iv PREFACE. Aryan and other languages. Subsequently, in following up an observation of Mr. Bath's as to the dualistic tendency of the prefix oma- (ama-), the sexual dual became clear to him. Now, these two fundamental principles, viz., the vowel-laws and the sexual dual, have already, at any rate as regards Bantu, met with the approval of two high philological authorities in England, and it is hoped that the other principles set forth in this little work — the result of over thirty years' patient research — will also commend themselves to the student as truths founded on sufficient evidence, and illustrative of the new and fuller light the study of Bantu is destined to shed on the Aryan family and on the origin of language universally. For whilst, in Aryan and other families of speech, the science of language must be content to " begin with roots as its ultimate facts," we are, in Bantu, where we find language in an earlier stage of development, enabled to discover the mry first laivs hy which lav^uage was formed, and to restore the original concord between language and nature, words and things. As to material for the study of Bantu, there is a vast deal of it already accessible, as may be seen from the " Index of the Grey Collection," by Dr. Th. Hahn, and the very opportune " Sketch of African Languages," by E. N. Oust, Esq. (Messrs. Trilbner & Co.) But still more should be done. The recent opening up of Central Africa and the Congo Eegions — all peopled with Bantu nations — should be taken advantage of to collect, with the aid of missionaries and others, the new philological treasures now placed within our reach. And this should be done without delay ; for the steady influx of Euro- peans as time rolls on must necessarily interfere with the primitive purity of the Bantu languages. Would not the Committee of the Imperial Institute take the matter in hand. PREFACE. V and devote a section of the library to African languages, and initiate, at the same time, the compilation of a Comparative Bantu Dictionary ? " Ever since the publication of the first part of Dr. Bleek's Comparative Grammar," says a leading philologist in a letter to the author, "the importance of the Bantu languages to the scientific study of language has been becoming more and more evident to every comparative philologist." Indeed, strange as it may seem that the science of language should have to go to the Dark Continent for more light, it is nevertheless true that " the origin of the grammatical forms of gender and number, the etymology of pronouns, and many other questions of the highest interest to the philologist, find their true solution in Southern Africa " (Bleek). A single glance at the appended comparative table of pronominal forms ought to suffice to convince the most sceptical of the truth of this. It is earnestly hoped that the present attempt may, in some measure, contribute towards raising the great Bantu family to the prominent place which it deserves to occupy in the science of language. F. W. K. Cape Town, September 19, 1887. CONTEl^TS. CHAPTER I. INTEODUCTOEY CHAPTEE II. PRIMITIVE ALPHABET — PRIMEVAL LAWS OF CONSONANTS . II CHAPTEE III. PRIMEVAL LAWS OF THE VOWELS . . . . .21 CHAPTEE IV. THE VOWEL-METHOD IN UNIVERSAL ETYMOLOGY . . ■ 3^ CHAPTEE V. NOTES ON THE PRIMITIVE BANTU PREPOSITIONS AND ADVERBS 4 1 CHAPTER VI. ROOT-FORMATION: ITS BEGINNINGS AND SUCCESSIVE STAGES . 45 CHAPTER VII. ON THE DEVELOPMENT OF THOUGHT IN AFRICAN BANTU . £2 CHAPTER VIII. ON THE HERERO PRONOUN — ABSOLUTE AND CONDITIONAL FORMS PRIMEVAL LAW OF THE PLURAL SEXUAL DUAL ......... 59 viii CONTENTS. PAGE CHAPTEE IX THE OPERATION OF COMMON LAWS, TEACEABLE IN THE BANTU AND ARYAN PRONOUNS . . . . . . 71 CHAPTER X. PRONOMINAL TABLES — THE PRIMITIVE PRONOMINAL SYSTEM RESTORED 84 TABLES or PRIMITIVE PRONOMINAL FORMS, ETC. . . -95 A LANGUAGE - STUDY BASED ON BANTU. CHAPTER I. INTRODUCTORY. What is language 1 Essays on this ever-recurring question abound : their number is legion. From the earliest times there have not been wanting speculative minds who have endeavoured to solve this problem. Various theories have been propounded, but not one has led to an absolute certainty as to the true beginnings of human speech. Indeed, such is the mass of conflicting opinions on the subject, and such the obscurity which still envelops it, that a Linguistic Society in Prance is said to declai'e in one of its first statutes that it will receive no communication concerning the origin of language. But a resolution like this is evidently premature and unscientific. Is it possible then, in any science or art, to determine beforehand what can be discovered, and what not ? And is it, not so that com- parative philology is a progressive science, and, compared with other branches of human knowledge, still in its infancy ? What if, after all, in some obscure part of the globe, a language or family of languages be in existence so primitive that the words can be traced to first elements, and that in it the first laws of universal speech can be discovered ? It appears to me that there is good reason for believing that the African Bantu family, and especially Herero, which may be called the Sanskrit of Bantu, has been preserved in such a primitive state as to make it possible to discover certain simple laws that guided the first man in creating the stock of radicals from which universal A 2 A LANGUAGE-STUDY BASED ON BANTU. language has sprung. Let the student for once divest himself of all preconceived notions on the subject, and carefully examine the facts that shall be laid before him. He will then be convinced that " the continent of Africa supplies new and wondrous forms, the examination of which wUl upset many favourite theories, based upon the very limited phenomena supplied by the Aryan and Semitic families " (R. N. Oust, " Languages of Africa "). Hrrors to he noted. — And here, at the outset, I must refer to what for many years has appeared to me to be a fundamental error in our modern science of language. Comparative philologists very fre- quently confound the terms " isolating " and " monosyllabic." In speaking of Chinese and kindred idioms, they call .this class of languages "monosyllabic or isolating, their words being, in fact, composed of simple monosyllabic roots, isolated and, as a rule, independent of each other." And again, " It maybe well to state at once that all linguistic systems have passed through this mono- syllabic period " (A. Hovelacque). ISTow, if this be true, if Chinese and other isolating idioms are really monosyllabic, it would be time and strength wasted to resort to an agglutinative and polysyllabic class of languages in search for the elements or ultimate roots of human speech ; for nothing can be more certain than the fact that true monosyllables have preceded polysyllables in the growth of language. But is it so ? are the above-mentioned languages mono- syllabic in the true sense of the word ? I think it can be proved that they are not; the majority of their respective words have merely been ground down to the/or»j of monosyllables, so that they are now pronounced as such, but originally they had more than one syllable. We may, for instance, call the English verbs to " send," to " find," to "bring," monosyllabic words, but then it must be clearly understood that they are not so in reality, but only ■pronounced as such, the wear of time having reduced them to their present defec- tive monosyUabic form. Thus, in speaking of Chinese and kindred idioms, this class of languages ought to be called simply " isolatino- " or if "monosyllabic" be added, it ought to be explained that by the wear and tear of ages, the bulk of the words in these languao-es have been reduced to such crippled forms that they have lost their "oriainal polysyllabic character, and are at present pronounced as monosyllables ■ as, for example, the Chinese words kuncj, ascend, tap, answer, which' INTRODUCTORY. 3 like the English send, hang, had originally two syllables, as they have still in the Swedish senda (though here the terminal a bears a grammatic character, and is only a substitute for an original i or e), and Icelandic hdnga, the consonants ng in hung, p in tap, nd in send, and ng in hang being remnants of an original second syllable. True monosyllables in Chinese are words like i, lean against, depend on, hi, lean on one side, ta, great, greatly, greatness ; but such true monosyllabic forms we also meet with in the agglutinative Herero, as, for example, pa, to give, ta, to reach, i, to go, ta, to perish, to die. Words, however, like Chinese yik, change, tap, tread, chung, straight, correct, horn, receive, receiver, cavern, were originally dis- syllabic, like Herero tyilca, to be oblique, yenda, to go, walk, travel, ramha, to chase, pursue, suta, to satisfy, pay. In searching for the elements and first laws of human speech, we may find, therefore, as much originality in agglutinative as in isolating languages, since the lexical stock of the latter is not of necessity more primitive than that of the former. Moreover, the hypothesis that those branches of human speech which present the greatest mechanical difficulty in pronunciation are the most original, is a delusion all the more deceptive for being mixed up with truth ; for error is never more dangerous than when it comes to us in the garb of truth. It is unquestionably true that generally those sounds which require the greatest exertion in pronouncing have a claim to priority. Thus, if a word is pronounced in one Bantu dialect thitha, and in another lila, we know at once that the former pronunciation is nearer the original than the latter. But to assert that those languages whose sounds are most uncouth and clicking, as Hottentot and Bushman, are on that account the most primitive type of speech, is evidently a misapplication of an otherwise sound principle. I doubt whether any linguist would be prepared to accept the Swabian ischt as the most original form of Ger. ist, Lat. est, Skr. asti, Eng. is.] There cannot be the slightest doubt that Bantu, which had briginally no clicks, is, on the whole, much more primitive in form than Hottentot-Bushman. If it indeed be true that Hottentot, from its very first outset, started with clicks, or, in other words, if clicking phonetics belong to the first stages of root-formation, how is it that, with the exception of the prefixed demonstrative particle in the third person 4 A LANGUAGE-STUDY BASED ON BANTU. (xei-), no clicks are found in that part of speech which in all languages is acknowledged to be the most primitive — the pronoun 1 In vain do we look for clicks in such primitive Nama words as tith, I, saKHUM, we two, sokvh, we, saz, thou, sakho, you two, saKO, you. This is significant, and tends to show that the so-called clicks are not primitive, but of later growth. There can be little doubt that they originated, after the separation of the several families of speech, in a struggle for existence. When symptoms of decay made their appearance, and the root-words, one after another, began to be reduced to crippled monosyllabic forms, not unfrequently homo- phones, the genius of the language roused itself to fresh exertions, and showed its determination to stem the downward course, and to make itself understood at all hazards. The result was that, besides the introduction of tones, those consonants which were left, especially the gutturals, were often strained to such a pitch that they became what we call clicks. But they remained real consonantal sounds for all that. To show, in Hottentot, how these so-called clicks were employed much in the same way as consonants are in Semitic roots, namely, to act as expositors of different shades of the radical meaning, would not be one of the least interesting studies in Afi-ican philology. Thus we have, for example, in Nama the verb uri, to spring, to jump, and hurt, to leap ; pronounced with the cerebral click q, qhuri, the same word means to frighten, terrify, properly to frighten up, startle up, " aufscheuchen." And the same principle we notice in the use of the clicks in Kafir. Zulu Muma (hi = h = Jc) is to spring up, shoot up, as plants in spring, to grow well, large ; but the change from M to the clicking sound q gives the word a different shade of meaning ; for quma is to spring up as sparks from the fire, to crack as mealies heated, to start, shiver; whilst a second modification, gquma, signifies to throb, as a wound, beat, as a pulse or heart. So it would seem that even the clicks are not excluded from the reign of law. As an instance how easily a guttural may be turned into a clicking sound, I may mention here, in passing, that once in my hearing a Dutch Cape farmer pronounced the pronoun gij (you) in a manner which strongly reminded me of the lateral click of the Hottentots. He had never been in contact with clicking tongues, so he did not imitate, like those Boers of whom Barrow remarks — I quote from INTRODUCTORY. 5 Gust's " Sketch of African Languages " — " that they affected similar (clicking) sounds in pronouncing words of their own language." There is a third error, equally misleading. It is this. Some scholars, in endeavouring to trace a word to its ultimate root, are satisfied if they can only explain the first part, leaving the rest to shift for itself. Thus they derive Latin poena, suffering, punishment, satisfaction, and purus, pure, from Sanskrit pil, to purify, quite dis- regarding the undoubtedly radical consonantal element n and ?• in these words. Now nothing could be more unsatisfactory. It is, therefore, gratifying to observe that the untenableness of this mode of proceeding has at length been exposed. " If we look, for instance," says Professor Max Miiller ("Selected Essays," p. 91), "as I did myself formerly, on such roots as yudh, yug, and yaut, as developed from the simpler form yu, then we are bound to account for the modification elements, &c. But what are these modificatory letters ? Every attempt to account for them has failed." Claims of Herero. — Yet we ought not to despair. There is a language still living, more primitive in form than Sanskrit, in which we can trace, in a convincing number of instances, every letter of a word back to its true primitive source. This language is Herero. Just one or two specimens as an illustration. The English adverb eke (A.-Sax. eac, Sax. 6c, ac, 0. Fries, dlt, oke, Goth, and Icel. auh, Ger. audi, Dutch ooh, Swed. och, Dan. og) means in addition, also, likewise. It is derived from the verb eke, to increase, enlarge, extend, add, supply. In A.-Sax. the word has the form eacan, and in O. Fries, aha ; but the majority of the parallel forms, as 0. Sax. ocan, Icel. auka, Swed. oka, Dan. oge, O. Dutch oeken, Lat. augeo, Gr. av^oi, av^dvca, all point to u as the radical vowel. We may thus accept the Dutch oeken and the Icel. auka as the most perfect form of the word in Aryan. But farther we cannot go ; all we can ascertain in the latter family is, that the root of English eke is something like UK-, which means to increase, add. In vain do we consult our etymological dictionaries for the reason why UK- has that meaning. Let us see whether we can find a clue in Bantu. With the above adverb eke (oc, oke, auk, &c.) we identify Zulu onke or oke, Herero he (for uhe), Konde ohe, Tshuana otlhe = all, altogether, every one, the whole number, the whole mass, radically 6 A LANGUAGE-STUDY BASED ON BANTU. identical with the Bantu (Herero) verb nJta {yulm, primitive form KU-KTJ) = to be full : otyityuma tye ulta, the vessel it (is) full, the literal sense being, as appears in Herero from kindred words of the same genus, the vessel is running over. And this meaning it has not by chance, but in accordance with a deiinite rule, which shall be hereafter stated ; the consonant k meaning to run, to go, and the vowel M conveying the by-meaning high, upward, over, &c. We may thus learn in Bantu that the original form of English eke is UKU, or more correctly KU-KU, a reduplication of the monosyllable KU (the initial k having fallen off), with the primary sense to run upward, to run over, to overflow, hence to be full, and to fill, to fill up, increase, add, supply. Closely allied to eke (auk) are the words hug (0. Eng. hoge, hugge, Icel. liuga, to care, think, hugga, to comfort, console, Swed- kdgna, to hedge, wall in, Dan. hygge, to guard, A. -Sax. liegan, to wall in, guard, Ger. hegen, licigen, O. Dan. Tiage, to fence, hug, cherish) and huge (0. Eng. Tiogge, liouge, Dutch hoog = high, 0. Sax. hoJi, Goth. Jidus, Swed. hog, Ger. hocJi). Hug and huge are found close together in the dictionary as near neighbours, but as to origin and meaning they seem to be as distant from each other as the east is from the west. And yet they will probably, by means of the vowel-method (Chapter IV.), one day be recognised as offshoots of one and the same root ; the only difference between them being that in huge the vowel u means up, high, whilst in hug it has the opposite meaning, from above, from on high, downward, bowed down, hence bent, curved, round. The primary meaning of huge would seem to be to run up, to be high, whilst hug appears to have the radical sense of to run or go round, to enclose, embrace, to surround, as a garden with a fence, a child with the arms. Both words can easily be identified in Herero. Here we have the roots hunga and honga (nasalised forms of Mika and hoka), which mean — 1. To go up (or before), to rise up, be high, end in a point, be prominent; hence o-honga, high point, point, top, and o-hunga, isolated hil], properly prominent point ; hunga-ma = " sich nach etwas richten," to go by a thing, keep a prominent point in view ; and 2. To go doivn, bow, bend, curve, go or be round, go round a INTRODUCTORY. 7 thing, put something round it (" umstellen "), protect, as ■with a fence, cover, thatch, &c. (Jiok-era) ; take care of, "pflegen, verpflegen," nourish, foster, cherish (hunga). Now, if the above identification is true, then we are able to trace in Bantu the very first monosyllabic origin of huge and hug. For in Herero the root-words huJca, hoJca, strengthened hunga, honga, can be proved to be modifications of Jcuka, to rise, start, travel, and Jcoka, to be crooked, curved. The primitive monosyllabic root is KU, redupUcated KU-KU = HU-KU = HU-NGU = to go up (be high, foremost, prominent), and to go down (bow, bend, curve, put round). Or let us take the verb to send, A.-Sax. sendan, Goth, sandjan, 0. H. Ger. santjan, sentjan, N. H. Ger. seiiden, Dutch zenden, Icel. senda, with which we identify Herero Mnda or sliinda (to send), allied to Goth, sinth, 0. H. Ger. sind, A.-Sax. sklh, way, journey, and to Herero tyinda (Idnda), to go out, remove, travel. According to Bopp, the primary meaning of Goth, sandjan is to cause to go, " ich sende, mache gehen " (" Vocalismus," p. 2 1 6). But a comparison with the identical forms in Bantu enables us to put the sense in a more definite shape ; to send means primarily to cause to go out, to make go from within one place to another, '' ausgehen machen." The first vowels, a and e, in sandjan and send are substi- tutes for a primitive i, which gives a root the by-meaning in and from within or out. This is very plain in Herero. Indeed, it is not always the vowel a that can be looked upon as most primitive in an Aryan group of roots in which several vowels compete. Herero hinda or shinda (to send) is the parallel «-form of handa, to go or run together, congeal, and hunda, to go or run over, be full, and its literal meaning is to mahe go out, whilst the cognate tyinda (kinda) signifies to go out, leave a place, travel as nomads, hence also to carry a burden. The A.-Sax. sendan is, therefore, nearer the original {KINDI or KHINDI) than the Goth, sandjan. In the Aryan words sinth, sind, sidh, way, journey, which correspond to Herero tyinda, Zulu sinda (be burdened, heavy, properly travel, carry a burden), the primitive i has been preserved. In Zulu the two Herero verbs tyinda (Aryan sind) and Mnda (Eng. send) have coalesced in the one form sinda, which means in that idiom — 8 A LANGUAGE-STUDY BASED ON BANTU. a. (to go out, to outgo, outrun), exceed, reach beyond ; b. (to go out), go out free, escape, get off, as from a punish- ment, escape from an illness, get restored to health, be saved, healed ; c. (to (JO out, leave a place, as nomads, carry a heavy burden), be heavy, weigh down, oppress with weight ; d. (to go out, do as in removing, take all the things out, clear out a native hut, in order) to smear the earthen floor with fresh cow-dung (sinda being in this case used jocularly, the clearing out of the hut preceding the smearing being compared to an exodus). In Tshuana the root has assumed the form of sita ( = sinda, nd being in this dialect frequently hardened to t), which means (to outgo, outrun), overcome ; causative sitisa (outrun, outdo), exceed, surpass. The identical sidi in Hausa denotes (to go out, leave, remove, hence) to bear, carry, endure. The study of Bantu enables us thus to demonstrate that the primary meaning of send is to (cause to) go out. But we can go even farther than this. We can analyse Herero hinda {hindi = KINDI = KI-TI, i-form of the genus Ka-Ta) and reduce it to its two monosyllabic elements Ka and Ta, both of which still exist in Bantu as independent true monosyllabic words — KA meaning to move, to run, to go, and TA to stretch, reach, extend, a combination of which we have in Ka-Ta = move-stretch (the legs, feet) = step, go, run, hence hata = go close together, stick to; — nasalised kanda = go or run together, congeal ; one of the I'-forms being hinda {KINDA) = (cause to) go oid, send. All this will be clearer hereafter, when the laws and beginnings of primitive speech are treated. Fundamental Philological Truihs.^-1 shall now proceed to state in a few brief theses some truths bearing upon the origin of language which, I believe, are discoverable in Herero. INTRODUCTORY. Language is the offspring of sight, not of sound. The ground- work of language was not formed in imitation of the cries of animals, nor were the first articulate sounds of an interjectional nature accidentally uttered, but roots were produced by rational observation guided by definite laws — laws founded on the beautiful harmony between the three motion- asjieds in nature (life-motion, wind- motion, rest) and the three organs of speech (guttural, labial, dental), as will be explained hereafter. Names like cuclcoo, pewit, formed in imitation of sound, and interjections as pooh, pshaw, are few in number, and do not belong to the organism of language. II. The beginnings of language consist of true monosyllables, with one consonant and one vowel, such as KA, TI, PU, with wide and general meanings, as to move, to go, to stretch, to wave, to fly. Words ending with a consonant, as, for example, Chinese tap, horn, are mutilated forms, and though at present pronounced as mono- syllables, they are not such in reality. III. By far the greatest number of roots in all languages (the isolating not excepted) are combinations of two primitive monosyllables which in most languages have lost the terminal vowel, as Chinese yi-Ti', Eng. fi-nd', loo-h\ rea-d', which, however, in Bantu have been preserved in their complete form, as ra-nda, buy, hi-nga (hi-ngi), drive, mu-na (mu-nu), see. No true primitive root ever exceeds two syllables. IV. The two first grand principles of language are motion (associated with rest) and space, the consonants representing motion (and the absence of motion or rest), and the vowels the various relations as to space, (time), and locality. The primeval laws which regulated in the beginning the use of the primitive consonants (guttural, dental, labial) and the primitive vowels (a, i, u) can still be observed in Herero. 10 A LANGUAGE-STUDY BASED ON BANTU. V. The differences between the several families of speech are, on the whole, not radical or material, but merely grammatical, formal, and conventional, each family having moulded what it possessed of the original common stock in its own fashion. The vowel-method (see Chapter IV.) will enable the science of language to demonstrate the origin of the several families of speech from one common source. ( II ) CHAPTEE II. PRIMITIVE ALPHABET— PRIMEVAL LAWS OF CONSONANTS. Language, like every other "good and perfect gift, cometli from above, from the Father of lights." But it must not be forgotten that reason is a gift even greater than language. God did not give language to man as He imparted to the nightingale her stereotyped inimitable music, but He gave him more ; He endowed him with reason and the organs of speech, and thus enabled him to create language. Man is the image of God ; language, in more than one aspect, the image of man. Very true is Herder's remark that the origin of language is really divine, inasmuch as it is human. The nature and measure of immediate divine assistance which Adam received in respect to language must of course remain a mystery ; but so much is plainly revealed in Scripture that the giving of names to animals was the work not of the Creator, but of the first man. "And out of the ground," we read (Gen. ii. 19, 20), "the Lord God formed every beast of the field, and every fowl of the air ; and brought them unto Adam to see what he would call them : and whatsoever Adam called every living creature, that was the name thereof. And Adam gave names to all cattle, and to the fowl of the air, and to every beast of the field." Now in this way, surrounded by the endless variety of the aiiimal creation, Adam practised lan- guage, and created those root- words which are still in the mouth of his children to this very day. Those who hold that Adam was created with a perfect philosophical language, do not consider that the intellectual labour of creating language with its accompanying daily discoveries and joys, must have greatly cheered our first parents in their solitude, and that, therefore, we have to look upon that task as a great boon bestowed on them by a loving and all- wise 12 A LANGUAGE-STUDY BASED ON BANTU. Creator. But enough. It is a fact which can be demonstrated that the present grand structure of human speech arose by intelligent effort, and in accordance with certain simple laws, from small, very small beginnings; and this fact, as it is in keeping with all the works of God, is also, like every other fact in true science, in perfect harmony with Eevelation. Priviitive Alphabet. — The primitive language would appear to have commenced with seven sounds, represented in the following scheme : — P- t k From these principal letters evolved the aspirates or stronger sounds, Icli, tli, ph, and, on the other hand, the medial or softer sounds, g, d, b, and the rest of the consonantal sounds down to the phantastic phonetic excrescences of Hottentot and Bushman. The great variety of vowel-sound can, as is well known, be traced back to the three primary vowels a, i, u, the pronunciation being as in German and Italian. The vowel a corresponds to the guttural Ti ; i to the dental t ; and u to the labial p. Primeval Consonantal Laws. — The guttural h (Jca) is the represen- tative of breath, life, and spontaneous motion; t (ti) means death, absence of motion, rest, hence also stretching, reaching ; and p (pu) is the natural and legitimate interpreter of wind, air, and motions caused by the wind or observed in the air, as the waving motion of the wings of a flying bird or the branches of trees. The letter m is unique ; only in the primitive noun (Bantu prefix and pronoun) it seems to be original, its meaning being mother, female, partner, mate, and, transferred to localities, inner, hence also present, place, properly mother, womb, cavern, grotto, house. In verbs, 7/1 will be found, in most cases, to be a substitute for an original p, as, for example, in Herero Jiama, press together, squeeze out, tama, stretch, pama, compress ; words which are contractions respectively of kamba {KA-M-PA), tamba {TA-M-PA), and pamba (PA-M-PA). PRIMITIVE ALPHABET. 13 Examples. — The following examples from Herero and Zulu will illustrate the powers of the consonants of the three different organs. The form in italic capitals is the approximate primitive one. K h, sh, s, 111, ty, y, g, ng, dy, Sfc. kauTca (Zulu), KA^-nJm), come to a stop, be broken off, interrupted, be stayed, as blood, lit. run back, be checked in running {Icauka being the inversive form of the obsolete ha, to run). Parallel transitive form ; kaula, KA(-ula), bring to a stop, terminate, put an end to, set a bound or limit to, staunch, as blood, lit. counteract running, stop running ; i (Herero), KI, to go, to go out ; M (Bantu), KU (pron.), he, she, properly the erect walking being, man ; kaka (Bantu), reduplication of KA, in Zulu bitter, pungent, properly hard, as in ixai-kaka (hardened) rings of a tree ; in Herero to become dry, hard, as a healing wound, to get a crust, the literal sense being to run together, congeal, become hard, or run on the ground, harden (the ground), beat a (hard) road. tyika (Herero), KI-KI, run out of the straight line, be oblique, slanting ; Icuka (H.), KU-KU, (run upward, start up), start for a trip, go on a journey, travel; hanga (H.), KA-KA, smoke, fumigate thoroughly, properly make dry, hard, as fumigated meat : literal meaning, like Icaka, to run on the ground, to harden, or to run together, congeal, (hence in Zulu hanga, draw or attract the eyes, look well, be attentive, watch, make the eye strong, hard), allied to Herero nyanga in nyangatara, to swarm, crowd, be numerous, properly run together for sight-seeing, as a crowd, the word being compounded of nyanga, run together, collect, crowd, and tara, look, see ; kangama (Herero), raise oneself, as in getting up from sleep, make oneself strong, properly hard ; frequentative form kangura, to burn, properly harden, bricks ; 14 A LANGUAGE-STUDY BASED ON BANTU. hanga (H.), KA-KA, to assemble, get together, collect, as people, warriors, properly make run together ; tyenga (H.), KI-KI, frequentative form tyengura, to upset, as a pot containing food, lit. cause to run out, throw out (the contents of a pot) ; henga (H.), KI-KI (run out of the way, shift, turn aside, i.e.), change ; hunga (H.), KU-KU, run upward from the stomach, vomit. th, z, s, d, nd, r, I, n, SfC. ta (Herero), to stretch, to reach ; ta (H.), to die, to perish ; tata (H.), to throw on the ground, to lay prostrate, properly stretch as a dead one, on the ground ; tandaura = tandavara (H.), to stretch, extend, from taiida, TA-TA, to stretch ; ram or Ma (Bantu), TA-TA or DA-DA, stretch (originally on the ground), lie down, sleep ; tim (H.), TI-TI, prop up, support, lit. place a dead thing, as a post or a stone (e-tize, a prop) against a house, a wall, in an oblique position ; a kindred form being teza (H.), TI-TI, to follow in a track, to pursue, properly to over- take, stop progress, bring to a stand, hold back, as a prop, a tottering wall, analogous to Lat. sustineo, to hold up, support, but also oppose, restrain ; allied to tiza and tea is tila (Tshuana), TI-TI or TI-DI, to avoid, get out of the way of anything likely to harm, turn aside, identical with Herero ti7'a, to fear (be kept back by something) ; tola (Zulu), toora (Herero), TU-TU or TU-DU, to pick up, take up (orig. lift up something lying on the ground), carry away ; lula (Z.), in lulama, TU-TTJ or DU-DU, rise up a little from a recumbent position, properly lie up, sit up (Tshuana), stretch upward; in Herero (ntrama) be straight, erect; tundura (H.) = tuiumuna, TU-TU, raise, lift up, as one in a fainting fit; in-tondo (Z.), TU-TU, heap of (dead things, as) stones, money, grain. PRIMITIVE ALPHABET. 15 P /, b, V, II), m, ^c. pa (Bantu), give, primarily make grasp, cause to take with the fingers, stretch the hand, fingers, the finger-rows being in Bantu looked upon and treated as wing-like objects ; papa (Zulu), flutter, fly as a bird : u-pape, a wing, a plume ; vava or papa (Herero), in vavera, papera, PA-PA, to spread out, as a skin on the ground (originally to spread out the wings) ; pepa (Z.), PI-PI, start (properly fly) suddenly aside, evade, escape by starting aside ; pepa (H.), PI-PI, blow, as in making fire : om-Sepo, wind ; pemba (H.), PI-PI, be smooth, pure, properly bright, shining, burn- ing, blown, stirred, as fire ; pema (H.), contracted from jiemba, to blow the nose; pupa (H.), PU-PU, to flow, properly drive on the water, buoy up, be light, as a wing, feather ; pupa = pumha (H.), PU-PU, to prune, lop (properly lighten) a tree. As to the pronominal roots or prefixes of the noun,^here too the working of the consonantal laws can still be traced in Bantu. In offering the following Herero nouns as illustrations, let me premise a brief remark on the meanings of some of the prefixes occurring in them. It will guard us against confusion and misunderstand- ings. The prefix oku- xv. in such nouns as dkn-tui, ear, &c., is difl^erent in character from the infinitive oTcu-. The former is a true prefix, but the latter is merely the preposition or directive hu ( = to) elevated to the rank of a prefix. Both are indeed radically one, but the difference is that the oku- in o^SM-tui is a true primitive noun and personal prefix, whilst the o/fw- of the infinitive is a. secondary form. Thus d^u-tui (okvi-THUVI), ear, is properly the hearing person, or the hearing one, i.e., ear, but olcu-zuva (oJcti-THUVA) means literally "the to hear" (infinitive) or "hearing." In the e-v (or e-) class, two (or three) classes of nouns have coalesced i6 A LANGUAGE-STUDY BASED ON BANTU. whicii are diametrically opposed to eacli other. There we have, first of all, e-(KA) for living, and then e-(TI) for dead things. Also in on- xiv. some classes are thrown together : the singular (PU), the plural (KHU), and perhaps another plural form (PHU). As to the prefix ozon- x., it is sometimes a true plural, as in ozon- gombo, goats, plur. of on-gomho, goat, and might then be distinguished as ozo-n- (EHA-KIATT), but in nouns like those we shall quote presently, it is one of the forms of the sexual dual, ozon- (primit. form TI-MI, TU-MV). K = living being or thing : onm-ndu {KU-MU-ndu), man (orig. human couple) ; on-nyanda (KI-MI-nyanda), cattle, properly cattle-pair ; on-gombe {KI-MI-komhe), ox or cow (orig. ox and cow) ; on-yama [KI-MI-yama), flesh (orig. suckling animal, " saugethier, " animal whose meat may be eaten) ; om-hua (KI-MI-hua), dog (male and female dog) ; on-gomho {KI-MI-lcombo), goat (he- and she-goat) ; on-Au (KI-MI-zu), sheep (ram and ewe) ; on-geama {KI-MI-lceama), lion (and lioness), and most other animals. oku-o/fo, arm, front-leg, oma-ofe {KA-MA-oJco), arms, properly male- female-arm, both arms; oTsM-rama, leg, oma-imna (KA-MA-rama), (both) legs; ols.n-tm, ear (as being fleshy, living, or moving, pricked up, as the ear of an animal), plural (orig. sexual dual) om.a,-tui (KA- MA-tui), (both) ears ; e-ke (KA-Jie), hand, oma,-Jce {KA-MA-he), hands, originally male- female hand, right and left hand ; e-vere (KA-vere), female breast, oma.-vei-e (KA-MA-rere), breasts (orig. the two breasts, looked upon, as all fleshy members of the body, as living). T = dead thing : e-yo {Tl-yo), tooth. The Zulu i-zm-yo, dialectic i-tia-yo, is the sexual dual TI-MI-yo = male-female-row of teeth, or the two rows of teeth, abbreviated in-yo / e-tupa (Tl-tupa), bone ; PRIMITIVE ALPHABET. 17 e-ue (Tl-ue), stone ; ozon-yara {TI-MI-yara or TU-MU-yara), (double row of finger) nails, Zulu (iz)m-tipo, nails, singular w(lu)-i!t>o (TU-tipo) ; ozon-ya (TI-MI-ya or TU-MU-ya), horns (properly pair of horns), Zulu i{z)im-pondo, horns, sing. uQia)-pondo {TU-pondo) ; ozom-himht {TU-MU-pumbu), eyebrows, properly the pair of eye- brows (r. pumha, to prune, to lop, to clip, from the shortness of the hair), sing. orn-{p)umhu {TU-pumbu), and also, but irregularly, om-bumbu , oiu-uua (TU-uua), rock ; ora-uma (TU-uma), dust. P = waving thing : ou-to (PU-ta), bow (PU = wing, branch, bough, tree, &c.) ; omTi-ii (PU-MU-ti), tree (from the branches resembling a pair of wings) ; onm-nue {PU-MU-nue), finger, lit. the waving pair, the two finger- rows, on account of their being like the wings of a bird or the branches of a tree ; omn-na (PU-MU-na), lip, originally the waving, flapping, or blowing pair, their motion being like that of the eyelids or wings ; omu-pepo (PU-MU-pepo), bellows, lit. the blowing lips ; omn-Jcova (PU-MU-Jcova), skin, cover, orig. probably pair of wings : omulcova u'elw, eyelid, cover of the eye, properly male-female- wing of the eye (-pair) ; oTod-tuha, levity, plur. of ovaxi-tuha {PU-MU-tuka ; r. tuha, to start up, to fly up), orig. probably pair of wings ; orsm-vare (PU-MU-vare), large sheet of water, lake, lit. the wing- like waving expanse (-vare = wide, expanded) ; omM-ramba (PU-MU-ramba), torrent, lit. chasing, pursuing lake (r. ramba, follow after, pursue) ; Kafir um-lambo ; omu-pupo {PU-MU-pupo ; v. pupa = to flow), stream, torrent (from the motion of the waves). From specimens like the above, which abound in Bantu, we arrive, by the safe method of induction, at the following general principles in the domain of the consonants : — I. Every living being and thing (man and animal kingdom), and B I8 A LANGUAGE-STUDY BASED ON BANTU. all spontaneous motion, is represented by the breathing throat-letter k and kindred gutturals and palatals. 2. Every dead thing (including the mineral kingdom), and all absence of motion or rest, is expressed by what may be called the dead or tooth-letter t (it being produced by contact of the tongue with the dead rows of teeth), and other dentals and Unguals, though, as we have shown, dentals and linguals are also, under certain con- ditions, applied to indicate motions. 3. Every waving thing (vegetable kingdom, &c.), and all waving motion, and the blowing action of the wind which causes it, is denoted by the blowing lip-letter p and other labials. This natural correspondence between Thing and Word is, of course, to a great extent obliterated in our modern languages. But even in the most developed, traces are left to show that originally it existed. With the addition of the word " originally," the following quotation from Trench's " Study of Words " (p. 29) appears to me true and to the point: "The words which we use are (originally) not arbitrary and capricious signs, affixed at random to the things which they designate, for which any other might have been sub- stituted as well, but they stand in a real relation to these." Reasons for the Consonantal Laws. — These are self-evident, as we have intimated just now; for no articulation could have been a better representative of breath, life, and spontaneous motion than the consonantal sound ha (and its modifications), which is produced by the tongue touching the soft palate in closest contact with the fountain of breath or the throat, and aided by the breath from the throat. I am aware that the appellation "guttural" for h is objected to by some, who prefer to speak of it as a palatal, restricting the term "guttural" to the stronger and harsher modifications of h, as we fi.nd them in Hebrew and other languages. But we are not dealing here with an artificial alphabet, but with the " alphabet of Nature;" and it is not likely that any one will contend that the sound ha was pronounced by primeval man merely by bringing the tongue to the palate, loithout at the same time emitting breath from the throat. In pronouncing ti and pii we can very well dispense with the throat, but we cannot say ha in a natural manner without the sound being accompanied by emission of breath. As to the dental ti, produced as this sound is by contact of the tongue with a PRIMITIVE ALPHABET. 19 double row of dead, hard objects, its fitness for describing the absence of motion, hardness, death, &c., is obvious. Nor will it be denied that the choice of the lip-letter p (and its modified sounds) for express- ing the notion of blowing, waving, flying, was the most simple and natural that could have been made. Indeed, there was no choice at all in the matter. If in any degree we succeed in bringing ourselves to look at God's beautiful world as if we saw it, like the first man, for the first time, our very first observation will be the threefold variety in reference to motion, namely, spontaneous motion (men, animals, living members of the body), waving motion (wings, branches, plants, waves of the ocean, lakes, and rivers, the eyelids, lips, &c.), and the absence of motion (teeth, bones, earth, stones, metal, (fee.) ; and we shall then be struck at the same time with the corresponding fitness of the representatives of the three organs of speech : of Jca to make breath, life, and voluntary motion audible ; of pu to turn wind, air, and the waving motions therein into articulate sound ; and of ti to be the audible sign for the absence of motion, dead matter, solidity, and rest. It is out of this intimate union of the threefold motion- aspect in Nature and the three corresponding organs of speech that Language was horn. All this will appear to some, I am afraid, as a mere play of fancy. But let it be remembered that the element of imagination figures as an important factor in the creation of language. We who are living in an advanced state of culture find it difiicult to realise the primi- tive simplicity, naivete, and even poetry of the world when it was young. It has been said that the last man who leaves this world will be a poet : be this as it may, it is certain that the first man was, and the immortal poem he produced is Language. In researches like those we are now engaged in, on the sacred paradisiacal ground of the origin of language, we must stoop down, and in a literal sense put off the shoes from off our feet ; for how can we otherwise discover that our feet, on account of the rows of toes, are wing-like objects ? We may smile now at such an idea, but it was in that light that the hands with the finger-rows and the feet with the rows of toes appeared to our first parents. Some time ago I listened with pleasure to an able lecture " on wings," in which the lecturer also called attention to man as a loinged being, confining his remarks, of course, to the soaring mental powers of man ; and it is not likely 20 A LANGUAGE-STUDY BASED ON BANTU. that any one of his hearers did find fault with him for not discover- ing wings among the members of the human body. It would, how- ever, have been different, I am inclined to think, if the lecture had been addressed to the first family of man ; for they would have considered it defective on the ground that no mention was made at all of the wing-like members of the body, viz., the finger-rows, the rows of toes, the pair of eyelids, and the pair of lips. Now, it is this primitive intuition of Nature which we must strive to get restored, if we would arrive at clear and true ideas regarding the origin of language. ( 21 ) CHAPTER III. PRIMEVAL LAWS OF THE VOWELS. The vowel-sounds had, in the beginning of language, an inherent power to modify the sense of a root, forming in this way from a single root a whole cluster of independent root- words, with one pervading general idea, but differing as to space and locality. As the consonants represent, regulate, and diversify motion, so the vowels are originally signs for the various positions and relations in space. These relations — as, for example, far, together, on the sur- face, within, without, in, between, aside, oblique, up, on high, over, through, &c. — at present denoted by prepositions and adverbs of place (or space), were originally indicated by the three primary vowels a, i, u. If, therefore, KA signified to run on the ground, to run together, the meaning of KI was to run in, or from within, out, to run in or between, aside, out of the straight line, oblique, and KU conveyed the sense of running upward, on high, over, and from above, down- ward, also ahead, before, &c. It is, therefore, wrong to say that "the vowel a is the great primitive vowel," and " that if an a-sound compete with another vowel-sound, the a-sound belongs to the primitive form." We may, indeed, assign to the vowel a the honourable place of being first among equals, but farther we cannot go ; for it can be proved that the other two vowels, i and ii, are equally primitive, and quite inde- pendent of a. Each of the three primary vowels has a distinct individuality, moves in a sphere exclusively its own, and performs its own peculiar functions. A. The vowel a gives a root the by-meaning — on or along the ground, on the surface, flat, level, horizontal, 22 A LANGUAGE-STUDY BASED ON BANTU. near the ground, not raised, not vertical, high, or full-grown, hence diminutive, common, low, small (in pronominal roots), wide, broad, extended, stretched out, abroad, far, distant, there, absent, past (in pronominal roots), straight, in a line, parallel with, together, toward, at — and the reverse, asunder, scattered, spread. Any action, motion, or condition naturally inciting a horizontal gesture with one or both arms, falls within the sphere of a. I. The vowel i signifies — in, inner, inside, hidden, in a place, present, here (in pronominal roots), within — and from within, out, without, outside, out of, forth, out(running), exceeding (in length or height, be big, tall), excelling, projecting, straight, stiff, (running) out (as liquids, hence also) over, up, out of the way, at the side, aside, to and fro, turning, circling, out of the straight line, oblique, across, athwart, in, between. U. The vowel m has the power of pointing upward to motions and actions in the air which would call forth, as a natural gesture, the lifting up of the arm. U means — above the ground, in the air, up, upward, vertical, perpendicular, high, over,— hence, as over a river, through, overflowing, full, before, in front, ahead — and the opposite, at the back, behind, following another, up, erect, rising, great, large, above — and from above, PRIMEVAL LAWS OF THE VOWELS. 23 downward, down, under, below, bowed down, bent, curved, round, crooked, twisted. These primeval powers of the vowels, therefore, naturally produce a great variety of meaning in one and the same radical. So we find, for example, in Hebrew, that the meanings to cry, call, to be astir, awake, to dig, hollow out, to go round, encircle, enclose, and several others are all centred in the one form ghur : the reason is because the consonantal skeleton of the root, GliR, means "to go," and the vowel u gives it the by-meaning of (i) upward, hence to go up, rise, and rouse, cry, call, awake, and be awake; (2) to go from above, downward, bow down, bend, curve, go round, hence to encircle or enclose ; but also (3) to hollow out, dig, properly go round with a digging instrument, as in widening a hole, make a round hole. Now the same meanings are combined in Herero koea. This verb means (i) to go up, rise up, run over and make run over, fill, feed up, nurse, as an infant, make grow; (2) to go round, as in a round native hut, when searching for something ; and (3) to go round with an instrument, hollow out. Nor shall we have to go far in looking for an example in our own languages : the Aryan root KUR (K.EU, K.RO, G.RO, &c.) at once presents itself for illustration. The vowel u (o) gives KR (GR) the by-meaning of (i) upward, up, high, in Gnow, Dutch ano'Eien, and in GU'EAt, Ger. geosz, Dutch GROOiy (2) from above, downward, bowed down, bent, curved, as in Lat. cvnvo, to CEOo/i:, allied to Ger. KRUmm, Dutch krojm ; hence (3) round, hol- lowed out, hollow, in crocZ;, CRUse, Dutch, keoes, krviJc, Swed. KEu/ia, Ger. Kuvg, Gael, cnog (the round, hollow thing, earthen vessel), not to mention a number of other words sprung from the same root. But it will, of course, not be expected that in all cases the primi- tive vowels should have retained their original purity ; a, in a num- ber of roots, has become e ; i, too, often sounds like e ; and not less frequently do we find u changed into 0. Nay, more than this. There are rare instances in Bantu in which a radical u has changed to e or a ; for example, Zulu tamba, to subdue, be tame, soft, mild, gentle, which evidently is a modified form of the original tumba, to bring into submission, take captive, capture, just as the identical English, to tame (Lat. domo), is a variation of the (as far as the radical vowel is concerned) more primitive doom (O. H. Ger. tuomjan), the primary 24 A LANGUAGE-STUDY BASED ON BANTU. sense being in both words the same, namely, to make go down, to bow down, to subdue (as animals, enemies), hence also to pronounce judgment upon captives of war, decree, punish, condemn. Similar cases of vowel-shifting will be often met with in the Aryan languages, though here the radical vowels have, on the whole, stood their ground much better than in the Semitic and Hamitic languages, Hottentot- Bushman included, where the radical vowel-element has been terribly convulsed. Nevertheless, even in the latter families the working of the prim.eval vowel-laws can still be traced. We have, for example, in Hottentot (ISTama) — xhua (for XKUNA, to go up, to rise), to dawn ; and the reverse xkua (for XKTJNA, to go from above, down), to descend, to come (properly bow) down (xkvA-glia) ; from this xkud-p {XKUNA-P, the bending one, or) the knee ; quni (the bending one), the elbow ; cf . Herero e-ltono (the bending one), the arm, hence also branch ; and Gr. yom, Eng. knee (the bending one) ; vhona (to stoop low, crouch), to beg ; cona, to beg ; cf. Tshuana kona, to bow down, to bend ; Herero hona, to stoop low, to crouch, creep ; vkan (variation of vkona), to ask, beseech ; vhan (allied to vkona and vkan), to creep, shrink ; qganu (to go up, over, or through, the a being a substitute for an original u), to cross a river; Herero konda ; qganu, prep, through ; qkAu (contracted of QGANU, Bantu konda, kondo, primitive form KU-TU= KU-N-TU, to go over or through), to ford a river, cross over ; qkau (for QKANU, to go up, over, through, go through with a knife), to cut (through) ; Herero konda, kondo ; cf. Eng. cut, nasalised sund-er ; qkau-qa, to cut asunder ; qkau-s {QKANU-S), circumcision; vhanu-vhanu (go up, over, through, cut through, decide), administer justice ; vhanu, straight, right, just ; vhanu-p, justice, rectitude; qanu-qanu (to go or run up, over, overflow, wash clean), to purify, PRIMEVAL LAWS OF THE VOWELS. 25 make holy; cf. Herero Icona, Jcono {KUNDV, to run up or over, as water), to flow over, to make clean, wipe ; qanu, pure, clean, holy ; ganu-p, purity, holiness. Now all these words, so different in form and meaning, can, with the aid and guidance of the vowel-laws, be recognised as near akin to each other, as branches of the same root and stem, namely — KUNDU nasalised form of KU-TU, which, in various modified forms, means — 1. To go up, to rise, to run over, be full, overflow, wash clean, &c. 2. To go from above, down, to come down, to crouch, creep, beg, &c. 3. To go over or through, ford a river, go through with a sharp instrument, cut, sunder, separate, decide, &c. In some of the above N"ama words the radical vowel u has changed into a ; in others the first consonant has been encumbered with the click element (c = dental, x = lateral, q = cerebral, and v = palatal click) and the nasal ra (remnant of nd) has disappeared — quite in keeping with the tendency in Hottentot to grind the root-words down to monosyllables — still all of them have retained so much of the original family likeness that they betray their close relationship to each other. In the Aryan languages the consonantal skeleton of roots is more perfect, and the primitive radical vowel has been more gently dealt with, so that, as a rule, if lost in one, it has been preserved in another idiom. But in which ? Here lies the difficulty. Com- parative philologists are often not a little puzzled at the variety of vowel-sound in many a group of Aryan root-words, not to speak of the vowel-changes in moods and tenses. They are at a loss as to wMcJi form ought to he placed at the head as the nearest approach to the original. Now, in endeavouring to settle questions of this kind, the study of Bantu is indispensable : it will render material aid to the student, and in many perplexing cases point out to him the way to arrive at a satisfactory result. The English verb to stand, for example, sounds in 0. Eng. stonde, A. -Sax. stondan, standan, Goth. standan, 0. Fries, stonda, Dutch staan, Ger. stehen, Skr. std. Now 26 A LANGUAGE-STUDY BASED ON BANTU. which of these is the most perfect and approximately primitive form? Again, the verb to bind is in Sax. and 0. Ger. Undan, Dutch linden, pret. lond, Ger. linden, pret. land, Skr. landli. Which of these forms is the most original ? If the student will take the trouble of examining Herero and kindred idioms in African Bantu and Polynesian, he will find sufficient reason to decide, in the former case, for the 0. English stonde or O. Friesian stonda, and, in the latter, for Sanskrit landh. For a comparison with Bantu and Polynesian leads to the discovery of tundu or tutu {TU-TU, nasalised TU-N-DV) as the primitive form of s-tand {stonda) ; Herero tunda, indicative tundu, " aufrecht stehen," tund-ama, to stretch up, be raised, elevated, stand high, &c. ; Fijian tu (probably abbreviated from tutu), to stand, ai-tutu, a stand or place to stand on or in, allied to donu, Tongan tonu (TONDU), straight, right, correct; whilst the full original form of to lind has been preserved in Herero paxda (pand-eka, to make go together, to bind; ortia-pando, fetters), nasalised form of pata, to go or cause to go close together, to shut, close ; Zulu pata (in several forms), to move, to draw together, to clasp, to shut close together, as an iron trap, to get close upon, engage in close fight, "handgemein werden," hence to touch, handle, pat ; allied to Tongan fatu (in which the second a has changed to u), to tie, as rafters of a house, to make go together, to fold, fcdui, to fold up, mata-fatic, hard, not easily made to cry, from ma-ta, eyes, and fatu, shut, literally eye-shut or eye-bound. The following scheme is intended to represent the three principal powers of the primary vowels, the source of the various secondary meanings as stated before : — a = on the ground, horizontal ; i — in, within (the body, earth, place, water, &e.); u = above the ground, high, in the air, vertical. M PRIMEVAL LAWS OF THE VOWELS. 27 The Voivel-Latos Proved. — I shall now endeavour to illustrate and exemplify the laws that have been stated, and for that purpose invite attention to a few groups of Herero root-words. For brevity's sake, and in order to present a clearer view of the several groups, I shall, as a rule, give the verbs only in their simple form. KAKA, TYIKA (kIKA), KUKA, kaJm ( = to go or run on the surface of the earth, to beat a road, make hard, or go together, congeal, shrink), be hard, dry, get hard, get a crust ; tyiJta ( = to go out of the straight line), incline to one side, stand oblique, be aslant ; xeJca ( = to go out of the straight line, out of its proper place), to be or go out of joint, to get loose (as a waggon), to relax (Zulu); kuJca ( = to go or run up), to start, as for a journey, to travel ; to overflow, sweep away, as a stream of water, clear off (Zulu (Icuk-ula) ; swell, expand, swell with passion, pride, &c. (Zulu hultu-mala) ; to appear above the surface (Tsbuana kuku-nya), to rise above the horizon, used of clouds (Tshuana kuku-mologa) , Icoka ( = to run or go up, over, and the opposite, to move downward, bend, curve; to go ahead, in front, before, &c.), to rise up or swell out, as food boiling (Tshuana koko-moga) ; to run over, or make run over, iill, satisfy, pay, render what is due (Zulu) ; to go before, to lead (Xosa), hence Zulu u-koko, ancestor, progenitor, grandfather; to go before something, drag it over the ground (Herero koka, koko-xm'a, Tshuana koko-tha) ; move from above, downward, bend, curve, stoop, &c. (Zulu koko-ha, crouch or stoop in walking; koko-heza, subdue, put down ; ama-koko-ma, stoop in the back ; Herero -koko; curved, crooked ; koko-vara, to be curved, crooked). — Modifications of the above are : — kaha ( = to run on the ground, harden the ground by running over it, or, to run together, congeal, become hard), to be hard, dry, firm, solid (-kahe) ; haka ( = run on the ground, run on, run fast), Jiaka-hana, make haste, oru-haka, rashness ; Mha ( = run out of the straight line, move to and fro, from one side 28 A LANGUAGE-STUDY BASED ON BANTU. to the other), be moved (with pain, pity), feel pain, feel pity (Mh-ama), be moved with kindness towards others, be kind and considerate, be anxious to provide for the wants of others {hiha) ; huka ( = run up, rise), rise, get up, go away (Makonde), hence Herero omu-huka, morning, properly the rising fire or light ; huha ( = i-un downward, be bowed down, be bent, curved, round, go round), to bewitch, properly bind round, fetter; e-hulia, loop, tie ; otyi-huha, state of being bewitched, properly state of being bound, fettered ; holca ( = run downward, go down, bow down, bend, curve, be round, and go round), put round, fence in, protect, as plants by a hedge, thatch a house (hoJc-era). Nasalised forms of this group are : — KANGA, TYINGA (kINGA), KUNGA. kanga ( = to run on the ground, to beat a road, make hard, or run together, congeal), to be hard, dry, or to make hard, dry ; tyenga ( = to run out of the straight line, or to run from within, out), to upset, as a pot containing food {tyeng-ura), to throw out; xenga-xenga ( = to run or go out of the straight line, to move to and fro), to move from one side to the other, be shaky, loose (Zulu) ; kunga ( = to run up), to throw up, from the stomach, vomit ; to run over, as liquids ; kunga ( = to run downward, bend, curve, go or be round), to bind round, as a string of beads round the arm, or to put a rope round the neck of an animal, tie up an animal (Zulu) ; konga ( = to go over or through, to go before, hence also to follow), to go through with a sharp instrument, to sever : to follow, lit. be fronted, run or go after something in front (kong- orera). TATA, TITA, TUTA. tata ( = to stretch on the surface of the earth), to throw flat on the ground ; Uta ( = to sink into), Konde titi-ma, sink into; Zulu titi-hala, be PRIMEVAL LAWS OF THE VOWELS. 29 wet thorouglily, soaked, confounded, helpless (be as a drown- ing one); titi-nya, to sound, as the depth of a river or a person's meaning or purpose ; teta ( = to put in or between, as the teeth or a knife, to divide), bite, chew, crack, sever, crop ; tiza ( = to turn out of the straight line, to place oblique), to lean against, to prop, support ; tuta ( = to reach up, to pile up), to carry and throw things together in a heap, make a heap ; tuta and tota ( = to go through, or to go round, as in making or widening a hole), to hollow or be hollow (-tutu, -toto). — Closely allied are : — tara ( = to stretch on the ground, stretch horizontally), to stretch, extend (tara-vara) ; tira ( = to turn out of the straight line, turn aside), to fear ; Tshuana tila, to avoid, get out of the way of anything likely to harm ; tera ( = to turn out of the straight line, from one side to the other), to stagger (tera-teroi) ; tura ( = to stretch upward, lift up), lift up anything (iur-ilca) ; Zulu tula in u-tuli (idu-tuli), dust, disturbance, lit. rising ; opposite sense : tula (be put down, be subdued), be silent, mute, quiet, calm, peaceful, be settled as water ; tura ( = to stretch upward, lift up, as the fist or an instrument for striking, crushing), to strike hard, pound, crush, as a bone ; tora or toora ( = to stretch upward, lift up), lift up from the ground, take up ; Zulu tola, pick up, take up, &c. Nasalised forms : — TANDA, TINDA, TUNDA. tanda ( = to stretch, reach horizontally), to aim at, fix the eye as in taking aim, have the eye on, hence to intend, devise mischief, threaten (in Zulu to love) ; tanda-vara, to stretch, extend, spread ; tinda ( = to put between, to intervene), to resist, refuse (Zulu tinta, to intercept, stop) ; tenda ( = to put between, as a sharp instrument, to divide), to cut ; tunda ( = to stretch up, reach high), to rise, climb, be erect, high ; tonda ( = to lift up, as a stick for beating, allied to the contracted io A LANGUAGE-STUDY BASED ON BANTU. form tofia, to beat), to stamp, as witli a pestle, to pound, to hate ; (identical with Zulu zonda, to hate, abhor ; have a fixed pain). PAPA, PIPA, PUPA. papa (= to squeeze close, together, as the fingers, wings, <&c.), to be firm, solid, compact ; papa ( = to spread out the wings, spread the wings on or near the ground), flap the wings, flutter (Zulu) ; pejpa ( = to fly out of the straight course), start aside, avoid (Kafir) ; p>epa ( = to blow into, or to blow out of the mouth), to make fire, blow up a fire ; pupa ( = to blow or fly upward), to be light, easy {-pupu), to float. — ISTear akin are : — ■ vava or papa ( = to spread, as wings, or a skin on the ground), to spread on the ground and fasten with pegs, as a hide to dry (vav-era, pap-era) ; otyi-vava, a wing ; viva (= to move out of the straight line, move to and fro, from one side to the other), to fan, wag, practise fencing, &c. ; viv-iza, to whet, sharpen (from the motion), get ready for a fight (applied to a bull) ; hoha ( = to quickly move downward), to stoop in order to hide, hide oneself behind something ; hoi-ela, to stoop forward, to sub- side, as a swelling (Tshuana) ; bopa ( = to quickly move downward, bow down, stoop, bend, curve, go round), bind round, fasten round the waist, as a belt, gird, wind a bandage round a wound, &c. (Zulu). Nasalised forms : — PAMBA, PIMBA, PUMBA. pamha (= to move, squeeze close together), to plait, properly put things close together ; piimha ( = to move out and step in, as in changing places or taking turns ; to make room for another, or step into the place of another), to exchange, barter, requite, retaliate ; pumba ( = to fly or rise quickly upward, or cause to move upward, to ease, remove a burden, as the superfluous branches of trees make light), to prune, to lop; to take off the point, to blunt. PRIMEVAL LAWS OF THE VOWELS. 31 Meaning of the Vowels in the Herero Prefixes. — Oa this topic I can only give a few hints here. A fuller statement of my views on the original signification of the Bantu pronominal forms the student will find in Chapters VIII. and X., and in the Introduction to my English-Herero Dictionary. The Roman numbers here and else- where belong to Bleek's arrangement of the prefixes, based on Herero, as being of all Bantu idioms as yet discovered in South and Central Africa "the richest in classes of nouns." The reasons of my partially departing from Bleek's classification I have intimated in the aforesaid dictionary (p. xxviii.) As to the primitive forms in italic capitals, they will be explained in subsequent chapters. omu- (KU-MU) I;— plur. ova- (KHA) II. The vowel u in omu- means up, upright, erect, this being the posture of man (omxL-ndu, Zulu umu-iitu). Originally the prefix on-, o(v)u-, Zulu w-bu- (KHU) XIV. = men absolute, corresponded to omu- as legitimate plural, but on assuming its present abstract meaning, as in on-ndu, humanity (orig. men), ova-, Zulu aba- (KHA) came into use. A in the plural prefix ova- means (men) in general, or spread over the earth, (men) abroad. on-, om- (KI-MI) IX. ;— plur. ozo-w-, o-zo-m- (KHA-KIMT) X. The fact that the prefix on.' (oin-, in-), which is the prefix for names of animals, and its corresponding pronoun should appear, in all Bantu languages, with i as the radical vowel, has always been a puzzle to me, until quite recently the true cause has, I think, become clear to my mind. The vowel i in in- IX (animal, &c.) stands to omu- (man) in somewhat the same relation as, e.g., the Herero verb yera (YIRA) to yura (see Chapter VII.). Both verbs mean " to raise," but there is this primary difference : yera literally means to oailift, " /jerawsheben," while the original signification of yura is to upliit, "aufheben." So the prefix omu- signifies a going- up, that is, a grown-wp, upright, erect-moving being, man, but on- (m-), in virtue of the vowel i, a grown-OMi, i.e., full-grown living thing ; hence also Bantu HI (orig. animal-) father, and NI (orig. animal-) mother, the vowel i signifying here " out," and answering exactly to the German " awsgewachsen sein," full-grown, in distinc- tion from young, immature, which originally was represented by the 32 A LANGUAGE-STUDY BASED ON BANTU. (now diminutive) prefix oka- ; the vowel a meaning in general (a living thing moving) on the earth; hence om.a,-paha {KA-MA-imha), a couple of children, twins (singular KA, at present e-) ; but omu- ndu {KU-MU-ndu), the grown-up, upright pair (full-grown, complete man, orig. man and woman, father and mother, husband and wife), and ozi-gombe {KI-MI-lwmbe), the grown-ot;^, i.e., full-grown (" aus- gewachsenes ") pair of cattle, orig. ox and cow, at present ox or cow : — KA = living thing generally on the earth, hence also child, young animal, living member of the body, &c. KU = raised, grown-up, upright KI = growa-out, expanded, full- (" aufgewsLch sener ") living grown (" az(sgewachsenes "), one (man, father), and what- living thing (animal, father), ever resembles the erect and any object resembling human body. an animal. The vowel a in the plural ozo-n- (KHA-KIMT), corresponding pronoun ze, za, signifies on the earth, abroad, or in general : — KI = full-grown animal ; ^/-il/7 = full-grown animal-pair ; KHA-KI2II = number of living things in general, or spread over the earth. omu- (FU-MV) III. ;— plur. omi- (PI-MI) IV. U in omu- means obviously up in the air, as the outstretched wings of a flying bird, the waving branches of trees ; hence the application of this prefix for waving things generally and whatever resembles them, as the branch-like finger-rows, the wing-like eyelids, the flapping lips, the waving river, &c. The plural omi- is possibly just a phonetic variation (umlaut) of omu-. e- (TI) and e- (KA) V. ;— plur. oma- (KA-MA) VI. The vowel i in e(ri-), Zulu Hi-, prefix for names of dead things, means in, as the dead teeth (e-yo) in the mouth, the bones (e-fupa), in the body, the stones (e-iic) and metals in the earth. With the above e(ri-) V., another prefix, e- (KE, KA), identical with oka- XIII., has coalesced, with the vowel-meaning on the earth, PRIMEVAL LAWS OF THE VOWELS. 33 KA signifying originally, as we stated just now, living thing on the earth, hence living thing generally, as e-paha, one of a twin-pair, e-hono, one of the living arms, &c. It is to this latter e- (KA) that oma- (KA-MA), originally a form of the sexal dual, now corre- sponds as plural. The original plural of e(ri-), which must have been something like TI {THI), has been supplanted by oma-. oru- {TU) XI. ;— plur. otu- (THU) XII. U signifies here up, rising upward, high ; hence the nouns of this class embrace high, long and lengthened, thin objects. Otu- is the original legitimate plural of oru-. otyi- {KI) VII. ; plur. ovi- {KHI) VIII. Otyi- (Kongo ki-, Zulu si-) is properly the singular of the sexual dual form on- (o-in-, in-, KI-MI), which see. Originally ^/signified full-grown animal (" aMSgewachsenes thier "), but at present it is a neuter prefix with the general meaning " thing." Ovi- is its primitive natural plural. oka- {KA) XIII. ;— plur. on- (KHU) XIV. The meaning of the vowel a here is on the earth. Oka-, at present diminutive in Herero and other idioms, must, as we said before, have meant originally living thing in general, child, young animal (hence diminutive), in distinction from KU (grown up, great, or erect living one) and ^/ (big, "atjsgewachsenes," full-grown living thing). The plural ou- {KHU), in Angola tu {THU = KHU), instead of ova- {KHA) or ozo- {THA = KHA), is an irregularity which pro- bably arose from the disturbance caused by the substitution of ova- {KHA) II. for the original plural (now abstract) prefix ou- {KHU), ou- {PU) XIV. ;— plur. oma-w- VI. The vowel « of the singular prefix ou- has the same meaning as u in omu- {PU-MU) III., namely, up in the air, high, as the boughs of trees. Thus ou- in ou-^a, bow, is properly the singular of the sexual dual omu- in o-mu-ti, tree ; its original meaning is one of the wings or branches of a tree, a bough, a bough for shooting (ou-^a), i.e., a bow. C 34 A LANGUAGE-STUDY BASED ON BANTU. The plural oma-u-ta means literally boih bows, that is, more than one, a number. oku- (EU) XV. ; — plur. oma- or oma,-7cu- VI. U in oku- is the same as u in omu- {KU-MU) I. Its force is up, upright, erect as man, pointed, high, hence also distant : okn-tui, ear (originally not any ear, but the pricked-up, pointed ear, as of a horse or an ass, hence also) oku-iya, thorn (one of a couple of those long straight thorns which resemble the pricked-up ears of an animal) ; dkn-oko, the perpendicular man-like living member, front- leg of an animal, hence also arm, and oku-ra?;ia, originally hind-leg of an animal, at present leg in general; oiiu-ti, field, woodland, probably tree with a head like the erect high head of man, or head high up in the air, bushy head of trees, bush, forest, country, to which latter noun the infinitive oku- ( = motion to a place or to an object at some distance) probably refers. The legitimate primitive plural of oku- is ou- {KHU) XIV., but as this form has been appropriated as an abstract and as plural of oka-, and as, moreover, the cori'esponding dual omu- (KU-MU) has been set apart for man {orsm-ndu), the a-form of the latter prefis, oma- {KA-MA), is now in use as plural (properly dual) of oku- : ols.n-tui, ear, oma-tui, ears, properly couple of ears; okn-iya, thorn, oma-Jcu-iya, thorns, properly a couple of pointed ear-like things. Irregularities like the above, in the correspondence between singular and plural, date from the remote period when the original powers of the vowels ceased to be known. The local prefixes opo-, oko-, and omo- are briefly treated in Chapter V. Reasons for the Vowel-Laus. — But how, it remains still to be examined, came the vowel a to mean on the ground, horizontal ; the vowel i inward, within ; and the vowel u upward, on high ? As we can still trace the original meaning of the primitive con- sonants I; t, and p in both pronoun (or primitive noun) and verb, such a question ought not to be regarded as presumptuous. "We know for certain that k originally meant to breathe, to live, to run , to walk, t to be dead, to lie, stretch, and p to blow, to wave, to fly. Now, it is evident that these letters, when first uttered by primeval man, were not pronounced vowelless as k', t',p' ; they must have PRIMEVAL LAWS OF THE VOWELS. 35 come forth, as it were, twin-born, with a vowel joined to them. And if we further inquire which were the twin-sisters of h, t, p respectively, we shall not hesitate to admit that the organs of speech point to a as the natural companion of the guttural Jc ; to i as most intimately connected with the dental t, and to u as being nearest akin to the labial p. These three twin-born primary articulate sounds are entirely independent of each other. " Man kann im allgemeinen sagen : die articulation des h beginnt da, wo die fiir das t aufhbrt, und umgekehrt" (Briicke). Ka naturally bursts forth from the throat (and palate), and does not require the aid either of the teeth or lips ; ti as easily takes its origin from contact of the tongue with the teeth, independent of the throat and lips ; and as to pu, it can be produced by the mere motion of the lips (as in blowing), without receiving help from either teeth, throat, or palate. In short, a trial with each of the three organs of speech (throat and palate, teeth, lips) separately will call forth from the throat (and palate) the sound ha, from the teeth (and tongue) ti, and from the lips pu. Now, as running, walking {ka, a) is done on the ground ; blowing, waving, flying [pu, u) is observed in the air, on high; and as the dead teeth and bones are within the mouth and body, it is perfectly intelligible that Adam may have been led to apply the vowel-sound a to living things and motions on the ground, u to waving objects and motions in the air ova'head, and i to dead things, and at the same time to any motions and conditions that are within or hidden, like the rows of teeth in the mouth, the bones in the body, and the stones and metals in the earth. ( 36 ) CHAPTEE IV. THE VOWEL-METHOD IN UNIVERSAL ETYMOLOGY. The primeval vowel-laws, whose operation in Herero is too plain to admit of doubt, furnish us, as I have shown in former papers, with a new method of word-comparison — a method which, unshackled by the different grammatical superstructures, goes in a direct way to the body of language, and which therefore may be called the direct metJiod of word-comparison, or also, as it springs from a knowledge of the laws of the vowels, the vowel-method. Its outlines are as follows : — 1. No true root-word stands isolated in language ; it is a member of a family, in close relationship to a more or less numerous group ; and each group of roots is, whatever individual difference there be, pervaded by one leading idea. We have thus, in universal ety- mology, to treat a root- word as in aiSnity and relation to the whole group. "We do not compare isolated words ia the several families of speech; we confront genera with genera, species with species. 2. The classification of root- words into genera and species is effected by the primary consonants k, t, p (with an additional primitive m in the pronominal roots), and the primeval vowel-sounds a, i, u. In a genus we have root-words with the vowels a, i, u, and kindred shades of sound, but a species comprises only roots of one primary vowel and its kindred sounds. Thus in Herero the genus PaTa would comprise roots as the following •.—■pata, paza, vaza; pita,piza, vira, vera ; pitta, puza, pura ; whereas a species would be confined to pata, paza, &c., a second species to pita, pitha, vera, &c., and a third to puta, puza, pora, kc, ; future researches are sure to suggest subdivisions, but so much may be established even now as funda- mental, that it is the consonants which mark off the genera, and the THE VOWEL-METHOD IN ETYMOLOGY. 37 vowels the species (and sub-species); or, in other words, motion gives the genus ; direction varies it into species ; thus : — Genus. Species. C hata, go on the ground, go together, &c. R-T = go . . I hita (EITI), go in, enter, &c. (. kuta (KUTU), go up, over, &c. 3. Now each genus is pervaded by one leading idea, generally to go, to move, to run, to stretch, to reach, to wave, and the like, the vowels modifying the sense, as has been stated before, producing, as a rule, seemingly opposite by-meanings, as together and scattered, upward and downward, in and out ; and wherever we find this pheno- menon in the several families of speech {however imperfectly the roots may have ieen preserved), there we discover true relationship and original unity. The very fact, for example, that the root PAT A means in Herero (jiata) to shut, in Hebrew (patah) to open or expand, and in Latin to be open, free, expanded (pateo), warrants our identifying these words ; for the root PATA means both to go or run together, to shut, and the opposite, to go asunder, to spread, to expand. The genus KaTa will serve as an illustration and specimen of universal etymology, in accordance with the rules of the vowel- method. In Bantu the root kata has the meaning to fold, to glue together, to cleave together, stick to, be attached to, &c. In Fijian we have " kata, a., close together, touching, as boards on a floor, so as to leave no crevice; va-kata, a., shut, close" (Hazlewood's "Fijian Dictionary "). Now as the Polynesian languages stand in a sisterly relationship to African Bantu, based on the principle of grammatic identity, I expect to find it so, and there is no doubt in my mind concerning the identity of the Bantu and Polynesian kata. It is different when I turn to other families of speech. The Bantu- Polynesian kata, to cleave together, to be close together, reminds me of the Aryan root KAT- in Lat. catena, Germ, kette, Dutch keten, chain ; and as k and g are closely allied, also the root GAT- {gad-, gath-, in Germ, gatte, Dutch gade, spouse, gader, gaderen, Engl, gather) occurs to me. Now the question arises, Is the 38 A LANGUAGE-STUDY BASED ON BANTU. Aryan root KAT- {gat-, gath-, gad-) identical with tlie Bantu- Polynesian kata ? Without some law to guide me, the answer to this question must be mere guesswork. But here the vowel-method steps in and regulates the inquiry, and I reason thus : Kata in Bantu-Polynesian is only a member of a genus of root- words whose leading idea is to go. Kata is a root of that species which, in virtue of the vowel a, means to go together, &c., closely allied to other species, as Mta (hita, sMta) and Jcuta (Jwta), which respectively mean to go in, and from within, out, to go over and through, &c. Now if I discover in the Aryan family the very same phenomenon ; if I find forms like KATA (cat-, Icet-, gat-, gatli-) with the primary meaning to go together, and parallel i- forms, KITI {hid-, shid-, shit), with the seemingly opposite meanings to go in and to go out ; as also parallel M-forms, KUTU, KOTO, denoting primarily to go up, over, through, and from above, doim : then I have evidence as strong as can be expected in philology that kata in Bantu-Polynesian and kat- {ket-, cat-, gat-, gad-) in Aryan are identical. Genus KaTa. First Species. Kata = to go on the ground, to go together. Bantu : kata (to go together, hence) to stick to, cleave to, attach to, as e.g. thorny grass to clothes, to smear, plaster, paste on, make to stick to (Zulu) ; from this in-kata (Zulu), on-gata (Herero), n-gata (Konde), coil, ring or knot of grass, a pad; n-gata (Tshuana), bound-up package, bundle; omn-kato (Herero), trunk of an elephant, lit. the coiling-up, roUing- up lips or skin ; on-gata-oJco (Herero), slo-\vness, unsteadiness in one's work, lit. arm-coil, close folding of the arms ; Jcata (Herero), to shrink together, of plants, dry up, wither, reduplicated kakatera, KA{TA)KATAIRA, to stick to, cleave to; allied to hata, reflex, rihata (Herero), to coil oneself up, as in sitting stoopingly on the ground, with crossed legs ; allied to sata (Zulu), draw together, embrace, have connection with a woman, sat{amsa), fasten on one thing to another, as the blade of an assegai to the haft. THE VOWEL-METHOD IN ETYMOLOGY. 39 A strengthened form of Imta is handa (Herero), to run together, to congeal. The meaning on the ground we have in yata (Herero), to go on the ground, to tread, step, nasal- ised or strengthened yanda, "fest auftreten," walk with firm steps, run, run fast, far, out of sight, cease, end. Polynesian : kata (Fijian), close together, touching, as boards on a floor, so as to leave no crevice. Aryan: hette (Germ.), heten (Dutch), catena (Lat.); gadde (Icel.), to press together; gador (A.-Sax.), gather (Engl.), gader, gaderen (Dutch); gatte (Germ.), gade (Dutch), spouse, con- sort, mate — words which appear to be radically identical with Bantu- Polynesian kata. With Herero yanda or anda, to cease, end, the Sanskr. anta, Goth, andeis, Engl, end may be compared. Second Species. KiTi = to go in, between, to go out. Bantu : Mta (Herero), to go in, to go between ; allied to tyinda (Idnda), to go from within or out, to leave a place, as nomads, remove, and to hinda, to cause to go out, to send (Herero) ; tyera (Herero), to go between, intercept, waylay, aim at one, try to seduce, allied to tyiza (Herero), perceive, discern, orig. go between. Polynesian : Mli in faka-hili-MU (Tongan), to intercept, turn into another route; hele in faka-ma-hele, to cut in two, helu, scissors, comb ; helu-helu, to comb, properly make a path, divide the hair ; kild (Fijian), to know, understand, regard, literally go between, discern. Aryan: shide (0. EngL ), [s/ac? (Icel.), splinter; scidan (A.-Sax.), to cleave, split, divide; scheiden (Germ, and Dutch), to go between, separate, part; scheitel (Germ.), schedel (Dutch), crown of the head, probably from the dividing of the hair ; scirian, scerian (A.-Sax.), to go between, to divide, to part among two or more, allied to shear, share; the seemingly opposite meaning from within, from between or out, being found in the nasalised or strengthened form send, Icel. senda, to go^or cause to go out. 40 A LANGUAGE-STUDY BASED ON BANTU. Third Species. KuTU = to go up, over, througli, down. Bantu : kuta (Herero), he filled with food, properly be running over ; suta (Zulu), eat or drink to one's satisfaction, be full, sated ; suta (Herero), pay one's debts, also moral debts, satisfy, atone, lit. make run over, fill up ; opposite meaning : kota = Icora (Herero), to go downward, bow down, bend down, be crooked, curved, bent ; nasalised or strengtbened form hunda (Herero), to go over, run over, be fuU (of numbers) ; allied to Iwnda, to go over or through, also to go through with a sharp instrument, to cut through, to saw. Polynesian : koro-horo (Kjian), heaps, as of sandbanks (lit. a running up or high, a great quantity, a mass, a meaning which -koro also has in Herero) ; lcoro-{niniuna), the prominent parts of the buttock on each side of the backbone ; gutu-va (Fijian), to cut off (go over, through); faka-goto (Tongan), to go down, sink, press under water. Aryan : sat (in which the original u has given place to a), A.-Sax. sad, sated. Germ, satt, Dutch eat, Lat. satis, enough (full, running over) ; leuta, to cut with a knife, kiiti, a small knife (Icel.); kotta (O. Swed.) = Engl. cut, allied to sunder. Germ. sondern, to go or cut through. The opposite meaning, to go down, be bowed down, bent, we have in such words as crook {KUR = KUT), Icel. crdkr, Swed. krok, a curve ; cf. cur in Lat. curvus, bent, arched. The above words form, of course, only a small part of the wide- spread ramifications, both in Bantu and Aryan, of the genus KaTa, but they will answer our present purpose, and be helpful, in some measure, in showing the scope and working of the vowel-method. 41 ) CHAPTER V. NOTES ON THE PRIMITIVE BANTU PREPOSITIONS AND ADVERBS. 1. The hlgli antiquity of the prepositions and adverbs of space, PA ( = at, by, near), MU ( = in, within, present), and KU ( = to, above, on, off, at a distance, &c.), is evident from the fact that they occur in a more or less perfect form in all Bantu idioms. 2. The reason why there are three is, because they have stepped into the place of the three primary voioels, A, i, u. When the reign of the latter ceased and their laws became obsolete, it was found necessary to choose other signs for expressing those relations of space which had been indicated before by vowel-sounds. 3. PA became the successor of a, MU oi i, and KU oi u. In the forms PA and KU the original vowel-meaning is obvious (a = at or near the ground, not high in the air ; u = upward, above, on the top, and from above, down, &c.), but in MU it is exclusively the consonant m to which the idea in, within (orig. mother, womb, inner and present place), attaches ; in other words, m as preposition, adverb, or particle means always in, present, whether it be pronounced with u (MU), a (MA), or i (MI). 4. The preposition (and adverb) PA is radically identical with the Herero prefix opo- (PA) in opo-wa XVI. = a place near, at hand, allied to the prefix ou- (sing., prim. r. PU), as in ou-ta (PU-ta), bow, properly bough, one of the wing-like branches of an omn-ti (PU-MU- ti), tree. PA, as a primitive noun, probably signified originally the human hand and foot, both being looked upon, as they are in Bantu to this very day, as wing-like objects, on account of the branch- resembling fingers and toes, which, therefore, in Herero are still denoted by one and the same name, omvi-nue (PU-MU-nue) = the wing-like member, PU-MU signifying originally a pair of wings (as 42 A LANGUAGE-STUDY BASED ON BANTU. of a bird, flying in the air, or the wing-like branches of a tree, waving high in the air) : hence the a-form of PU : PA = wing or branch of the human body on or near the ground = hand (or foot) = at hand, at, near, close by. The vowel u in the Herero preposi- tion pu, and in the adverb po are substitutes for an original a. 5. The primitive Bantu noun PA XVI. runs parallel with the verb pa, to give, properly make grasp (with the branch-like finger- row or hand), and with pa in Bantu p)apa, vava (fly near the ground, flutter, stretch or flap the wings, spread, as a skin on the ground, &c.), from which Herero e-pa, branch, and otyi-vava, wing. 6. As, therefore, the Bantu monosyllable PA means a. (at) hand, near, close by, and h. to (cause to) grasp, give, it is probably identical with the Chinese " classifier "pa, which means " to seize, to grasp, or lift up with one hand, being applied to many things held in the hand by a handle when used," and also with our preposition and adverb ty, A. -Sax. he, bi (near to, by, of, from), Goth, hi, Germ, hei = at hand, near, at, on, &c., originally the same as the prefix he- (A.-Sax. he and hi, Goth, hi, Germ, be and hei) in Sefore, because, beware. 7. The preposition MU is radically identical with the Herero prefix omo- {MU) XVIII. in omo-na, cavern, grotto, house, inner or present place, the primitive noun MU signifying, as we said before, mother, womb = in, present. It has also been stated already that in this preposition the notion in is transferred from the vowel i to the consonant m, which means in, loithin, present, even when pro- nounced with u {MU) or a {MA). The Herero adv. and conj. nu { = now, and) is an offshoot of mu = in, present. Compare our adv. and conj. now, Goth., Dutch, Dan., Swed. nu, 0. Sax., 0. Germ., and Icel. «i2 = at the present time. The Bantu preposition na = present, near, with, and, &c., is a modification of MA, a-form of MU. 8. The preposition KU is radically identical with the Herero prefix oko- {EV) in o'ko-na XVII. = distant place (orig. head or high place, top-place, also bush, woodland, extent of country, &c., height and length or distance being the same thing in Bantu). The primitive noun KU signifies the great, erect, or high Hving one, hence man; but in the primitive language it was also probably applied to the high living member of the body, the head, as we may infer from the Herero noun o'ka-ii XV., field (wood, bush, PRIMITIVE PREPOSITIONS AND ADVERBS. 43 forest), land, country: — ti = wood: omn-ti (PU-MU-ti), tree, properly winged or branching wood, tree, the branches being com- pared to a pair of wings ; orvi-ii, a long stick, as a long whip-stick, properly a liorn-tvee, a piece of wood, long and thin, as a horn ; oku- ti, bush, forest, field, country, literally 7isac^-wood, i.e., high and great wood, or tree with a high bushy head or crown, hence bush, forest, country, distant place. KU was, therefore, well fitted for expressing the notion " on the top, high," and also " long, distant, in front, &c." The oku of the infinitive is a secondary form, derived from the primitive KU as occurring in dk.W.-ti. 9. As the Bantu pronoun kv, (thou, thee) is in Aryan tu, so is probably also in our preposition and directive to (Sax. and O. Fries. to, Goth, du, tu, Germ, au) the dental a changed h. The Bantu Jcu is used in exactly the same way as our to ; ma vanga okuya hu ami (Herero) = he-wants-to come-to-me; up to heaven, Veyuru (ku-eyuru). That our to originally not only meant "motion toward, extent," but also "high, up, over, on," like the Herero Jco, appears from the emphasised too (A. -Sax. to. Germ, zu, as in da-7M, zu grosa, zugebeii), over, more than enough, in addition. The demonstrative so (coupled with al : also) too seems to be a modified form of to, and if so, it betrays the existence, at a very remote period, either of a mono- syllabic noun KU (TU), or of a noun which had KU (TU) for a suffix or prefix (as in the Bantu xv. KU class), and to which noun so corresponded as a demonstrative pronoun. 10. The result thus of our inquiry into the nature and original meaning of the three principal Bantu prepositions is as follows : — P.4 = wing, branch, or branch-like object near the ground, human wing or branch-like member, i.e., hand (foot) = at hand, by, near, close by, not high up in the air, but on or near the surface of the earth, corresponding to the primary meaning of the vowel a. i(fC= mother = womb, womb-like place, cavern, grotto, house = in (and out), inner or present place, present, noio (nu), cor- responding to the primary meaning of the vowel i. KU= the high living one = head = top, bush, height, length, dis- tance, extent, motion to (and from), on high, on the top, in 44 A LANGUAGE-STUDY BASED ON BANTU. addition, too (t6), upon (and from upon), off, &c., cor- responding to the primary meaning of the vowel u. Accordingly, the three local Herero nouns to which the three primitive Bantu prepositions and adverbs correspond, appear to have the following primary meaning : — opo-wa {PA-na), hand-place, i.e., at hand place, a place which is near, close iy. omo-na (MU-na), mother- or womb-like place, inner place, hence also present place, in which one now is. dko-na (KU-na), head- or top-place (high, hence also) distant place, to which one has to move. ^ For some further remarks on the nature of the Bantu pre- positions, see § 32 in the Introduction to the author's English - Herero Dictionary. ( 45 ) CHAPTEE VI. ROOT-FORMATION : ITS BEGINNINGS AND SUCCESSIVE STAGES. The first utterances of man were monosyllabic, regulated by the simple laws we have stated in former chapters. They consisted of one consonant and one vowel, the former taking in all cases the precedence, as Tea, ti, pu. Now we may observe that such primitive monosyllables served in a double capacity, namely, as a name for a person or thing (noun, pronoun), and as a term for a motion, action, or condition (verb). The monosyllable KA, for instance, means the living, running, walking one, but also to move, to run, to go. KA embodies thus originally subject and predicate ; it conveys the meaning of a whole sentence ; noun and verb proceed from it in the following manner : — KA = the runner runs. ha = (the) runner. ha = (is) running. Here we have the first or germinal stage in root-formation. The first beginnings of language must have been something like the following : — 1. Throat. KA = the breathing, living, running one, animal, moving spon- taneously ; to breathe, live, move spontaneously; run, go, strike the ground, &c. 2. Teeth. T/=the (inner) dead one, tooth, bone; to be dead, motionless, die, lie, aka will lead us on the right track, and clear up the matter at once. DEVELOPMENT OF THOUGHT. 53 Palca oma-tui means in Herero folk-lore to prick up the ears as an animal, properly, to move the ears together, to attend closely, to listen sharply, of which the later phrase pakera m'omatui (to put into one's ears, to listen) is a corruption. The origin and chain of thought as to Zulu paha-ma may therefore be traced as follows : — Pa-Ka = to move quickly. PA-KA = to move quickly together, hence •paka (omOrtui), to move the ears quickly together, as an animal, to raise the ears, listen attentively, " die ohren spitzen ; " from this om-haJca-tui, attention ; e-patye (for e-pake), an observant one ; paka-iza, to look searchingly (like an animal with pricked-up ears ; cf. with paka the widespread Aryan root pas, spas, spek, spcih, spell, to look, to look searchingly, to examine, to spy) ; paka-ma (Zulu), to be raised into a point, to stand erect, be high, elevated. The same line of thought has been followed out in forming the Zulu nouns in-taba, mountain, hill, and in-daha, news — two words which seem to be perfect strangers to each other, but which never- theless are closely allied, or, in fact, identical. For the Bantu root taba or tava primarily signifies to stretch horizontally, as the hand ; to stretch or strain generally ; to stretch or strain, as the udder or teats in milking (Herero tava); to stretch the head or the neck, as in close looking (Dutch reikhalzen), inspect closely (Herero tav-iza) ; to stretch the neck and prick up the ears, as an animal, to be attentive (to a call), to respond (i-tav-era) ; to listen to news, hence in-daba (Zulu), story, tale, news, report ; further, to raise into a point, be high, elevated, as the stretched neck and pricked-up ears of animals : hence Zulu intaba, hill, mountain, properly point, allied to Herero on-davi or oru-tavi, point (of a plant or branch), ear of corn. Or let us take the notion straight, right, just. The word for these conceptions is in Zulu lunga, primarily to move upward, raise erect. In Herero, where ronga ( = lunga) has been employed for to prepare, get ready, equip (rong-era), the name for straight, right, just, is seniha, primarily to outrun, to run as in a race, to 54 A LANGUAGE-STUDY BASED ON BANTU. run in a straight course, hence to be straight, right, just. In the Herero root seka, too, we have the vowel e (i) for a similar concep- tion. Seka means now to equal, but originally it signified to outrun, to rival, as in a race, the primary meaning being still found in ha-seka, to distance, properly to run fast, far, to leave behind, as in racing. This combination of thought, namely, to race and to be straight, right, is very plain in Tshuana sia, identical with or allied to Herero seka. Sia means to run away (run fast, literally outrun), hence to win a race; sia-na, to race: hence sia-ma, to be running in a straight course, to be straight, right, just. So it may also seem strange that in the Herero word for to drag (koka), to drag along the ground, the vowel o (u) should have been applied, and not a. The reason is, because some one or some thing must go before the object which is to be dragged, hence koka (indicat. koko), lit. to run or go in front, before, make follow, make a thing follow, drag it. Precisely the same intuition underlies the Konde synonym uia, to draw, and uta-nga (stronger form), to drag. Here we have the very same primary meaning, namely, to go before. Uta means in Herero to begin, properly to step in front before another, be the first, as in founding or cultivating a place. It is the parallel u-ioxxa of yata or ata, to step, tread : — ata = to step, tread (on the ground) ; uta {utu), to step before, to go ahead, before, to begin, be the first, found, create ; but also (as in Konde) to go before a burden, i.e., to drag. Now the same line of thought we observe in the afore-mentioned Herero verb koka (koko). Koka (primitive form KU-KTJ) is the M-form of A-a/if( = to run o« the ground, to beat hard, as a road, or to run together, to congeal, become crusty, hard : — kaka = to run on the ground or together (be or become hard, get a crust) ; koka (koko), to run or go before (a burden), to make follow, drag. In the Herero translation of the Psalms we read n'otyinuino iyandye tyi iika, and cup-mine-it-runneth over (Ps. xxiii. 5). The DEVELOPMENT OP THOUGHT. 55 passage might also be translated n'otyinuino tyandye tiji Iwna, for kona (contracted form of Iconda) means to run over, to flow over (hence also to wash, wipe, cleanse, "reinigen"). Now the latter word (Jiona) substantiates our statement that the vowel u (0) primarily means up, over; but how can tika, whose radical vowel i signifies in and from within, out, convey the same sense as kona ? For the simple reason, because a ilowing out may also be a flowing over. N'otyinuino tyi tika means literally "and my cup is running out,'' which is tantamount to "is running over." Hahn, in his " Herero Wdrterbuch " says correctly, " Tika = aus- und iiberlaufen." For the primary sense is out, hence also over. But tika, if doubled, means also to tickle (Zulu tikatika, Herero tikatik-isa). At first sight there does not seem to be any conceivable connection between the notions to run out or over, to overfiow, and to tickle. But on closer examination we shall judge differently. Brincker leads us on the right track when he says, " Tikatikisa = kitzeln, eigentlich zum lachen reizen " (to tickle, properly to excite to laughter). Viewed in this light, the difiiculty is at once removed. The literal meaning of tikatika (causative tihatikisa) is, to cause a pouring forth (of excessive laughter, as in tickling), hence to tickle. Again, in Herero we have for the notion to lift up, to raise, both yura {ura) and yera (era). But how can the vowel e (i) in yera give the root the by-meaning up ? Obviously because its original force is out, to outgo, exceed, hence also to owMift, to w^-lift, to raise : — ■JURA, to go uj>, to lift up, raise, " aufheben ; " YEEA (yiea), to outgo, exceed, lift out or up, " herausheben, empor- heben, aufheben." In the following instances — predicative roots with the primary vague idea to run, to go — the evolution of thought is, on the whole, clearly seen, and speaks for itself. Hekebo. Pbimaky Sense. Development. xnga run together assemble, form an alliance, enter into a covenant, make peace. 56 A LANGUAGE-STUDY BASED ON BANTU. Heebeo. yanga Pkimaet Sense. run togetlier handa go or run on the ground yanda go or run on the ground pata go or run to- gether panda vanda mana (modified of vanda) hilia go or run to- gether go or run over the surface. go or run over the surface move to and fro henga run aside tyiza go in, between heza step out of the straight line kenda go out of the straight line, be oblique Development. collect, as water in a peri- odical river or in holes ; float on, as alluvial matter ; collect, as one's thoughts, stand still, reflect, "sich zusammennehmen, sich sam- meln." tread firmly, stem. run fast ; cease, end (run out of sight, disappear). shut (as the two parts of a door, a trap, a box, &c.), catch (get close upon, be engaged in close fight, hence also) quarrel, contend, deny. bind, fetter; work hard, be industrious (as a tied or tamed animal, om-handi, or a bondsman). level, flatten, smooth, &c., spread (as a covering), cover. plaster (Konde inata), finish (originally a building by plastering), complete, bring to an end. be moved with pity, sym- pathise, be desirous to pro- vide for the wants of others, &c. shift, change. discern, perceive. glide out, commit a mistake, fail. be ambiguous (oimi-liendi) DEVELOPMENT OF THOUGHT. 57 Hbeeko. Peimakt Sense. tenga outrun zenga go or run in, be- tween vera (allied to run out (as a veto) missile out of the hand) Tivka run up koha run over lidha run up Mha run round Development. be first, be respected, begin, (intercept, as with a snare), entangle, (throw), beat, punish. start, travel. wash, cleanse, purify, be pure, chaste. heap up, add. (run round, as a rope), bind, fetter (by witchcraft), be- witch, bring misery upon one. teach, good or bad, instigate. be full (of food), be satisfied. satisfy, pay, atone for. bow down, reverence, worship. be subdued, tame, calm, cool (as to temper), just, pious. Herero yazema, to lend, to borrow, means literally to go or draw close together, to know each other well, as relatives or friends, to be on terms of intimacy, the root yaze or aze being identical with Zulu azi, to know (well, intimately), be kind to, regard, respect, be intimate : mave yazema (lit. they go close together, are closely connected, are intimate friends, hence) they lend, borrow. The Herero verb lama, to sow, to plant, primarily expresses the general and comprehensive idea to go before as a pioneer, to cultivate the ground, to found a place, to farm. For Icuna is a contraction of kunda, which has in Herero the following meanings: (i.) to go or run up or over, to overflow, to be full (applied to numbers) ; (2.) to go before (as a herald), to make known in the villages, to announce (properly to herald) ; modified form : liuna = to go before, as a pioneer, be the first in cultivating a place, cultivate the ground, hence "to sow, to plant." With Herero kuna we identify the Aryan root KUN- (KON-, honga run or go before Jcida go or run over sufa (cause to) run over kota go down, bow pora go down, bow 58 A LANGUAGE-STUDY BASED ON BANTU. QAN-) in the 0. Sax. kun-ing, Dutch kon-ing, Sansk. ganaka (in which the original u is changed to a), Engl, king, — and in Goth. huni, O. Sas. kunni, Dutch kunne, Lat. genus (genero), Engl, kin, relationship, family, &c. If the identity is admitted, we can in Bantu trace the primary meaning of Icing and hin as follows : — KUNfls (Herero ; contracted of kunda ; indicative form kunu {KUNDU) = (to go before, as a pioneer, be the first in cultivating a place), to farm, to plant, to sow ; (found a clan, beget, originate), create (Sansk. gem for a primitive KUN-): hence KVTS-ing (king), first cultivator, founder of a place and family, progenitor, chief of a race or tribe, father, king (of. Herero onm-hona, chief, lord, the stem Itona being possibly a modi- fied form of kuna) ; Kvm (kin), generation, family, relationship. We may here also, in conclusion, again point to the Herero verb uta (utu), most probably identical with, or allied to, the German ur- (in ur-bar, arable, being in a state of cultivation ; ur-heber, first beginner, author), in which we observe the very same train of thought, namely (i) to sfej> before, to go before ; (2) to be the first, to begin ; and (3) to found, to originate, to create. ( 59 ) CHAPTER VIII. ON THE HERERO PRONOUN — ABSOLUTE AND CONDI- TIONAL FORMS—PRIMEVAL LAW OF THE PLURAL- SEXUAL DUAL. We observe in Herero the following facts in connection with the pronoun : — I. The pronouns, identical with the formative prefixes of the noun (treated in the Introduction to my English-Herero Dictionary), are primitive nouns, and mean — a. the living one (man, person, animal) ; h. the blowing, waving one (tree, branches, wings, waves, &c.); c. the dead one (tooth, bone, stone, earth) ; d. mother, female, mate (applied, in the first place, to living things, but also to dead things). Examples. U M = he or she (person) falls ; u M = it (the tree) falls ; ri u = it (the stone) falls ; ru M = it (the rock) falls ; tu u = they (the rocks) fall ; mu u = (she) it (the grotto, house) falls. Primitive form and literal sense. KU 'u = the erect moving one (man) falls ; PU u = the high waving one (tree) falls ; TI II = the dead one (stone) falls j TU u the high dead one (rock_) falls ; THU u = the high dead ones (rocks) fall ; MU u = th6 mother (womb-like place, grotto, house) falls. 6o A LANGUAGE-STUDY BASED ON BANTU. 2. Pronouns being thus in reality nouns with the meaning man, person, &c., the same pronoun may be used for representing the third and second, or even the first person, analogous to the Chinese " servant says " for " I say ; " as, for example : — u z = he knows ; u i = thou knowest. Primitive form and literal sense. KU i = (he) man know(s) ; KU i= (thou) man know(est). The above rule applies, however, only to those forms which repre- sent man or woman, and whose consonants were originally Tc, Jch, and m, mil. The originally dental forms {ri, ru, tu), and those who had for their primitive consonants j3, pTi, can, as they are names for inanimate things, only occur in the third person. 3. We distinguish two kinds of form in the pronoun — (a) the primitive, natural, or absolute form ; and (6) what we shall call the conditional form, because, under certain conditions relating to space and time, it is modified by the vowel-laws. The conditional is derived from the absolute form by change of vowel, the vowel a giving the pronoun the by-meaning iliei'e, yonder, distant, absent, abroad (third person) ; whilst the vowel i, by virtue of its meaning in, in loco, present, here, fits the pronoun for representing the first person or person present. Examples. KU (u) = man = he (absolute) ; KA (a, e) = man = he there, absent, abroad (conditional). Plural. KHU (vu, u XIY.) = men = they (absolute, at present in use for abstract nouns) ; KHA (va, ve, ba II.) = men = they there or abroad (conditional, at present used for men, people generally). Now, as the conditional form for the third person is derived, by means of vowel-change, from the absolute ground-form, so is also the form for the first person obtained by changing the vowel u to i, as the following scheme shows : — THE HERERO PRONOUN. 6i KU (u, ku) = man = lie ; thou, thee (absolute). KA (a, ye, e) KT (redyi, ngi, i) = man tliere, absent, abroad = man in loco, present, here = he (conditional). = I (conditional). All the singular forms of the first person have, in Herero, the vowel i, or, contracted with the particles a and ma, e, viz., mdyi («gi), i, e ( = a-i), me ( = ma-i), ami (a-mi), mbi (m-vi). The latter form, though at present in general use for /, is originally first person plural, running parallel with the third person plural va, ve; — KHU (vu, u in o\i-ndu, humanity, originally) = men, people = they (absolute). ' KHA EHI ^ (va, ve, ?Bba, mbe) = (vi, mbi) = men here, present men there, abroad = they ^ = we (at present in use for (conditional). first person singular I). 4. Primeval Law of the Plural. — The plural of the Bantu primi- tive noun or pronoun was formed in accordance with the following simple law : — In order to indicate the plural or a number of persons or things, aspirate the consonant of the singular, and pronounce the word with greater force. Thus if ka meant the living one, kha denoted a number of living ones ; if ti was the name for a dead body, as, e.g., a tooth, a bone, a stone, thi conveyed the idea of a number of such dead bodies ; and if pu signified the waving one, as a branch, a wing, a wave, the plural must have been something like phu. Later, when the primeval law of the plural became obsolete, a reaction took place ; the massive plural forms collapsed by their own weight, and softened down — kha (or khha) to tya, ta, za, ya, wa, va, la, a, &c. ; hhu (or hhhu) to tyu, tshu, tu, yu, wu, bu, u, &c. In these and similar shapes the plural forms appear at present in Bantu. Thus the plural of.ru, lu (TU) XI., originally thu or thhu, is now (in Herero) tu XII., and in the corresponding demonstrative pronoun su ; both the singular and 62 A LANGUAGE-STUDY BASED ON BANTU. the plural have lost a grade of their original power : tu having been weakened to ru (lu), and tlm or Mm to tu or su. The present plural prefix vi- VIII. (wi-, hi-, iy-, 1-), to add another instance, is only a weak remnant of an original khi or Jchhi, the plural of ki- VII. (tyi-, si-, y-, i-). See the pronominal tables in the Introduction to my English-Herero Dictionary, and at the end of Chapter X 5. This leads us to call attention to the modification of consonants in the various forms of the primitive noun (prefix) and pronoun, especially in the plural — changes which can still be traced in Herero to their natural cause. We observe that gutturals have changed, on the one hand, to dentals, and, on the other, to labials, as e.g. the primitive plural form KHA : — KHA = living beings = they. va II. = they (men). za X. = they (animals, &c.) In Kihiau the forms of the VIIL class of nouns are vi (wi) and hi, the latter betraying its descent from an original KHI. In Ndonga we have j30-A«, on the earth, i.e., beneath (Herero p-e-hi), but e-vi, earth (Herero e-hi), h having changed to v. As to the transition in Bantu from Jch, tij, dy, to z, this is the same as the change in Greek from dy to z, — Sanskrit Dyaus, for example, being in Greek Zeus. In the same way as va II. and za X. from KHA, the plural pronoun tu (Tshuana oho, tsho), we, us, was derived from an original KHU, the connecting link (tlm, tshu, tsho, tyu) being still in existence in Herero and other Bantu idioms : — KHU = living, erect moving beings = men (absolute form). vu, u, bu (orig. = men) = they tu = men = we, us (absolute (absolute form). form). Here we see that the third and first persons (plural) are repre- sented by forms essentially one. But it must be borne in mind that in the XIV. class of nouns and pronouns two (or perhaps three) classes have coalesced, viz., the primitive form PU (u, vu, bu, sing. XIV. ), and KHU {vl, vu, bu, plur. XIV.) = men (absolute), THE HERERO PRONOUN. 63 at present in use as a diminutive plural and also as a singular for abstract nouns. All the abstract nouns with the prefix on- (vu, bu XIV.) must be regarded as adjectives of ou-nclu {ja-ndu = KHU- ndu), at present denoting humanity, but originally men, people (absolute form), of which the conditional form ova.-ndii {\)a,-ntu = KHA-ntu) is an offshoot : — oVi-ndu (bu-ntu — KHU-ntu) humanity, properly men, people (abso- lute); ova,-ndu (ba-»f2« = KHA-ntu), men there, abroad (conditional). The original meaning, therefore, of Herero on-ninga-ndu (happi- ness) is happy men ; of ou-Jiaze-ndic (negligence), negligent people ; of on-pore (gentleness, righteousness, piety), righteous, good men, &c. , just as the ending -head, -hood (in manhood, knighthood) also appears to have primarily signified " person." Shifting of Nouns. — It is particularly interesting to observe in Herero how a concrete noun, after coming into use as an abstract, was replaced by another. Sometimes the new concrete noun was merely a modification of the old one, but in other cases an entirely new name was coined. Thus on-ndu denoted, as we have seen, originally men (absolute), and on becoming an abstract (viz., humanity), it was replaced by the conditional form ova-ncZw, men, people (properly people abroad). Omi-hiTca (r. tuJca, start up, fly) was one of the original names for wings, but on assuming the char- acter of an abstract in "levity," another word (otyi-vava, outspread thing) was chosen to take its place. On-dyoae, originally the twist- ing, spinning animal, probably spider, from yoza, to twist, spin (allied to on-goze, cord), means at present phantom, vision, " traum- bild," literally twisted thing, answering exactly to the German " (hirn)gespinnst. " Its substitute is oty-auvi, spider. The original Bantu name for goat is (Zulu) im-buzi (identical with the Herero abstract om-huze, news, rumour, report, inquisitiveness) = the prying, inquisitive, curious animal, from buza (Herero pura), to inquire, ask about news, be inquisitive. Now when om-buze ( = im-buzi), the inquisitive, curious animal, came to mean "news, rumour, report (curiosity)," the present name for goat, on-gombo, was adopted. 64 A LANGUAGE-STUDY BASED ON BANTU. Hebbeo Nouns. Oeiginal Con- Abstract Sense. Pkesent Subsii- OEETE Sense. TDTE. on-ndu men humanity ovor-ndu, men. omi-tuka wings levity- otyi-vava, wing. on-dyoze spinning ani- mal (spider ?) phantom oty-auvi, spider. om-buze goat news (curio- sity) ou-gmibo, goat. on-dyoura elephant abundance an.-dyoit{ra), ele- phant. After this digression it may be convenient to briefly sum up the evidence for the primeval law of the plural. (i.) Dr. Bleek says ("Comparative Grammar," p. 145): — "The whole system of substituting a plural prefix for a singular one " (as oru-Di'o, knife, otu-fio, knives) " is certainly older than that of adding a particle indicating the plural to the form of a singular prefix " (as ou-ta, bow, oma-u-ifa, bows). Now if this be so — and no true Bantu scholar will contradict it — then we are forced to the conclusion, considering the perfect harmonious regularity we everywhere meet with in Bantu, that there must be some innate connection between the singular form and its plural substitute. (2.) The existence of such a connection is plain from the fact that, for example, a singular k in Herero, in whatever vowel-colour it may appear in the domain of the pronoun, invariably takes the consonants v (tu) and z (whose easy interchange is obvious in in- stances like Herero ovz- VIII. and Zulu izi- VIII., Herero o?j-(d)zm and Zulu im-wu, sheep) for its plural substitutes, whilst the singular r (Z) changes into the cognate dental t : — omu-ndu (ku-mu-ndu), man ova-ndu, men. u (ku), he, she (man) ve, they (men). otyi-puka (ki-puka), wild animal ovi-ptika, wild animals. tyi (ki), it vi, they (things), ke, it (diminutive) u (vu), they. i (ki), he, she, it (animal) ze, they (animals, &c.). oTu-vio, knife vtu-vio, knives. ru,it tu, they. THE HERERO PRONOUN. 65 (3.) Now it is especially the latter correspondence, r in ru XI. taking for its plural substitute the stronger dental t in tu XII., which throws light on the nature of the original connection between the singular and plural forms. It consists simply in this, that in indicating the plural, the consonant of the singular was pronounced with stronger emission of breath and greater force. Thus the singular ru {TU) became the plural tu (THU), and the singulars 1m, tyi (KI), u (KU) changed into plural forms by assuming such consonants as IcJi or Jchh, which in course of time degenerated to sounds like v, b, w, z. The easy transition from a guttural to a labial is obvious from examples like Herero -haxe (negligent, dis- orderly) and the identical Zulu -vazi (scattered, neglected) ; Bantu -kad (female) and Kafir (um)-fazi (wife, woman), also from English words ending in gJi, as trough, rough, enough, which are now pronounced trof, ruf, enuf. Here a word of caution may not be out of place. Care must be taken not to confound the prefix tu-, which in Angola and some other idioms corresponds as plural to ka- XIII., with the above- mentioned tu- (THU) XII., the original and legitimate plural of ru- {TU) XI. The Angola prefix otu-, as plural of oka-, is only a phonetic variation of the identical Herero prefix on (ovu-) XIV., both being derived from the primitive absolute form 9 KHU. Angola otu- stands to Herero o(v)u- in the same relation as Zulu izi- VIII. to Herero ovi- VIII., and ought to be marked, according to Bleek's classification of the prefixes, otu- {KHU) XIV., and not otu- {THU) XII. The Bantu pronoun ist. pers. plur. tu (orig. KHU = persons, or loe) is therefore identical with the Angola plural prefix otu- {THU=KHU = va = livmg things, answering to the singular ka- XIII., orig. living thing), whilst it is radically different from the homophonous Herero otu- {THU) XII., which corresponds to oru- {TU) XL, and means, as we stated before, rising, high, long dead things. Angola o(v)u- and otu-, like Herero ova- and ozo-w-, are variations of one common root, the dental element having been introduced in Angola otu-, in order to distinguish this prefix from the abstract o(v)u-. As a consequence, however, the real Herero and Kongo tu- XII. was lost in Angola. (4.) We have already drawn attention, in Chapter VI., to the analogy of the consonantal change in the verb, greater force being E 66 A LANGUAGE-STUDY BASED ON BANTU. expressed by a stronger letter, intransitive verbs being made transitive or causative by substituting an aspirate for a tenuis, though these aspirates were afterwards reduced to fricatives, — M, for example, to V and z, exactly as we find it in the pronominal roots. The deriva- tive (frequentative) verb rambuha, for instance, means to be thin, lean (r. ramba, drive away, pursue, chase), but rambuza (orig. ram- butJia = rambuklia) is to malce thin, to cause to be lean ; hutuka means unbound, but kutura ( = Icutuea = Jcututha = hutukha), to unbind. Now here we have the operation of the same law which formed from the Zulu singular isi- VII. {KI) the plural izi- VIII. (Herero ovi-, orig. KHI), or from the Herero singular on- (Zulu in.-, orig. KI-MJ) the plural ozo-n- (Zulu i-zi-n-, full form KHA-KI-MI). (5.) Additional collateral evidence we have in the numerical correspondence in the pronouns of our own languages, the guttural It Qi) in the pronouns of the first and third persons / {KI) and he {KU or KA) having changed to plural we (wi, orig. KHI) and they (orig. THA = KHA). It would thus seem that the restoration of the following numerical Herero correspondences to their primitive full form is warranted by sufficient proof : — Singular. Plural. omu-ndu (ku-mu-ndu) man. ova-ndu (kha-ndu), men. u (ku), he, she (man). ve, va (kha), they (men). on-gombe (ki-mi-kombe), ox or ozon - gombe (tha - ki - mi - kombe = cow. kha-ki-mi-kombe), cattle. i (ki), he, she, it (animal). ze, za.(tlia = kha), they (animals, (fee.) otyi-rongo (ki-rongo), a habit- ovi-rongo (khi-rongo), habitable able place. places, tyi (ki), it. Yi (khi), they. oka-na, a little thing. ou-wa (khu-na), little things, ke, ka, it. u, vm (khu), they. oru-vio (tu-vio), knife. otu-vio (thu-vio), knives, rw (tu), it. tei (thu), they. 6. Gender — Sexual Dual. — There are three different kinds of gen- der in the primitive noun or pronoun, viz. : — (a.) The common personal gender, denoting living beings, without reference to sex, also used for the masculine gender : THE HERERO PRONOUN. 67 ^0"= living one, man = he, person (common gender and mas- culine). (6.) A distinct femimne gender to distinguish' females or mates of beings and things appearing in pairs : MU= mother, woman, mate = she (feminine gender). (c.) The sexual dual, male and female united as one, couple, pair : ^J7-iWC7^= man-wife = they, the two in one, the human pair (sexual dual). It would appear that in the primitive language all breathing, living beings, whether male or female, were treated alike. They were spoken of indiscriminately as living beings or creatures. Viewed separately, a young man and a young woman, a young male and a young female animal, might be designated by the same name, meaning the living one, the living thing. It was only when the living beings or creatures, in mature age, appeared in pairs that the grammatical distinction of gender was resorted to, and the forms of the sesual dual were used, denoting male and female united. Now, if one of the united pair had to be named separately, the form of the common gender {KU, KI, KA) was used for desig- nating the male, whilst the distinct feminine form (MU, MI, MA) was applied to the female. And as to inanimate things, primeval man, in viewing and naming them, did not ask, " Is the thing in any way like a male — big, strong, hard, active; or like a female — smaller, weaker, soft, passive 1 " — questions belonging to a later period ; but, " Is the object like a living thing (with life, blood, as arm, leg, ear) ? " or, " Is it waving like wings or branches moved by the wind ? " or, "Is the thing motionless, dead (like bones, horns, stones)?" The single horn, for example, of a one-horned rhinoceros would not suggest to him the idea of gender ; he would simply call the isolated horn a dead one or a dead thing; but observing on many other animals two horns standing together, the second one would appear to him as the female or mate of the first, and thus comparing them to a married couple, he would accordingly classify the pair of horns, though lifeless objects, like pairs of living beings, in one of the classes of the sexual dual. 68 A LANGUAGE-STUDY BASED ON BANTU. ' Now when, in the long lapse of ages, the original meaning of the sexual dual, "the being (or thing) and its mate," or "male and female," was lost, the idea " male or female " was substituted, and so it happened that the dual form was applied to males or females indiscriminately, in the same way as also our man, mensch, denotes man in general, "male and female," but also "male o?' female." We have in Bantu the germ of what is more extensively developed in the so-called sex-denoting languages. Evidence for the Sexual Dual. — (Cf. §§ 10-25 o^ *'^^ Introduction to my English-Herero Dictionary.) (i.) The existence, in Bantu, of double formative prefixes, pro- perly compounded primitive nouns (something like the Chinese fu mu = father-mother = parents), which originally must have denoted a double object, a couple. (2.) The duaHstic tendency of the compound prefix oma-, KA-MA, first observed in Herero by the Kev. J. Eath : oma- being applied as plural for most things which appear m pairs, as ome-Jio, eyes (properly the pair of eyes, male and female), oma-ke (the two) hands, oma-oJco (the two) arms, oma-rama (the two) legs, &c. (3.) The radical identity of the two prefixes for those classes of nouns in which natural gender is observed, omu-, KU-MU, I. (man), on, ora.-, KI-MI, IX. (animal), with the dualistic prefix oma-, KA-MA, VL, the radical identity of these three forms being placed beyond doubt by the corresponding demonstrative pronouns : — Pbefix. Noun. Pees. Pkon. Dkmoksts. Peon. oma- {KA-MA). oma.-oTco, arms (orig. e {KA). in-ga, {INI-KA). sexual dual, male and female, [i.e., right and left arm), on- (KI-Ml). on-gombe, ox or cow i (KT). zw-dyi (INI-KI). (orig. sexual dual, ox and cow). omu- {KU-MU). oran-ndu, man, per- u {KU). in-sai{INI-KU). son, male or female ' (orig. sexual dual, male and female). THE HERERO PRONOUN. 69 (4.) The identity of the Bantu primitive nouns EI, father, and NI, mother, with the compound formative prefix on- or m- (KI-MT) :— in- (IX. prefix) = in MI HI, father, male. NI, mother, female. (5.) The peculiar character of the Herero noun oiaan-tena = brother of a sister, or sister of a brother, which originally must have denoted brother and sister, " geschwister-paar." (6.) The fact that omu- III., though at present a singular prefix, is used for representing the two finger-rows or the number ten. OmM-rongo means ten in Herero. The word is compounded of the singular prefix omu- III. and the adjectival stem -rongo, from ronga (in rongera), to make straight, get ready, prepare, equip. The prefix omu- evidently refers here to the same object as omu-nue, namely, finger. The proper meaning of omu-rongo is, therefore, the ready or skilled omu- or finger. I think there can be no doubt that omu-rongo is one of the original Bantu names for finger. But would the word have been employed for representing the number ten if originally it denoted only one finger ? N'ot even if its mean- ing had been one finger-9'ow, for that would have only been five. The fact is, the full form and original meaning of omu-rongo is PU-MU-BONGO = th.e ready, skilled, wing-like (or branch-like) pair, that is, the tivo rows of fingers, that is, ten. (7.) It is also worthy of note that in Suto (and probably other dialects) the right hand or arm is called the male (letsogo le letona) and the left the female hand (letsogo le letshegali), which seems to indicate that, although the sexual dual, as such, is extinct in Bantu, yet the primitive intuition underlying it is not quite obliterated from the mind of the people. 7. Some of the primitive pronominal forms have, through the wear and tear of ages, been reduced to single vowels. Of the primi- tive dual form KA-MA, for instance, only the first vowel is left in the Herero personal pronoun a {ka,-ma), and even that is changed 70 A LANGUAGE-STUDY BASED ON BANTU. to e, though in the corresponding demonstrative pronoun nga, the first part, Jca (ka-ma), has been preserved. Some of the pronouns of different classes or genders of nouns have coalesced ; others have shifted from one class to another, but their wanderings can still be traced. Also, by tampering with number and case, the grand primi- tive system has suffered : dual and plural forms have come into use as singulars, and genitive forms have taken their place in the nominative. Mhi (vi), for example, at present = /, is properly, as we saw already, the parallel t'-form of mha, (va) = these, those (people), and meant originally we ; and ami = of me, mine, me, stands now in Herero in the nominative for I. For a complete statement of the Herero pronominal forms (pre- fixes and pronouns), the student is referred to my English- Herero Dictionary (Cape Town and London), and to Table III. at the end of Chapter X. ( 71 ) CHAPTER IX. THE OPERATION OF COMMON LAWS TRACEABLE IN THE BANTU AND ARYAN PRONOUNS. It is, I believe, generally admitted that the pronouns belong to the most ancient forms in language, and that, for this reason, their etymology is enveloped in much darkness and doubt. It would seem there is not a single personal pronoun in the Aryan and Semitic families whose etymology has been established as perfectly certain. All the labour bestowed and all the learning brought to bear upon this subject appear to have resulted in nothing but the unanimous confession of those best able to judge that "the etymology of the Aryan personal pronouns is doubtful, — that they are words which for the present must remain without a genealogy." And so they would have to remain for ever, were it not that a family of languages has been discovered in Africa which has pre- served about twice, if not three times, as many genuine pronominal roots as are found either in the Semitic or Aryan languages. The time is probably not far distant when it will be accepted as a fact that the pronouns of the latter and other families are based on the very same principles as the pronominal forms in Bantu. The changes and shiftings of sound are, of course, considerable, but as to the first and second person, we have the advantage of knowing that, in all families of speech, they can only be derived from those absolute forms which mean, in the primitive language, the erect moving living one, man {KU, plur. KHU), and the erect moving mother or female {MU, plur. MRU). A dental or lingual in Aryan, Semitic, and other families, may be original in the third person, but if found in the first and second, as, e.g., Hottentot ta, tita, I, Lat. tu, Germ, du {thou), we know that the dentals t, d, th have been substituted for the original guttural k, plural Ich. Or 72 A LANGUAGE-STVDY BASED ON BANTU. when we find that in Hebrew the separate pronoun" of the 2nd. pars. sing, is ATAH = thou, we know that this is only a variation of the more primitive form retained for the accusative (or verbal suffix) KA = thee, radically identical with or closely allied to hua = he, Hu = him. The identity of Herero ?zdyi (wgi), i, KI ( = man present, ist. pers. sing.) and English / (Goth. iJc) ; of Makonde we^a, KRI ( = men present, ist. pers. plur.) and English we (Swed. vi) ; of Herero eye, ye, KA, KU (3rd. pers. sing.) and English he, and other forms, can hardly be doubted. But let us proceed methodically, in accordance with the principles set forth in the preceding chapters, and tentatively trace the iden- tity of some of the most important forms of the pronoun in Bantu and Aryan. Prind_ples of Comparison. — It wUl be expedient, at the outset, to briefly premise the points which in any attempt to identify pronouns of different families must form the basis of research. 1. No pronoun stands isolated in any language j it is in all cases a member of a group, and must be treated as such. 2. A personal pronoun is properly a primitive noun, meaning man, person, &c., or, in feminine forms, mother, loovian, female; one and the same absolute form may stand, therefore, for the second and third, or even for the first person. In the primitive language, the sentence " man goes " signified both " he goes " and " thou goest." That, for example, the Hebrew -nu means our, and the Herero -nu signifies your (plur. ), is no reason why the two pronouns should not be identical. 3. There are, however, conditional forms for the first and third person, so called because they have by-meanings referring to space and locality, the form for the first person assuming, as a rule, the vowel ?' = here, present, whilst in the third person we find the absolute w-f orm changed to a (e) = there, at some distance, abroad. 4. All pronouns representing man have originally the guttural Jc (common gender and masculine), plural M ; and for the feminine gender the labial m, plur. mh. To these primitive letters the various consonants of the first and second person in universal speech, how- ever adventurous their career may have been, must be traced. The third person, including as it does inanimate things, has also the other primitive consonants t, plur. th, and^, plur. ph. COMMON LAWS IN THE PRONOUNS. 73 " 5. Originally the feminine gender was distinguished, not only in the third and second, but also in the first person, and sometimes the feminine form, if lost in the nominative, reappears in other cases. Sanskrit 7nd (-mi) and aJiam, for instance, are not variations of one common root, as som^ hold, but two distinct pronouns, the first of the originally feminine and the latter of the common personal (and masculine) gender. Bopp points out clearly that the -cmi in ahatn is merely an ending and inorganic. " Das am von ahmn," he says, " ist endung, wie in tvam, du, ayam, dieser, und svayam, selbst, und wie im plural vayam, wir, yuyam, ihr. Der nominativ I. pers. sing, ist von anderm stamme als die obliquen casus." In Herero we have an analogous case : besides me {ina-i = ma-Kl), I, we find also the form ami I, me, in which the labial is radical ; in me it is inorganic. M& is common personal and masculine, but ami, though at present also common gender, was primarily feminine. Now as Herero me {maA, ma-KI) corresponds to Skr. a-h-am, Goth. i-ls. ; so the Skr. accusative rat, mtm (ma-am), and the nominative verbal suffix -mi, I, correspond to^Herero a-mi, I, me : — KU = man, person (absolute form). Common personal and masculine, at present common, gender. Bantu. Aryan. ki(Tshuana),«dyi,i, me (Herero) «k (Goth, and Dutch), jch = man here, person present = (Germ.), ego (Lat. and Gr.), ];_ a-h-awi for agam (Skr.), I. Feminine, at present common gender. a-mi (Bantu), I, me: MI sig- -mi (Skr.), as in as-mi, Engl, nifying originally mother, a-m, I am ; md, mdm, ma-am woman, or female here = I, (Skr.), me, mi-^ (Goth, and me. Icel.), mi (Low Germ.), me. 6. The primeval law of the pronominal plural is the same as that of the primitive nouns (or Bantu prefixes). It is, as we saw before, of an extremely simple nature : the consonant of the singular is aspirated and strengthened, the singular k, for example, becom- 74 A LANGUAGE-STUDY BASED ON BANTU. ing plural Jih or kJiJi, the latter originally strong plural consonant appearing in its present collapsed state in the several families as w, V, b, th, z, &c. 7. Sometimes the plural is substituted for the singular, as an original loe for I. 8. The question of case is of no moment in investigations into the nature and identity of pronouns. Originally any pronoun, in virtue of its character as a primitive noun, could stand in any case. The present fixed use of certain forms for certain cases is purely con- ventional. 9. In some pronouns, as we know them now, two or more forms have coalesced. I o. I would also draw attention to what I consider an interesting discovery in Bleek's " Comparative Grammar " (pp. 150, 151), viz., that the vowel (or article) which precedes a Bantu formative prefix, as in Zulu wlu-, «li-, is in its origin a 'pronoun and identical with the prefix wMcIi it precedes. In other words, the primitive article in Bantu is formed from the prefix of the noun (or pronominal root) which it precedes, by suppressing the consonant and retaining only the radical vowel, as li-lu- XI. from lu-lu-, aba- II. from 6a-ba-, &c. ; and, we further add, if the prefix was a compound, as M-mu- I. {KU-MU), «-mu- III. (PU-MU), the second syllable, -mu-, -ma-, &c., was ehded in the article. Thus not kumu-KTJMTJ- 1., but ku-KUMU (afterwards «i-umu-, «-mu- I.); not pumu-PVMU-, hut pu-PUMU- (at present w-umu-, M-mu- III.). Now it appears to me that there are some traces left in Bantu and in the Aryan languages to show that the primitive article, dis- covered by Bleek in the Bantu noun, also may be found to precede and emphasise a pronoun in all genders, numbers, and persons. Bleek has shown that Zulu i-si- VII. = a thing or the thing, was originally si-si = thing-nnmQ (this or the thing), that u-{n)iw- l. = a MAN or the MAN, was in the ancient language ma?i-MAN (this or the man). Now, it would seem possible, or even probable, that, for example, the vowel i which precedes the radical k in Goth. iK (I), and the prefixed u in Goth. u-QK-is (us two), may have been originally identical respectively with k(i), I, and gk(u), us, so that the full form and meaning of Goth. IK and UGK- may have been — COMMON LAWS IN THE PRONOUNS. 75 ki-KI = man liere-MAN HERE = i-KI = ik = this man here (I). khu-KHU = men-MEN = u-KHU = mgk- = these men (a number, or only two, we, us). Another illustration. The Kongo prefix eye- (in eye-kala Y., human being, plur. a-7cala II.) and the Herero separate pronoun eye, he, she, properly this or that person (with which the Skr. a-y-am, this one, may be compared) are radically identical. Now the initial e- (Skr. a-) is the primitive or (as we may call it in honour of its discoverer) Bleeh's article. Its force is " this, that, the : "—ye = he or person there : e-ye, {ij)e- ye = (A:)e-ke (primarily lio-KA) = that (this or the) person there, the-h.e or the-&he ; — eye-Jcala (Kongo), human being, lit. that (this or the) human being ; eye (Herero), he, she, properly the he, or that person there ; a-y-am (Sanskrit), this one. It would therefore appear that there was a period in language when there were as many articles (or demonstrative pronouns) as there are primitive nouns (or personal pronouns), namely, forty- three (see Table II.). But it is self-evident that this state of things could not last : in course of time the primitive articles lost their power, and became part of the primitive noun or pronoun, as, for example, Zulu mIu- (for lu lu), Goth. 2^(1), I (for hi ki) ; and only one or two of the large number survived, which were now generally applied in all cases. These forms, as, e.g., Bantu a-, the Semitic ha, hal, al, and the Aryan ta, sa, the, we shall call secondary articles. Very frequently the primitive and the secondary article have blended in Bantu into one sound, as Herero oru- (for a-u-iu.-), otu- (for a-M-tu), Kongo e-ki- (for a-i-ki), eri- (for a-«-ri), &c. II. But there are other particles besides the primitive article which stick to the pronoun as limpets to a rock, as, for example, the prefixed in-, {i)n-, in Herero (»)?jdyi, 1, ingvd, this one, Hebrew an-, en-, in andk.i, I, ewhu, him, and the Sansk. ending -am in abam, I, ytyam, you. In Bantu these particles fall into three classes : (a.) demonstrative or emphasising particles, as in- in Herero {i)nd'yi, I, properly this self-I, I myself, ingvd, this one, this self-same one ; (&.) case-particles, as a- in ami, I, of I, of me, mine, hence also me and I ; (c) tense-indicatiag particles, as ma- in main, we (present and 1(> A LANGUAGE-STUDY BASED ON BANTU. future tense) = we on the spot, we here present; tua, we (past tense) = we there, at a distance, in the past. Some of the latter kind can still be separated from the root, as ma- from matu, a- from atu, ave, but as to the emphasising particles, they have, as in other families, grown together with the pronoun into one word, as, e.g., Konde mipa (mi-jja), I, properly I (near or) here. Affinities between the Aryan and Bantu Personal Pronoun. — Let us now briefly glance at the personal pronouns I (Goth, and Dutch ik, Skr. aham), plural we (0. Sax. wi, Swed. vi, Skr. va,ya7n) ; thou (Lat. and Pers. tu. Germ, du, Skr. tvain, tVLam), plur. you (Dutch u, Skr. ydyam); he (Sax. he, 0. Engl, ha, a), plur. they (O. Sax. tht, Goth, thdi), and compare them with corresponding forms in Bantu. We begin with the singular pronoun of the second person, thou (Lat. tu). Here the primitive vowel u has been preserved, but the dental consonant is a changed 7c, the original form being the Bantu KU= the living, erect moving one = man = thou. Thou, therefore, is pro- perly (thou) man. The original It we have in the first person ik (I), and approximately {h being so near akin to A-), in the third person he. Tu {thou), like Bantu ku (thee), by aphaeresis U (thou), is the absolute form of the common personal pronoun (singular), simply meaning man, without reference to person or place. Thus "thou sayest" is properly "man says," just as the Herero "u tya" means both "he says" and "thou sayest," because its literal sense is "man says." Now, from this form iu = ku is derived, by aspirating and strengthening the consonant, the plural you {guw, yil-yam), origin- ally KHU or KHHU, which in Bantu has assumed the forms tu, tyu, tshu (ist. pers. plur.) and vu, u (3rd. pers. sing, and plur., orig.) = ^iZC/"=men (absolute form) : — SiNGULAK. PlUEAL. thou (Lat. tu) = you (0. Engl, guw) = KU= man KHU = men (absolute). (absolute). COMMON LAWS IN THE PRONOUNS. yy Bantu. ku ; u : thee ; thou, tu (tyu, tshu) = he, properly KHU= (we) men KU= man (absolute), (absolute). We further observe that from the above two absolute forms KU and KHU proceed by change of vowel — (a.) the first person I, Goth. «k, Zulu ngi, Herero mdyi, i (in me = ma-i), Kafir Kdi, Tshuana ki = KI= man here, man present = I — From this the plural, in accordance with the law stated before : we, 0. Sax. and Low Germ, wi, Swed. vi, originally KHI, the corresponding Bantu forms being Konde wepa, Herero mbi ( = m-vi, I, but properly we), Zulu ti, tsi, si, Herero ete (= THI=KHI) = we. (6.) The third person, or person there, at some distance, abroad, absent, he, Bantu eye, e, ka, a, primitive form KA — From which is derived the plural they, O. Sax. tht, Goth. %lai,i = KHA = (they) men there, abroad, absent = Bantu va (ve) ba (they II.), and collaterally za, ze (they X. Herero), as the subjoined comparative table shows : — 78 o -p M s ^ 1 H J2 a s iz; q M c« M g a « '5" o o Ph rfl M P iz; o o O 3 M A LANGUAGE-STUDY BASED ON BANTU. I 5^ cS ^ ^^ o So fH a a s _r g =2 ,J4.^r3 ■a tuo 3 ,2 § ^-^ -J3 a CO ;h ^ o cS e^-H M 0) o ^ a '^ - -■ — ^ "w t- ^ o Si. £3 .2 ■■ II w- !^ -*j II I*' be o O c8 ffl o t^ .S ^^ O a a S ^ ^. « 1^ -3 "§ ^ ^ I — I l> •J3 O o m tM -2 g d -S o m rt ,r! Ph nS- ij Pj <« cS o a a II p^ b w t*^ m II 13 I H E3 O rd r^ a !^ 8 !,§ h- 1 cS O r« a a o II ^ O tq ff> .2 4i O o II ^s a COMMON LAWS IN THE PRONOUNS. 79 Review of the English Personal Pronoun. I (KI). Fuller form Goth, and Dutch iJc, German ich, Lat. and Gr. ego. As no primitive root can end with a consonant, the original form of I must have been iKI, analogous to Hebrew ara-o-ki, (Bantu) Tshuana ki, Zulu ngi, Herero «dyi, Xosa wdi, in which latter form the guttural is changed to a dental, as also in Nama (Hottentot) tita. The radical vowel i, which has become mute in «'k', Jk(i), has been preserved in the corresponding plural we, Dan. and Swed. wi or vi, originally KHI. The ^initial vowel i in Goth. «k we take to be the primitive or Bleek's article. The full form and meaning of ik, I, is therefore probably ldKI= iki = this man here (or self), 7a KI being the conditional i form of ']KU=t\iB grown-up, erect moving one, man (absolute). The predilection of Sanskrit for the a-sound has probably been the cause of changing the primitive i, so essential in the first person, into a. Aham (a-h-am) appears to be a combination of the following three parts : — (a.) The radical h.', a remnant of 7a KI=iia&n here;- (&.) Bleek's article a- (for an original i) = this or the (man here); and (c.) The ending -am. ME (MI). Goth, and Icel. mik, 0. Germ, mih, Skr. mam, ma, objective case of I, myself. The ending -ik, -k, -ch, -h in the Teutonic lan- guages, and -am (yam, sdm) in Skr., seem to add to the pronoun the force of "here," "same," or "self." Me is radically different from /. Originally it represented the first pers. sing, feminine. It is identical with Bantu ami (m,' n'), I or me (properly mine), the prefixed a being demonstrative and the sign of the genitive. In Herero, ami stands at present both in the nominative (just as the English me is also sometimes used for I) and in the objective case : ami me i, I-I-(shall) go, and hungira ku ami, speak to me. In the primitive language 32a ilfJwas the conditional i-form of 32 MU, mother, woman, and signified mother here, or woman here, I or me. 8o A LANGUAGE-STUDY BASED ON BANTU. In Sanskrit the primitive form has been preserved in the verbal suffix -mi = I : os-mi (Lith. es-mi, Goth, i-m, Engl, a-m) = I am. In raim (msb-am, abbr. m&), as in aham (a-iL-am), the original ^ is lost. The Hebrew ani (an-ni), 1, is also probably identical with me (MI). WE (KRI) is the natural corresponding plural of // primitive form ga KHI. In we, Swed. wi or vi, 0. Sax. wi, Germ, loi-r, the original radical vowel has been preserved. We is derived from J by a stronger or aspirated pronunciation, just as the Herero plural ovi- (6 KHI) has been formed from the singular otyi- (4 KI). Bopp indeed says, " Der plural der ersten person ist vom singular stammhaft ver- schieden, well das Ich eigentlich keines plurals fahig ist. Denn es gibt nur Ein Ich." Quite true, if it could be proved that / (Skr. aJiam) had, from the beginning, the abstract meaning which it has now. But I believe philologists are at present pretty well agreed that in no case language commenced with abstracts. Besides, the meaning of the first person singular can, in Bantu, methodically be traced to have been "man here, person here," or, in the original feminine gender, "mother here, woman here." This removes the difficulty at once. If / means " man here," there is no reason why there should not be a corresponding plural, signifying " men here." And this corresponding form has been preserved in the shape of ?oa, wi, vi (for the primitive KHI). It is interesting to observe that in Konde (Bantu) the word has assumed exactly the same form as in English, namely, we-jia = we-near = we-here, present, the demon- strative Bantu particle pa denoting nearness, hence also presence. And in Herero we meet with the Danish form vi (we) in mhi (im-vi), originally the plural 2ee, but at present in use for the singular I : inbi tare (orig. that we look, but at present) that I look, just as tu tare, that we look, is also used for that I look, or let me look ; tib pa 0, give me, please ; properly, give us, please. The parallel form in Skr. is vayam (ve + am). "Vayam, in accord- ance with the law of the primitive plural, is properly a strengthened form of aham. But it is more especially the dual (radically identical with the plural) which corresponds to aham in every particular, inasmuch as both the singular aham and the dual dvdm appear to COMMON LAWS IN THE PRONOUNS. 8i have the primitive or Bleek's article prefixed to them, which is not the case in the plural vayam : — aham (i-hi-am) = this man here, self, I ; dvdm (i-vi-am) = these men here (them)selves, we (two) ourselves ; vayam (yi-am) = men here (them)selves, we. US {KHV). Though the subjective ioe and the objective us are at present totally different in sound, neither consonant nor vowel betraying the remotest relationship, yet they appear to be radically one, us being the absolute (9 KHU), and we (ga KHI) the conditional form of the same root. That the primitive root of us is 9 KHU appears to be plain from the Goth, ^lg^^^s {uglc{u)is), us two, igqvis, you two. The dual form of the second person is evidently only a modification of that of the first person. From the two forms we can easily reconstruct the primitive form. Taking from the first person ugh- and adding the « or m of the second, we have u-gku, primitive form khu-KHU {g) = these MJEN=we or us, the initial u {&hu) being the primitive or Bleek's article. The radical s in Goth. unsis, us, is only a weak trace of the stronger consonant gk or gq {Till or Mil). The nasal in unsis, Germ, uns, seems to be inorganic, and the ending -is probably means, like -is in veis, (we) " selves." The difference thus between the plural pronoun of the first person us {Jchu-KHIJ) and that of the second person you {KHU) seems to consist simply in this, that the first person is emphasised by the primitive article, which is wanting in the second : — Ichu-KHU = M-gku = M-s(u) = us = these men (absolute) = us ; KHU= you = men (absolute) = you. Us and you, therefore, stand in a similar relation to each other as Tshuana ro-na or tsho-?ia ( = THU= KHU), we, us, and lo-na { = THU=KHL), you. The ending -na is demonstrative. The Herero form for us is tu { = THU=KHU), which, in some instances, is also, as in Tshuana, pronounced tyu or tsliu, e.g., tu-ende (irregular imperative), go, which is generally pronounced tiju-ende or tshu-ende, literally (that) we go, or (let) us go. F 82 A LANGUAGE-STUDY BASED ON BANTU. THOU (KU). Thou, Lat. tu, Gr. ad, Germ, chi, is a variation of the primitive form 7 ^f7 (Bantu m = thou, /cm = thee), and means simply man, person (absolute). The labial v (iv) in Skr. ^t^am is properly u : tvam = tu-am, thou. The objective ^7iee seems to be an abbreviation of A.-Sax. fhec, 0. Sax. tliic, Goth, tlwk, probably a contraction of Viy\i-ih. The original full form and meaning of thee appears to be KU-iki = TU-iki = thou-self, thyself. The natural plural of thou is YOU {KBV), which literally means men, persons (absolute). Skr. y6-7/am {KHU + am). The vowel e {i) in the nominative ye (0. Dutch ghi) signifies " here, present : " — KHU = men (absolute) = (you) men = you ; KHI= men here, (ye) men here = ye ; KEI being applied in the first and also in the second person, so that ye and loe appear to be only variations of the same primitive root (96 KHI). In the Skr. dual yuvdm {u-v{-a)-am, u-KHU-am) the initial u (yu) appears to be the primitive or Bleek's article, the radical u being absorbed in d : — yuvd7n (jou ty^o) = khu-KHU{-am) = uKHU {-am) = these MEN, these very (two) men, i.e. you two. HE {XA), and its modified form she, appear to be identical with Bantu a (ka), e, ye (he), eye (cf. Skr. a-y-am, this one), he and she— all variations of the primitive conditional form 76 KA =ma,n there = he or she. The neuter IT (AT), Goth, ita, 0. Germ, -iz, N. H. Germ, es, Dutch hef, Skr. it, is perhaps identical with the Bantu neuter form tyi (Herero), si (isi-) in Kafir, M (Kongo), se (Tshuana), ez- (Mpongwe). Primitive form 4 KI=it, the living one, animal, but also generally it, the thing, place, &c., the initial i being probably (as in the Zulu prefix isi-) the primitive or Bleek's article. COMMON LAWS IN THE PRONOUNS. 83 THEY (KHA), A.-Sax. th^, Goth. tha«, seems to be the natural plural of he (KA), namely, the conditional form 96 KHA = men there, though it is possible that the absolute form 3 KHA (living ones) may have coalesced with it. It may also be that in 96 KHA a conditional a-form of 6KHI ( = THI) is included. CHAPTER X. PRONOMINAL TABLES— THE PRIMITIVE PRONOMINAL SYSTEM RESTORED. Such is the wealth of pronominal forms in Bantu, especially in Herero, that an attempt to restore the primitive pronominal system, comprising the formative prefixes and suiSxes of the noun and the pronouns of universal language, ought not to be regarded as hopeless. I have therefore ventured to draw its outlines in the Introduction to my English-Herero Dictionary, and at the end of this chapter similar tables will be found, more complete in so far as they contain the conditional as well as the absolute forms whose characteristics have been explained in Chapter VII. In the restored pronominal system of Bantu — wMcli I hold to he the primitive p-onomindl system of universal language — there are thirty-three absolute and at least ten conditional forms, thus altogether forty- three. (See the appended tables.) After what has been said in the preceding chapters on the nature, laws, and original meaning of the pronominal roots. Tables I. and II. will, on the whole and in their outlines, be clear and explain themselves. The forms in thick type have in some shape or other been preserved in Bantu; either as formative prefixes of the noun or as pronouns, in most cases as both ; those printed in ITALIC CAPITALS are hypothetical. Table III. — The table of the Herero prefixes and pronouns shows that of the forty-three original pronominal forms nearly thirty can still be traced in Herero. There are two or three forms about the identification of which I am not quite sure. Omi- IV. may be identical with 23 PI- MI, but as omi- only occurs as correspond- ing plural of omu- III., it may possibly be a phonetic variation PRONOMINAL TABLES. gj (" umlaut ") of tlie latter form. Also the Nano plural ovi- (in ovi-ta, bows, plur. of n-ta, bow) may not be identical with 24 PHI, but likewise an "umlaut" of u-, on- (ovu-) XIV. (25 PU). Nor is it quite certain that the original form 27 PHU has coalesced with ou- (uhu) XIY. This uncertainty, however, does not interfere with the fact that these threeforms, viz., 23 PI-MI, 24 PHI, and 27 PHU, actually existed in the primitive language. The prefix o- I. seems to be a blending of the primitive noun or pronoun VL (y KU = man, person) and the secondary article a : a-u (a-KV) = = thg person, the he (she). Originally 0- I. was prob- ably the singular of the sexual dual form omu- (KU-MU) I., hence we find it prefixed as a kind of article to proper names and to the names for father (o-iate, o-ihe) and mother (p-mama, o-ina), who, considered separately, could, of course, in the beginning of the language, not have been represented by a dual form. The plural 00- II. which corresponds to the singular 0- I. is probably a con- tracted demonstrative form of the XIV. prefix ou- (9 KHTJ = men, persons), identical with the corresponding separate pronoun ouo XIV. = they (orig. men, persons, absolute). Go- II. (9 KHU) would therefore appear to be the original and legitimate plural of o- I. (7 KU). Of. the form 00- XIV. (9 KHV) in Kongo (Bleek's Comp. Grammar, p. 224). U- 1, is evidently the personal (absolute) pronoun u ( = person, he, she) in the genitive case, the sign of the genitive (a) being afiixed to it. Thus, e.g., the literal meaning of VL-a-mbangu (stranger, alien) is he-of-the separation (om-6fflwg'M = difference, separation), or the separate one. Ov- II. (plur. of u- I.) is a demonstrative form of the genitive pronoun v' (ve, va, separate form owo, ovo) II. : ov-a- 7niangu = they (or those)-of-the separation, i.e., strangers. It will be noticed that in the Herero objective pronoun 2nd. pers. sing, ku, the radical Jc, lost in the subjective u, has been preserved. This k in ku must not be mistaken for a remnant of the preposition k(u). Me ku sutu, for example, is not me k'u sutu (I-to-thee-pay), but, in analogy with all the other objective pronouns, none of which has a preposition, me ku sutu, I-thee-pay. So also in the Zulu Orke (of him), his, her, the k appears to me to be radical ; thus not a-k-e (of-of-him), but a-ke (of him). Bantu KU signified origin- ally both thee and him, and MU thee (feminine) and her, but when 86 A LANGUAGE-STUDY BASED ON BANTU. the original meaning (KU =ma,n, person, ^^^7"= mother, woman) became lost, KU and MU were employed, without distinction of gender, respectively for thee and Mm, her ■ me hu sutu, orig. I-man- pay, at present, I-thee-pay (male or female) ; me mu sutu, orig. I- woman-pay, at present, I -her or him-pay. The objective pronoun of the ist. pers. sing, m or n (see " Me " in my English- Herero Dictionary) appears to be a mutilated form of (a)mi (32a MI), I. Table IV. will require a more detailed explanation. The Hottentot (Nama) Pronominal Forms Review'ed. — In comparing the pronominal roots in Bantu and Hottentot, special caution is needed against the danger of being misled by mere similarity in form and sound. Superiicially viewed, we should be tempted to give to most of the Hottentot forms a different place from what they occupy on Table IV. Sa-rum, we two, e.g., seems to be nearer 17 TU-MU than to 8 KU-MU. But we learn in Bantu that in the first and second person only the gutturals Jc and Jch, and the labials m and mh are possible, as only these can represent living beings. Thus if the modern pronouns of the first and second person appear with such consonants as t, d, r, v, b, or n, we know that these sounds are not original, but modifications of 7c, Teh, or in, mh. In the third person there is the possibility of t or p being the original consonant, though it will probably be found that in the Hottentot, Semitic, and Aryan families all the primitive forms (on Table IV.) from 10 TA till 27 PHU ure lost, at any rate as far as the personal pronoun is concerned. A few of them may indeed still be recognised as pronominal particles, prepositions or adverbs, as, for example, English by, which in Chapter V. we identified with 19 PA, but as true pronouns they seem to have altogether disappeared, except in Bantu, where of the said seventeen primitive forms about half the number has been preserved. Surveying, in the light the study of Bantu affords, the whole of the Hottentot pronominal domain (suflixes of the noun and pronouns), we observe that the primitive compound form 8 KU-MU (= Nama sa-khum, we two) has been admirably preserved here, perhaps better than in any other language. For in the Arabic hum, ^thiop. humu (they), originally the same as sa-khum (and sa-kum, we), the guttural has been changed to a spirant. But most of the Hottentot PRONOMINAL TABLES. 87 pronouns are terribly mutilated, so much so that they appear to be the very opposite of those Bantu forms with which we venture to compare them. But it is just this absence of similarity of sound which strengthens our position. The fact is, some of the present Bantu and Hottentot prefixes, suffixes, and corresponding pronouns represent only AaZ/ the original compound form. Now, whilst in the prefix-pronominal Bantu family the second half, as a rule, has been preserved, we find in the sufiix-pronominal Hottentot the first part retained. In other words, in prefix-pronominal and in sufiix-pro- nominal languages the pronouns are identical with the prefixes and suffixes of the noun. Now, it is a rule that a compound pronominal root in its capacity as prefix or sufiis is reduced to a monosyllable. Here lies the secret : the dissyllabic prefix naturally loses the first, and the suflSx the second syllable. Thus 2 KA-MA, as prefix in Bantu, will, for brevity's sake, throw off the first member, and assume the form of {KA^ilLA. , whilst the same primitive form, as suffix in Hottentot, will drop the second syllable, and survive in the shape of KA(lfA). Now let the Hottentot sufiix ka and the Bantu prefix ma be joined together, and we have the full primi- tive form 2 KA-MA restored. I quote from the Introduction to my English-Herero Dictionary, p. xiv. : — " If we compare the Khoi- khoi (Hottentot) nominal sufiixes with the Bantu prefixes, we observe that in the former the first part of the full form exists whilst the second part is dropped (-kha or -ka instead of KA-MA), and that in the latter (Bantu) the second part of the full form has been retained, whilst the first syllable has been elided (ma- instead of KA-MA ; mu instead of KU-MU), as, for example : — Herero — ome-7io [KA-MA-iho, the pair of) eyes ; Hottentot — wiM-ka or mu-kha {mu-KA-MA), masc. dual, two eyes (from mu, to see) ; and we observe further that, as in Bantu, so also in Hottentot, ilnQ first syllable of the originally compound form reappears in the corresponding demonstrative pronoun : NOUN. Bantu — (ka-)ma.-iho, eyes ; Hottentot — mM-ka(-ma), two eyes. A LANGUAGE-STUDY BASED ON BANTU. DEMONSTRATIVE PRONOUN. Bantu — in(i)ga, . . these (orig. two) ; Hottentot — {ne)ka, . . (or ?je-klia), these two. For brevity's sake, KA-MA, as prefix in Bantu, naturally dropped its first, and, as suffix in Hottentot, its second syllable. It is as if in one museum the head and front part of some curious animal were found, and in another the back part and tail. The head is, of course, very unlike the tail, but if both parts are brought together, the whole animal is restored. So the Hottentot dual suffix -ka or -kha, just now quoted, is, viewed in itself, very different from the Bantu prefix ma- (dual and plural). And yet both are originally one. But in their present state they are imperfect, Hottentot -ka being the first, and Bantu ma- the second member of the compound primitive form. Let them be joined together, and their oneness as 2 KA-MA is clearly seen. We observe further that the feminine forms in Hottentot have been derived from the masculine or common gender by modifying either the consonant or the vowel, or both. Sd-Tshnm, we two (masc), for example, is changed to sa-im {sa-KIM), we two (fem.), on the same principle as the Hebrew feminine form Ma, she, is derived from hua, he, and the Hausa feminine ke, ki (thou), from kai, ka (thou, masc.) ; and the radical s in the Hottentot feminine plural sa-so, you, is only a modification of h, as found in the corresponding masculine plural sa-ko (you), analogous to the consonantal change from m to n in forming the Hebrew feminine aten, ye, hen, they, from the corresponding masculine forms «tem and hem. Other important points to be borne in mind here, as, for instance, the shifting of person, number, and case, the easy transition from guttural to dental consonants, &c., we have noticed already in Chapters VIII. and IX. It is especially the vowel i or e which affects the guttural, and changes it to ty, tsJi, f, s, &c. Thus we have in Herero the pronominal roots 7m, ku, and (not /,■/, but) tyi, hi, just as in Italian c (/t) has become ch (tsh) before i and e. Xei- ku or iKsi-ka is in Nama they (masc). The feminine is derived from this, like Hebrew hia from hua, by substituting the vowel i, and the result is that the h is changed to d : xei-Ai (for xei-KI), they PRONOMINAL TABLES. 8g (fern.); «oi-kum, we (masc.) : sa- si(?M), we (fern.) ; whilst in sa-im, we two (fern.), the consonant Ich {It) is entirely suppressed. The radical t in tita, I, is a changed Ic. The relation of Hottentot ti or t« (I) to Hebrew aw&(ki) is the same as that of Xosa (Kafir) n6.i to Zulu mgi or Tshuana ki, I. Now, whilst the vowel i is bent on turning a guttural to a dental, u, on the other hand, shows the tendency of changing h to a labial (of. English rough, ruf, enough, enuf) ; hence Hottentot xeiAi (Old Egypt, entot, suffix -f), he (for xei-KU) : but xei-B (0. Egypt. e?iios, suffix -s) she (for xei-KI.) We shall now proceed to review the Hottentot (Nama) forms in detail, as we find them in H. Tindall's " Grammar and Yocabulary of the Namaqua-Hottentot Language," Tita (affix -ta), I. The root is ti, primitive form 7a KI = man here, I. Tita appears to be a reduplication of ti, the a in the second syllable being demonstrative and identical with the a of other objec- tive pronouns, as ba, him, sa, her. The full form is therefore probably TI-TI-A ( = ^/-^/-^) = I-I-there = ti-ta = I-me = I, the Nama thus reversing the order of the Herero ami-me = me-I = I (present and future tense). (Sa-ts (affix -ts), thou (masc), r. ts {tsa, tsu) = 7 KU= man, person (absolute) ; modified feminine form sa-s (affix -s), thou (fern.) : sa- si = sa-KI = sa-KU = person (thou, he or she). Xei-s (xei-la, suffix -p or -b), he (masc), r. 'b(i) = BU (BA), primi- tive form 7 KU (conditional 76 KA) = man, person, he; modified feminine form xei-s (suffix -s), she : xei-s(i) = xei-KI = xei-KU = man, person, he or she. Xe(i)-i (suffix -i), it (com. gender), originally xei-KI, primitive root 4 KI, which is the first member of the common plural xe(i)-in (suff. -n, -in = 5 KI-MI), they. The relative pronoun Ma (the only one left in Nama) = that, which, who, is probably allied to, or rather radically identical with, the suffix -i and the pronoun xei. Compare the Herero pronoun i (he, she, it), the neuter pronoun tyi (it), and its corresponding demonstrative form hi (this, that). (Sa-khum (affix -khum), we two (masc), r. khum (aspirated to distinguish it from kum, now in use as plural) = 8 KU-MU (sexual dual) ; modified forms : sa-im (sa-KIM), we two (fem.) ; sa-rum (= sa-TUM= sa-KUMU), we two (com.). "Sa-kum, we (masc), plural, but originally dual, primitive form go A LANGUAGE-STUDY BASED ON BANTU. 8 KU-MU; modified forms sa-si {sa-SIM, sa-KIM = sa-KUMU), we (fern.) ; sa-da (sa-da?» = sa-KAM = sa-KUMU), we (com.). /Sa-kho, you two (masc), abbreviated from sa-khum = sa-kum = 8 KU-MU (sexual dual) ; modified form sa-ro (abbreviated from 6'a-rum = sa-TUM = sa-KUM), you two (fem. and com.). /Sa-kho is only a somewhat stronger pronunciation of Sa-'ko, you (masc. plur.), from wMcb is derived sa-so, you (fem. plur.), and sa-du, you (com. plur.) radically identical with Xei-'kw. ovxei-'ka. (suffix -ku, KUM), they (masc.) and xei-61 (suffix -ti or -di, orig. KI, abbreviated from KIM = KUM), they (femin.), all modifications of 8 KU-MU. In Xei-')sh& (suffix -kha or -ka, abbreviated from KAMA), they two (masc. and com.) and xei-ia, (suffix -ra = TA = KA, mutilated form of KAMA), they two (fem. or com. gender), the primitive forms 2 KA-MA and 86 KA-MA may have coalesced. As to the suffixes of the noun, their identity with the correspond- ing pronouns is self-evident, and, as far as our present purpose is concerned, they do not require a special treatment. In whatever light the suffixes may be viewed, either in primary state as primitive nouns, or as roots arrived at pronominal stage (see Tables I. and II.), nothing can be clearer than that the Hottentot terminations of the noun and the corresponding pronouns are identical. The Hebrew Personal Pronouns Compared. — It is obvious that the study of the Nama pronominal element throws a good deal of light also on the Hebrew pronouns, which evidently have been shaped and adapted in much the same fashion. As in Hottentot, so also in the Semitic languages, the primitive plural forms (with the exception of perhaps one) are extinct, and forms of the sexual dual, variously modified, have been substituted. The feminine forms are not original, but clearly modifications of the masculine (properly personal) gender — in two cases by vowel-change : at, ati, from ata, thou (person) ; ilia from ima, he (person) — and in two others by changing one of the consonants : aten, from atem, ye ; hen, from hem, they. But the dual being not represented in the Hebrew pronoun, there was no occasion for carrying the process of modifying and curtailing so far as in Nama. We notice that in the Hebrew pronouns, just as in Nama (and indeed in our own languages), only those forms have stood their PRONOMINAL TABLES. 9 1 ground which originally represented living things, namely, pronouns which originally had the consonants k (modified h, i, &c.) and m (modified n) : the exchange of k Qch, k, s) and t (th, d, r) being equally easy in both languages. The plural pronouns hem, they, and atem, ye, are originally forms of the sexual dual (pair, hence more than one, a number), and closely allied to the sign of the dual (-aim) and the plural {-im), whose primitive form I hold to be 2 KA-MA or 5 jK'/-ilf/=the two (living) things, or also more than one, a number. In anoki, I, the primitive "ja KI= man here, person present, I, appears to have been well preserved. Anoki is the ground-form of the nominal suffix -i = my, and is radically identical with the ^thiop. suffix -ku = I (gra&ar-ku, I made). But supposing the vowel u to be radical, and not a colouring of i, there is this difference : the ^thiop. -ku is the absolute form, and means simply man (hence he, thou and also I), whilst the Hebrew (awo)ki is the conditional form, with the by-meaning (man) here, thus more definitely I : — KU, modified TU=ma.iD., person (absolute), hence he (she), thou; KI, modified 2^J=man here, T^evson present = 1. It is not impossible that the strong guttural ch in fmachnu, we, is the original plural of anoM, and identical with ga KRI=men here, we, analogous to the Gothic ugkis, us (two), where, in accordance with the primitive plural law, the stronger consonant gk represents "more than one," two or a number of 4k(i), man here, I : — Hebrew araoki, I. awa-ch(-nu), we (-we). Gothic i-k(i), I. u-gk-is, us (two). But when the primitive plural became obsolete or was deemed wanting in emphasis, the originally feminine plural form anu (33 MHU), on losing its definite feminine character, was, for the sake of clearness or emphasis, added; so that possibly the fuller form and literal meaning of the double pronoun awachnu may be not ana-KI-NU, ana-KI-MHU (yet + 33) = I-we, but ana-CHI-NU or ana-KHI-MHU (ga + 3s) = we (com. pers. and masc.) -we (fem- inine) = " we here- we," the prefixed ana- being demonstrative and inorganic. 92 A LANGUAGE-STUDY BASED ON BANTU. An analogous case of two originally distinct genders blending into one form we have in Herero ami-?)ie = I, and ami-wdyi = I, forms in whicli two genders, the common personal and the originally feminine, combine : the original meaning of ami-me and ami-wdyi being I (femin.) -I (com. pers.) — a double 7-7, just as a«a-ch-nu appears to be a double emphatic we-we, in which two originally dis- tinct genders amalgamate. The other form for I, am, from which the verbal suffixes are derived, is probably not a contraction of anoJci, but the originally feminine form 33a MI (Skr. -mi, Engl, me, Bantu ami, Hebr. a-ni {an-MI=an-NI, a-ni), I (orig. femin., but afterwards com. gender). In Herero, a prefixed n regularly changes m to n. The fact that in Hebrew the simple form of the verbal suffix of the first person singular is not -i, but -ni, is certainly in favour of the assumption that the nasal in ani is not demonstrative, but radical. There is ground to believe that in an earlier stage of Hebrew, when the distinction of gender was still observed also in the first person, the forms ani, I, and anu, we, corresponded to each other as feminine singular and plural; whilst an-KI, I, and an-CRI (the first part of aMach-nu) were in use as singular and corresponding plural for the mascuhne or common personal gender : — I. Pers. Sing. I. Pers. Plur. a?j-ki (prim. r. KI) = person an-c\i (prim. r. KHI) = persons here, I (masc. and com. per- here, we (masc. and com. per- sonal), sonal). a(n)-ni (prim. r. Ml) = female a{n)-mi (prim. r. MRU, absolute) here, I (feminine). = females, we, us (feminine). AtA, thou, modified at or ati, fern., appears to be only a varia- tion of the objective ka, both forms being radically identical with the primitive pronominal root 7 KU= man, person. We have remarked already that the exchange of 7c (kh) with t (th), or even a labial, can still be traced in the Bantu pronoun. Such an inter- change was, in the beginning of language, impossible. TU, for KU (living one, person), would have meant an erect dead one, a statue, a raised one ; and PC a flying one, an airy one, a spirit. But when by degrees the special characteristics and original powers PRONOMINAL- TABLES. 93 of the consonants faded_ away, and only expediency and euphony were consulted, the consonantal interchange between the three different organs of speech came into play— first, it would seem, in the domain of the pronoun, and afterwards also in the verb, espe- cially in the Semitic languages. " Je weiter die sprachen von ihrem ursprunge sich entfernen, desto mehr gewinnt die liebe zum wohllaut an einfluss, weil sie nicht mehr in dem klaren geftihl der bedeutung der sprach-elemente einen damm findet, der ihrem an- streben sich entgegen stellt " (Bopp). As we mentioned before, the plural forms otem (Arabic an-%nva), ye, modified fern, aten, appear to be adapted from the sexual dual form 8 KTJ-MTJ = human pair, hence ye (two), or ye generally. The Arabic has preserved the original u, which at the same time shows that the primitive form of ata (thou) and of ka (thee) was something like {an)TU, KU (Herero ku, thee). For the first member of atem (a-te-m, anAM-m) is evidently identical with ata, just as ke in ke-m (you) and the objective ka (thee) are the same. In hua, he, of which ata (thou) and ka (thee) are only variations, we find the primitive m of 7 KU (man) uncoloured, as also in the Arabic plural hum, f. hunna, they, identical with the blunted Hebrew forms hem, f. hen, they. The primitive form is 8 KU-MU, they (pair), more than one person, hence a number, they, and the terminal u is preserved in the ^thiop. humu, homu ( = hem, hen, they), as also in the Hebrew verbal sufiixes -mo, -amo, -emo, them. The feminine form hia, she, is derived by vowel-change from hua, he, as at(i), thou (fem.) from ata, thou, (masc), Nama di (they, fern.) from ku (they, masc), sa-iva. (we two, fem.) from sa-khum (we two, masc). In the Pentateuch the masculine (originally personal) hua is, with some rare exceptions, common gender, standing for both he and slie, like the personal Bantu u (he, she) and ku (thee, masc. and fem.) : an archaism which in itself alone affords sufficient evidence for the high antiquity of the books of Moses. General Remarks. — Taking a general survey of the pronominal forms in the Bantu, Aryan, Semitic, and Hottentot families, as represented on Table IV., we observe the following distinctive features and peculiarities in their relation to the common original stock. In Bantu, the primitive correspondence between singular and 94 A LANGUAGE-STUDY BASED ON BANTU. plural has been wonderfully well preserved, also the fm-ms of the sexual dual, whilst the idea of the dual, except in one case, has been lost, the originally dual forms being now used for the singular and plural Some original feminine forms are still extant, but they have assumed a common personal and local meaning. Real grammatical gender is, therefore, wanting in the present state of the Bantu languages, no effort having been made by the ancestors of the African nations to keep it alive by substituting conventional feminine forms derived from the common personal gender, as has been done in the Aryan, Semitic, and other families. The personal, neuter, and local meaning are at present the chief features of the Bantu prefixes and pronouns. The Aryan nations have, in all the three persons, retained the primitive natural plural, whilst few, if any, traces seem to be left of the sexual dual. For the Aryan dual is merely a modification of the plural : in Gothic and other idioms it is evidently, as Bopp has pointed out, a composite consisting of the plural pronoun and part of the numeral two, meaning literally we two, ye two, as Gothic vi-t (we two), Lith. yu-du (ye two), &c. Also in Sanskrit the dual seems to be radically identical with the plural. The original femiaine forms, sing, and plur., MU and MHU, &c., have been preserved, as, e.g. in Skr. -mi, I, Engl, me, and plur. Skr. nas (us), Lat. nos (we, us), but the original feminine meaning is lost. Later the feminine was formed from the masculine (or common personal) by change of consonant, as Engl, she from lie, or in other ways. In the Semitic languages the original correspondence between singular and plural is, except in one or two cases, extinct, forms originally belonging to the sexual dual being in use now for the plural. A few primitive feminine forms have been preserved, but their signification as such has been lost. The feminine of the third person singular is formed from the masculine or personal gender by changing the vowel u to i. The Hottentot family, too, has lost the primitive correspondence between singular and plural, but has made the most of a few retained forms of the sexual dual, which have been modified by aspiration, abbreviation, or change of consonant and vowel, to serve as plural and dual pronouns. The feminine is derived from the masculine or personal gender by consonantal and vowel changes. PRONOMINAL TABLES. 95 THE PEIMITIVE PKONOMINAL SYSTEM EESTORED. Table I. — Absolute Poems. Roots in Primary State as Primitive Nouns. (Forms in thick capitals extant in Bantu.) SiNGULAB, Sexual Dual, Plukal, common, personal, and male-female, couple, common, personal, and masculine. pair. masculine. Living things. I KA, living thing. 2 KA-MA, living 3 KHA, livingthings. pair. 4 KI, „ „ 5 KI-MI, ,, ., 6 KHI, 7 KU, „ „ 8 KU-MU, ,, „ 9 KHU, „ Dead things. 10 214, dead thing. 11 2MM4, dead pair. 12 TS4, dead things. 13 TI, ,. „ 14 TI-MI, „ „ IS THI, „ 16 TU, „ „ 17 TU-MU, „ „ 18 THU, „ Waving things. 19 PA, waving thing. 20 PA-MA, waving 21 PHA, waving pair. things. 22 PI, „ „ 23 PI-MI, „ „ 24 PHI, 25 PU, „ „ 26 PU-MU, „ „ 27 PHU, „ Feminine gender. 28 MA, mother, fe- 29 MHA, mothers, 1 male. females. 30 MI,'. „ „ 31 MHI, „ 32 MU, „ „ 33 MHU, , 96 A LANGUAGE-STUDY BASED ON BANTU. ^ a . a Q H m W ■a .a ^3^ si § " -, - 3 '*-' 1^ cS t: 'o nT m r-' r=i 'i^ s-^Sg ^. o f^vO 0\ M IT) »-l M HI i-« N M PRONOMINAL TABLES. s. 8 =l< s O t) p H fe rt s O ^ w. o H rt g 5^ H £' H .S cc SS l>^ k: 02 8 ^ S i-H C^ g E o ^ § M 03 P a a s 9T S 60 ja ,c O 4a 1 -ciS g-go ^^es„. g-^- r4J .-.« . J --.= £ .-,=1 CO ^O tTi tD .S 0) 1m >^ M- - I> M ft? o .!3ja -^ a ■ s s CO <» W 1/1 00 H Ht E-i "8 X *t5 isg-s-^ ^ ■B a o ^< ft °.a »-^ °= a S si) o 5 M 3 o 5 o Vrt I _3 "ft Oj K s B 1^ ^ [S -p ^ < 02 PM pq M H 17H t D3 H £ o « i-i a S e3 .Ss" ai>-g §"°M Mr o 03 tow . t* o< >H t> .w ^ d ft ffl 03 d3 ^="3 ■- 03 . 03 ■• ' ^ fc> 6h gi s°igi .§ S S S s . H 0) i_r . F-&(ji_i rifles 00 O N w cn CO CO en, G THE PRIMITIVE PRONOMINAL SYSTEM RESTORED. Table IV. — Heeeho, Kama, Hebrew, and English Pronouns. Hereru Primitive Forms. Living things. 1 KA, living thing (generally) ; it. 2 KA-MA, living pair ; it-she. s 3 ff!ff.4, living things (generally); they. 4 KI, living thing (animal) ; it. 5 KI-MI, animal-pair ; it-she. 6 KHI, living things ; they. 7 KU, man, person (absolute); thou; hej (she) ; the great living thing ; it. \ ja EI, man here ; I. ^b KA, man there ; he (she). Formative Prefixes. oKA-XIIL,3ing.; E- V. sing. oMA-VL, dual, plur., sing. oZO-n-X.,plur. oTYI-Vn.,sing. oN-, oM- IX., sing. oVI- VIII., plur. oKU-XV.,8ing.; oKO- XVIL (local) ; o- 1. Personal Demonstrative Pronouns. ' Pronouns. Kama (Hottentot). Hebrew. Suffixes of the Noun. Personal Pronouns. Personal Pronouns. Personal Pronouns. ke, ka, it. tn^a, this. ye, e, ya, a, . t>,ps(, . they (two); "^a, these, they; it. *'"^- ze, za, they, tyl, it. inda (indza), these, ihi, this. i, he, she (ani- indyl, this, mal) ; it. vl, they. 8 KU-MU, man-woman, human pair (ab- [ oMU- I., sing, solute) ; he-she. ( ! u- 1. 8a KI-MI, pair here ; we two. 86 KA-MA, pair there ; they two. imbi, these. idgUi, this. 9 KHU, men ; they ; you (absolute). ga KHI, men here ; we ; ye (here). 96 KBA, men there ; they. Dead things. 10 TA, dead thing ; it. 11 TAMA, dead pair ; it-she. 12 THA, dead things ; they. 13 TI, dead thing ; it. 14 TI-MI, dead pair ; it-she. 15 THI, dead things ; they. 16 TU, dead thing ; it. 17 TU-MU, dead pair; it-she. 18 THU, dead things ; they. Waving things. 19 PA, waving thing ; it. 20 PA-MA, waving pair;. it-she. 21 PHA, waving things ; they. 22 PI, waving thing ; it. 23 PI-MI, waving pair ; it-she. 24 PHI, waving things ; they. 25 PU, waving thing ; it. 26 PU-MXJ, waving pair ; it-she. 27 PHU, waving things ; they. Feminine. 28 MA, mother, female (generally) ; she. 29 MHA, mothers, females ; they. 30 MI, mother, female (animal) ; she. 31 MHI, mothers, females ; they. 32 il/C7'(human)mother, woman; she(ab3ol.) 32a MI, mother, woman here ; I ; me. 326 MA, woman there ; she ; her. 33 ilf.ff£7,mothers, women; they; you(ab3ol. 33a MHI, women here ;'we, us ; ye (here). 336 MHA, women there ; they, them. oU- XIV., plur.; sing(abs tract) 00- II. ku, thee ; ku, it. ndyl, I, me. eye, e ; ye ; he, she; him, her. u, he, she ; thou ; ouf, o've, thou ; i'/igui, tliis. mO (ma-u), thou. i(inme=ma-i) I. e, a (in ma= ma-ka), he, she. u, they ; it ; tu, we, us. oVA-II.pl.; ov- II. (ozo-n-X..). e- V. sing. oZON-X.(dual), plur. mbi, I (orig. we) ; cte, we, us. ve, va, they ; ouo(oWo),they, oRU- XI. sing. (ozon-) X. oTU- XII. plur. oPO-XVI.(local). ri, it. ze, za, they. oMl- IV. plur. (oi;j- vin.) oU- XIV. sing. oMU- III. sing. (ou- XIV. plur.) -ma- (in oma- VL) n-, -m- (in ore-, om- IX.) -mu- (in oinu- I. andIIL);oMO- XVIII. (local). ru, it. tu, they. pe, pa, it. vl, they. u, it. u, it. mu, him, her; mu, it. aml,oiiami,I, me. mu, ye, you ; (-enu, yours). ene, ye, you -KHA (-KA), masc. dual. ■RA, com. dual. ' xeikha,, they two, mas. or com. ; xeixa,, I" they two, fern, or com. gender. -I, com. sing. -N, -IN, com. plur. xel, it, com. gen. xeln, they, com. gen. -B, mas. sing : -S, fern. sing. tniba, these. mdi, this. j'reda {indza], these. iniXLi, this, isut, these. imba,, this. inibi, these. imbui, this, imbui, this. mU), this. -KtJ,mas. plu. ; -TI, or -DI, fem. jjlur. iflibUi, these ; ! this. sats, thou, mas. ; sas, thou, fem. ; sats or sas, com. xnb, he; xeis, she. tita, I, me. it (Goth, ita, O. Germ. iz). ata, thou, mas. ; at, thou, fem. hua, he (she) ; Ma, she. ka, thee. anoki, I. sakhum, we two, mas. ; savn, we two, fem. ; gg niTTI, we two, com. gen. sakum, we, mas. ; sasi, we, fem. ; sada, we, com. gen. sakho, you two, mas. ; saTO, you two, fem. and com. gako, you, mas. pi. ; saso, you, fem. pi. ; sadu, you, com. pi. ajeiku (or xeika,), they, mas. ; xeidi, they, fem. {xeikha, they two, mas. or com. ; xeira, they two, fem. or com. gen.) atem (Arab, an- turn), ye, mas. ; aten, ye, fem. kem, you, mas. ; ken, you, fem. hem(Arab.lium, .Sthiop.humu), they, mas.; hen (Arab, hunna), they, fem. anoiChnu, we. I tliou(Lat. tu, / Skr. twam), I (Goth, tk, Skr. aham). he (ha, a, Skr. ajam, this one), she. i you (Skr. ytl- yam), «B (Goth. tigkiV, us two). we(wl,Skr.va- yam) ; ye (0. Dutch ghi). they (tha,). anl, I anu, we; anach- nu, we; uu, ua. me (Goth. mU-, Skr. m&m, m& ; ml, I). (Lat. nos, we, us; Skr. nas, us).