Full text of "Introductory sketch of the Bantu languages"

See other formats

:CD

"co

INTRODUCTORY SKETCH OF THE
BANTU LANGUAGES

ALICE WERNER

Reader in Swahili, etc., School of Oriental Studies,
London Institution.

Author of Language-Families of Africa, Native Races
of British Central Africa, etc.

LONDON :
KEGAN PAUL, TRENCH, TRUBNER & CO., Ltd.

NEW YORK: E. P. DUTTON & CO.

1919

RAP
L 3.3-Mt

1097921

MAGISTRO ET AMICO
CM.

Inter arma silent artes et amiciticz
ne sileant in czternum !

IX. Kal Sept. MC MX VIII

ERRATA.

22, 1. 4 : for ili-zwe read i(li)-zwe,

55, 11. 18-19 : delete "(which does not occur in
Zulu.) "

59, 1. 2: for "nasal" read "labial."

99, 1. 17 : for naba read nampa.
114, 1. 7 : for nanku read nangu.

for nabo read nampa.
185, 1. 12 : for Hereo read Herero.

222, 1. 21 : "in" . . . . to end of page,

delete and substitute ", while I have heard satu,

which may be borrowed. Krapf gives both satu

and chatu, the latter as a quotation from Steere.''

CONTENTS.

PAGE

PREFACE ... ... ... ... iv

CHAPTER

I. INTRODUCTORY ... ... ... i

II. THE ALLITERATIVE CONCORD ... ... 20

III. THE NOUN-CLASSES ... ... ... 31

IV. THE NOUN-CLASSES (contisnied) ... ... 54

V. CASES : THE LOCATIVE ... ... 70

VI. THE PRONOUN ... ... ... 86

VII. THE COPULA AND THE VERB ' To BE ' ... 109

VIII. THE ADJECTIVE ... ... ... 118

IX. THE NUMERALS ... ... ... 133

X. THE VERB ... ... ... 143

XI. THE VERB (continued) — MOODS AND TENSES 156

XII. ADVERBS AND PARTICLES ... ... 182

XIII. WORD BUILDING ... >.. ... 199

XIV. SOME PHONETIC LAWS ... ... 218

APPENDIX I. TEXTS — i. ZULU ... ... 232

„ „ „ 2. HERERO ... ... 248

„ „ „ 3. ILA ... ... 264

„ „ „ 4. NYANJA ... ... 272

„ „ „ 5. SWAHILI ... ... 276

„ „ „ 6. GANDA ... ... 295

APPENDIX II. BIBLIOGRAPHY — i. GENERAL ... 307

„ ,, „ ii. SPECIAL LANGUAGES 309

INDEX ... ... ... ... ... 343

PREFACE.

IT is well to 'state at the outset that this little book
makes no pretensions to originality. It has not, in all
cases, been possible to give detailed references for
statements which may be recognised as derived from one
or other of the standard authorities (they are not
numerous) on the subject. Sometimes, in the course of
studies covering, intermittently, a period of some thirty
years, one assimilates an idea so thoroughly as to forget
where one first picked it up ; sometimes, too, doing
first-hand work at a language, one may, unknowingly,
arrive at facts or deductions already recorded. In all
such cases, the originaf owners are requested to believe
that no misappropriation was intended.

It may be as well to state that the languages at which
I have worked in situ, that is to say, in the countries
where they are spoken, are Nyanja, Swahili, Zulu, and,
in a lesser degree, Giryama and Pokomo. Some others
I have, to a certain extent, studied from the inside, with
the help of books.

It will, I hope, be sufficiently clear from the title that
the present work is only an " Introduction " and does
not in any respect seek to enter into competition with
those of Bleek, Professor Meinhof, Father Torrend and
Sir Harry Johnston. I am under great obligations,
more or less, to all four, though compelled to differ,
occasionally, with all respect, from each ; but I venture
to think I have occupied some ground not completely
covered by any of them, yet important from the beginner's
point of view.

If I might venture to appeal to my own experience, I
should say that my feeling on first introduction to Bleek's
Comparative Grammar was one of mere bewilderment,

PREFACE V

caused, I think, partly by the highly technical character of
the first part, which presupposes a considerable acquaint-
ance with phonetics, and partly by the use of Lepsius's
alphabet, which, though not very difficult, involves a little
preliminary training if one is to use the book profitably.
Moreover, this alphabet has been considerably modified
(and, in my view, improved) by Meinhof, so that there
is a slight additional difficulty involved for those who
have already made the acquaintance of the latter.

It is superfluous to say anything in commendation of
the Lautlehre and the Grundzuge einer vergleichenden
Grammatik der Bantusprachen ; they are as yet
practically the only works of their kind,1 with the
exception of Mile. L. Homburger's highly specialised
study, which is of comparatively limited scope. But
experience has taught me that they are of very little use
to at least three-quarters of the students, whom it has
been my lot to induct into one or other of the Bantu
languages. For one thing, there is as yet no English
edition of either, and — in spite of recent improvements
in this respect — the number of English people who can
study a subject by means of a French, German or
Italian book (which is a different thing from gathering
the drift of a novel or a newspaper article) is still
deplorably small. For another, like Bleek, they pre-
suppose a kind and degree of philological knowledge
which few of the people who take up some Bantu
language at short notice have had time or opportunity to
acquire.

Here, parenthetically, at the risk of seeming to attempt
the impossible feat ascribed to " Old Man Hyena," who
split in two with the effort— I want to say a word about
two opposite errors.

I have repeatedly insisted, in the following pages, on

1 Father Torrend's book, valuable enough in some ways, has
to be used with caution, not only because of the irrata unavoidable
in a pioneer work, but because the learned author has not been
proof against that temptation to unbalanced theorising which is apt
to beset the African philologist. De Gregorio's Cenni di Glottologia
Bantu, so far as I have examined it, does not seem to go beyond the
material furnished by Bleek and Torrend.

Vi PREFACE

the danger of being misled by preconceived notions of
grammar into erroneous treatment of Bantu speech.
Therefore it would seem as if a knowledge of Latin and
Greek were no help towards the acquisition of African
languages, and indeed, as we shall see, such knowledge
has in some cases given rise to positive stumbling-blocks.
But the fact remains that those who have enjoyed a
sound classical training are best fitted to cope with the
unfamiliar prefixes, affixes and infixes of the " Lingua
Bantu."

The truth is that — at any rate till quite recently — the
classics have been the only subject taught in our schools
and universities which provided a thorough grounding in
the principles of comparative philology. Neither Latin
nor Greek will by itself throw any light on the structure
of, say, Zulu or Ganda — nor, for that matter (except for
its greater approximation, in some points, to primitive
characteristics) will Hebrew. No very great amount of
classical scholarship is needed to discriminate between
roots and formative elements, to distinguish and compare
the functions of the latter and to ascertain and apply the
laws of sound-shifting. But it is the method and the
principle which make all the difference, and those who
follow them will never go astray over fruitless compari-
sons with Akkadian, Tibetan or what not. So that, if
some parts of my book should seem to be needlessly
elementary, I may be allowed to point out that I have
found them by no means superfluous in practice.

I have not dealt with theories of origins or conjectures
as to the successive Bantu migrations. Neither have I
attempted a classification of the Bantu languages into
" branches " (Bleek) or " clusters " (Torrend). I cannot
help thinking that it would be premature at present and
will be for some time to come. Father Torrend
perceived that new facts had (to some extent) disturbed
Bleek's arrangement; and there are still so many languages
of which little or nothing is known, that we can
scarcely regard his own as other than provisional.
The queerest isolated links of affinity are continually
cropping up in unexpected places and upsetting one's

PREFACE ' vii

most cherished prepossessions ; and, for my part, I am
perfectly content, to take the languages as we find them,
leaving the questions of how they came to be where
they are, and whether they have a right to be there, to
more competent heads and a future stage of inquiry.

Neither have I attempted to treat of Bantu phonetics
from the strictly scientific point of view. This branch
of the science, which is still more or less in the pioneer
stage, is safe in the hands of Mr. Daniel Jones and
Professors Meinhof and Westermann — I would only
take this opportunity of emphasising its importance.
The time has passed when the practical linguist or the
research student could afford to rely on his ear and a
certain amount of theoretic knowledge gained from the
older works on the subject. It is one which can never
be satisfactorily studied from books alone, and everyone
intending to proceed to Africa ought to avail him or
herself of the excellent practical courses now open.

I have tried to explain in the text the various shifts
and compromises I have been forced to adopt in order to
arrive at a working orthography for my own immediate
purpose. Both Meinhof's diacritic marks and the
alphabet of the I. A. P. have raised endless typographical
difficulties, and I have found it best in the end to fall
back on Steere's rule-of-thumb, explaining, where they
occur, such symbols as he failed to provide for. It may
be necessary to repeat that kh, th, ph, stand for the
aspirated consonant and not (except where specially
pointed out, as in the case of Herero th) for the sounds
which we associate with those symbols.

The aim of the book is not to furnish all details with
regard to any particular language, but to depict the
broad principles underlying the structure of all belonging
to the Bantu family, in such a form as to facilitate the
subsequent study of the one specially chosen. Their
grammar is of so homogeneous a character that it is
unusually easy to construct such a general outline. Nor
need the student be afraid with any amazement when
he finds that his own chosen idiom fails to conform in
one or more particulars to the outline here sketched.

Viii PREFACE

By the time he has advanced so far as to discover this,
he will know enough to fit the differences as wdl as the
resemblances into the framework.

It has, of course, been impossible to provide for every
contingency, for instance, I have just become aware that
Chaga possesses an infixed adverbial (intensive) particle
for which I know no parallel elsewhere. I shall be
grateful to anyone who can give me any information
throwing light on the distribution of this particle, if it is
not an isolated phenomenon— but no doubt we shall be
able to read all about it in Sir Harry Johnston's great
book, when the present distress permits of its publication.
Very likely it contains the answers to many other puzzles
here suggested ; but, all the same, I venture to repeat that
there are some who will read it to better purpose, when
it comes, after making use of the humble stepping-stone
here offered them.

It only remains to thank those who, by word or letter,
have contributed information, advice and encouragement,
both in former years and recently. Chief among these
I wouldmention Professor Meinhof ; Sir H. H. Johnston;
Dr. Cook, Miss Allen (of Gayaza), and the Rev. H. K.
and Mrs. Banks (of Mbale), all of Uganda ; the mission-
aries of the C.M.S. at Mombasa ; the Rev. Dr. Hether-
wick, Blantyre ; Mrs. Lloyd of St. Faith's, Rusape,
and the Rev. H. Buck (Rhodesia), the Rev. H. B.
Barnes (Penhalonga, Rhodesia) ; Miss Nixon-Smith,
U.M.C.A., Likoma; the Ven. Archdeacon Woodward
and Miss Woodward ; the Rev. W. A. Crabtree (late of
Uganda) and others. I hope anyone accidentally
omitted will not think me ungrateful.

Any criticisms or comments — especially coming direct
from Africa, will be heartily welcomed.

Wasalaam !

A. WERNER.

School of Oriental Studies,
Finsbury Circus,

London, E.G. 2.

INTRODUCTORY SKETCH OF THE
BANTU LANGUAGES

CHAPTER I
INTRODUCTORY

THE Bantu family of languages is spoken
throughout Southern and Central Africa, as
far as the Gulf of Cameroons on the north-
west, and the Tana river on the north-east.
This area is interrupted by the following
islands or * enclaves ' of speech belonging to
other families :

The Galla : between the Sabaki and Tana.

The Masai : to the east and south-east of Lake
Victoria.

The Jaluo (' Nilotic Kavirondo ') : at the north-
eastern corner of Lake Victoria.

The Hottentots and Bushmen : in South Africa.

Also several small and little known tribes (Mbugu,
Sandawi, etc.), in the depression south-east of
Kilimanjaro, whom we need not specify more
particularly.

2 INTRODUCTORY

In the Cameroons and along the southern
edge of the Congo basin, the line of demarca-
tion between Bantu and non- Bantu (in this
case Sudan or ' Nigritian ') languages is not
very easy to draw. In the former territory
we find several languages classed as ' Semi-
Bantu,' or ' Bantoid,' which share certain
characteristics with the family, though not
apparently belonging to it. But these, and
the exact delimitation of the frontier, need
not concern us for the purposes of this book.

The number of known Bantu languages is
well over 200 ; but as there are probably
others yet to be recorded, and as some names
may have to be omitted (being synonyms, or
denoting mere dialects — if not altogether
erroneous), this figure must be regarded as
merely provisional.

The principal features of the Bantu lan-
guages are so clearly marked in all, that, as
far as grammar goes, a knowledge of one
materially facilitates the acquisition of the
rest. Most of them differ from each other no
further than do French, Spanish and Italian ;
in some, the resemblance is even closer.
Natives of one tribe cannot, in general,
understand the language of another, without

INTRODUCTORY 3

learning it (though they pick it up very easily),
nor can the European expect to do so ; but
the second language should cost him far less
labour than the first. And an acquaintance
with the framework of Bantu grammar,
comprising, at least, those features which all
the languages have in common (and which, to
those who know only the idioms of Europe,
are so striking and novel as to impress them-
selves readily on the memory) is a useful
preparation for taking up the study of any
particular language in Africa.

The name Bantu was first introduced by
Bleek (1827-1875), who may be called the
father of African philology. It is simply
one form of the word for ' people,' whicl^ is
used throughout the languages of this family.
Various objections have been raised to this
name, but no better one has been proposed,
and it has now so far gained currency that it
would be extremely difficult to displace.
As its meaning is perfectly clear, and as it
is easily pronounced, there seems to be no
sufficient reason for rejecting it. We shall
therefore continue to speak of the Bantu
family.

Though the name was not introduced till

4 INTRODUCTORY

the middle of the nineteenth century, the
existence of this language-family was at any
rate conjectured as early as 1808, when the
German naturalist, JLichtenstein (who had
spent four years travelling in South Africa),
published -a paper entitled Remarks on the
Languages of the savage tribes of South Africa,
with a short vocabulary of the most usual dialects
of the Hottentots and Kafirs. The two Bantu
languages of which * he collected specimens
were ' Kafir ' (Xosa) and Chwana. Many of
his words are recognisable, in spite of a curious
orthography ; but he does not seem to have
grasped the system of prefixes, and sometimes
confuses the singular and the plural of a word.
However, he had no doubt as to the relation-
ship of these languages to each other and the
fundamental difference between them and that
of the Hottentots He says : ' All the idioms
of the South African savages must be regarded
as dialects of one or the other of these two
principal forms ' ; and the information he was
able to obtain respecting the more northerly
tribes led him to the conclusion that ' we are
justified in considering all the inhabitants of
the East Coast of Africa, from 10° or 12° S.
to the frontiers of the Dutch Colony, as one.

INTRODUCTORY 5

nation to which further research

may perhaps compel us to add the inhabitants
of the South-west Coast.'

A similar conclusion was reached inde-
pendently, a few years later, by our own
orientalist, William Marsden (1754-1836).
In 1816, he drew up a paper of instructions
for collecting words and sentences, to be used
by the members of Captain Tuckey's ill-fated
expedition to the Congo, in which he remarks
on the similarity between the vocabularies
previously obtained in Angola and Loango and
the specimens of the Mozambique language
dictated by a native of that country who had
been Marsden's servant in India. But the
study of the Bantu languages singly, and
without reference to their place in a system,
goes back to the middle of the seventeenth
century. In the library of the British Museum
is a curious little book — with Southey's
autograph, dated ' Keswick, 1810,' on the
title-page — printed in 1642 and containing a
short exposition of elementary Christian
doctrine, under the form of a dialogue, in the
language of Angola, with a Portuguese version
on the opposite page, and a few introductory
hints (in Portuguese) on pronunciation and

6 INTRODUCTORY

grammar. It was the work of a Jesuit
missionary, P. Francisco Pacconio, but was
revised and edited after his death by P.
Antonio Do Couto, to whom it is generally
attributed. The language is that now called
Mbundu, arTd, though somewhat disguised by
the Portuguese spelling, appears not to differ
appreciably from that spoken to-day. Some
years later, in 1659, an Italian friar, Giacinto
Brusciotto, published in Latin a grammar of
the Congo language, to which we shall have
occasion to refer more than once in subsequent
pages. Cust remarks: ' The book is very small,
and the author was not a linguist ' ; which
seems to me unduly severe. He certainly
grasped the characteristic features of the
language in a way some later writers failed to
do : Cust himself says, ' he remarks the use
of prefixes, and he classes the nouns.' We
have just seen that Lichtenstein did not
understand the system of prefixes ; — it is, of
course, not surprising that a passing traveller,
picking up, in the short time at his disposal,
what linguistic information he can, should be
unable to do more than record words and
phrases without penetrating very far into their
grammatical relations. But it does seem

INTRODUCTORY 7

strange that Dr. Van der Kemp, whose help
he acknowledges with regard to the Xosa
language, should not have called his attention
to peculiarities so striking and so unlike
anything that could previously have come in
his way.

But the great advance in the knowledge of
African languages followed the remarkable
development of missionary activity which
characterized the end of the i8th and the
beginning of the igth century. Moffat's
translation of the Bible into Sechwana was
begun in 1831 ; Archbell's grammar of the
same language appeared in 1837, Boyce's
Xosa grammar in 1844 ; while at the same
time Casalis, Arbousset and the other French
missionaries were active among the Basuto
and marking their progress by valuable
linguistic work. About the same time, Krapf,
on the eastern coast of Africa, was practically
the first to make the Swahili language known
to European scholars : — for, though two or
three vocabularies had been collected (chiefly
by the praiseworthy exertions of British naval
officers) they do not seem to have attracted
much attention. It was the material sent
home by Krapf which first made possible

8 INTRODUCTORY

anything like a scientific study of the subject,
and the beginnings of this may be seen in
three remarkable essays contributed by Ewald,
Pott, and Von der Gabelentz to the first and
second volume of the Zeitschrift dtr Deutschen
Morgenlandtschen Gesellschaft — work produced
in the golden age of German scholarship,
before it had begun to lose itself in over-
spscialisation. These essays were the pre-
cursors of Bleek's Comparative Grammar, the
first part of which appeared in 1862.

Bleek's book, though of course it has been
supplemented by later research, and, as might
be expected, requires some correction in detail,
remains the foundation of everything that has
been done since. I shall not attempt to give
any account of this more recent work, though
I shall frequently have occasion to refer to
the outstanding names of the last thirty or
forty years — men who have not merely given
us grammars and dictionaries of separate
languages, but examined their structure from
a scientific point of view and done something
towards determining their relationship to each
other and to the other speech-groups of the
world. Such have been Miiller, Lepsius,
Meinhof, De Gregorio, and others. In this

INTRODUCTORY 9

country, Sir Harry Johnston is, sad to say,
almost the only writer who has occupied
himself with the Bantu languages not merely
in detail but also from the comparative point
of view.

Bleek confirmed Lichtenstein's view (which,
considering the data he had to go upon,
almost deserves the name of a brilliant intui-
tion) that all the languages of South Africa
fall into two groups,1 and he was able, as
Lichtenstein was not, to account for the
differences on philological grounds. The one
crucial distinction between them, he considered,
lay in the fact that the one group — the
Hottentot — has grammatical gender ; the
other — the Bantu — has not.

This difference, Bleek thought, was based
on a fundamental difference of organization,
and from it he deduced an ingenious argument,
proving that people whose speech has no
grammatical gender were not merely at present

Bleek was uncertain whether to reckon two groups
or three. He felt that not enough was known about the
Bushman language to pronounce definitely as to its
classification, but was inclined to think it of a distinct
type from the Hottentot. Recent research goes to show
that he was probably right and that it is allied to the
Sudan family.

10 INTRODUCTORY

incapable of personifying nature, but that they
could never in the future advance beyond a
certain limited range of ideas. However, as
fuller knowledge has shown many of his
premises to be untenable (he thought, for
instance, that the kind of animal-stories so
well known to us through Uncle Remus was
confined to the Hottentots and unknown to
the Bantu), we need not occupy ourselves
with his conclusion.

It is certainly remarkable that the three
great inflected families of language — the
Aryan, Semitic and Hamitic — corresponding
to the three divisions of the ancient world and
the civilizations (broadly speaking) of Europe,
Assyria and Egypt, should possess gram-
matical gender and the rest be without it.
But we need not think that the possession of
this characteristic draws a hard and fast line
on one side of which no progress, is possible,
for (setting aside the case of Japan and China),
recent research has thrown a good deal of
light on the way in which gender arose, and
we find that some languages, classed with the
Hamitic ' sex-denoting ' family, only have it in
a rudimentary form ; some Bantu languages
show signs of a tendency to acquire it ; and

INTRODUCTORY 11

languages at a very advanced stage, as
English, tend to lose it.

* Absence of grammatical gender,' it may
hardly be necessary to say, means, not that no
account is taken of sex-distinctions, but that
they are not in any way shown by the form of
words. All languages have words for * man '
and 'woman,' ' male ' and ' female ' ; but those
of which we are speaking have nothing corre-
sponding to ' he,' ' she,' ' his,' ' her ' ; nor can
they indicate a feminine noun by any change
in the word : if it is absolutely necessary to
distinguish the sex a word is added, as in our
1 he-goat,' ' she-goat,' ' buck-rabbit,' ' doe-
rabbit,' etc. Still less do they attribute sex,
by a grammatical convention, to inanimate
objects, as is done in Latin, French, German,
etc.

The Sudan languages (which include Twi,
Ga, Ewe and others, spoken in Western and
Central Africa) have no grammatical gender ;
but neither have they, properly speaking, any
grammatical inflections at all. The Bantu
languages, however, do indicate number, person,
and, in a limited sense, case ; and, for verbs,
in addition, voice, mood and tense.

They are usually reckoned as belonging to

12 INTRODUCTORY

the class of Agglutinative languages. These
are distinguished from the Isolating languages
on the one hand and the Inflected on the other
by the fact that, while they indicate gram-
matical relations by particles prefixed or
suffixed to -- the root, these particles are
recognizable as independent words and can be
used as such. This, as we shall see, does not
quite apply to the Bantu languages, where
some of the ' formative elements ' (prefixes and
suffixes) can no longer be used separately,
and sometimes we even find internal changes
in a word, comparable to those by which in
English we form the plural of a noun like foot
or the past of a verb like run.

So that it would be nearer the truth to call
them ' partially inflected languages,' or ' lan-
guages in course of acquiring inflection.' For
we must remember that the three classes just
mentioned are not hard and fast divisions, like
water-tight compartments ; but a live language
is continually growing and changing and will
sooner or later pass from one class to another.

The first point which strikes one on
beginning to examine these languages is the
employment of prefixes where we should
expect to find suffixes — e.g., to indicate the

INTRODUCTORY 13

plural of nouns, the agreement of adjectives,
etc. We shall find that -suffixes are also use_d
in certain cases ; but the system of prefixes is
so characteristic and peculiar that Bleek rightly
regarded it as a distinguishing feature of this
family, which — before finally adopting the
designation ' Bantu ' — he called the ' prefix-
pronominal languages.'

It was also noticed by Brusciotto who, at
the very outset of his Grammar, says : ' In
the first place it must be observed, in general,
that in this language we have to attend, not to
Declensions (i.e., terminations), but rather to
Principiations (i.e., Prefixes).'

For want of acquaintance with this principle
we sometimes give a double plural to an
African word, as when we speak of ' the
Basutos,' '.the Mashonas,' or use a plural for a
singular, as ' a Basuto,' ' a Bechwana ' — the
singular in these cases being Mo-suto, Mo-
chwana. Besides these prefixes indicating
singular and plural, there are others indicating
the language (as Ki-swahili,Lu-ganda, Se-chwana,
Chi-nyanja) and the country (as U-kami, Bu-
ganda, etc.) — varying, of course, with different
tribes. It may be well to note in this -place
that we shall uniformly throughout this book

14 INTRODUCTORY

use the names of languages without prefix, as
Chwana, Ganda, Swahili, etc.

Though Bantu nouns have no gender and
so cannot be classified as masculine, feminine
and neuter, they are divided into several
classes — usually eight or nine, distinguished
by their prefixes. These prefixes are repeated,
in one form or another, before every word in
agreement with the noun ; and this method of
indicating agreement (which will be fully-
explained and illustrated in the next chapter)
is called the Alliterative Concord.

These three^ points : the absence of gram-
matical gender, the system of prefixes, and
the Alliterative Concord, may be called the
principal characteristic features of the Bantu
family.

We may mention a few others, put on
record long ago by Lepsius, as distinguishing
the Bantu family from the Sudan languages
on the one hand and the Hamitic (Berber,
Galla, Somali, etc.) on the other.

(1) Personal Pronouns are always prefixed to

verbs, never suffixed — as they are in Hebrew,
Arabic and the Semitic languages.

(2) The Genitive always follows its governing

INTRODUCTORY 15

word. That is, they always say ' the house of
the man,' never (as in the Sudan languages)
' the man's house.'

(3) The usual (but not invariable) order of words

in the sentence is : Subject + Verb + (Noun)
Object.

(4) The object-pronoun is inserted (' infixed ')

between the subject pronoun and the verb-root.
> Thus, in Zulu, ngi-ya-tn-bona, ' I see him,'
is made up of ngi = ' I,' ya (tense particle),
m = ' he,' bona = ' see.'

(5) Syllables always end in a vowel.

Here it is well to say a word about stress
(accent) and intonation.

In many Bantu languages it is an invariable
rule that the accent — at any rate the accent
most readily noticed — falls on the penultimate
syllable, and, if a syllable is added, the accent
moves forward. Thus, in Zulu, we have
bona, ' see,' which becomes, in the causative,
bonisa, ' make to see.' In Swahili, nyilmba is
' house,' but, the locative, ' in the house,' is
nyumbdni. This is called the 'rhythmic
stress,' but there is also an 'etymological
stress,' — viz., one on the root syllable. In
words like bona, nyumba, these coincide. ; but

16 INTRODUCTORY

otherwise, in Zulu and Swahili, the rhythmic
stress seems to be much more strongly marked.
In Ganda, it is the stress on the root-syllable
which is noticed. There are a few languages
which have the rhythmic stress on the ante-
penultimate.

Intonation, or pitch, is a very important
feature in some languages, as in Chwana,
where it serves to distinguish many words
otherwise similar. It exists in Zulu, Xosa,
Nyanja, etc., — probably to a much larger
extent than has hitherto been observed. All
learners are advised to attend to this point
very carefully.1

This book being devoted to the grammatical
structure of the Bantu languages, it does not
enter into my plan to discuss their sounds from
a scientific point of view. Indeed many of
them have not been examined at all in this
respect, and others very imperfectly. Almost
the only comprehensive work on Bantu
phonetics at present in existence, Professor
Meinhof's Lautlehre dtr Bantu-Sprachen, is
not yet published in an English edition,

1 It is possible that in some cases, pitch and stress
have been confused. This, also, requires particular
attention.

INTRODUCTORY 17

though a translation is being prepared. Some
of the Bantu languages are being phonetically
analysed by Mr. Daniel Jones, Reader in
Phonetics in the University of London, who
has published some provisional results of his
studies in Le Maitre Phonetique, in his pamphlet
The Pronunciation and Orthography of the
Chindau Language (Rhodesia), and more recently
in the Sechuana Reader (see Bibliography in
Appendix).

The sounds of the Bantu languages are,
superficially, not very difficult, except in a
few cases which at once strike the newcomer
by their strangeness, such as the clicks in
Zulu (which, however, do not properly belong
to Bantu), the 'laterals' in this and some
other South African languages ; the Thonga
and Venda ' labio-dental,' the very common
' bilabial ' f and v, etc. But there are subtler
gradations, both of vowels and consonants,
which are at once perceived by a trained
phonetician, and which an untrained linguist
with a good ear will consciously or uncon-
sciously adopt without being able to define
them, but which frequently escape the notice
of the average person. Thus, perhaps, in
Nyanja, the learner will be in doubt whether

18 INTRODUCTORY

the word for ' five (people) ' is asanu or asano ;
and an old resident who knows the language7
fairly well will tell him that ' these endings
are very uncertain, and the people themselves
sometimes say one and sometimes the other.'
The truth is that the* sound is intermediate
between u and o, the mouth-opening being
wider than for the first and narrower than for
the second. As this is not a treatise on
phonetics, I shall make no attempt to spell
the words quoted as examples according to
the system of the International Phonetic
Association, more especially since the sounds
of so few Bantu languages have been sufficiently
analysed to make this possible. For my
purpose, the spelling introduced by Bishop
Steere for Swahili and generally used in
Swahili books is, in general, sufficient. Its
principle may be stated thus : the vowels are
pronounced with the sound they have in
Italian, the consonants (including the com-
pound symbols ch, sh, th) as in English — each
symbol standing for one sound and no more,
and no sound having more than one symbol.
On this system, c, q, and x are superfluous, so
are sometimes used to denote sounds not
provided for in the Roman alphabet, as the

INTRODUCTORY 19

clicks in Zulu.1 (C is often used for the Sound
of ch in " church " but may also stand for the
somewhat different " palatal plosive."

As, however, some languages have sounds
not found in Zanzibar Swahili, on which
Steere's Handbook is based, a few extra
symbols will be needed in our examples, and
these will be explained where they occur.
But it must be repeated that this can give
only a very general idea of the sounds, and
that anyone who has to acquire a Bantu
language for practical use cannot do better
than take a course of general phonetics, which
will enable him to accomplish very useful
work in recording correctly the sounds of
unwritten, or, as is sometimes the case,
hitherto atrociously mis-spelt, languages.2

1 Dinuzulu (late Chief of the Zulus) used to say that
the English alphabet Deeded ' several more letters' in
order to write Zulu satisfactorily. The Europeans who
first recorded the language have in some cases expressed
two entirely distinct sounds by the same symbol.

a See D. Jones, The Pronunciation of English and
Noel-Armfield, General Phonetics. A table of the Inter-
national Phonetic Association's Alphabet, and also of a
script largely used on the Continent (Meinhof's moderni-
zation of Lepsius's Standard Alphabet] will be found
'n Language-Families of Africa.

CHAPTER II

THE ALLITERATIVE CONCORD

IN Latin we say, * Equus albus currit ' — ' the
white horse runs'; in the plural, ' Equi albi
currunt.' The termination of the noun
indicates the declension, case, and number ;
of the adjective, the gender, case, and number
in agreement with the noun ; of the verb, the
tense, number, and person. The terminations
of the noun and adjective are the same ; that
of the verb is different and has no relation to
them.

This arrangement is somewhat different
from that of the Alliterative Concord in the
Bantu languages, but will help us to understand
it, if we try to imagine the endings all alike1
and transferred to the beginning of the word.
Let us take a specimen sentence in Zulu.

1 They are not really as much alike as the word
' alliterative ' might imply, but they are all recognizable
as derived from the prefix.

THE ALLITERATIVE CONCORD 21

Uw/ana omubi uyatshaya inkomo yomfundisi
wami, ngiyakumlungisa. ' The bad boy is
beating the ox of my teacher : I will punish
him.'

This sentence contains nouns of two different
classes and words agreeing with them. Um-
fana is a noun of the first or ' person '-class :
the root isfana, the prefix um-, shortened from
umu- (as seen in umu-ntu, l person '). Omu-bi,
* bad,' is an adjective agreeing with umfana
the prefix assumes the form omu because it
was formerly preceded by a demonstrative
particle fl, and a-\-u coalesce into o (i.e., the
broad o, pronounced like ou in ' ought ').*
This means that, when the adjective is used
attributively (that is, as in ' the bad boy ' — not
predicatively, as in ' the boy is bad '), it is
really a relative construction that is employed :
— ' the boy who is bad.' We shall be able to
make this clearer in the chapter on relative
pronouns. In the same way 'good ' is omu-hle
(a+umu-hle).

The equivalent for an adjective can never
be given in its complete form, unless the noun

1 When it is necessary, in this book, to distinguish
this o from the narrow o (as in ' stone ')• it is printed
with a line under it, as in Meinhof's notation.

22 THE ALLITERATIVE CONCORD

with which* it agrees is known. 'A good' (or
1 handsome ') ' person,' is umu-ntu omu-hle ; ' a
fine ox,' inkabi en-hie; l a beautiful country'
ili-zwe eli-hle ; ' a fine cattle-kraal,' isi-baya esi-
hle ; 'a beautiful face,' ubu-so obu-hle ;
1 beautiful language,' uku-kuluma oku-hle. For
this reason, adjectives must be given in the
dictionary under their root only : -hie, -bi,
-kulu (large), etc. But these roots are flever
found standing by themselves in any Bantu
language. They are always used with the
prefix of the class to which they belong — viz.,,
that of the noun with which they are in
agreement.

U-ya-tshaya. Tshaya is a verb meaning
' beat.' The bare root in this form is never
found alone, except in the second person
singular of the imperative. Everywhere else
it has some addition. Even the second person
plural of the imperative takes a suffix, -ni :
tshaya-ni = ' beat ye.' The other moods and
tenses all take prefixes.

U- is the personal pronoun of the first (or
' person ') class. It will be recognised at once
as part of the prefix umu. (The prefix, except
in some languages which have departed
considerably from the original type, is not in

all cases identical with the pronoun.) This
is the subject-pronoun : the second part of the
prefix, -mu (usually contracted to -;») is used
as the object-pronoun, as we shall see presently.
It should be noted that this subject-pronoun,
«, can never be used apart from a verb or its
equivalent. There is a separable, or inde-
pendent, pronoun, of quite a different form,
which will be considered in the chapter on
Pronouns.

-ya- is a tense-particle : originally the
auxiliary verb ya, 'to go.' It imparts a kind
of habitual continuative force : uy*atshaya is
rather ' he is beating,' or ' he is in the habit of
beating,' than simply ' he beats.' In Zulu,
the -ya- tense is the present most commonly
used, though it cannot always be translated
as above.

In-komo, ' a cow,' is of the class which has
the prefix i, or in- originally ini-. Nouns
have no indication of case (except that they
suffix -ni for the locative), so that they have
no distinct form for the objective, though
some pronouns do.

Yomfundisi is for ya umfundist* Ya is the
particle corresponding to * pf,' which expresses
the genitive case and varies its initial according

24 THE ALLITERATIVE CONCORD

to the noun with which it agrees — which is
always the thing possessed, not the possessor.
/, the initial vowel of inkabi, becomes y before
a vowel : i-\-a=ya. In Zulu, a before u
amalgamates with it to form o, which is an
intermediate position of the mouth between
the two. In many other languages this
amalgamation does not take place, because
the initial vowel has been lost ; thus, in
Nyanja, we say ya muntu, not yomuntu —
ya-\-umuntu.

Um-fundisi, a noun of the person-class
meaning ' teacher,' derived from the verb
fundisa, ' teach.'

W-ami — ' my.' The roots of the possessive
pronouns are : -ami, l my ' ; -ako, l thy ' ; -ake,
1 his, her ' ; -etu, ' our ' ; -gnu, ' your ' ; -abo,
* their.' They take as prefixes the pronoun of
the class with which they are in agreement :
in this case the thing possessed is of the person
class (umfundist), and the pronoun will be u.
But u before a vowel becomes wt u-\-ami—wami.
Similarly, ' thy teacher ' is umfundisi wako
(u+ako), and so on.

Ngi-ya-ku-m-ltmgisa. Lungisa (causative
of lunga) is properly ' make right,' * straighten,'
and so ' correct,' ' punish.' Ngi- is the

THE ALLITERATIVE CONCORD 25

inseparable subject-pronoun of the first person
singular. Ya- is the tense-particle already
mentioned, but, in combination with the
following particle — ku — it indicates the future,
-m- is the object-pronoun of the third person
singular^' him.'

All these prefixes change for the plural.
Supposing we take as our English sentence :

' The bad boys are beating the cows of our
teachers ; we will punish them."

The Zulu will be :

Aba-fana aba-bi ba-ya-tshaya izin-komo zaba-
fundisi betUy si-ya-ku-ba-lungisa.

This needs no further analysis ; but we
may call attention to two points : the plural
pronoun (inseparable) of the first person, si-,
and the double plurality, if one may say so,
of the possessive betu. It must be 'our,'
plural of * my,' in order to agree with the
possessors, ('we,' understood), but the initial
must be b-, not w- in order to agree with the
things (or persons) possessed (abafundisi).
This double concord of the possessive is an
important point, to which we must recur later
on.

The same sentence would read in Ganda
as follows :

26 THE ALLITERATIVE CONCORD

Omu-lenzi omu-bi a-kuba en-te yomu-igiriza
wa-nge ; n-na-mu-kangavula.

Aba-lenzi aba-bi ba-kuba en-te zaba-igiriza
ba-ngf tu-na-ba-kangavula.

Here, though the roots are mostly different,
the identity of the formative elements will be
evident on examination. The chief differences
are : a instead of u for the pronoun of the
third person singular (which will be noticed
in the chapter on the pronouns), and e instead
of i as the initial vowel for the in- class. The
possessive of the first person is -nge instead
of -mi, but this is evidently connected with
the Zulu subject-pronoun of the first person,
ngi-, which, in Ganda, has been reduced to n-.

This sentence affords a very good illustration
of the fact that, in comparing languages, one
should take into account the grammatical
structure rather than the vocabulary. All the
noun-roots are entirely different from the Zulu
ones ; so are the two verbs, kuba and kangavula.
If we looked to these alone, disregarding the
prefixes, we might come to the conclusion
that there was no sort of relationship between
the two languages. But we should not be
justified in doing so, for a comparison of
single words may very easily lead us astray.

THE ALLITERATIVE CONCORD 27

Take the case of four European languages,
which we know to be closely related : English,
German, Dufcch, and Danish. Here are four
words which cannot possibly be derived from
the same root :

Boy ; Knabe ; jonge ; Dreng.

Yet the Dutch word exists in English as
the adjective ' young,' and is used in German
(junge) side by side with Knabe, which is our
1 knave ' — an instance of the way in which the
same root may assume different meanings.
Dreng is found in Anglo-Saxon in the sense
of ' warrior,' and the old Icelandic use of it to
mean ' a valiant youth,' supplies the connection
between the two.

' Queen ' is the same word as the Danish
Kvinde, ' a woman,' and therefore has nothing
to do with the German Konigin, which is the
regularly-formed feminine from Konig, or the
Danish Drottning, which, though used as the
feminine of Konge (' king ') is really that of an
obsolete word Drott, meaning ' lord.'

Or take the French word cheval : it has
nothing in common with the Latin eqmis, but
is derived from a different word, caballus, not
used by the classical writers, but existing in
the language of the people. Again, the

28 THE ALLITERATIVE CONCORD

Spanish comer, ' eat,' cannot possibly come
from the same root as the French manger;
and th$ classical Latin is edere, which, at first
sight, does not seem to be connected with
either. But comer is derived from comedcrct
properly ' to eat up ' — a more colloquial and
popular word than edere — and manger comes
from manducare, properly * to chew ' ; whence
also the Italian mangiare.

Why one language should choose the first of
these two words, and another the second, is a
question which, in the present state of our
knowledge, cannot be answered — or only in
the same way as Moliere's doctor explained
why opium sends people to sleep by saying
that ' it has a definitive virtue.'

So the roots, lenzi, kuba, iga (' learn,' from
which are derived iglriza, l teach, and omu-
igiriza, * teacher ') and kanga (frown,7 of which
kangavula, ' rebuke ' or 'punish ' is a derivative)
are probably to be found in other Bantu
languages, though I have as yet been unable
to trace them. Ente, I believe, is not Bantu,
though I cannot say whence it has been
adopted. -bi, ' bad,' will be recognised as
identical : it is found in most Bantu languages.

Let us now take, from Swahili, an example

THE ALLITERATIVE CONCORD 29

of the concord in another class, which has the
prefix ki-, in the plural vi-.

Ki-ti cha-ngu cha m-tl ki-me-vundika, ni-me-
ki-ona ki-ki-anguka.

' My wooden chair is broken ; I saw it when
it fell.'

Ki-ti, ' chair,' is originally * a wooden thing '
— ti being a root which, with the prefix m-,
means ' tree.' (The most primitive form of
seat, after the mere stump or fallen log, is the
stool cut out of a solid block, the cross-section
of a tree.) 'Of in this class is cha, because
ki becomes ch before a vowel. The possessive
pronoun consists of cha prefixed to the pro-
noun-root, which is for the first person -ngu
(cf. Ganda, -nge : the subject-pronoun for the
first person is ni). Cha mti, 'wooden,' —
literally ' of wood,' or ' of tree.' Ki-me-vundika,
' it is broken ' : ki-, subject-pronoun of the ki-
class, agreeing with kiti ; -me-, a particle
denoting the perfect tense of the verb ;
vundika is the neuter-passive of the verb vunda,
' break.' Nimekiona : ni- subject-pronoun of
the first person singular ; -me-, tense-particle ;
-ki-, object-pronoun agreeing with kiti ;
ona, verb, meaning ' see ' (in Zulu, bona) ;
anguka is a verb, meaning ' fall ' ; the first

30 THE ALLITERATIVE CONCORD

ki is the subject-pronoun agreeing with
kiti ; the second a tense-particle equivalent to
' if ' or ' when,' often giving the verb a. kind of
participial force.

The plural of the above is :

Vi-ti vy-ttu vya mti vi-me-vundika^ti-me-vi-ona
vi-ki-anguka.

This needs no further explanation.

We thus see that the prefix of the noun is
repeated, in a form more or less recognizable
before every word in grammatical agreement
with it. The way in which it enters into the
composition of pronouns other than the simple
subject and object prefixes, will be explained
later.

CHAPTER III
THE NOUN-CLASSES

WE have already referred to Giacinto
Brusciotto as the author of the first attempt
at a Bantu Grammar. He was an Italian
Capuchin, Prefect of the Apostolic Mission to
the Kingdom of Congo, about the middle of
the seventeenth century. Judging from his
book (published at Rome in 1659), his
linguistic aptitudes were of no mean order,
and no doubt he had profited by many years'
residence in the country. It is remarkable, at
least, that he succeeded in grasping the
principle of the noun-classes, which eluded
more than one of his successors. We have
seen that Lichtenstein missed it ; and — even
more unaccountably — Burton, writing about
1860, with the work of Krapf and Rebmann
before him, could speak of ' the artful and .
intricated system of irregular plurals ' in

32 THE NOUN-CLASSES

Swahili.1 In Cavazzi's History of the Kingdom
of Congo? first published in 1671, it is stated
that a missionary, after six years spent in
trying to learn the rules of the language,
only found out that there were none ! It is
strange that this book takes no notice whatever
of Brusciotto or his grammar.

The first section of Brusciotto's manual3
has the following heading : ' Of the Declen-
' sion of Nouns, or, as it is better expressed,
' their Principiation, and their Rules ; wherein
' it is shown what articles are to be attributed
' to each noun, both in direct and oblique
' cases, for their correct construction in them-
' selves, or when they are joined to other
1 words ; and generally this is first to be noted
' that in the present tongue we must not look
' for declensions but rather principiations, for
' which we have the following Rules.' . . .

1 Zanzibar, I, 443.

3 Istorica Dcscrizione de1 tre Regni, Congo, Matamba
et Angola, sitnati nell 'Etiopia Inferiore Occidentals
e delle Missioni Apostoliche esercitatevi da Religiosi
Ca^wccj'm'^accuratamente compilata del P. Gio. Antonio
Cavazzi da Montecuccolo. (Milan, 1671.)

8 Regulae quaedam pro difficillimi Congensium
idiomatis faciliori captu, ad Grammaticae normam
redactae. (Rome, 1659.)

THE NOUN-CLASSES 3

Later on, having reached the end of the
' Principiations,' he says, once more :

' As has been said above, the language of
1 the Congos and others of Negro lands is not
' founded, nor forms its rules upon the
' declension of words, but on their principiation;
' therefore the rules which are distinguished
' and marked in this idiom are chiefly taken
' from the various principiations of the sub-
' stantives and varied accordingly.' From
this it appears that he duly appreciated the
importance of the noun-classes as a feature of
the language.

The first thing we have to do in studying
Latin is to master the declensions — the classes
into which nouns are divided according to
their terminations and genders.' Such classes
exist, though to a less extent, in German ;
they have almost disappeared in Dutch, and
entirely so in English. When we think of
declensions, we also think of cases, each
having its own case-endings.

It was quite natural that anyone educated,
like Brusciotto, mainly on the classics, and
more especially on the Latin grammar, should,
in trying to discover the laws of an entirely
strange language, look first for the declensions.

34 THE NOUN-CLASSES

He soon recognized that the plural of .nouns
was formed in different ways, according to
distinct rules, but that the inflection came at
the beginning of the word instead of at the
end, so that he invented, as we have seen,
the name of ''principiations ' for the different
classes so distinguished. Of these he enu-
merates eight, which can be identified without
difficulty in present-day Kongo,1 allowing for
differences of dialect and for some mistakes
and confusions. It is curious that he does
not notice the person-class, but makes ' gentile
nouns ' exceptions to his first and second
principiations. At the end of his chapter, he
quaintly adds :

' Note, with regard to the preceding, that
' there is no rule so strictly observed as to be
' without many exceptions, all which by
' practice and the Spirit of God inspiring, will
' be easily understood and by continuous and
' unwearied labour overcome.'

Exceptions are the refuge of the imperfect
grammarian, and a knowledge of the Bantu

1 This spelling is preferred in modern books when
referring to the particular language treated by Brusciotto,
while ' Congo ' is retained as the name of the river or
its adjacent territories.

THE NOUN-CLASSES 35

languages, unattainable by our pioneer (though
not, in his case, for want of ' continuous and
unwearied labour ') would have shown that
they usually exemplify rules not immediately
obvious.

Brusciotto may have been led astray partly
by his belief in the existence of an article — a
part of speech which, as we understand it in
English, is not found in Bantu. He is not
alone in giving this name to the initial vowel
of the prefix — a point as to which we shall
have more to say presently ; — but it is less
easy to see why he should have extended
it to the possessive particle (wa, ba, ya, etc.).
We shall return to this point in the fifth
chapter.

The number of noun-classes, as the ' prin-
cipiations ' are now generally called, varies in
different languages, but is mostly eight or ten.
There is some uncertainty about the original
number, and Meinhof s theoretical table is, as
he points out, not complete, since some
languages have anomalous forms only to be
interpreted as survivals of lost classes, and
more of these may yet be discovered.

Meinhof, following Bleek, counts singular
and plural classes separately, thus arriving at

36 THE NOUN-CLASSES

a total of twenty-one. Some singular prefixes
have no corresponding plural, while some
plural prefixes are attached to two or more
classes having different prefixes in the singular.
As the order in which these classes are
arranged is hardly the same in the grammars
of any two Bantu languages, it seems most
convenient in this book to follow Meinhof's
arrangement and refer to the prefixes by his
numbers. Some advantages are secured by
placing singular and plural in the same class,
and in drawing up a practical grammar it
might be better to follow that arrangement ;
but the want of uniformity makes reference very
difficult in a comprehensive survey. When we
find, e.g., most Zulu grammars giving as the
second class what Steere, in S\vahili, calls the
fifth, the French Fathers in Ganda the sixth,
and Madan in Lala-Lamba the ninth, one is
ready to ask why we cannot adopt some
uniform system. But, when we remember
how many classes have been dropped by one
language and another (Duala, e.g., having only
seven in all) we see that it is impossible to
number them always in the same way, though
we may keep in every case the same relative
order. Some writers, conscious of the difficulty,

THE NOUN-CLASSES 37

have frankly given up the numbers and simply
designate the classes by their prefixes (' the
mu-ba class,' ' the mu-mi class,' and so on).
But as the prefixes assume different forms,
and are sometimes lost, this arrangement is
useless for comparative purposes unless some
standard form is agreed upon. The following
table contains the forms which Meinhof has
arrived at as probably the original ones-
Even if this view should be erroneous in some
cases, it is at least possible to see how all the
forms actually in use could have been derived
from them ; and, in any case, this does not
affect their use as a means of reference.

The prefixes are given in Prof. Meinhof's
orthography, as to which the following points
should be noted : v is the sound called
' bilabial v ' which is very common in present-
day Bantu, though in some languages it
has become b or w.1 i may here be
disregarded, merely remembering that in
Meinhof's opinion the vowel of the vt-
prefix originally differed from that of the ki-
prefix, also the li of the tenth class from that
of the fifth. The etymological importance of

1 See Noel-Armfield, General Phonetics, p. 71.

38 THE NOUN-CLASSES

this ' heavy i ' is shown in his book on Bantu
phonetics. 7 stands for the ' voiced sound '
(which does not occur in English) of Scots
ch in 'loch.'1

It may be as well to state here that the
orthography used in this book for specimens
of Bantu languages is that adopted in the
printed texts available for each particular
language. Where it has been found necessary
to depart from this orthography, or where
any symbol needs special explanation, the
fact is mentioned in a note. No attempt has
been made to unify the various systems : the
only satisfactory uniformity would be that
obtained by transcription into I. A. P. charac-
ters, and for this the study of Bantu is not
sufficiently advanced. The reader not already
familiar with this script, is referred to Mr.
Noel-Armfield's General Phonetics (1915).

He will also find a table of it facing page 30
of the author's Language-Families of Africa.

1. mu-

2. va- Plural of 1.

3. mu-

4. mi- Plural of 3.

5. li-

For which, in this book, we use the Greek character )(.

THE NOUN-CLASSES

9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.

ma- Plural of 5 and 14.

ki-

vl- Plural of 7.

li-ni Plural of 9 and 11.

lu-

tu- Plural of 11, 13 and 19.

ku No plural.

21.

<ya

Locatives. No plurals.

Plural of 20.

Other classes and prefixes of which occa-
sional traces survive, willjbe discussed later on.

It will make matters clearer if we subjoin
to the above skeleton table of prefixes, which
are mere abstractions and, as such, difficult to
grasp and remember, specimen nouns from
eight fairly typical languages, showing the
forms in actual use for each class, with their
concords as exhibited in the adjective and the
possessive particle. As far as possible, words
have been chosen which are found in all the
eight languages, so as to facilitate a comparison
of roots.

THE NOUN-CLASSES

No. 1 NOUN-

Zulu

Chwana

Herero

Class

Human Being

uniu-ntu

mo-tho

otnu-ndu

Do. pi.

aba-ntu

va-tho

ova-ndu

Tree

umu-ti

mo-re

omu-ti

Do. pi.

imi-ti

me-re

omi-ti

Tooth

i(li) zinyo

le-ino

e-yo

Do. pi.

ama-zinyo

ma-ino

oma-yo

Ch-st (thorax)

isi-fuba

se-huba

[otyi-na=
thing]

Do. pi.

izi-fuba

li-huba

[ovi-na]

Elephant

in-dhlovu

tlou

on-dyou

Do. pi.

izin-dhlovu

li-tlou

ozon-dyou

Wand

u(lu)-ti

lo-re

oru-ti

- 12

Do. pi.

otu-ti

Little stick

oka-ti

Human nature

ubu-ntu

vo-tho

o-undu

Death

uku -fa

Xo shwa

oku-ta

Place, at

—

opona

— to

okona

— in

omona

[Diminutive]

[Found in Duala

, e.g., i-seru '• d

warf antelope"

Clumsy person

—

20A

Do. pi.

Giant

—

•<M»

2lA

Do. pi.

NOTE. — Words in square brackets are inserted when

THE NOUN-CLASSES

CLASSES.

Nyanja

Swahili

Ganda

Gisu

Kongo

mu ntu

m-tu

omu-ntu

umu-ndu

mu-ntu

a-ntu

wa-tu

aba-ntu

baba-ndu

a-ntu

m-tengo

ffi-ti

omu-ti

[kumu-ba =
sugar-cane]

[mu-nse=
sugar-cane]

mi-tengo

mi-ti

emi-ti

[kimyuba]

[nii-nse]

dz-ino

j-ino

eri-nyo

li-sino

d-inu

ma-no

m-eno

ama-nyo

kama-sino

in-enu

chi-fua

ki fua

eki-fuba

[kiki-ndu =
thing]

[ki nzu=pipe]

zi-fua

vi-fua

ebi-fuba

[bibi-ndu]

[i-nzu]

njobvu

ndovu

en-jovu

i-tsofu

nzau

njobvu

ndovu

en-jovu

tsi-tsofu

nzau

[u-konde=net]

uti

[olu-ga=cane]

[lu-hingo=
bow]

[lutnbu=
fence]

—

[otu-dzi =
drop of water]

—

[tumbu]

[ka-ntu=
little thing]

—

aka-ti

[ka-busi,
little goat]

—

u-untu

u-tu

obu-ntu

bubu-ndu

uwu-ntu

ku-fa

oku-fa

ku-fwa

(ku)iwa

pa malo

[nyumbani (pa)]

wa-ntu

ha-ndu

v-uma

ku malo

[nyumbani
(kwa)]

[ku-mpi, near]

ku-ndu

k-uma

m'malo

[nyumbani
(mwa)]

[munda=
the inside]

nui-nclu

m-uma

—pi. lo-seru 12,

andNy-wcnia 0-

ulu "bird," pi.

tufulu 12]

[fi-mbele—
little knife]

ogu-ntu

aga-ntu

ki-ji-tu

gugu-ndu

—

ini-ji tn

—

gi ini-ndu

—

that with the same meaning has a diflerent root.

42 THE NOUN-CLASSES

Zulu and Ganda — both too well-known for
further comment — indicate, approximately, the
two extremities of the Bantu area. Herero—
called by South African colonists Damara — is
spoken in the south-western territory which till
recently was German. It is a language, in
many respects, of the highest interest ; but its
speakers are now sadly reduced in numbers.
Nyanja extends, more or less, from the north
end of Lake Nyasa to the Zambezi and is
closely cognate — if not virtually identical —
with the main speech of Southern Rhodesia —
usually, though not very correctly, called
1 Mashona ' or ' Chiswina.' The range of
Swahili is roughly from Warsheikh on the
Somali coast to Cape Delgado (though the
people themselves limit the name ' Swahilini '
to the coast north-east of the Tana mouth), but
it has been carried as a trade language far
into the interior, and even to the Congo.
Gisu, sometimes called ' Masaba,' is a very
remarkable tongue, whose principal speakers
live round Mount Elgon in the north of the
Uganda Protectorate : it has, perhaps, pre-
served more ancient forms than any other.
Kongo (sometimes called Fiote) is spoken by
large numbers of people on both sides of the

THE NOUN-CLASSES 43

Lower Congo, as far up as Stanley Pool, and
in the old ' Kingdom of Congo ' south-east of
the river, where the Portuguese missionaries
laboured.

Originally, we may suppose that some
definite meaning attached to each class, just
as, in languages possessing grammatical
gender, the masculine and feminine termina-
tions corresponded to a real distinction of sex.
It would be difficult now to give any reason
why hortus should be masculine, or mensa
feminine ; but no doubt, when it had once
become an accepted fact that nouns in -us
were mostly masculine and nouns in a mostly
feminine, words which happened to end in
these ways were ranged under one category or
the other, without regard to their meaning.

Sometimes it is still possible to say that a
class consists of nouns denoting a certain kind
of objects, such as the first (or ' personal ')
class, the fifteenth, which contains verbal
nouns (infinitives), and the diminutive class
found in some languages. Again, certain sets
of nouns may be found in one class— e.g.,
trees in the third, though it contains others as
well. Much ingenuity has been expended —
and, I cannot but think, wasted— in drawing

44 THE NOUN-CLASSES

up definitions of the classes : the attempt
seems to be hopeless at the present day,
because prefixes originally distinct may have
become identical in form, through elision or
contraction, and so two or more classes have
been merg'ed into one. We know this to have
happened in Swahili, where n (lu) and 14
(vu) have alike been contracted into u and are
now treated as one and the same class.

Class 3-4 (mu-mi) contains, besides trees, a
number of the parts and organs of the body,
which may, at one time, have formed a
distinct class.1

It should also be noted that the same word
is found in two or more languages with
different prefixes, e.g., ' year ' ; Zulu, umnyaka,
2 ; Nyanja, chaka, 7, etc.

The Fulfulde language of West Africa has
a most remarkable system of noun-classes,
much fuller and more clearly defined than
anything now to be found in Bantu. Professor

1 It is quite possible that they were originally locatives
with the prefix mn (mu-kouo, really ' in the hand ')>
which became confounded with this class owing to the
similarity of the prefix. To understand how this might
have happened, we may refer to Nyanja, where the noun
kamwa, ' mouth,' is never used by itself — only pa-kamwa
and m-kamwa, ' at ' and ' in the mouth."

THE NOUN-CLASSES 45

Meinhof's theory as to this language and its
possible connection with the Bantu family
is set forth in his Introduction to the Study of
African Languages.1

In some- cases, the distinction of mean-
ing implied by the prefix is quite clear.
There is no doubt about the diminutives, nor
the infinitive used as a noun (15) ; i indicates
persons ; 7, (sometimes) collectivity ; 14,
either abstractions, or some substance taken
in the mass, such as grass, wool, flour, etc.
And we find, over and over again, that the
same root may take different prefixes and
have its meaning modified accordingly. In
Zulu itmii-ntu, i, is ' a person ' ; isi-ntu 7, ' the
collectivity of beings ' — ' the world ' ; ubu-ntu,
1 human nature.' Umu-ti 3, is 'a tree ' ;
u(lu}-ti n, 'switch' or 'wand' (this has
suggested to some writers that Class 1 1 con-
sists mainly of long, thin objects — which is
scarcely borne out by the facts) ;2 ubu-tl 14,

1 Pp. 99, 100. See also Language-Families of Africa,
Chapter VI.

46 THE NOUN-CLASSES

1 poison ' ; and in Swahili, we have ki-ti 7,
* chair ' — i.e., ' the tiling made of a tree,' or
perhaps 'the little tree (thing)'1 — (see ante
p. 29).

Though, as we have seen, Bantu knows no
linguistic distinction of sex, a very definite
line is drawn between the living and the life-
less— or rather, perhaps, between the human
and the non-human. The first (;»«) and
second (ba) classes, in every Bantu language,
consist pre-eminently, if not exclusively, pf
names denoting human beings.

As a rule, even ghosts and other preter-
natural beings are not placed in the same
class. In Zulu i-ziuiu (usually translated
'cannibal,' but in reality a kind of ogre or
goblin), ama-to-'igo and ama-dhlozi (ancestral
spirits) are of 5 and 6 ; in Nyanja, mzimu, the
most usual word for spirit, is 2, and so is
Mulungit, which sometimes designates an
ancestral ghost, though at others it seems to be
used in a sense almost implying a ' High God.'
Some languages include animals in the first
cla??, but this is -evidently an afterthought.

The African mind, in general, inclines to treat

1 Ki- may here be the diminutive prefix — see next
Chapter.

THE NOUN-CLASSES 47

animals as persons ; — we see in their folk-tales
that the distinction is kept up with difficulty.
(Uncle Remus is sophisticated enough to be
conscious of the confusion, and reminds his
hearer that, once upon a time, ' creatures had
sense same like folks.') While most of them are
usually of the gth (ni) class (cf. in Z-ulu, imbuzi,
1 goat,' inkomo, ' cow,' ingwc, ' leopard,' indhlovu
'elephant,' etc.), it seems to have been felt that
this was quite illogical, so some languages (as
Swahili) removed them into the first class with-
out changing their form — i.e., treated them, in
respect of all their agreements, as first-class
nouns ; others gave them a special plural, by
placing the second prefix before their own
plural one. The few names of animals which
in Zulu and Chwana are included in the first
class are treated in a special way, suggesting
that they did not always belong to it. Their
plural prefix is o-, not aba-, which is also
taken by certain nouns denoting degrees of
relationship (as u-yise ' father,' pi. o-yise,
u-nina ' mother,' pi. o-nina) and all proper
names.1 Perhaps there was once a separate

Proper Nouns are often used in the plural, to mean
' a person and those with him.' See Colenso, First Steps
in Zulu-Kafir, § 29.

48 THE NOUN-CLASSES

class for names of relationship (which in most
Bantu languages are treated exceptionally in
some way or other) and it is worth noting
that most (not all) Zulu names of animals
coming under this heading are compounded
with uno- (= unina) . The corresponding nouns
in Chwana take the prefix bo- (vo-).

This first class has the prefix mu some-
times heard as mo-, sometimes worn down to
m, and sometimes changing to un-, especially
before t or d. In Zulu it is umu- or urn-,
sometimes contracted to «-. One might
be tempted to think that the longer form is the
more primitive one, and that languages like
Chwana and Nyanja have lost the initial
vowel. There seems reason, however, to
think that this initial vowel is not really part
of the prefix, but the remnant of a demonstra-
tive which has become amalgamated with
it. Gisu — a language in a very archaic stage —
supplies the clue : not in this class indeed,
but in the second, third, and fourth, where we
have baba-ndu l people,' kumu-kono (elsewhere
umkono, mukono, mkono), 'hand,' plural
kimi-kono.

Meinhof thinks that this demonstrative
originally had the form ya and assimilated its

THE NOUN CLASSES 49

vowel to that of the syllable which followed it,
while the y passed, sometimes into ng, some-
times into k. Thus ya nimmtu would become
ngumuntu, Gisu kumundu, and, the consonant
being dropped, the prefix would remain as
umu-. This pronoun survives as the * copula '
(to be treated in a later chapter), and in Zulu
we have ng'umuntu, ' it is the man,' ng'umuti,
1 it is a tree.' This is pointed out by Bleek
in his Comparative Grammar.1 On this theory
of its origin, it is quite intelligible that he
should call the initial vowel an article,
especially as it is chiefly absent in cases
where the employment of an article would
be inadmissible — always in the vocative,
and in certain negative sentences. However,
as it is impossible by its means to make any
distinction corresponding to that indicated by
the use of ' a ' and * the,' the initial vowel
can hardly be called an article in our
sense.

This original ya probably accounts for the
fact that in some languages (e.g., Swahili,
Nyanja, etc.) the pronoun of the third person

1 Pp. 150, 151. Of course he knew nothing of the
Gisu language, which has so fully justified his
deductions.

50 THE NOUN-CLASSES

(see Ch. VI.) is a- instead of ul. Forms like
om-untu, omuti, perhaps show that sometimes,
instead of the u assimilating the vowel which
followed it, contraction took place :
(y}+aumuntu=omuntu.

Names of tribes may be either of the second
or the sixth class: Aba sutu, Bechwana, Barolong,
Amazulu, Makaranga, Waswahili, Agiryama,
Abaganda. Those of the second have corre-
sponding singulars of the first : Umsutii,
Mochwana, Morolong. But sixth class nouns
do not always have a singular of the fifth :
I -zulu does not seem to be used for an
individual of the Zulu nation — perhaps because v
it is the same as the word for the ' sky ' — but
we have i-Bunu, ' a Boer,' i-Lawu, ( a
Hottentot/ i-Xosa, ' a Cape Kafir.'

The concords of Class 3 are almost identi-
cal with those of Class I. There are, however, J
two differences in the pronouns, which we
shall notice in a later chapter. Class 4 has
the same pronouns as Class 10, though its \
prefix (and consequently its agreement with the
adjective) is different. Besides the names of

1 Also for gu, gi and ga being occasionally found as
pronouns in the third, fourth and sixth classes, instead
of the more usual «, i and a.

THE NOUN-CLASSES 51

trees and parts of the body (usually : ' arm/
* finger,' ' leg,' ' foot,' ' heart,' — sometimes
' head,' etc.) this class contains some words
not easily placed under any special heading,
such as ' village ' (i.e., a ' kraal ' or collection
of huts of any size, up to what can be fairly
described as a town) — umu-zi, mo-tse, mu-dzi,
m-ji, mu-nzhi, mu-ndi.1 (Herero, curiously
enough, places it in n : oru-dhe.) Words de-
noting streams or rivers (though differing in
form) very often belong to this class. So
do, many proper names of rivers : Zulu
Utngtni, Umzimkulu, etc. ; others are found in
Class ii.

The fifth prefix is /*-, in Zulu ///-, usually
contracted into i-, as i-tshe (for ili-tshe), l stone,'
• i-zwij 'word,' etc. In some languages this
prefix has assumed the form di- ; in some it
is lost altogether, except before roots beginning
with a vowel, where it is sometimes represented
by dz or j : Nyanja dz-ina, 'name,' which is
in Svvahili j-ina, (Yao l-ina). In Swahili, the
pronoun // and the possessive particle la, used
with nouns like neno, 'word,' ziwa, 'pool,'
would be something of a puzzle if it were not

1 Bleek thinks these may have originated as locatives
in mu-. See his Comparative Grammar, p. 130.

52 THE NOUN-CLASSES

known from other languages that these words
must originally have begun with li.

Names of fruits are found in this class,
differing only by prefix from the name of the
tree which produces them : Swahili m-buyu,
1 baobab,' buyu, ' calabash ' (which grows on
it), m-kuyu, ' fig-tree,' kuyu, ' a fig,' m-chnngwa,
1 orange-tree,' chungwa, ' orange ' ;x Zulu
um-kiwant, ' wild fig-tree,' i-kiwane, etc.

Many words denoting liquids, or substances
handled in bulk, which are not individualised
and therefore have no singular, belong to the
sixth class; e.g.t 'water': Zulu anta-nzi,
Swahili ma-ji or ma-i, Nyanja (and other lan-
guages), ma-dzij Kongo ma-za, Duala ma-diba,
etc. ; ' milk ' : Swahili ma-ziwat Ganda ama-ta,
Chwana ma-ski; 'sour milk': Zulu ama-si,
Herero oma-ere ; 'millet': Zulu ama-bele,
Nyanja ma-ere, ma-pira, etc., etc.

Here we may mention a feature which at
first sight seems puzzling in Chwana and some
other languages. Nouns of Class 9, besides
their own plural sometimes take a second in

Nazi, ' coco-nut,' tende, ' date,' ndizi, ' banana," are
exceptions, being of the ninth class. But none of these
three seem to be original Bantu roots. The names of
the trees are quite regular : tn-nazi, m-tendt ; but the
banana-tree is m-gomba, not m-dizi.

THE NOUN-CLASSES S3

ma-, indicating that there are very many of the
thing in question ; e.g., nku, l a sheep,' — plural
li-nkit, but ' many sheep ' = ma-nku. These
may be regarded as collectives with no singular,
and they are possibly connected with the lost
augmentative class, to which we shall recur
at the end of the next chapter.

It will be seen by reference to the Table
that the form assumed by the sixth prefix in
Gisu is kama-, in accordance with the principle
already explained. I have not met with any
other example of this form ; but the original
ya prefixed to the ma explains why the pronoun
in some languages (Giryama, Ganda, etc.)
should be ga. It is usually a or ya.

It is less easy to see why the prefix should
appear in Herero as oma-. But that language,
as the Table shows, has a tendency to make
all the initial vowels o : the only exception is
the fifth prefix, which has assumed the form t-.

CHAPTER IV
THE NOUN-CLASSES (continued)

THE forms of the seventh and eighth
prefixes vary considerably. Besides those
shown in the Table, we have : —

Duala . . . e-, be- : e-koto, ' cap,' pi. be-koto.
I la (Middle Zambezi) . . . chi- shi- : chi-bia,

' pot,' pi. shi-bia.
Kikuyu (British East Africa) . . . ki-t i- :

ki-hato, ' broom,' pi. i-hato.
Konde (north end of Lake Nyasa) . . . iki-

ifi- : iki-tala, ' bedstead,' pi. ifi-tala.

In Ronga (Delagoa Bay), and in some dia-
lects of the language spoken in Rhodesia the
eighth prefix has a very peculiar sound, written
in the International Phonetic Alphabet a- and
by Meinhof s. It is produced by getting ready
to say ih (as in * thin ') and then rounding the
lips, as if for w, and somewhat resembles the
whistle which results from blowing into a key.
Venda (North Transvaal) gives this prefix the

THE NOUN-CLASSES 55

voiced sound, which is written p and by
Meinhof z.

This class, like some others, consists of two
or three originally distinct ones, which may
account for the prefix conveying several
different senses. Sometimes it has a kind of
instrumental force and is then used to make a
noun out of a verb, implying the thing by
means of which the action is accomplished, as
isi-hUlo (Zulu) ' a seat,' from hlala, ' to sit ' ;
Nyanja chi-psero, ' a broom,' from psera, ' to
sweep ' ; chi-peta, i a winnowing-basket,' from
P&ta. Sometimes it seems to convey a
collective sense, as in Zulu, isi-ntu (already
remarked on), isi-Znlu, ' the whole of the
Zulus, the Zulu nation,' isi-zwe, ' tribe, nation,'
from i-zwe, l country.' This is distinct from
the use of the prefix to express language (which
doec not occur in Zuk^: Chi-Nyanja, Ki-
Swahili, Se-Sutu ; or, more accurately,
' likeness, fashion, manner,' as its application
is not confined to language. Again, it forms
a kind of abstract noun, expressing, not a
quality, but the action of a verb, as Swahili
ki-lio, ' weeping,' from lia ; Nyanja chi-funo,
' wish,' {romfuna.

In Swahili, a great many nouns in ki- are

56 THE NOUN-CLASSES

diminutives, hi- having taken the place of the
thirteenth prefix ka-, as it has also done in
Kongo.1 In Nyanja, curiously enough, the old
augmentative class (20) has become merged in
the seventh, exchanging its prefix yu for chi-.

On the other hand, nouns which have
dropped their prefixes, and whose stems begin
with chi, ki, ski, etc., are liable to be mistaken
for nouns of the seventh class. These are
fairly numerous in Nyanja, as chinga, ' fence/
pi. ma-chinga, chindu, l roof,' pi. ma-chindu.

This class also contains some nouns meaning
persons. In Zulu these are derived from
verbs as isi-hambi, l a traveller,' from hamba,
1 go,' isi-gijimi, ' runner,' homgijima, ' run.' In
Swahili we find in this class personal nouns
implying some defect, as ki-pofu, ' a blind
person,' kl-ziwi, ' a dumb person,' etc. Pro-
bably these originally belonged to a
' depreciative class,' which will be mentioned
later on.

The ninth prefix is usually found as in- or «- :
in Ronga yin-, Ganda, en-, Herero on-, Makua
usually i- without the n, as i-kuo, 'cloth,' which

In Kongo, however, diminutives are distinguished
from the nouns in ki- by reduplication of the root, as
ki-mwanamwana, ' a little child.'

THE NOUN-CLASSES 57

is equivalent to the Swahili or Yao nguo.1 n
becomes m before stems beginning with a
labial, as im-buzi (Zulu), 'goat,' for in-\-buzi.

The tenth class properly has the plural
prefix in addition to the singular (Zulu izi-n-
dhlu, izl-m-buzi), but the former often disap-
pears through contraction, so that singular and
plural are alike, as Kongo nzo, Swahili and
Nyanja, nyumba, ' house.' Some dialects of
Nyanja, however, have zi-nyumba 10, and
Herero, ondyuo 9, odho-ndyuo* 10, Ronga yin-dlu
9, tiyin-dlu 10. Ronga sometimes contracts
both prefixes, as ndlebe, * ear ' (Zulu in-dhlebt),
plural tin-dlebe, and may drop even the 11, as
in huku, ' fowl,' plural tihuku. Chwana Sutu
and Venda have the tenth prefix li : pull, ' goat,'
pi. li-puli ; k\omo 'cow,' li-kypmo. This
suggests that the primitive form may have
been li-ni-.

1 Both in Chwana and Makua there is a tendency to
substitute the voiceless stop for the voiced stop and
nasal (i.e., p for tub, t for nd, k for ng). Cf . pttli,
'goat,' Nyanja, inbuzi ; itotwa (Makua) 'star,' Yao,
ndoudwa.

a Dy is probably the nearest equivalent to this sound
in ordinary English spelling, but it may be that
represented in the International Phonetic Alphabet by
inverted f. Instead of 2, Herero has the sound of th in
' there ' — here represented by dh.

58 THE NOUN-CLASSES

So far as any rule can be laid down for them,
most names of animals belong to the ninth
class. Many of these are the same through-
out the greater number of the Bantu lan-
guages. We give a few of the most striking
examples : ^

Cow or ox : Zulu, inkomo, Chwana, kxoino,
Ronga, homily Herero, ottgombe, Nyanja (and
many other languages), ng'ombe,1 Bobangi
(Middle Congo), ngombo. Even where the
root is different, as Ganda ente, the word still
belongs to the same class.

Goat : Zulu, imbuzi, Ronga, mbuti, Chwana,
puli, Nyanja (and many others), mbuzi,
Bobangi, mboli, Duala, mbodi, Herero, ongombo
(a different root), Kongo, nkombo, and in some
dialects, ntaba.

Sheep : Zulu, imvu, Chwana, nku, Nyanja,
nkosa, Swahili, kondoo, Giryama, ng'ondzi,
Ganda, endiga.

The dog is usually mbwa or imbwa — quite
recognisable as representing his bark, — but in

1 Ng' thus written has the same sound as in ' ring,'
* sing,' etc. Meinhof writes it n. Where no apostrophe
follows, the sound is that heard in ' finger.' The former
does not seem to occur in Herero or Pokomo. In the
latter language ' cow ' is ngombe, which would be
ridiculed as a mispronunciation by Swahilis or Anyanja.

THE NOUN-CLASSES 59

Zulu this has become inja, because w cannot
in that language follow another-nasal: ^td^i^

The eleventh prefix is found in most present-
day Bantu languages, though its presence is
frequently disguised by the fact that it has
been contracted into u- ; as Zulu u-ti for ulu-
ti. Here it is still easily distinguishable by
its concord — when we hear uti olude Iwa leyo
'nkosi, ' the long staff of that chief,' — we know
that uti cannot be one of the first-class nouns
which have u- for their prefix (u-baba, etc.).
Moreover, when used as proper names, (as
any noun may be in Zulu, keeping its own
prefix, and changing only the initial vowel),
they appear in the uncontracted form ; thus
we have uLuzipo, a man's name, from u-zipo,
' a claw,' uLutulii from u-tuli, ' dust.'

In Swahili, the distinctive concord is no
longer seen, the pronouns, etc. being identical
with those of 3 : uti m-refu u-me-anguka, ' the
long stick has fallen.' Moreover, 14 has
undergone a like contraction, so that these two
classes are merged in one. The /^-class still
exist separately in Yao and Ila : in Nyanja
there are a good many nouns beginning with
this prefix, but they are treated as belonging
to the fifth. In Ronga, though still distinct, it

60 THE NOUN-CLASSES

has modified its prefix to H-, while the fifth
prefix is ri-. As will have been noticed in
Herero, some languages sound this prefix with
r rather than /. Sometimes it is used (instead
of 7) to express language : Lu-ganda, Lu-nyoro,
Lu-gisu : this seems to be confined to the
northern region of the great lakes.

There is considerable divergence in the
plural prefix appropriated to this class. The
most usual appears to be the tenth, but some-
times we have the sixth. In Herero, Ndonga,
Kongo, it has retained its original one, the
twelfth (tu-)t which also belongs to the diminu-
tive class. Sometimes we find that, as in 10,
the plural prefix is added without rejecting
the singular ; thus in Swahili, uti makes
nyuti = 111 + uti ; in Mbundu (Angola), we
have lubango, ' stick,' — pi. ma-lu-bango1.

The twelfth and thirteenth classes have
dropped out in a good many languages. Ka- is
properly the sign of the diminutive, and is still
so used in Herero, Nyanja, Yao, I la, Ganda,
Pokomo, Giryama, Zigula, etc.

As we have seen, Swahili and some others
have substituted the seventh prefix for it.

1 Meinhof thinks this was at first the rule with all the
classes and that the second prefix gradually dropped out.

THE NOUN-CLASSES 61

Zulu, Ronga, Chwana, and their cognate
dialects have got rid of it altogether, and
express their diminutive by a suffix — perhaps
under the influence of the Hottentot language.
E. g., Zulu, indhlwana (iudhlu-ana), 'a little
house ' ; Chwana pitsa, ' pot,' pits-ana, ' little
pot.'

Though hi- seems to be the original plural
prefix attached to the thirteenth class, a variety
of others have taken its place, and the learner
should be prepared to meet with pi- (a form of
8), u-, bn-j (14, as in Herero), ma-, vi-t and tin-

(10).

In Ganda, where the plural of aka-ntu, ' a
little thing,' is obu-ntu, the twelfth prefix has a
distinct and very curious use — otu-dzi, l a
single drop of water,' from ama-dzi, tu-nyu
1 a little salt,' from mn-nyu.1 That is, it is
used to individualise a single particle of
something which has to be looked at or
handled in the mass, like liquids, flour, grain,
etc. These, as we have seen, when belonging
to Class 6, have no singular; when of Class
14, they have no plural — or, it would be more
correct to say, they have neither singular nor
plural. This formation is also found in
1 Conversely, lu-nyu ' a lot of salt.'

62 THE NOUN-CLASSES

Pogoro, a language spoken east of the northern
end of Lake Nyasa.

In Nyanja, the diminutive plural prefix is
ti- (which, Meinhof thinks, may have been a
distinct class) : ka-mbalame, ' little bird,' plural,
ti-mbalame. ••

Duala has a different diminutive class,
traces of which occur elsewhere, but only in the
western part of the continent : it is the one
numbered 19. Its prefix i- is supposed to
have been originally pi-, and its plural prefix
is the twelfth, which has here assumed the
form lo- : i-scru, a small kind of antelope — pi.
lo-seru. In Kongo the prefix has assumed
the form fi- and constitutes an additional
diminutive class, which, however, has no
plural of its own. This class is also found in
Nywema and some other languages of the
central regions between the Great Lakes and
the Upper Congo.

The fourteenth class, as already stated,
includes abstract nouns (e.g., Zulu ubu-ntu,
Swahili u-tu, etc., ' human nature '), and names
of materials which are not properly speaking
either singular or plural : nbu-si (Zulu), u-chi
(Nyanja), 'honey,' utshwala for ubwala =
nbu-ala (Z.), 'beer,' utshane for ubu-ane (Z.),

THE NOUN-CLASSES 63

'grass,' ub-oya (Z.), nbwea (Ny.), 'hair of
an animal.' If nouns of this class are ever
put into the plural, they usually take the sixth
prefix (ma): thus 'night,' itbu-suku, has no
plural in Zulu, but the Chwana bo-sixp has the
plural ma-si\o.

But there seems reason to suspect that
most fourteenth class nouns which take a
plural do not originally belong to the class at
all, as they denote concrete objects which
there seems no reason for including here.
Probably the same thing has happened as
with the ki- and u- classes in Swahili. Uta,
' bow ' — originally bu-ta or vu-ta — is found in
nearly every Bantu language. In Nyanja it
has the concord bu, which stands almost
alone : uta bu-funa kutyoka, i the bow is likely
to break' (lit. * wants to break '). The same
is the case with boa, ' mushrooms,' said in
Scott's Dictionary to be of Class I (no
explanation is offered of the anomalous con-
cord, but the existence of the bu- class in other
languages makes it quite clear) : boa bwanga
bu-li apa (li is here the verb ' to be ') — ' my
mushrooms are here.1

Then we have ba'ato, ' canoe,' stated to be
of the fifth class, with plural ma-bwato ; but

64 THE NOUN-CLASSES

forms like Pokomo waho (u-aho), pi. m-aho,
seem to show that bw- is the prefix (bu-ato).
Nyamwezi has vato 14 and Konde ubwato
(ubu-ato). This last has its plural of Class 4 :
imyato (imi-ato), which may point to a still
further confusion. The little group of north-
western Bantu languages, of which Duala is
the chief (Benga, Dikele, Lsubu), make the
plural of the fourteenth class in mi-: the same
word is in Benga bwalu, mi-alu, in Duala
b-olo, mi-olo. Duala has bw-ele, ' tree,' pi.
mi-ele : a curious exception, as regards the
singular.

These last can scarcely be explained as
collectives, or nouns of material, though
* mushrooms ' might : being considered in the
first instance as food they would be thought
of by the basketful. Even as to uta I am not
quite clear : uta in Swahili seems to mean a
bow with quiver and arrows all complete (the
' artillery ' that Jonathan carried), while upindi
is a bow pure and simple. The plurals nyuta
and mata are given in the dictionaries, but
they may be later formations. In Nyanja the
plural mauta apparently has the same collective
sense — at least, as applied to the three stars
in Orion's belt, I understand it to mean ' The

THE NOUN-CLASSES 65

Bow and Arrows/ not ' The Bows.' Ubu-so
(Zulu), ' face,' is found almost everywhere and
has the same root as i-so, ' eye ' — as though it
were a sort of abstraction : the ' eyeness,' if
one might coin the word. In fact one does
sometimes hear ' eyes ' used to mean ' face,' in
Nyanja and probably elsewhere.

The fifteenth prefix, £?/-, denoting infinitives
of nouns, offers little difficulty and has the same
form (except for the occasional presence of an
initial vowel) almost everywhere. In Chwana
it is modified to x°> and several western
languages are without it, though it has left
traces in Kongo.

There are a few nouns in ku with plurals in
ma-, which are certainly not infinitives and
are perhaps remnants of a lost class. They
sometimes retain the ku- in the plural, after
the ma- prefix : Nyanja kutu, l ear,' pi.
ma-ku-tu ; but Herero oku-twi, oma-twi, Kongo
ku-tu, ma-iu. Several of the western Bantu
languages have the words for ' arm ' and ' leg '
similarly formed, and Herero has a Tew more
besides. Meinhof thinks these are locatives,
a theory which will be more fully explained in
the next chapter.

Classes 16, 17, 18 are not found in Zulu or

66 THE NOUN-CLASSES

Chwana, though slight traces of them exist.
They can better be discussed in the next
chapter, in connection with the locative ; here
it may be sufficient to say that they differ
from most other classes by having a movable
prefix : a preposition is added before the usual
prefix and entirely changes the concord. The
prepositioji and its noun are treated as a single
word. It is somewhat as though, instead of
saying 'The house is near a field,' we said
' Near-trie-house n-is a field,' or ' By-the-house
6-is,' etc.

The locative prefixes have entirely dis-
appeared in Swahili, but the concord remains,
as we shall see.

The nineteenth class, already referred to,
was not recognized by Bleek.

The remaining two classes — more properly
four — were also unknown to Bleek and only
survive in a rudimentary condition. Ganda
has a prefix ogu-, which seems to convey a
notion (i) of size, (2) of depreciation : ogu-ntu,
' a clumsy thing,' pi. aga-ntu, ogu-nyo, ' a large
quantity of salt ' (omu-nyo), as opposed to ' a
small quantity ' of the same, otu-nyo. In Gisu
there is a class denoting large things, with the
prefix gu- in the singular, gi-mi- in the plural :

THE NOUN-CLASSES 67

gu-koko, 'a giant fowl,' pi. gimi-koko, with
a prefix now virtually equivalent to the
fourth. It might be preferable therefore to
enumerate them thus : 2.0 yu 21 ya 22 yu 23
yi (yimi}.

But as. a matter of fact, no language to-day
seems to have kept both 20 and 22 inde-
pendently. Both classes survive in Swahili,
though one is completely merged in the fifth
and sixth, and the other would be but for its
anomalous plural. To express unusual size,
a noun loses its original prefix in the singular
and takes ma in the plural, as dege, ' a large
bird/ (from n-dege), pi. ma-dege. If the stem
begins with a vowel it prefixes j, zsjoka, ' large
serpent,' from nyoka. But if it is implied that
a thing is not only large but monstrous, or
ill-conditioned, the form is the same for the
singular, but the plural has the prefix mi-: as
jombo (from ch-ombo), 'a big ugly vessel,' pi.
mi-jombo, vua, ' heavy rain ' (implying some-
thing abnormal), pi. mi-vua. (Some of these
forms in mi- have no singular corresponding
to them). But these distinctions are becoming
blurred, and a further source of confusion is
the insertion of -ji- between the prefix and
stem of monosyllabic diminutives. We may

68 THE NOUN-CLASSES

add that ki- sometimes has a depreciatory

sense.1

It has already been pointed out that some
seventh class nouns in Nyanja (beginning with
chi-} are really relics of an old augmentative
class. In Kinga (spoken among the moun-
tains E. of the north end of Lake Nyasa)
there is a class with the prefix ugu- which
does not seem to have a plural and conveys
either an augmentative or a depreciative
sense : it is no doubt an amalgamation of 20
and 22. Traces of the same are also found
in Venda. Ganda uses both the ki- and lu-
prefixes with augmentative force. There is no
need to follow out the matter any further, but
the student should keep it in mind as a
possible explanation of seemingly exceptional
forms.

Hints of several other classes are found :

(1) We have already said that Meinhof
considers the Nyanja diminutive plural //- as
a distinct prefix, parallel with tu-, as li- is with
lu-

(2) There are in Nyanja one or two words,

Steere's Handbook of Swcftiili, p. 19, ' ki-buzi, a
poor little goat." We have already referred to the
depreciatory sense of ki- when applied t« persons, as
kipofu, kiziwi, etc. „

THE NOUN-CLASSES 69

tulo, ' sleep,' tubsi, ' dung,' which seem to stand
in a class by themselves. The concords
(lido t-ambiri, 'much sleep,' tulo t-ache, 'his
sleep,') indicate that tu- is the prefix. We
find.otu-lo, ' sleep/ in Ganda, but it does not
seem possible to place it, as the French
Fathers1 do, in the tu- class which denotes
1 small quantities.' However, they may have
been originally thought of as plurals and, as
such, would belong to Cl. 12, like tuvia ' fire '
(Kongo) which, though generally used in the
plural, seems to have a singular luvia.

(3) Some languages have personal nouns in
ka- which are not diminutive — this is especially
marked in Herero — and we might compare
Ganda ka-baka, ' king,' ka-tikiro, ' prime
minister.' And we might suggest the large
number of animal names in Nyanja, which
begin with nci and nanka (nadzikanibe, ' chame-
leon,'nakodzwe. 'water-buck,' nankabai, ' hawk ').
But enough has been said to show that the
number of classes was once probably far
greater than it is now, and to show that in
languages not yet fully studied we are quite
likely to come upon traces of extra classes.

1 Manuel de Langue Luganda. Einsiedeln, 1894,
p. 31.

CHAPTER V
CASES : THE LOCATIVE

•

IF I begin by saying that Bantu nouns have
nothing which can, properly speaking, be
described as case, it will appear as if this
chapter, being of the same kind as the famous
one on * Snakes in Iceland,' had better be left
unwritten.

However, as we have already seen, there is
a Possessive — if of a somewhat peculiar
character. There is no difference in form
between the noun-subject and the noun-
object, but some pronouns have distinct forms
for the accusative, as we shall see in the next
chapter. There is something like a vocative.
Perhaps the dropping of the initial vowel in
Zulu, as ' Zatshttke,'' when addressing a man
(instead of ' u Zatshuke ') is too slight to be
mentioned in this connection ; but Chwana

CASES: THE LOCATIVE 71

(at least in some dialects) has a different
terminal vowel for a noun, according as the
person referred to is spoken to or spoken of.
And Duala prefixes a to nouns in the Vocative.
Finally, the Locative in -ni, though confined
to a comparatively small number of languages,
is a feature which must be taken into account,
and it can hardly, for the purposes of this
sketch, be classed under any other heading
than that of Case.

If we limit the term 'case' to those
relations which are expressed by inflexions of
the noun-stem (declensional endings), we shall
have to admit that English nouns are entirely
without it, except when the possessive is
indicated by 's. Case-endings are becoming
obsolete in Dutch, though they still exist in
German ; they are better exemplified in Latin
and Greek, and still more so in Sanscrit,
which has seven cases.

The Latin declension, for instance (mensae,
' of the table,' rei, ' to the thing,' horto, ' from
the garden '), indicates by means of the
termination what we express by a preposition,
or by the order of words in the sentence,
which is our only way of distinguishing sub-
ject from object. The same relations are

72 CASES: THE LOCATIVE

expressed in the Bantu languages by means of
prepositions.1

Indo-European prepositions are invariable.
1 From,' ' to,' ' by,' ' with,' never change their
shapes, whatever nouns they may precede or
follow ; and neither do the Latin ab? de, ex,
pro, super, etc.

This is not the case with Bantu prepositions,
though the difference is perhaps more apparent
than real. We have already seen, in the
second chapter, that the equivalent for ' of '
assumes different forms according to the noun
it follows. It is as though we said in English :
1 the house Aof the man,' ' the child c/zof the
house,' ' the door dot the room.' This is
because the initial of the noun-prefix is
combined with the root -#, which, whatever
its original force may be — we can for practical
purposes assume to mean ' of.'

We can now see why Brusciotto called this
wa, ya, etc., an article. He saw that one part

This word is used for convenience. Meinhof points
out that there are, strictly speaking, no prepositions in
Bantu, the words serving as such being really pronouns
or possessive particles — except na 'with,' which might
equally well be rendered and,' and called a conjunction.

The alternative forms a, ab, e, ex are not variations
in the sense here intended.

CASES: THE LOCATIVE 73

of it meant ' of,' and — reminded of the way in
which di (' of ') combines with the article in
Italian (del, dello, delta, dei, degli, delle)1 con-
cluded that the other part of the word might
be an article. The combination seemed to
carry out that reversal of European rules
which had struck him so forcibly in connection
with the prefixes, and the conclusion he came
to was a very natural one under the circum-
stances and does credit to his linguistic
insight. He might have represented the
whole thing in a diagram, thus :

D-EL

W-A

— the Italian article il having originally been
the Latin demonstrative pronoun ille, while u
(which becomes w before a] is the prefix-
pronoun of its class. With nouns of other
classes, we should have ya, la, za, etc., just as
in Italian with a feminine singular noun we

This combination (the ' partitive article') is better
seen in Italian than in either French or Spanish, where
it is not carried through consistently (dn, de la, des ;
del, de la, de los). In Italian too, not only di and a,
but the other prepositions con, in, per, are combined
with the article and constitute something like a declension.

74 CASES: THE LOCATIVE

have dflla, with a masculine plural del or
degli, and so on. But wa fulfils no function
of the article as we understand it, the pro-
noun in it being purely representative and
not demonstrative.

This possessive particle is closely connected
with the possessive pronoun, to be considered in
a later chapter. But I mention it here,
because the next point can only be illustrated
by the help of possessive pronouns in European
languages. In English, as we know, possessive
pronouns are among the few parts of speech
which have grammatical gender — which show
by their form the sex of the nouns they
represent — or the absence of it. Possessive
pronouns agree in this respect with the noun
which stands for the possessor : ' his mother,'
( her father,' — while the reverse is the case in
French : son pere, sa mere, leave the sex of the
possessor quite uncertain, though beginners
insist on translating ' her father ' by sa pere.

Bantu nouns follow, with a difference, the
French principle ; ' the man's child ' is in
Zulu : um-ntwana wendoda (for wa-indoda) and
not, as it would be if the concord followed the
class of the possessor : um-ntwana yendoda
(ya-indoda}.

CASES: THE LOCATIVE 75

I say ' nouns ' advisedly, for we shall see,
when we come to treat of them in the proper
place, that possessive pronouns agree both
ways : that is, the first part follows the class
of the thing possessed, the last that of the
possessor.

It is scarcely necessary to say, after giving
the above examples, that the thing possessed
always precedes the possessor, as in French
(I'tnfant de Vhommt). In English, we have it
both ways, according as we use the inflected
possessive or not — ' the man's child,' and ' the
child of the man.' The Sudan languages1
put the possessor first and say ' man child ' —
having nothing corresponding to article or
inflection, though sometimes a particle
indicating ownership is suffixed to the first
word.

There is no way of showing whether a noun
is subject or object except by its position in
the sentence, the subject coming before the
verb and the object after,2 just as in English.

1 See Language Families of Africa, p. 40.

a Arabic, which uses case-inflections, usually puts the
verb first, the subject next, and then the object; but as
the two latter are sufficiently distinguished by their
endings, it really does not matter in what order they are
put.

76 CASES : THE LOCATIVE

Sometimes, where there can be no possibility
of mistake, inversion is used for the* sake of
emphasis — as by us in rhetoric or poetry ; but
the outsider had better not meddle with
figures of speech such as this, and it is scarcely
necessary to mention them in a general
outline.

We now come to the Locative ; and this
has to be treated under two different headings :
the locative formed by prefixed prepositions
and the suffixed locative in -ni, which seems
to be a later development. There are not
many Bantu prepositions, as we shall see in
the chapter on Particles : the principal ones,
which (or, at any rate, traces of them), are
found in every Bantu language, are pa, ku,
mu.

Pa, roughly speaking, conveys the notion of
' at ' or ' upon ' ; mu, of ' in,' and kit, of
' motion to and from,' though it sometimes
has the meaning of ' outside.' It also serves
as the sign of the infinitive ; but here its
function, if not its origin, is different, so that
we are quite right in treating 15 and 16 as
separate classes.1

Meinhof thinks that the use of ku as infinitive prefix
was a later development from its locative function.

CASES : THE LOCATIVE 77

Pa, ku, mu are the prefixes of Classes
16, 17, 18 respectively; but they differ in one
important respect from the other prefixes.
'These, as a rule (with the exception, in some
cases, of augmentatives and diminutives), are
attached to the bare root, which cannot be
used without them, while pa, ku, mu are
prefixed to the whole noun, root and all, as in
Nyanja pa-chilindo, ' at the look-out,'1 ku-
chilindo, mu chilindo ; not pa-Undo, ku-lindo,
mu-lindo. But in all other respects they are
true prefixes and take their own concord,
entirely superseding the one properly belonging
to their noun when used by itself. For
instance, chilindo, being a noun of the seventh
class, would take the possessive particle, cha
and pronoun chi :

Chi-lindo ch-a Pembweka chi-ri cha-bwino :
(The) watch-hut of Pembereka it is good.

But Pa-chilindo pa Pembereka pa-li pa-bwino
would mean, ' At Pembereka's watch-hut it is
(a) good (place).'

So we may have also :

1 Chilindo, also called nsanja ('staging' or ' plat-
form ') is a small temporary structure raised on poles,
open or roofed, erected in the fields so that watchers
can overlook the ripening crops and scare away birds,
monkeys or other depredators.

78 CASES: THE LOCATIVE

Ku-chilindo kwa Pembereka ku-li kwa-bwino,
with much the same meaning as the last
sentence, except that the place is thought of
from the standpoint of one who is at a distance
from it and going, or thinking of going, towards
it ; while m'chilindo (for mu-chilindo) mwa
Pembcreka mu-li mwa-bwino, means : ' the
inside of Pembereka's watch-hut is good.'

Some nouns, as kamwa, * mouth/ are never
found without the locative prefix — we have
pa-kamwa, ku-kamwa and m-kamwa, but never
kamwa alone. The word is evidently connected
with mwa ( drink,' — perhaps ka is the prefix of
a lost class distinct from the diminutive.

Here are some further illustrations, also
from Nyanja :

A -li-ku-nka ku munda kwa Champiti : he is
going to (the) garden of Champiti.

Ku-mudzi kwanu ku-li kwa-bwijio : at our
village it is good.

Ku-Mlanje ku-li mpunga, koma ku-KabuU
kuno ku-libe : at Mlanje there is rice but
at Kabula here there is none.

Kuno, l here,' is really a demonstrative,
agreeing not with ' Kabula,1 but with ' ku-
Kabula,' as though, instead of ' here at
Kabula,' one had said * at this at-Kabula.'

CASES: THE LOCATIVE 79

A mistake to be found in some of the older
grammars is illustrated in Steere's Handbook
of Swahili (p. 22) by the statement that a class
(the seventh, in his arrangement) contains ' the
' one word mahali, place or places, which re-
1 quires special forms in all adjectives and pro-
' nouns.' Mahali is really a borrowed Arabic
word, which is sometimes Bantuized (oftener,
I think, at Zanzibar than at Mombasa) by
substituting pa for the first syllable, as if the
latter were a removable prefix, and so making
it into a noun of the sixteenth class. Steere
mentions the locative concord a little later on,
but does not call it a class (though including
it in his ' Table of Concords '). The change
undergone by the Swahili locative (which we
shall discuss presently) has so obscured the
relation between it and mahali or pahali that
without a fuller comparative study than was
possible when Steere wrote, it was not likely
to be perceived.

Several languages have a word for * place *
which is either pa-ntu or some cognate form
and may have the same root as mu-ntu. In
Ganda, where primitive Bantu p becomes wt
we have the nearly obsolete wantu, which was
at first thought to constitute a class by

80 CASES: THE LOCATIVE

itself.1 But wa, as well as mu and ku, is prefixed
to other nouns, which are also used as adverbs:
wa-nsi, ' the ground ' or ' below ' ; mu-nda,
' the inside,' ' or within.' There is also a
prefix e, not generally found among the
locatives, but which may possibly have some
connection with the peculiar Zulu form.

In Pokomo we havefantu (bilabial/, for/>),
in Giryama hatu, elsewhere hantu, handu, etc.
(A great many East African languages
substitute h for />, except in certain particular
cases : e.g., Giryama has hendza, ' love,' for
Swahili penda.)

Some languages have this word also with
the prefix kit : kutitu, kundu ; but mu-ntu,
mu-ndu, if used at all is less common, perhaps
because it would have the same form as the
word for ' person.'8

' The tenth class contains the single word wantu,
"place"; this word is obsolete except in the single
expression buliwantu, " everywhere." But its influence
in the language is great, because adjectives, pronouns,
verbs and adverbs are all formed with the prefix wa-,
referring to this disused word, e.g., wano wa-lungu,
"this is a pretty spot." ' — Pilkington. Wano walttngu
exactly corresponds with Nyanja pano (pa It) pabwino.

In Zigula 'the prefix appropriate to the word hantu

place " is ha, and as it belongs exclusively to this word,

its mere presence is sufficient to indicate that place is the

CASES: THE LOCATIVE 81

Kongo has vuma (v for p), kuma, ultima, as
three different forms of the word for ' place.'
These take the locative concords as we find
them elsewhere, but other nouns preceded
by the corresponding prepositions keep their
concords instead of taking a locative concord.
Muma seems sometimes to be treated as the
plural of vuma : this may arise from a confu-
sion of the m- with the sixth prefix (as though
it were contracted from ma-tima) and possibly
indicates that the whole system is dying out in
some of the western Bantu languages. Bentley
remarks concerning vuma : ' In most Bantu
' languages a corresponding word for place
* will be found standing in a separate class of
' its own, and wearing a prefix va, pa, or
' something similar.' This is the same mistake
adverted to just now, and no doubt one reason
for it is that the prefixes of these words are
not, as a rule, removable ; but they really
come under the same heading as the locatives

substantive referred to, so that the word hantu is often
omitted. The same rule refers to the more indefinite
ku-ntu and its appropriate syllable ku-, and also to
tun -nt u, ;;m-.' — Kisbey.

Archdeacon Woodward, though mentioning the inn-
concord, does not refer to the word niu-ntn, ' in a place,'
which, however, certainly occurs in the Likoma dialect
of Nyanja.

82 CASES: THE LOCATIVE

which are made up as wanted and simply put
the preposition before the ordinary noun-
prefix. We draw a distinction between the
two — but the Bantu speaker feels none.

We now come to the suffixed locative in
-ni, which seems to be confined to Bleek's
South-Eastern Branch of the Bantu family
and to a few languages in East Africa, of
which Swahili, Bondei, Hehe, and Makua are
the chief, if not the only ones — and in Makua
it is combined with the prefixed locative —
mashi = water; va-mashi-ni = a.t the water;
m-mashi-ni = m the water.

But the prefix has quite disappeared- in
Swahili, and the locative is simply formed by
suffixing -ni, which may mean ' in,' ' at,' ' on,'
' from,' ' to,' etc.

nyumba-ni = in the house. The accent is always

tnji-nt = to the town. shifted forward by the

kiti-ni — on the chair, locative suffix, seep. 15.
etc.

But the concord differs according to what
is implied in the suffix. ' In my house,' is
nyumbani mwangu, ' to my house,' nyumbani
kwangu ; ' he is sitting on my chair,' anakaa
kltini pangu, and so on.

CASES: THE LOCATIVE 83

A-li-anguka mlangoni pangu.
He fell down at my door.
Atakwenda shambani kwake.
He will go to his plantation.
Amelala nyunibani tnwako.
He is lying down in your house.

This concord is not found in Zulu, where,
however, the rules for applying the suffix are
not quite so simple as in Svvahili. We may
suppose that it originally had the form -ini :
this is rendered probable by the effect it has
on the final vowel of the noun. A become^,
e and i remain unchanged ; o becomes we, and
u, .wi (or sometimes i, eliding the final vowel
instead of changing it into w). The initial
vowel is changed into e (in a few cases, in 1 1
and 14, into o), fop which I can suggest no
reason, though it may possibly have some
connection with the locative prefix which
appears in Ganda as e-.

intaba — mountain makes entabeni.

ixwe = country ,, ezweni.

isi-hlalo = seat ,, esihlalweni.

in-dhlu = house ,, endhlivini or end-Mini.

In Chwana this locative is found in the
form ng, as in the well-known place-names
Mafeking and Shoshong. The same termina-
tion is used to form a kind of participle, to

84 CASES: THE LOCATIVE

which we shall refer later on, in connection
with the relative pronoun and the verb. In
Ronga and other languages of the Delagoa
Bay district, the locative suffix is -ni, though
the final i is frequently dropped : tikwen, ' in
the country, ' from tiko.

The preposition ku still exists in all these
languages (in Chwana under the form ^o), but
pa and mu are no longer found independently.
That they did exist in Zulu is shown by the
adverbs pansi1, pezulu (= pa + izulu — ' on the
sky' = 'above'), pa-kati, 'in the midst,'
pa-ndhle, ' outside,' etc. When used as pre-
positions these are followed by kwa, as pakati
kwcndhlu (kwa-indhlu), ' inside the house,' —
which, properly, should only go with ku- but
has quite usurped the place of pa- and mu-.
The latter, however, survives in Chwana, in
combination with the sufHx, as mo motseng,
1 in the village ' (motse).

Some nouns whose presence in the ku- class

1 -nsi is no longer used in Zulu as a noun, but it is
found in Swahili as nti, nchi, meaning ' land,' ' earth.'
Instead of the adverb pa-nsi, ' down,' below,' ti-ni or
chi-ni is used. In Chwana the root and prefix (ha-tse)
have become so closely welded together that they are
looked on as inseparable and have been given a fresh
prefix, le-hatse 5.

CASES: THE LOCATIVE 85

is difficult to understand and is rendered still
more perplexing by the fact that they have a
plural in ma-, are to be accounted for as
Locatives. The Nyanja kul'u, ' ear,' pi.
ma-kutu, might be taken if it stood alone as a
fifth-class noun which lias dropped its prefix
li- ; but it cannot be dissociated from the
Herero oku-twi, pi. ouia-twi, and Ndonga
oko-tshwi, pi. oma-kotshwi. Properly, the
word means ' to the ear,' ' the place of the
ear,' and then, the word being generally used
with the locative prefix, the separate nature of
the latter was forgotten, as was the case with
the Chwana hatse. The same applies to the
Herero oku-oko, 'arm,' pi. oma-oko ; Kongo
k-oko, pi. m-oko. Other examples in Herero
are ohu-rama, ' leg ' ; oku-ti, ' veld, open
country,' oku-ruo, ' hearth,' oku-apa, 'armpit.'
It is evident that all these may have a locative
sense, and that, wherever similar words are
found, they should be placed in Class 17.
The uncertainty about the prefixes shows
that their original meaning is almost, if not
quite forgotten, and that they tend more and
more to be regarded as part of the stem.

CHAPTER VI

THE PRONOUN

THE Pronoun is one of the most important
features of the Bantu Languages. I do not
even add ' next to the Prefixes,' because the
two are so intimately associated that it is
difficult to say which should have the priority.

It used to be thought that the Pronouns and
Prefixes were, in the last resort, identical,
whatever their origin — whether they were
nouns which had lost their distinctive
character and become mere formative elements
(like the suffixes of ' king-dom,' ' man-hood,'
1 lord-ship,' which have long ceased to have
any independent value) — or whether they
already, even as separate words, had the force
of pronouns. (Bleek's discussion of this
subject — Comparative Grammar, pp. 123-131 —
should be carefully read, though time has
shown that it requires some modification.)

Meinhof (Grundziige, p. 35) points out that

THE PRONOUN 87

the prefix and pronoun are not identical, but
that the latter is really the demonstrative
particle, discussed in Chapter III., which
became incorporated with the prefix and was
then, in many cases, lost, but survives, e.g. in
Zulu, as the 'initial vowel.' It was, as we
have seen, originally yd, but assimilated its
vowel to that of the prefix which followed it.

Thus we have :

3 ya + mu = yumu = umu, and the pronoun yn=gu,u

4 ya+mi = yimi, =imi „ ,, ,, yi = gi, i

6 ya-\-ma^=yama — ama „ „ ,, ya = ga,ya,a
9

Where the prefix does not begin with a
nasal, e.g., in 2, 5, 7, 8, the demonstrative (or
' article ') disappears without leaving a trace,
and the noun-prefix only remains, to serve as
pronoun — so we get the pronouns ba, ki, li,
etc.1

1 Class 1 has been omitted here because, as Meinhof
says, ' it contains all sorts of irregularities, as is not
surprising in a class so much used.' One would expect
its pronoun to be yu, like that of Class 3, and in fact u
and o, which are found in Zulu and Chwana, evidently
come from that form. But the Swahili yu cannot have
come from it, so Meinhof thinks there must have been a
second pronoun yyu, of which he explains the formation
on p. 36 of the work cited in the text, whence the
substance of this note, and of the paragraphs imme-

88 THE PRONOUN

This simplest form of the pronoun is never
— or, if at all, very seldom — used by itself,
but is always attached to a verb, or to an
adjective with the verb ' to be ' understood.
We shall refer to it, henceforward, as the
' Inseparable Pronoun.' Its use as the
' copula ' — where, in some cases, it assumes a
different form — will be discussed in the next
chapter.

The following Table shows. the ' Inseparable
Pronoun ' in our eight typical languages, as
compared with the noun-prefixes. Some
languages — especially Duala and Kongo — tend
to use the unaltered prefixes of all classes
before verbs by a merely mechanical repetition,
having lost their feeling for the pronoun
as such. This explains the occurrence of
such forms as mi- and ma-, where we should
expect i- and a-.

The pronouns of the first and second

diately preceding it in the text, is taken. As for the
alternative pronoun a, which has almost displaced yu in
Swahili, it is no doubt a remnant of the unaltered ya ;
but it is not clear why it did not assimilate its vowel to
the first prefix, or why, in Herero and Chwana, for
instance, a is sometimes used in dependent sentences,
while in principal sentences we have u or its equivalent.
This last fact may possibly be connected with the
function of a as relative particle.

THE PRONOUN 89

persons have, of course, no noun-prefixes
corresponding to them and stand outside the
framework of the classes. They are therefore
placed by themselves at the head of each
column in the Table.

The use of these pronouns is illustrated in
Chapter II. They are used both as subject
and object ; the first class and the second
person, singular and plural, are in almost all
cases the only ones with separate objective
forms. And, even of these, the second person-
plural is not common. (Swahili has -wa-,
Zigula, -;;//-, Ganda, -ba-). Whatever other
particles are prefixed to the verb-root, the
object-pronoun must always come next it.

E.g., u-ya-ngi-bona (Zulu), 'he sees me';
wa-ku-tskaya (Zulu), 'he beat thee ' ; a-li-m~
fundisha (Swahili), 'he taught him.'

Duala and Kongo have no object-pronoun
before the verb — though it exists in some
cognate dialects such as Isubu. The object
is expressed by a separable pronoun following
the verb.

The reflexive pronoun, which is alike for all
persons and numbers, is placed in the same
position as the object-pronoun. In Zulu it is
zi, as in uku-zi-tanda, * to love one's self ' ; in

s" : £

•° * , •- •- S

A i I B ? g 5 '£ .A' :i = » ' i J

1 1 1

i r

rl ,.'

"A ?

c/3 : ft.

SA aega^-£ccis i il

1 1

3 3

:*s. §

r rt 3 .A i rt .A .A "•; A i <i 3 3

• 4- Ji JM si ji 9 2.. Jk 2^ ^ t jS A Ji

2 |

53 * £

• i 3 i rt

n -2 s i i . , , ,

B « 3 .5 iA ^ .A .A £ -3 A, | rt 3 3

A 1 1

rt
a

*i|

cd*c CkCbOnbo^vO v 's . t* .^ -^ **i

s M a

. ° a)
•— - t-
(/) 0,

3J.3.A rtii ''AAi
E^ BE'aSSS s c 2 3 £ 2 Jx

Ort on) ortuu vvoortO O

0 J! a

S = 1

rt ^ 3 .A S 5. 32 '> .A 'N 3 | I 3 ^i

i J I

.— • u

v> ~ a,

BSE'E i a •£ & A ^ 1 1 4 JB

1 I

"c"

il g

— • " u

, , A 13 .A . .A • i rt 3 3
k t 3.A^5>^ON.AN3 I.M.a.*

rt 3 2

C. ^! B

•S i

5" 1

BASE 1 1 i c c 3 1 ^ 3 J

rt 3 2
3, Jtf H

3 g e

S*= £

3 "liTl'C >> fr > 1 kA "O u 2^! 3 ^

is3

3 u;

S>iE'a S D "^t^ i "9 S S M if . S

oooo uooo oooooo o

1 ' 3

5" o o

rt
c

^~ 8

6*1 ' ' i i i d: ' ^3 o 1 1 ? i ^
EI O v »S II • ^3 fl)T3^^ ' I ^1 "^

1 1 1

A6 ^
55 s fl«

° A ° *> ' * ai .A .A A 1 1 6 P.

a^iaa^ES-5 ^°li>,x.

o 4

S x a

«'S |
S = &•

i ^ i .A = « -a -s .A -s 2. \ \ Z -3

1 I 1

|=1

II 1 1 S 1 i i s i 1 ' 'II

I I 1

g .

in ,. .

S t- 00

w e

THE PRONOUN 91

Nyanja, dzi ; in SwahiliyV; in Herero ri ;
elsewhere i, yi, etc. Meinhof thinks the
original form was yi.

There are longer forms of the pronouns,
which can stand by themselves and need not
be used with the verb except for emphasis.
The different languages form these in various
ways, and in some of them it is difficult to
trace any resemblance to the Inseparable
Pronoun. In fact they are built up rather on
what is called the * prepositional form ' of the
pronouns (though some of them depart
considerably even from this) ,which, accordingly,
it will be better to take first.

This is a form which is suffixed to
prepositions : expressions like ' with me ' ; ' to
him,' etc., being treated as one word. There
is a form for every class — just as there is of
the Inseparable Pronoun — as well as for the
First and Second Persons. Thus we have (in
several languages) nami (=na + mi), 'with
me,' ' and I,' kumi, ' to me,' etc. These forms
are also used in connection with Relative
Pronouns, as we shall see presently.

Like the Inseparable Pronoun they are
never found alone — if not attached to preposi-
tions, they are suffixed to the possessive

92 THE PRONOUN

particle, in a way which will be explained
presently.

The separable, or independent, pronouns
are usually — if not always — built up out of
these forms. They»exist in most languages
for the three persons singular and plural, and,
in some, for all the classes. But some, like
Swahili and Nyanja, have none for any
classes after the first and second, using the
demonstrative pronouns instead.

The Possessive Pronoun consists --of two
parts and has to be considered under two
aspects.

It is made up of :

(1) The possessive particle of the class to which

the thing possessed belongs (wa, ba, ya, etc.),
and

(2) Either (a) a special pronoun-root for the first

and second persons, or (6) the ' prepositional
form ' of pronoun indicating the class of
the possessor. The first class is usually
exceptional in this respect, having a different
suffix for the Possessive.

We will now take the six possessive pronouns
of the three persons singular and plural in the
same eight languages as before. Chwana
seems to be exceptional in having some of the

Zulu Chwana Hereto Nyanja Swahili Ganda Gisu

IstPers.
2nd ,,

Si. PI.
-mi, -ti
-we, -ni

Si. PI.
-no, -ro
-o, -lo

Si. PI.
-ami,-ete
-ove,-ene

Si. PI.
-ne, -fe
-we, -nu

Si. PI.
-mi, -swi
-we,-nyi

Si. PI.
-nge, -fe
we,-mwe

Si. PI.
-se, -fe
wo,-nywe

Class
1

-ye

-e

-ye

-y«

-bo

-vo

-awo

-0

-bo

-we

-wo

-0

-awo

-wo

-wo, -o

-gwo

-kwo

-yo

-avyo

-yo

-gyo

-kyo

-lo

-aro

-lo

-lyo

-wo

-o

-ao

-wo

-go

-ko

-so

-so, -sho

-atyo

-cho

-oho

-kyo

-z«

-cho

-avyo

-zo

-vyo

-byo

-yo

-ayo

-yo

-zo

-cho

-adho

-zo

• zo

-tso

-lo

-arwo

-wo

-Iwo

—

-atwo

-to

-two

—

-ako

-ko

—

-ko

-bo

-vyo, -yo

-awo

-bo

-wo

-bwo

-ko

-%o

-akwo

-ko

-kwo

-apo

-po

-wo

-ho

-akwo

-ko

-kwo

-amo

-mo

-mu

—

This form of pronoun does not appear to be used in Kongo.

SUBSTANTIVE PRONOUNS.

Zulu Chwana Herero Ganda Kongo

IstPers
2nd „

Si. PI.
mina, tina

wetia nina

Si. PI.
nna rona

wona lona

Si. PI.
owami,
owete
ove,
owena

Si. PI.
nze fwe

gwe mwe

Si. PI.
mono y«to

nge yeno

Class
1

yena

ene

eye

yandi

bona

rone

owo

y«u

wona

one

owo

gwe

wau

yona

cone

ovio

eye

miau

lona

yone

oro

lye

dlau

wona

one

owo

mau

sona

shone

otyo

kye

kiau

zona

chone

ovio

bye

yau

yona

cone

oyo

yau

zona

chone

odho

zau

l(w)ona

lone

oruo

Iwe

luau

otuo

two

twau

oko

bona

yone

owo

bwe

waa

kona

Xone

okuo

•kwe

kwau

opo

vau

oko

kwe

kwaii

—

omo

mwe

tnwau

—

fiau

Nyanja, Swahili and Gisu use the Demonstratives for all classai but the first.

Gyanja. 1st Pers Si. ine i2nd iwe 3rd lye

Swahili. ,, mimi ,, wewe ,, yeye

Gisu. ,, is* ,, iwe ,, niye

Nyanja. 1st Pers. PI. ife 2nd Inu 3rd awo

Swahili. ,, sisi ,, ninyi ,, wao

Gisu. ,, ifwe ,, inywe ,, abo

THE PRONOUN

forms reduplicated (-axaxomste3id of -axo=ako).
There are also forms in some dialects which
are simply these separable pronouns with the

possessive particle prefixed to them a rona,

1 our ' (lit. ' of us ') ; -a lona, ' your ' ; -a vona,
1 their.'

Zulu Chwana HereroNyanjaSwahiliGanda Gisu Kongo

-ami

f -ame
1-aka

-andye

-anga

-angu

-ange

-ate

-ame

THY

-ako

-axaxo

-oye

-ako

-o

-owo

-aku

HIS

-ake

-axa*we

-e

-ache

-ake

-e

-ewe

-andi

OUR

-etu

-eshu

-etu -

-atu

-etu

-afwe

-efe

-eto

YOUR

-enu

-eno

-enu

-anu

-enu

-am we

-enywe

-eno

THEIR

-abo

-avo

-awo

-ao

-awe

-au

These, if the thing possessed is of the
first class, have the possessive particle wa
prefixed to them : wami, warne (some Chwana
books*print o ame), w andye, wanga, wangu, etc.

In Zulu, ' my child' is umntwana wami ; in
Nyanja, mwana wanga, and so on. ' My
children,' would be abantwana bami, ana anga ;
1 my village,' umuzi wami, mudzi wanga, ' my
country,' izwe lami, dziko langa. There is no
need to multiply examples.

In all these pronouns, the second part of the

96 THE PRONOUN

word does not vary, but if the possessor is of the
third person and of any class except the first, the
suffix has often to change as well as the
prefix.

In Zulu, ihashi lake is ' his horse,' supposing
that ' his ' represents a noun of the first class—
[ say umu-ntn, um-fanci, etc. But it might stand
I for a fifth-class noun : i-Bunu, ' a Boer,' — or a
seventh : isi-hambi, ' a traveller,' — or a ninth :
in-doda, ( a man ' ; in-kosi, ' a chief.' In these
cases we must say :

His (the Boer's) horse = ihashi lalo.
„ (the traveller's) ,, = ihashi laso.
„ (the chief's) ,, = ihashi layo.

That is, the first part of the word is the
possessive particle agreeing with the thing
possessed, and the second the pronoun
agreeing with the possessor.

In this way, the number of classes multi-
plied by itself will give the number of possible
possessives — or would, if some of the forms
did not coincide, so as to make them less
numerous. There is a neat diagram of Gisu
forms on p. 34 of the Rev. J. B. Purvis's
Lumasaba Grammar. We need not give a
table, as, the principle being known, it is quite

THE PRONOUN 97

easy to combine any form wanted from the
previous tables.

The double agreement seems to be confined
to the more archaic Bantu languages. It is
found, as we have just seen, in Zulu and Gisu ;
also in Chwana, Herero, Ganda, Kinga and
others; but not in Swahili, Nyanja or Kongo.

Demonstrative Pronouns. — These are usually
three in number ; one, equivalent to * this,'
denoting what is near the speaker ; a second,
what is somewhat farther off (in some cases,
what has been referred to before) ; and the
third, what is at a distance. They are built
up, in different ways, from the Inseparable
Pronoun ; a very common modification is that
the first demonstrative ends in «, which is
changed in the second to o, while the third is
formed by suffixing another syllable to the
first, or to its latter half. This process is
most clearly seen in Swahili and Nyanja.
Sometimes the first half appears to be taken,
as in Kongo (o-yu, and o-na, a-ya, and a-na).
Li (Ganda) and la (Gisu) may be the same
element as le, which, Meinhof thinks, may be
connected with the root -le, -de, ' long,' and so
suggest distance. This and other points
relating to the origin of the demonstrative,

which it is no part of my plan to discuss, may
be found in the second chapter of the Grundzilge
einer vergleichenden Grammaiik dcr Bantu-
sprachen.

Zulu Chwana Herero Nyanja Swahili Ganda Gisu Kon|

THIS

e n

ingui

uyu

hu-yu

ono

uno

oyu

THAT

lowo

eouo

ngo

uyo

hu-yo

oyo

uyo

oyo

THAT
YONDER
THESE

lowaya
laba

eole
vano

nguini
imba

udya
awa

yu-le
ha-wa

oli
bano

ula
bano

ona
aya

THOSE

labo

vauo

mbo

awo

ba-wo

abo

owo

THOSE
YONDER

labaya

vale

mbeni

adya

wa-le

bali

bala

ana

There are other demonstrative forms built
up from these — e.g., the two ' emphatic
demonstratives ' in Kongo, which we need not
notice here.

In Nyanja we have two other demonstrative
roots which may be mentioned here, because
they are used in a way which illustrates the
transition from the demonstrative to the
relative. They are -mwe, ' the same,' and
-mene, ' this same,' ' that same,' ' that very
one,' etc., with their compounds, formed by
suffixes corresponding to the three degrees of
the demonstrative already given. -Menc,

THE PRONOUN 99 '

when used without these suffixes, simply means
'who,' or 'which,' — as

mu-ntti a-mene a-na-gwira nchito.
The man who did work.

zi-ntu zi-mene ti-na-ni-sautsa.

The things which grieved me.

But

muntu ameneyu = this same man.

muntu ameneyo and amene udya = that same man.

chi-ntu chi-mene-chi = this same thing.

kasn li-menc-lo = that same hoe, etc., etc.

We shall return to these two pronouns in
the course of the next section.

A special form of demonstrative — sometimes
called ' adverbial demonstratives,' and mean-
ing ' Here he is,' ' here they are,' — is especially
noticeable in Zulu — nangu, nanku, nafa, etc.
They need not be further noticed here.

The Relative Pronoun. — This constitutes
somewhat of a difficulty in many Bantu
languages, though some cannot be said to
have any relative at all. The relative, as we
understand it, hardly belongs to the earlier
stages of speech. It implies a co-ordination
of ideas — a fitting of separate notions together,
whereas children, and primitive people, think
of one thing at a time and express it in a

100 THE PRONOUN

sentence by itself. The child will say, ' I saw
a man. The man had a dog,' — putting the
two ideas, as it were, side by side. The next
step is — ' I saw a man ; he had a dog ' ; and
then we come to — ' I saw a man who had a
dog." In the second case, we have two
co-ordinate sentences, of equal importance ; in
the third, a principal and a subordinate
sentence, which together make up a complex
one. Many Bantu languages cannot form
complex sentences at all, and those which can,
only do so to a limited extent.

In Nyanja there is no true relative. The
typical form of sentence runs thus :

mimtu a-na-dwala dzulo wafa.

The man (who) was ill yesterday is dead :

— literally, ' The man, he was ill yesterday :
he died.' But, to make the reference of the
second clause more definite, a demonstrative
is inserted. One could say, uyo wafa, or udya
wafa; but more commonly either -r,iwe or
-mene is employed.

muntu yennve anaihvala dzulowafa. Or
muntu atnene. anadwala dzulo wafa.
ntbalatne zimene zinadia mbeu za-gwidvca.
' The birds which ate the seeds have been caught.'

THE PRONOUN 101

Pamene (Class 18), * the place which,' is used
for ' where ' — and, by an extension of meaning,
for ' when.'

In other languages the relative is rendered
by_a particle prefixed to the verb and the
' prepositional form ' of a pronoun placed after
it. The simplest form of this is found in
Swahili : a-sema-yc (or, in Mombasa dialect,
a-senia-e), l he who speaks ' ; li-anguka-lo, * that
(fifth class) which falls ' ; ki-waka-cho, ' that
(seventh class) which burns.' This, when
analysed, is seen to be really equivalent to
1 he speaks (that is) he ' ; 'it falls (that is) it.'
This seems to be nearer the mark than to
speak £>f a ' relative pronoun expressed by a
syllable formed of the letter -o, preceded by
the initial consonants proper to its antecedent '
(Steere's Exercises, p. 22) ; but the construc-
tion is exceedingly difficult to make clear,
except in the light of comparative grammar.

When the relative is the object, it may be
expressed by using the same form, but insert-
ing the proper object-pronoun before the verb
and making the suffixed pronoun agree with
the object, not the subject.

Thus ' (the knife) which I want,' is (kisit)
ni-ki-taka-cho — literally, ' I it want*(that is) it.'

102 THE PRONOUN

~Po, -ko, and -mo, as relatives, indicate the
notion of ' where ' or ' when ' — ni-lala-po,
1 where (or when) I sleep ' ; a-taka-po, ( when
he wishes.' As we see, this relative is inti-
mately combined with the verb — so it is in
many other cases ; and this once more
illustrates the difficulty of applying our
received grammatical classification and
arrangement. In Steere's Handbook o/Swahili,
the treatment of the verb has in some degree
to be anticipated in the chapter on pronouns,
while that on the verb has to include the
application of the relative pronoun to certain
tenses.

While the use of the accepted nomenclature
is, up to a certain point, convenient and even
necessary ; we must never allow ourselves to
think of its definitions as rigid boundaries, as
though words could be isolated in closed
compartments, like specimens in a museum.
This applies even in English : if children are
taught, for instance, in parsing a sentence like
* Tell him that he must not do that,' to call
the first ' that ' a conjunction and the second
a pronoun, they will be apt to lose sight of the
connection between the two. But if we treat
grammar as a kind of unchanging framework

THE PRONOUN 103

into which every language must be fitted, we
get such absurdities as conjugating a verb ' to
have ' which does not exist, or ' declining ' a
Bantu noun, which, as we have already seen,
cannot be done.

A more elaborate form of the Swahili
relative combines the two pronouns with a
tense-particle as well as the verb and thus
forms three tenses ; in the simpler form no
tense-distinction is possible.

Present : a-na-ye-piga, ti-na-o-piga, li-na-lo-piga.
1 he (it) who (which) strikes.'

Past : a-li-ye-piga, u-li-o-piga, li-li-lo-piga.
' he (it) who (which) struck.'

Future : a-taka-ya-piga, u-taka-o-piga, li-taka-lo-

' he (it) who (which) will strike.'

Na, //, and taka must be reserved for
discussion in the chapter on verbs.

If the relative is the object, the pronouns
are changed as before indicated, the only
difference being displacement of the suffix
(since the object-pronoun must always come
next to the verb-root).

' The thing which I like.'

Kitu ni-na-cho-ki-penda (penda = like).

104 THE PRONOUN

' The house which we bought.'

Nyitmba tu-li-yo-i-nunua (nunua = buy).

In neither of these forms do we find anything
like a special relative particle, different from
the pronoun-s which, as has been said, may be
met with in other connections. In Zulu, we
J have two such particles : a-, which is prefixed
and usually combined with the Separable
Pronoun as subject, and -yo, which is suffixed
— in all cases, whatever the class of the ante-
cedent. Perhaps we can trace a similar
tendency at work elsewhere, for in Zanzibar
Swahili ' there is a disposition to make -o
'the general relative' (Steere), as alio- for
aliye-, lilio- for lililo-, etc.

Examples of the Zulu relative are :

nmuntu o-bona-yo (for a-u-bonayo) = ' a man who

sees.'

aba-ntu a-ba-kala-yo — ' people who cry out.'

indhlela e-lungile-yo (for ct-i-lungile-yo) = ' the
right path.'

(LungiU is the perfect of the verb lunga, ' to
be straight ' or ' right ').

The object is inserted in the same way as
already shown :

THE PRONOUN 105

umutttit a-tn-bonayol = ' a man whom he sees.'

We shall again have to notice this relative
particle a when we come to the Adjectives.

In Ganda, the principle of the Relative
formation is that of prefixing a- to whatever
other pronoun comes before the verb. Thus
abantu a-ba-laba, ' people who see.' (The
singular, omu-ntu a-laba, is indistinguishable
from that which means ' a man sees,' because
a-\- a coalesces with a.) This prefix appears as
o or & according to the class of the subject :
omu-ti o-gu-gwa, * the tree which falls ' ; emi-ti
c-gi-gwa, ' the trees which fall.'

Chwana indicates the relative by suffixing
the locative termination to the verb, without
any change in the pronoun.

' He who has come ' — eo o tsileng, (tsile, perf of tsa)
this one he has-come.'

Perhaps the most literal rendering of tsileng
would be ' is-at-having-come ' : the perfect
indicating a state of completed action. In
fact, this form of the verb is often called a

1 When the object is in the relative, with a subject of
Class I, a- is used without the pronoun (M-) — i.e., the
relative prefix is a- and not o-.

106 THE PRONOUN

participle, and is used as such : mo x° yeng —
' in eating ' ; and, looked at closely, it is easy
to see that the idea of the participle and that
of the locative may run into one another.

In Ronga the relative construction consists
of:

(1) the demonstrative, followed by the Inseparable
Pronoun.

(2) -ka suffixed to the verb, if present, -iki, if past.
mhunu liveyi a-famba-ka= the man who walks.'

(man that he-walks.)

Tihomu leti ti-famba-ka = ' the cattle which walk.'
Tihotmi leti hi-ti-shab-iki — ' the cattle which we
bought.'

(shaba = ' buy' ; hi = ' we.')

M. Junod thinks this ka is originally an
auxiliary verb.

Herero seems to come nearest to our
conception of the relative. There is a special
form of pronoun, different from the demonstra-
tive and used exactly as we use ' who ' or
' which ' — though, of course, it varies with the
class of the antecedent.

(1) oinundu ngu muna = the man who sees.'

(2) ovandtt mbe muna = ' the people who see.'

(3) omuti mbu ua = ' the tree which falls.'

(4) omiti tnbi ua = do. (plural).

THE PRONOUN 107

(5) eho ndi muna = ' the eye which sees.'

(6) omeho nge muna = do. (plural).

The above is the ' participial present ' tense,
which is of simpler formation than the
1 present.'

There are variations for other tenses, which
need not be given here.

Finally, Kongo has no relative, properly
speaking ; ' the relative pronouns are identical
in form and usage with the demonstrative.'

We might enumerate other varieties ; but
the above are sufficient to show that various
stages of evolution from the simple to the
complex sentence are illustrated in different
parts of the Bantu language-field.

This is, perhaps, the best place to mention
the Interrogatives, some of which, by function,
are pronouns, some adjectives, and some
adverbs. Some are invariable ; others take
the class-prefixes ; and of the latter, some,
which are used as adjectives (and also the
words for ' all ' and ' only '), are inflected like
pronouns. (This point will be more easily
made clear when speaking of adjectives.)

The following Table shows how the treat-
ment of these words varies, even when the
roots are cognate.

108

THE PRONOUN

Zulu Chwana Hereto Nyanja Swahili Ganda Gisu Kongo

WHO?

ubani ?

-mar.g ?

-ani ?

ndani ?

nani ?

-ani ?

nana ?

nani ?

WHICH ?

-pi?

-he?

-ne?

-ti?

-pi?

-ki?

—

nkia ?

WHAT?

-ni?

-ng?

tyike ?

chi ani ?

nini ?

ki?

kina ?

nki?

WHERE?-
WHEN?

"-pi?
nini?

kae ?
leng?

pi?
rune ?

kuti?

liti?

wapi ?
lini ?

-wa?
di?

hena?
lina?

-j kwe> i ?
(mweyi:

HOW
MANY?

-ngaki ?

-kae?

-ngapi ?

-ngati ?

-ngapi ?

-ineka ?

-enga ?

-kwa?

OFWHAT
KIND?

-njani ?

-ang ?

-ke?

-tani ?

gani?

-tya ?

-rye

The forms without hyphens are invariable.

There is a set of pronouns sometimes called
the ' Indicative Form,' meaning ' It is I,' ' It
is he,' etc. But, as they are a combination of
the Pronoun and the Copula, it will be better
to reserve them for the next chapter.

CHAPTER VII
THE COPULA AND THE VERB 'To BE '

IN most European grammars, the first thing
learned is the conjugation of the verbs 'to
have ' and ' to be.' In Bantu there is no
verb ' to have,' and 'to be ' is relegated,
comparatively speaking, to the background.
' Have ' is expressed by ' be with,' or simply
by ' with,' with the ' be ' understood. ' I have
a house ' is in Swahili nina nyumba : literally
' I with house ' ; in Zulu ' we have maize '
is sinombila (si-na-nmbila) . This one fact
shows how necessary it is for those who draw
up grammars to take the language as they
find it, instead of trying to fit it into the
framework of any pre-conceived scheme. The
late Dr. Henry began his Chinyanja Grammar
—in many respects an excellent piece of
work — by conjugating the non-existent verb

' to have.'

109

110 THE COPULA AND THE VERB ' TO BE*

Most — if not all — Bantu languages have a
verb 'to be,' but it is not often used in more
than one or two tenses, and, in many cases,
does not appear at all just where we should
expect to find it. Thus its place may be
supplied by the inseparable pronoun, as, in
Zulu : Vitanga ' it is a pumpkin,' I'ulwandhlc
1 it is the sea,' si'sitsha l it is a dish,' etc.

Or it may be omitted altogether.

SWAHILI : Hamisi mpagazi : ' Hamisi (is) a porter.1
ZULU : ngi-lapa : ' I (am) here ' (lapa = here) ;

Ku-njalo-ke ' il (is) so.'
HERERO : Owami omuhona : ' I (am) a king.'

Sometimes ' is,' ' are,' are rendered by an
invariable particle : Swahili ni, Nyanja ndi,
Chwana ke.

NYANJA : Nyalugwe ndi chirombo choopsya :
1 The leopard is a terrible beast.'

CHWANA : Boshwa jwa tau ke letlalo :

1 The lion's inheritance is the skin.'

(Proverb.)

SWAHILI : Dalili ya mvua ni mawingu :

1 The sign of rain is clouds.' (Proverb.)

As stated above, the inseparable pronoun
of the class to which the noun belongs can be
substituted for this invariable copula, as Hii i

THE COPULA AND THE VERB ' TO BE * 111

nyumba (instead of hit ni nyumbd) ' this is a
house ' ; hizi zi nyumba, ' these are houses ' ;
mil n mzuri, 'the tree is fine' (Swahili) ;
I'itanga, 'it is a pumpkin ' (for li(li)tanga) ;
Pulwandhlc (lu(lu)lwandhle) ' it is the sea' ;
b'utywala ( it is beer ' ; zinkomo (contracted from
z'izinkomo), ' they are cattle ' (Xosa) ; lo' muntu
I'idaka, ' that man is a sot ' ; waba I'ukuni, ' he
was (like) a log ' (Zulu). But we sometimes
find forms which cannot be thus accounted
for, as in Zulu : nguwena, ' it is you ' ; nguyena,
'it is he ' ; y'imina (or umina) 'it is I ' ;
ng'umuntu, ' it is a person ' ; ng'amehlo, ' they
are eyes,' etc.

The truth seems to be that this copula is
the old demonstrative root supposed by
Meinhof to have been originally 70, which
being placed before nouns gradually assimi-
lated its vowel to their prefixes, became ngUj
nga, ngi, etc., and finally dropped its consonant
or became a mere duplicate of 'the prefix
(as in Gisu ba-ba-ndu). In Swahili ngu
survives in the form yu as a copula, in such
phrases as yu mzuri, ' he is handsome ' ;
though before the verb it has generally been
replaced by a. Ila has retained the copula to
a greater extent than many other languages,

112 THE COPULA AND THE VERB ' TO BE'

and it may be of interest to give the forms for
the different classes here :

(1) Ngu muntu : ' it is a person.'

(2) Mbo bantu : ' they are people.'

(3) Ngu munzhi : f it is a village.'

(4) Nji tninzhi : ' they are villages.'

(5) Nd'isamo ;l ' it is a tree."

(6) Ngu masanio : ' they are trees.'

(7) nchi chintu : ' it is a thing.'

(8) nshi shintu : ' they are things.'

(9) nimpongo : ' it is a goat.'

(10) nshimpongo : ' they are goats.'

(11) ndu lumo or ndumo : ' it is a razor."

(12) ntu tushimbi : ' they are girls.1

(13) nku kashitnbi : 'it is a girl.'

(14) Mbuxane, or ntbu buzane : ' it is meat.'

(15) nku kufuna: ' it is love.'

(16) (not found).

(17) nku kutwi : ' it is an ear.'

(18) (not found).

Most languages combine the copula with
the personal pronouns (in the * prepositional '
or ' enclitic ' form) for such expressions as ' it
is I,' * it is he,' etc. In Swahili ndi is used
instead of ni for this purpose.

1 This is a very exceptional word for ' tree,' as
puzzling as the Chwana sctlare. In the plural it is
hard to see why assimilation has not taken place : one
would have expected nga masamo.

113

Swahili Nyanja Ila Giryaraa Gisu

' It is I'

ndimi

ndine

ndinie

luiiinl

sono

1 It is this '

ncliwe

ndiwe

lit: i we

niwe

' It is he'

ndiye

inguwe

nde>e

liye

' It is we '

ndiswi

ndife

ndiswe

i diswi

nifwc

' It is you '

ndinyi

ndinu

ndimwe

ndinwi

i in) we

' It is they '
Class
3

nclio
ndio

ndiwo
ndio

imbabo
inguo

ndo

nibo
nikwo

ndiyo

injiyo

ndoyo

nikyo

ndilo

indidio

ndoro

nilyo

ndiyo

ndiwo

ingao

ndcgo

niko

ndicho

ndicbo

inchicho

ndocho

nikyo

ndivyo

ndizo

inshisho

ndozho

nibyo

ndiyo

injio

ndoyo

niyo

ndizo

inshisho

ndozho

nitso

ndio

i ___

induljj

ndolo

nilwo

ndito

intuto

niko

—

ndiko

inkako

ndoko

niko

—

ndiwo

imbubo

ndo

nibwo

ndiko

inkuko

ndoko

nikwo

ndipo

ndoho

niho

ndiko

ndoko

nikwo

ndimo

ndomo

nimu

114

This form does not seem to be used in
Ganda, where ' it is I " is nze, l it is we,'/£ —
the same as the pronoun standing alone.

The above must be distinguished from what
is sometimes called the ' adverbial demonstra-
tive,' meaning ' here he is,' etc., as in Zulu I

G- * 54

nanjfu, 2 natto, 3 nangu, 4 nansi, etc., with

three forms, corresponding to degrees of
distance, like other demonstratives.

The copula is sometimes prefixed to
adjectives used predicatively (that is, in
sentences like ' the man is good ' as distin-
guished from ' the good man ') as in Ila bantu
mbabotu 2 ' the people are good.' Most of the
other prefixes, however, have dropped or
absorbed it, as masamo malamfu 6 ' the trees are
tall,' or ' the tall trees.' This point is worth
noting in connection with the difference (to
which we shall refer in the next chapter)
between the treatment of adjectives when used
as predicates and as epithets. Another, and
somewhat unexpected use of the copula is to
introduce the agent after passive verbs. We
find, in Zulu, e.g., kutshiwo ng'u Ngoza loko,
1 that is said by Ngoza ' ; and the obvious
explanation is that ng' is the preposition nga,
which usually indicates instrumentality. Or

THE COPULA AND THE VERB ' TO BE ' IIS

•

it would be the obvious explanation, were it
not for the disturbing fact that nga u Ngoza
should normally become ngo Ngoza, instead of
eliding its final vowel and leaving the u intact,
as is done here. Furthermore, if this were so,
why should we find ngilibelc y'imisebenzi, 'I have
been delayed by works,' and not ngemistbcnzi ?
In Svvahili, the construction which the foreigner
would expect, and which is sometimes heard,
is ninupigwa na huyu, ' I have been struck by
this (man) ' (na, literally ' with ' or ' and ').
But the more usual and idiomatic form is
nimtpigwa ni huyu ' — i.e., literally : * I have
been struck — it is this man.' So the Zulu
sentences given above are really equivalent to :
' It is said — it is Ngoza (who said it).' * I
have been delayed — it is works (which have
done it).'

As already said, there is an actual verb
equivalent to 'to be ' in most, at any rate, of
the Bantu languages, though its sphere is
much more restricted than a knowledge of
European speech alone would suggest. There
are several roots common to a number of
languages, which do not, however, all use
them in the same way. They are all mono-
syllables, and therefore classed by most

116 THE COPULA AND THE VERB ' TO BE*

grammars among 'irregular verbs' — though
that is hardly a satisfactory way of describing
them.

Thus, in Swahili, we have ku-wa, used in
the past (a-li-ku-wa ' he was ') and the future
(a-ia-ku-wtr ' he will be ') but never in the
present. This is the same root as the Zulu
nku-ba, which also is not much used in the
present, except as an auxiliary. Nyanja
prefers // (or n'),1 which in Swahili is only
found as a particle indicating the past tense.
Ganda has both ba and //, and Herero has ri.
These seem to be the two commonest forms.
Kongo uses what Bentley calls ' the defective
verb ' na? as in kina vava ' it (Cl. 7) is here,'
and also kala, which ' is much more definite,
and .... means to be habitually or
generally.'

But kala is also found in Nyanja and (as
kaa) in Swahili. Its primary meaning is ' sit,'
and thence ' stay ' or ' live ' in a place : in
Nyanja its use is so extended that sometimes
it is really equivalent to a verb ' to be.' The

1 The pronunciation varies with the preceding vo\vel :
ndiri, uli, ali, tiri, etc.

Nna is 'to be ' in Chwana.

THE COPULA AND THE VERB ' TO BE ' 117

Zulu hlala1 seems to be the same word, but
its meaning is not quite so widely extended.
This verb is an excellent illustration of the
way in which the abstract notion of ' being '
may be developed out of such a simple
concrete one as ' sitting ' or ' staying.' Kara
is similarly used in Herero.

1 As a rule, Zulu hi corresponds to s in Nyanja;
-Mann, -sanu, five'; in-hlatu, n-satit 'python':
hlamba, samba ' bathe.' There is, however, a Nyanja
word sala ' remain ' (of which the use is somewhat more
restricted than kala) but it is more properly tsala, and
sala is also found in Zulu (as in the parting salutation
sala kalifc.) Ila has kala 'sit,' and shala 'remain,'
side by side .

CHAPTER VIII
THE ADJECTIVE

THERE are very few real adjectives in Bantu.
Their place is often supplied by nouns and
verbs. Thus Nyanja has no adjective to
express * bad ' or ' black ' ; but there are verbs
' to be bad ' (ku-ipa) and ' to be black ' (ku-da},
and the place of the adjective is taken by a
kind of participle formed of the infinitive with
the possessive particle prefixed to it. ' Black '
is wa ku-da * of being black,' or, more literally,
* of to-be-black,' and ' bad ' wa ku-ipa, usually
contracted into woipa.1 This construction,
which has a genitive or partitive force, as the
Chwana participle in ng has a locative force,
sometimes replaces a relative pronoun. We
can say, for instance, mnyamata wosaka (for
wa-ku-saka) nyama, ' the youth who hunts

1 Monosyllabic verbs do not contract, so it is wa -ku-da,
never woda.

1x8

THE ADJECTIVE

game,' mzungu wosakala ' the white man who
never sits down.' (Sa is a negative particle,
to be explained in a later chapter.)

Or the idea may be expressed by a tense of
the finite verb. ' He is fat ' is a-li-ku-nenepa,
from ku-neiiepa ' to be fat ' ; ' there are too
many fowls ' nkuku zi-churuka (ku-churuka ' to
be too many '). So, too, in Zulu : uku-lamba
1 to be hungry,' uku-tshisa ' to be hot,' uku-godola
' to be cold,' Hkv-lunga ' to be straight ' (and
thence ' upright,' 'good'), etc.

Nouns are usually made to do the work of
adjectives by having the possessive particle
prefixed to them. So, in Nyanja, wa mpamvu
is ' strong ' (literally * of strength ' ) ; ' good '
wa bwinOj ' many ' wa mbiri, ' new ' wa tsopano.
Bwino and mbiri are not at present used by
themselves;, tsopano is an adverb of time
meaning ' now ' (so that ' a new thing ' is,
literally, * a thing of now ') ; but all three may
once have been nouns.

Zulu seems to prefer a relative construction
in similar cases ; ' a strong man ' is itmnntu
onamandhla (a-u-na-ainandhhi) ; literally* a man
who he (is) with strength.' And a very
common and curious idiom in Swahili is the
use of mwenye ' owner,' in the sense of

120 THE ADJECTIVE

'having': mwenye nguvu 'strong,' literally
' owner (of) strength.'

The genuine adjective roots (to be distin-
guished from the derivative adjectives, which
will be mentioned presently) are few in number
and should' probably be reckoned among the
most primitive elements of Bantu speech.
Some of them can be traced through many if
not most of the Bantu languages hitherto
studied ; others might seem to be confined to
one or two ; but it would be very rash to
dogmatize when so much still remains to be
known. Sometimes, when present-day forms
seem quite unrelated, the parallel is found to
have existed in an older stage of the language ;
and sometimes the cognate word is found to
have .different senses in two languages, like
-kulu, which is used almost everywhere for
' large,' but in Kongo and Herero has come to
mean ' old.' The following list is not com-
plete but comprises the most important of these
root-adjectives.

Adjectives derived from verbs have various
endings, of which, perhaps the commonest
are -u (-fu, -vu) and -<?. So, in Swahili, we
have nyama-vu ' silent ' from ku-nyamaa, and
-tnli-vu ' gentle ' from ku-tulia. With regard to

THE ADJECTIVE 121

ere-vit l cunning ' and vi-vu ' idle,' it does not
seem certain that they can be traced to verbs,
and -refu ' long, compared with le and -de in
other languages, suggests that it is the same
root, with the termination -fu suffixed to it.

In Herero, we have -poiu ' blind,'1 dhorodhu
' black ' (from the verb dhorera), taradhu ' damp '
(from tarara).

Of adjectives in -e, Herero has -kohoke
' clean,' from the verb kohoka, and' -pore 'just,'
'gentle,' 'kind,' from pora, of which the
primary meaning is ' to be cool/2

In Ila, a language of the Middle Zambezi,
there are a large number of adjectives ending
in -ski, usually derived from verbs in -ka :

-dimbushi ' foolish,' from ku-dimbusha' to be foolish.'
-komoshi 'broken,' ,, ku-komoka 'to be broken.1
•zapaushi ' ragged,' „ ku zapauka ' to be ragged.'

Some of these adjectives can scarcely be

1 Phonetically the same as the root of the Swahili
noun ki-pofu ; but -pofu is not used in Swahili as an
adjective, in the sense of ' blind.'

a A widely distributed root, which usually has the
secondary meaning ' recover ' (from illness) — probably
with reference to the reduction in temperature. But in
Herero, the notion of ' cooling ' seems to suggest that
of being, or becoming ' moderate' — and so 'reasonable,'
' just,' gentle,' etc.

M
^» rt

•- .0

a a

eg —

THE ADJECTIVE 123

1 Probably connected with the verb lunga, which (t.g. in Znlu)
means ' to be straight,' and so ' to be right,' ' good,' etc.

1 Old Swahili has -wi (-bi, vi).

I Diminutive of -fupi, which appears in the adverb hufupi.

* This word is also found in Nyamwezi, Shambala, Bondei and
some other East African languages.

8 Instead of an adjective, Nyanja has the verb kaltmla ' to b«
old,' and -a bait, which means ' of long ago.'

6 -kukuu, in the sense of ' worn out ' applied to things. Of
persons, -zte is used, or in some dialects -zima, properly ' whole.'
and so 'grown up.' Kali is also sometimes used, as in Mji wa
halt, ' the old town,' at Mombasa.

T ida 5, a noun, meaning ' age ' has perhaps the same root as
-daLi, but there does not seem to be an adjective of this form.

6 Used in some dialects.

9 This modification returns to its original form after a nasal, as
mpya in Cl. 9.

»o Preferred to -a mbiri in some dialects.

II This is found, e.g., in Swahili, with the meaning 'thick,'
'stout.1 I doubt whether the Zulu -nine ' generous ' is from the
same root.

11 More commonly used in a figurative than in a literal sense ;
the usual word for the latter is -kubwa.

11 -kulu is used in the sense of ' mature ' or ' important,' etc.

u When used with the simple prefix -ntnt means ' too large '; to
make it mean merely ' large ' it requires another prefix. This very
curious point in Kongo grammar will be touched on later.

16 In the Lamu dialect -titi and -toto are used.

16 Perhaps the root which we find in the other columns exists
in the Zulu u-bisi and Chwane It-vesi, 'fresh milk.' With -t*!<t
compare Herero taradhu ' damp ' ; — ' wet ' is one of the meanings of
-wisi in Nyanja.

17 Only found as a suffix, in ittkosi-kazi, etc.

18 Herero has no s or t : the former is represented by the
sound of th in 'thin,' the latter by that of th in ' there ' (here
written dh).

19 Only found as a suffix in one or two words ; the root -A*
has takeu its place.

124 THE ADJECTIVE

distinguished from passive participles, as the
Ila komoshi, ' broken,' given above, and, in
Sango1 :

fi-nhn fi-tele%e2 ' cooked food,' from telega ' cook.'
unni-pi-^i nni-hongole 'a hewn tree,' from hongola
hew.'

The Concord of the Adjective is often some-
thing of a puzzle. A priori, nothing could be
simpler : you have your adjective root, and
you place before it the prefix of the noun with
which it is to agree. This happens", in fact,
with most of the classes in Swahili.

1. m-tu ni-znri 'a handsome man.'

2. wa-tn wa-Kiiri 'handsome men.'

3. m-ti in-zitri 'a fine tree.'

4. mi-ti wi-zuri ' fine trees.'

5. tunda zuri 'a fine fruit.'

6. uia-tiinda ma-zuri ' fine fruits.'

7. ki-ti ki-zuri a fine chair.'

8. vi-ti-vi-zuri ' fine chairs.'

9. nyumba n-zuri ' a fine house.'

10. nyumba n-znri ' fine houses.'

11. u-pindi m-zuri 'a. fine bow.'

15. kii'shona ku-zuri ' fine sewing.'

16. pahali pa-zuri 'a fine place.'

1 The Sango (or Lori) people live to the north-east of
the Konde, some distance north of Lake Nyasa.

The Greek ^ is used to indicate the Scottish sound
of ch in ' loch.'

THE ADJECTIVE 125

Adjectives do not seem to be used, in
Swahili, with the locatives of the seventeenth
and eighteenth classes? thougli they are e.g.
in Nyanja.

The above is perfectly plain sailing, with
the exceptions of Class n which has taken
the concord proper to Class 3, the contracted
form u being doubtless associated with that
class through its pronoun, though the u has
disappeared from the third prefix. (Of course
the old form for 1 1 would have been lu-pindi
lu-zuri.) Phonetic laws have produced some
modifications in Classes 9 and 10 (such as the
change of ;/ into m before a labial and its loss
before A1, /, and some other sounds) but these
need not concern us here.

In Nyanja, the case is different. Here the
principle seems to be that the inseparable
pronoun is prefixed to the adjective root and
the Possessive Particle to it as chachikuln 7
zazikulu 8, etc. It is not quite consistently
carried out in the First Class, for there the
noun-prefix takes the place of the pronoun :
wa-ni-kulu, not wa-u-kulu ; perhaps in order to
preserve the distinction between it and Class 3,
which is wokulu (contracted from wa-u-kulii).
This applies to all real adjectives in Nyanja :

126 THE ADJECTIVE

any which do not take the concord as above
are treated either as nouns or as verbs. But
in Zulu a distinction is observed, to which we
shall now come.

The real adjectives, in Zulu, prefix (i) the
relative particle a, (2) the noun-prefix, (i)
coalesces with the initial vowel, i.e., when
followed by n, it makes o, when followed
by i, e. The contracted prefixes return to
their original form.

Thus we get omu-hle (a-umu-hle), eli-hle
(a-ili-hle), olu-hle (a-ulu-hle), etc.

But there are some other adjectives, which
take shortened prefixes in Classes i, 3, 4 and
6 (i.e., o- e- a-, instead of omit, emi, ama) as
umu-ntu o-nsundu ' a brown man,' imilomo
e-banzi 'wide mouths,' ama-hashi a-mhlope
* white horses ' — not omu-nsundu, emi-banzi,
ama-mhlope. The reason for the distinction is
not very clear, but some at least of the
adjectives so treated are originally nouns, as
-lukuni 'heavy' (u(lu)-kuni 'a log of wood'),
-luhlaza ' green ' (u(lu}-luhlaza 'green grass').

Then, in Chwana, both the Pronoun and
Noun-Prefix are added to the Adjective, but
in the reverse order from that in which we
find them in Nyanja.

THE ADJECTIVE 127

It will be sufficient to illustrate this by
examples from these three languages and
Ganda. In dealing with a Bantu language
which has not been much studied, the learner
should pay special attention to this point, as the
system followed may be different from any of
those which have been enumerated. We
must not too hastily assume — having studied
the theory of the Alliterative Concord, not
wisely but too well — that we can apply the
noun-prefixes, as they stand, to the adjec-
tives ; which, so far as it has taken place, is
probably a late development.

The adjective selected for the illustrations
is -knlu, which is found in most Bantu
languages, though in Ganda it does not seem
to be used quite in the sense here implied.

Many languages make no distinction
between the form of an adjective when used
as an epithet or as a predicate ; but some, as
Zulu, Xosa and Ganda, drop the initial vowel
in the latter case.

ZULU : Utnu-ntn otnu-file 'a good man ' — but

tunu-ntu tnu-hle 'the man is good.'

GANDA : ebi-gambo ebizibn ' difficult words,' — but

ebi-gatnbo bizibu ' the words are
difficult.'

"= .2

s s i a » -a a

*•( •*•» « B C* 7 a

5 3 -± •-

^ c -—

^ S - i ^ -9
•S s § « 3 =

C. M

£ ?

'-- o
"

' a

-
v f a 1 jg- 5 a

o • "T 8 " T-^

a S^ E S ^ S.

vo-^osi jo vo-^
^o-rata mo ^o-

a .5

L^ to t- CD

THE ADJECTIVE 129

I -kulu in Ganda is generally used in the sense of ' grown-up.'
a Contracted from a-a-kuln.

8 The usual word for ' tree ' is se-tlhart, but more is sometimes
used with the meaning of ' h«rb ' or ' medicine,' or in a figurative

sense.

4 Contracted from va-n-huln.
6 Contracted from ytt-i-kulu.

6 The particle jV seems anomalous here, like tse in 8 and 10 and
jo in 14. Meinhof thinks these forms may be relatives (Grundziigt
t'mtr vtrgltichenden Grammatih der Bantusprachtn, p. 32), but does not
fully explain them.

7 The full prefix is only found in Ganda with some mono-
syllabic adjectives, such as cri-ngi ' many,' eri-mpi ' short,' etc.

8 Shortened from a-a-kulu. In the Likoma dialect this class has
the pronoun ya. and the adjective has the for/n of ya-i-hulu. The
V is an almost extinct remainder oi the initial consonant to which
the Giryama ga is a nearer approach.

9 Isi-ntu does not mean 'a thing,1 as the other words in this row
do, but has been inserted because it is the same word, though
changed in meaning.

10 \olo hardens into h\plo after the di- (li-) prefix, which is the
same as zi in many other languages.

II The same hardening (see last note) takes place after $, which
also is contracted from e-e.

18 The occurrence of the forms lulime, liilimi. along with Iilimt,
in Nyanja, shows that the nth class is not quite merged into the
5th, though in process of disappearing. Pronunciation seems to
fluctuate, as in lipotga 'trumpet,' which is sometimes heard as
lupenga and in Yao definitely belongs to the In- class. In Nyanja
words beginning with lu have their agreements according toCl. 5,
as is the case here.

•• M Tula still survives in these two languages, and its adjectives
would agree as above, if they were used.

14 Of course -kulu cannot be used with this class.

15 Some dialects have ro vo-^olo or t>> o-t

1(1 Some concords of iita in bu- are given in Scott's Dictionary.
and though the above may not be in use, this would be the correct
form.

" I do not know how to explain this mo-.
»• See the remarks on the Locative Class in Chapter V.

130 THE ADJECTIVE

The distinction may seem a slight one, but
it must not be overlooked.

(The copula is not used in Zulu before
adjectives, as it is before nouns and pronouns.)

We mentioned on a previous page a number
of adjectives which are derived from verbs. But
there are also verbs derived from adjectives —
at least it is difficult to see how the Yao
kulungwa l to be great ' can be anything else ;
though, curiously enough, the adjective -kulu
is not found in this language. And, again,
there are some cases where it is difficult to tell
whether the verb or the adjective should have
the priority. Mr. E. W. Smith, in his Handbook
of the Ila Language, says (p. 61): 'Many of
the adjectives proper have corresponding
verbs which may be used in place of them
as predicates,' and gives a list which we need
not reproduce in full. Some of them seem to
be formed with the suffix -w, as -lemu ( heavy '
(verb kn lemci), -botu ' good ' (verb ku bota), but
ku fwimpa ' to be short ' seems just as likely to
be formed from the adjective -fwafwi (or its
rootfwi) as vice versa. We are reminded of
the Nyanja verbs fini-mpa l to be short ' and
tani-mpa ' to be long ' ; but there are no
adjectives -fini and -tani. There is the root

THE ADJECTIVE 131

ta in -tari, however, and^z may be akin to the
fu mfupi. At any rate the possibility suggests
itself that either the verb or the adjective, or
both at the same time, may be derived from
one of those ' interjectional roots,' which will
be discussed in a later chapter.

There is no need to waste any time on the
Degrees of Comparison. They do not exist,
as grammatical forms. There are various
ways of expressing comparison — the com-
monest, perhaps, is the use of some verb
meaning * pass,' ' excel ' or the like.

NVANJA : Ndi ichi ndi icho cha-pambana ndi icho :
' that is better than this.' (Literally : ' it
is this it is that — that (which) excels is
that.')

ZULU: indoda idhlula umfana emandhleni : 'a
man is stronger than a boy ' (' passes a
boy in strength ').

Or, kuna is used (the pronoun of the eighth
class followed by na, equivalent to ' there is ') ;
indoda inamandlila kunomfana (kuna umfatia).
The idiom is not quite easy to explain, but the
idea underlying it may be somewhat similar to
the Swahili kuliko ' where there is,' as in nyumba
hii ni nzuri kuliko He * this house is finer than
that ' — literally, ' is fine where that is ' — i.e., so

• 132 THE ADJECTIVE

fine that it would attract attention when the
other was in view, and therefore superior to it.

In Kongo, the simplest form of the adjective
implies that the quality is possessed to excess;
an additional particle has to be inserted for the
ordinary or what we should call the positive
form.

Sometimes it almost seems as if the notion
of comparison were absent till imported into a
language by European speakers. Thus, the
author of Elements of Luganda Grammar,
after mentioning the use of the word singa
(' surpass '), says (p. 58) ' Singa in this sense
is rarely heard among the peasants until they
have come in contact with European thought
. .. . thus . . . they would say ' — for

* Bring a longer stick,' ' This stick is short,
bring a long one,' and so on.

What we mean by the Superlative is
expressed either by some equivalent to ' very,'

* exceedingly,' — or by some such phrase as
' surpassing everyone else,' ' excelling all.'

CHAPTER IX
THE NUMERALS

NUMERALS, of course, are a kind of adjective ;
but, in Bantu, their agreements are not always
the same as those of other adjectives, and in
any case they are important enough to deserve
a section to themselves.

They are so convenient for the purpose of
comparing different languages, that perhaps
more attention has been given to them than to
any other part of speech ; and being among
the easiest words to ask for, they are found in
the vocabularies of all the early travellers.

The numerals from one to five, and the
word for ten are, with few exceptions, common
to the whole of the Bantu area. The numbers
six, seven, eight and nine present considerable
differences. Some have no separate words for
these numbers at all, but call six ' five and
one,' seven ' five and two,' and so on. This
does not facilitate arithmetical operations and
children in mission schools are usually taught

133

134 THE NUMERALS

the English names of the numbers before
entering on the mysteries of addition and
subtraction. ' Eighty -seven ' is certainly easier
to deal with, at least for the instructor, than
' five tens and three tens and five and two.'

Where the numerals from six to nine exist,
they are sometimes nouns, with an unmistak-
able reference to the practice of counting on
the fingers. Thus, the Zulu for six is isi-tupa,
1 the thumb ' — showing that the counting
begins with the little finger of tho left hand —
seven is isi-kombisa, ' the forefinger.' Eight
and nine are expressed, rather cumbrously, by
' leave two fingers ' (or ' bend down two
fingers ') and ' leave one finger ' respectively,
It is curious that Xosa, which is so closely
related to Zulu, has -tandatu for six, which is
also found in some of the Eastern languages —
Pokomo, Giryama, Nyamwezi, etc.

This is probably a modified reduplication of
-tatu ' three ' (contracted from tatu na tain),
just as -nane, sometimes used for 'eight,'
seems to be a doubling of -ne, ' four.'

The numbers up to five sometimes (as in
Zulu) agree like ordinary adjectives, some-
times (as in Nyanja) they take the inseparable
pronoun. ' Ten ' seems to be a noun — it is

THE NUMERALS

135

usually invariable, but sometimes, when it has
kept its prefix, it is treated as a noun, and
preceded by a connective particle, as in Zulu :
abantu abay' ishumi — literally ' people who are
ten.'

The following table shows these six numerals
in ten languages. Konde is spoken at the
head of Lake Nyasa, on the eastern side.

Zulu ChwanaHerero Nyanja Konde Swahili Ganda Gisu KongoDuala

-nye

-nwe

-mwe

-inodzi

-mo

-moja

-niu

-twera

-moshi

-wo

-bill

-vedi

-vari

-wiri

-bili

•\vili

-biri

-ole

-bm

-tatu

-raro

-tatu

-satu

-taru

-tatu

-lalo

-ne

-nai

-na

-ne

-na

-ne

-ya

-nei

-hlanu

-tlhano

-tano

-sanu

-hano

-tano

-nano

-tana

-tanu

ishumi

shome

omu-
rongo

kunii

mlongo

kumi

ekuuii

kikum

kumi

doni

Zulu (but not Xosa) omits the initial vowel
in the prefixes of nye : mu-nye, li-nye, si-nye,
\ not omunyt, elinye, etc. (which would mean
' some,' ' other ').

The roots as given here are sometimes
modified when preceded by noun-prefixes, e.g.,
in Swahili, -will becomes mbili when agreeing
with a noun of the tenth class. We may also
notice that there is often a distinct set of
numerals without any class-agreement, used
in counting where no particular things counted

136 THE NUMERALS

are specified. Thus, in Swahili, we count :
mosi, pilij tatu, nne, tano, whereas the same
numbers applied to people would be : (tntii)
mmoja, (watii) wawili, watatu, wane, watano ;
to trees : (mti} mmoja, (mitt) miwili, mitatu,
mine, mitano ; to nouns of the seventh and
eighth classes : kimoja, viwili, vitatu — and so on.

Yao (an important language occupying a
considerable area in Nyasaland and the
Portuguese territory) has mcheche (invariable)
for four, the root of which is found in Makua
as -cheshe. Yao has another peculiarity, in
treating five (msanu) as invariable.

Some dialects of Chwana use mphecho
' completion ' instead of -tlhano, that is ' the
whole hand ' — the five fingers.

The root -rongo or -longo sometimes serves
to form multiples of ten : e.g., in Pokomo
1 ten ' is kumi, but ' twenty ' mi-ongo mi-wii.
In Swahili mwongo survives, meaning ' a
decade ' ; in the older reckoning (now mostly
superseded by the Muhammadan Calendar) a
month was divided into three miongo of ten
days each. Twenty, etc., are usually ex-
pressed by maknmi followed by the number
required ; but sometimes, though rarely, there
is a special word for twenty. Such is du in

THE NUMERALS 137

Isubu,1 which seems, however, to be borrowed
from the Sudan languages. Konde occa-
sionally, along with amalongo mabili? has
umnndu ' a man' — i.e., both hands and both
feet. Swahili uses the Arabic word for
' twenty ' — iskirini.

Sometimes there are distinct words for
' hundred ' and ' thousand,' but in other cases
these are only treated as multiples of ten.
The Lower Kongo people and the Baganda
have the completest systems of numeration,
because they have been used, for many
generations, to deal with a cowrie currency,
and the latter in particular have an ingenious
plan of varying the prefixes for tens, hundreds,
thousands, tens of thousands : thus, 10 is
kumi, 100 ekikumi, 1,000 olukumi, 10,000
akakumi, beyond which this form of numera-
tion does not seem to go. At least I find in
the Rev. G. R. Blackledge's Luganda Vocabu-
lary a word for ' a million,' which is quite
distinct — akakada. Kongo does not use this

1 Isubu is spoken in the Cameroons delta, by people
living between the Duala on the south and the Bakwui
on the north.

2 Or imilongo mibili, as would be expected from the
usual singular. There is also the curious form tu-lon&o
tu-bili.

138

THE NUMERALS

system of prefixes, but has words for 100,
1,000, 10,000, 100,000 and 1,000,000.

The numbers in the following table, if not
preceded by a hyphen, are invariable, except
in so far as they are treated as nouns, and
behave like ishnmi in Zulu. These are
marked*. Those with a hyphen prefixed
agree like tKose in the first table.

Xosa Chwana Jl-rsro Kongo I .'mala lla

-tandatu

-rataro

hanibomwe*

sainbanu5

nmtoba

*chisaniboniwis

-sixenxe1

-shupa2

hamboinbari

nsainbwadi

sainba

*chiloba

-sibozo

hanibor.datu

naua

lombi

*lusele

-litoba

—

inuviu

vwa

dibua

*ifuka

Continuation of above.

Hehe7 Nyamwezi Giryama Kikuyu Gisu Ganda

niutanda

-tandatu

-handahu11

-tandatu

.
-sesaba

omukaga

niufun-
gate8

mptin-
gati

-fungahe

mugwanja

musafu

omusamvu

munane

mnar.e

-nane

-nana

kinane

omunana

igonza0

ken-
da10

chenda

kenda

kycnda

omwenda

THE NUMERALS 139

I Though the forms for 7, 8 and 9 look identical with ihe
prece 'ing ones (aba-hlanu, aba-tandatu.) they are really nouns
[isi-Xtnxt, isi-bozo, i (ii)-tol>j) and the pre6x is preceded by the
relative particle. Otherwise it would be akx-\enxe. not aba-si-
xenxe and so for the others. -xtnxt (the x stands for the
' lateral click ') and -bozo seem to be borrowed Hottentot roots.
i(li)-toba is evidently a noun formed from the verb toba • bend
down ' (ef. tob'umunwemunye for 9 in Zulu.)

a From the verb shupa ' show,' ' point.'

3 These numbers are not given, as they are similar to th« Zulu :
' b^nd down two fingers.' ' bend down one finger.' There is, how*
ever, in some dialects an almost obsolete word for 8. sesu-iii. of
which the derivation is curious: swaya means 'to mark' (with
paint), and as this is usually done with the middle finger of the
right hand, it comes to be synonymous with ' eight.'

4 hutub.i (the same word as the Zulu for ' go ') means ' jump over '
(i.e., from the thumb of the left hand to the thumb of the right)

— hamlo-ntu't 'jump over (and take) one.'

5 Kongo numerals have a double system of agreement (for the
details of which see Bentley, pp. 567-570) ; in the ' primary form '
7, 8, 9, and 10 are invariable, in the ' secondary ' they take prefixes.

6 This looks like a variant of the Hcrero word ; but the only
meanings given for samba in Mr. Smith's vocabulary are ' wash,
bathe, swim.1

7 The Wahehe are to be found some distance N.E. of Lake Nyassa
and to the south of the Wagogo.

» Fungate is still used in Swahili, meaning ' a period of seven
days' — but only in connection with a wedding (see Krapf, s.v. and
Steere's Handbook, p. 91). It was, no doubt, the old word for 7, but
has long been replaced by the Arabic saba'a.

n I have found no other example of this form.

10 Also found in Swahili, though not s* often as the Arabic tiss*
or tisia.

II 'Cerebral t' becomes h in Giryama. The difference between
the two t's is very important in Swahili: -tdt:i, with ' cerebral /'
becomes in Giryama -Kithu. but -tans, with dental t, t$*no.
' Cerebral '< is pronounced by p'essing the tongue against the
hard palate, ' dental ' by pressing it against the teeth ; our ordinary
English t is between the two, being ' alveolar ' — i.i., the tongue
touches the gums or 'tooth-ridge.' The two t's in Swahili may
be distinguished, if necessary as t (-tatu) and -t (-tano) or the
cerebral, as the commoner, may be left unmarked. The Rev.
\V. E. Taylor, in his African Aphorisms, prints the dental t in
italic; but in his version of the Psalms it is underlined. The
difference is more important at Mombasa than at Zanzibar, where
m >st of the words which at Mombasa have dental / are pronounced
with ch, — mato = macho 'eyes' ; teka = chckj 'laugh.'

140

THE NUMERALS

Some of the words for
' thousand ' are as follows :

hundred ' and

Zulu Herero Kongo Duala Ila Nyanja Hehe Kikuyu Gisu

100

ikulu

ethere

nkama

ebwea

mwanda

dzana2

igana*

igana

litondo

1000

inkuluns;-
wane1

eyovl

ezunda

ikoli

chulu

chikwi3

imbirima

ngiri

The Ordinal Numbers are usually expressed
by turning the cardinal number into a noun
preceded by the possessive particle of the
noun with which the number is to agree.

Thus in Nyanja muntu wa chi-modzi, wa chi-
wiri, wa chi-tatu ' the first, second, third
person.' Chiutu cha chi-modzi ' the first
thing' ; nyumba ya chimodzi * the first house,'
etc., etc.

But the first ordinal is not always an actual

This looks like a diminutive of ikulu, but I do not
know how to explain it.

Also in Ronga. In some dialects zana.

Kikwi was formerly used in Swahiii, but is now
seldom if ever heard. The usual word is the Arabic
elfu (mia for 100).

4 This (or gana) is also used in Nyamwezi, Shambala,
Zigula, Giryama, Pokomo, etc.

5 Gisu has no special word for 1,000, kaniatonda
kikuini ' ten hundreds ' being used.

Konde expresses 'a hundred' by 'five people.'
Xosa has the same word for ' hundred ' as Zulu ; but
thousand' is iwaka. Nyamwezi lias kilnttnbi for
' thousand.'

THE NUMERALS 141

numeral. In Swahili mlu wa kwanza is
literally ' the man of beginning,' from kwan^a
(ku-anza) ' begin ' ; and similarly in Zulu
umuntu wokuqala (wa uku-qala}.1

Invariable numerals, as a rule, simply have
the possessive particle prefixed to them, and
in Ila this particle is prefixed directly to the
stem even of the variable ones. In Herero a
somewhat curious system is adopted : the
inseparable pronoun followed by the verb tya
1 say,' is prefixed to the stem of the numeral :
' the second man ' is onmndu utya vari — liter-
ally ' the man he says two,' — ' the third tree '
oinuti utya tatu, ' the fifth name ' ena ritya tano.

The way in which the variable numeral is
changed into a noun is not everywhere the
same, and no general rule can be given. Zulu,
like Nyanja, uses the seventh prefix for this
purpose ; Chwana and Ronga the fourteenth,
Ganda the fifteenth ; and sometimes, as in
Swahili, the isolated forms of the numerals
(those which serve for counting when no
objects are specified) are used. In this language
4 the first* man,' as already stated, would be
mtu wa kwanza. ^

1 Q represents the ' cerebral ' click.

142 THE NUMERALS

The second tree mti wa pili.

The third name jina la tatu.

The fourth thing kitu cha nne.

The fifth house nynmba ya tano.

' Twice,' ' thrice,' etc. are formed in many
languages by prefixing ka-, which will be
noticed later on, as it forms adverbs from
other adjectives as well as numerals.

Special features to which attention should
be directed are the dual pronouns and the
distributive numerals in Ganda, and the forms
in Zulu expressing ' both,' ' all three,' etc. :
bobabili, bobatatu. But these belong to the
study of particular languages, and cannot be
dealt with here.

Some Bantu grammarians include the
numerals among the adjectives; others
(because of the difference in their agreement,
already referred to, observable in some
languages) place them among the pronouns.
This difference usually extends to the words
for ' all,' ' only,' and one or two others, some-
times called ' indefinite adjectives ' or
'indefinite pronouns.'

The most logical plan appears to be to give
the numerals a separate chapter as we have
done.

CHAPTER X
THE VERB

THE Bantu verb normally consists of two
syllables and ends in rt, e.g. ;

ZULU : lima ' cultivate ' ; haniba ' go ' ; tanda

1 love ' ; lala ' lie down.'
CHWANA : lema ' cultivate ' ; eta ' go ' ; rata ' love '

roma ' send.'
NVANJA : manga ' tie ' ; enda ' go ' ; konda ' love ' ;

teni'a ' carry.'

There are a few monosyllabic verbs, most of
which are used as auxiliaries : some are now
only found in composition, as tense-particles.
They are seldom fully conjugated, and have
some other peculiarities which have led to
their being described as ' irregular verbs.'
Sometimes, as we shall see more fully later
on, it seems probable that they have been
worn down from a dissyllabic stem. In other
cases they may be original roots, perhaps

143

144 THE VERB

traceable in the monosyllabic Sudan
languages.

Verbs of more than two syllables are practi-
cally certain to be either ' derived forms ' or
foreign importations (as Swahili fikiri ' con-
sider,' kubcdi l agree,' which come from the
Arabic). In the former case, the fact is some-
times disguised by the loss of the simple form.
In Zulu there is a verb kumula ' untie,' ' undo ' ;
this has the * reversive ' termination -ula,
showing that it is the opposite of a verb kuma
1 fasten ' — but there is no such verb now to be
found in Zulu. A very common verb in Swahili
is simama ' stand ' ; now in other languages we
have ima, yima, yema,jima, zhima ema (or ma)
with this meaning; and ima is even found in
old Swahili. -Ama is a termination implying
' to be in a position,' as ang-ama ' be sus-
pended,' in-ama * stoop ' (be in a stooping
position,)1 etc. «

1 But sometimes we may get a verb which looks like
a derived form, though it is not really one. Meinhof
gives an instance of a Konde word hov-ela, ' hope,'
which would naturally be taken for the applied form of
hova. But there is no such verb as the latter, and the
word is ultimately derived from the Arabic through the
Swahili subiri ' be patient.' Other verbs of more than
two syllables, formed direct from adjectives, nouns or

THE VERB 145

Verb stems beginning with a vowel are not
very common, and usually produce some
modification of the prefix, owing to the contact
of two vowels, which necessitates a special
paragraph or section being devoted to them
in most grammars. Comparative study makes
it appear likely that these " vowel verbs "
once began with a consonant, and Meinhof
thinks this consonant was the voiced velar
fricative, 7. This is not an easy sound to
pronounce at the beginning of a word, and
would very soon tend to disappear, or at
least to become modified. In the above
examples, where it has not been dropped
altogether, it is represented by _y, jt or zh
(pronounced like z in ' azure ').

In Zulu we find several verbs which may or
may not have an initial e : ema (or ma) ' stand,'
eza (or za) ' come ' ; emba (or mba) l dig,' epa
(or pa) 'pull up' (as weeds, etc.). These,
we can see, are reduced to monosyllables
by dropping the vowel, after the loss of
the original initial consonant. The vowel

the invariable roots called ' sound pictures ' or ' vocal
images' will be noticed later. Some diacrifcic marks
have been omitted from hovela — most Konde words
bear more than could be printed here without confusion.

146 THE VERB

being retained where it happens to be more
easily pronounced, keeps the real state of
the case before us ; otherwise it might be
thought that these were true monosyllabic
verbs.1

Verbs which do not end in a are very rare
(unless borrowed from other languages) and
chiefly monosyllables. Ti ' say ' is found in
nearly every Bantu languages, and so is li * to
be,' in composition if not independently. The
Zulu hlezi from hlala ' sit ' being a perfect,
does not count in this connection, yet even
as a perfect it is irregular, since it should end
in -e not -i. I have never seen it satisfactorily
accounted for.

The ' Derived Forms ' of the verb, to which
we have already referred, might perhaps be
most accurately described as ' Voices.' We,
in Europe, have the Active and Passive, to
which, in Greek, is added the Middle : we
also have traces of a Causative, as in ' fall ' —
'feH'( = make to fall) 'sit ' — set ( = cause to sit),
etc. The Bantu languages have all these,
and several others as well.

1 It is possible that in some of these cases the e may
have been adopted by analogy — e.g. in eza.

THE VERB 147

The PASSIVE is formed by means of the
suffix -wa : pig-wa (Swahili) from piga ' strike/
bon-wa (Zulu) from bona ' see.' Sometimes
the suffix is -iwa, (as in Konga), -edwa or -idwa1
(Nyanja), -tbwa or -ibwa (Ganda), -igwa
(Konde). Duala has the very peculiar form -be.
The suffix maycause considerable modification
in the stem of the verb, as in Zulu, where w
cannot follow/), b, or m.

The NEUTER-PASSIVE, usually ending in
-tka or -ika (sometimes in -tika, -aka or -akala)
is distinguished from the Passive by expressing
a state, or the possibility of being subjected to
an action, rather than the actual undergoing
of the action on some definite occasion. Thus,
in Swahili, kamba yafungitkcP is : ' the rope is
(in a state of being) unfastened,' but kamba
yalifunguliwa is: 'the rope was unfastened'
(by some person or persons). In Zulu,

1 Whether it is edwa or idwa depends on the rowel
contained in the verb stem. This ' Law of Vowel -
Harmony ' will be noticed in a later chapter.

1 Fung-uka is really a compound form, being the
intransitive (or neuter passive) of fungna, the reversive
(see p. 150 below) of funga 'fasten.' Fungnliwa, the
passive of fungna, is formed from the original fting-ula,
I between two vowels being usually dropped in Swahili,
and verb-stems ending in / making their passive in -iu-a.

148 THE VERB

inkanyezi ya-bonakala is: 'the star was visible,'
but 'the star was seen' *(by A. or B.),
inkanyezi ya-bonwa.

The APPLIED (sometimes called the ' Rela-
tive ' or ' P-repositional ') form of the verb
gives rise to numerous idioms, some of which
have no exact European equivalents ; but the
most general rule which can be laid down for
its use implies that the action is done with
reference to some person or thing other than
the direct object of the verb. If the verb is
intransitive, and therefore has no direct
object, this form makes it transitive, and
enables it to take one. The ending is
usually -ela (-ila) or -era (-ira) ; in Swahili -ea
(-id). Ex. :

ZULU : hamba ' go ' ; hainbela ' go to ' anyone, and

so ' visit.'

hlala ' wait ' ; hlal-ela ' wait for.1
lima ' cultivate ' ; lim-ela \ cultivate for '

some one else.

NYANJA : dula ' cut ' ; dul-ira ' cut for ' anyone.

netta ' speak ' ; nen-era ' speak to ' or ' for, '
etc.

The CAUSATIVE, as a rule, has the ending
«isa or -isha, or some easily recognisable modifi-

THE VERB 149

cation of the same1. Its meaning needs no
further explanation.

ZULU : vala ' shut,' val-isa ' make to shut,' hatnb-
isa ' make to go,' tand-isa ' cause to
love,' etc.
NYANJA : dul-itsa ' make to cut,' lim-itsa ' make to

cultivate,' nen-etsa ' cause to speak.'
SWAHILI : funda ' learn,' fund-isha ' teach ' (».«.

'cause to learn'), soma 'read'.sow-

esJia ' make, or help, to read.'
HERERO : rara ' sleep,' rar-itha ' make to sleep,'

thtira ' swell,' thur-itha ' cause to swell.'

An INTENSIVE form is sometimes found,
identical in form (though not^in origin) with
the Causative. Thus, in Nyanja, mang-itsa
(from manga, ( tie ') may mean, either ' cause
to tie ' or ' tie tightly,' tnd-ttsa either ' make to
walk ' or • walk far.' This is also the case in
Zulu, but here, the intensive sometimes
reduplicates the causative termination and
ends in -isisa : buza l ask,' buz-isisa ' inquire
thoroughly.' There is another intensive, in
Zulu, ending in -ezela, which belongs to the

1 It has not been thought necessary to take any
notice here of the causatives in -za and other varia-
tions arising from the presence of certain consonants
in the stem. The causative in -yu is a distinct form,
sometimes found side by side with the others.

150 THE VERB

applied form. In Rundi1 and probably else-
where, the Intensive is a combination of the
Applied and Causative endings : rira ' weep,'
riririsha ' weep continually ' ; saba ' ask,' sab-
irisha ' ask persistently.'

In Luganda, the Applied termination is
reduplicated : tonya l drip,' ' rain,' tonyerera
1 drizzle incessantly.' Sometimes the root of
the verb is wholly or partly reduplicated, to
convey an intensive, or sometimes, rather, a
repetitive force, but this is not the same thing
as the verbal forms we are considering.

The REVERSIVE form has the ending -ula
(-ura, in Swahili -no) — sometimes -ulula, e.g. :

NYANJA : tseka ' shut,' tseg-nta ' open,' pinda ' fold,'

pind-ula ' unfold.'
GANDA : simba ' plant,' siinb-ula ' dig up,' jema

' rebel,' jem-uhda ' submit.'
KONGO: kanga, 'tie,' kang-ula 'untie.'
ILA: aniba 'speak,' ainb-itlula ' retract' (un-

speak), yala 'shut,' yal-ula 'open,'

sonia ' sheaihe,' som-onona ' pull out.'2

1 Spoken in the country near the north end of Lake
Tanganyika.

3 Ila and Hereto both have two additional reversive
endings, -ona and -onona. These are found when the
stem contains a nasal (;;/ or //). Kongo also has -ona
and -una.

THE VERB 151

HERERO: pata 'shut,' pat-ttrura ' open,' yonya ' be
crumpled,' yony-onona ' smooth out,'
etc.

This form is made intransitive by changing
/ to k : tseg-uka ' hempen,' simbuka * be dug up.'
The Reversive form is not usually enumerated
in Zulu grammars, but certainly exists in the
language : jaba is * be mortified, disappointed/
etc., jab-ula ' rejoice,' and there are words in
-nla like kum-ula ' unfasten ' which distinctly
have a reversive meaning, though the primitive
verb may have been lost.

The RECIPROCAL, in -ana, implies, as may be
gathered from the name, an act done by two
or more people to each other :

SWAHILI : pend-ana ' love one another.'

NYANJA : meny-ana ' fight ' (' beat each other,' from

menya ' beat ').
ZULU : ling-ana ' vie with one another,' ' be

equal,' from linga ' strive.'

There are some variations in the ending.
Kongo has -ajiana, or -asajiofUL, as well as -ana ;
Ganda -agana or -ang'ana, as kyaw-agatui * hate
one another,' wulir-agana l hear one another,'
etc. ; and Herero -asana, as mun-asana ' see
each other,' from muna 'see.'

152 THE VERB

The idiomatic uses of the Reciprocal form
are curious : we may give some examples.

ZULU : sa-bon-ana nomgani wanii, ' we saw each

other (I) and ray friend.'

NYANJA : akulu a-bvut-ana mlandu ' the headmen
''contend in a quarrel ' (bvutana, reciprocal

of bvuta ' be difficult.')

In Swahili this form enters into several
expressions where its force is very difficult to
render in English : kupatik-ana ' to be obtain-
able,' kujulik -ai m ' to be knowable.' These are
not quite the same as kupatika and kujulika,
and the difference, probably, is in the implica-
tion that something is obtainable or knowable
by everybody, the acquisition or information
being, as it were, mutual.

The STATIVE form in -ama has left traces in
most languages, even if it is not expressly
recognised in the grammars. Verbs in -ama
usually express an attitude :

NYANJA : er-ama 'stoop,' kot-ama ' be in a crouching
position.'

SWAHILI : in-auia ' stoop ' (in-na, the reversive of
the same root, means ' lift up ') ang-aina,
' be suspended ' from anga ' float ' (in
the air) — angua, the reversive, means
'take down,' andang-uka,its intransitive,
' fall.'

THE VERB 153

Rot-ama is found in Zulu, with the same
meaning as in Nyanja, and we also find
hd-ama l rise up a little from a recumbent
position, ' fuk-ama l sit, as a hen hatching 'and
pak-ama, ' be elevated,' which may be verbs of
the same kind. Compare,

HERERO : themb-ama ' be straight,' pik-aina ' bp
aslant ' (from pika ' pull to one side ').

CHWANA : el-ama (or al-aina) ' sit on eggs.'

KONGO : lal-ama ' be afloat,' lamb-ama ' be
clenched ' (said of a nail), kok-ama ' be
hooked on to,' etc., etc.

Some languages have a REPETITIVE form in
-ulula — others express the same idea by wholly
or partly reduplicating the stem. Ila has ula
1 buy, trade' (cf. Nyanja gula), ul-nlula ' trade
a thing over and over again ' ; nenga ' cut,'
neng-nlnla 'cut up again and again ' ; Kongo :
sumba ' buy,' suinb-ulula ' buy again.' Kongo
also has the suffixes -unwia, -olola and -onona.

These two languages have, in addition, a
4 Persistent Repetitive,' which in Kongo has
the suffix -ujiola, with various modifications.
Ex.:

Tunga ' build,' tnng-njio'ii ' keep on rebuilding.'
Kntia ' plant,' knn-njiona 'keep on replanting.'

154 THE VERB

Ila has no suffix for this form, but inserts a
before the final syllable of the verb.

sotoka ' jump,' sotaoka ' hop, as an insect.'
sandula ' turn over,' sandaula turn over and over.'

There are-^some other endings of which the
functions do not seem as yet to be very clearly
ascertained : -ala, -ata, -nga (found in Herero)
and a few more.

The PERFECT IN -ILE is sometimes reckoned
among the Derived Forms of the verb,
because it is not a tense, strictly speaking —
that is, it does not refer to time, but to ' the
condition or progress of the action ' (Bentley),
and because, unlike the real tenses, it is
formed by a suffix.

Verbs formed from adjective-stems (as
mentioned in a previous chapter) by the
addition of -pa or -mpa, cannot be reckoned
among the Derived Forms. Such are the
Zulu de-pa l be tall,' Nyanja (and Swahili)
nene-pa ' be stout ' ; probably the Nyanja i-pa
1 be bad ' is so formed from the root bi,
originally vi, which has dropped its initial
consonant. In Zulu we have a second form
-pala, as kulu-pala * be fat ' (or ' big ').

In conclusion, we may remark that all

THE VERB 155

these forms of the verb can be compounded
with each other to almost any extent. So in
Zulu : hamb-ela ' visit,' hamb-cl-isa ' came to
visit,' hamb-el-is-ana' cause to visit one another,'
hamb-el-is-an-wa, passive of the last named.
Extreme instances of this kind of cumulative
composition are given in Bentley's Dictionary
and Grammar of the Kongo Language, pp. 640,
641. There is no need to say more on the
subject here.

CHAPTER XI

THE VERB (continued)

MOODS AND TENSES

IF we ask ourselves what we mean by the
term ' mood,' and find that it may be explained
as ' manner of being,' it might seem that the
distinction between the Derived Forms
discussed in the last chapter, and Moods is
not very clear. However, on considering
some examples of each, it becomes evident
that moods are the various conditions under
which some particular act is manifested : the
action, say, of writing is contemplated as
actually taking place (whether in past, present
or future time) — or as possibly taking place
under certain conditions — or as being desirable,
and so on. But it is always the same action
of writing. In the Derived Forms, the action
itself is in some way modified : it is looked on
from the point of view of the sufferer instead
of the doer, or as reversed, caused, intensified,

156

THE VERB 157

applied to someone or something, etc., etc.
And each separate form is carried unchanged
through all the moods and tenses.

Moods are only marked to a limited extent
in English. We have the Indicative, Infinitive,
Imperative and Subjunctive, though the last
is going out of use (that is, as shown by the
form of the verb itself : ' if I be,' ' that he
love,' etc.). In Latin, the distinctive inflec-
tions of the Subjunctive are more strongly
marked, and in Greek we have an additional
mood, the Optative.

The definition of ' Tense ' is simple enough,
if we keep to European languages, where the
word can be used in its strict etymological
meaning. It refers to the time at which an
action is performed — past, present, or future,
with the sub-divisions of ' complete ' and
* incomplete,' or ' perfect ' and ' imperfect,'
etc. But even here the matter is not quite so
simple as it seems — should we, for example,
call the French conditional a mood or a tense ?
For practical purposes, no doubt, the distinc-
tion matters little — yet it is worth thinking
over in connection with our present inquiry.

When we leave Europe, we find — e.g., in
the Semitic languages— that the word ' tense '

158 THE VERB

no longer applies, or rather, it has to be used
with a somewhat conventionalised meaning,
for the distinction of time is not kept in view
so much as that of completed and of incom-
plete or continuous action. We saw in the
last chapte* that the Bantu so-called * Perfect '
tense does not necessarily imply a past state
of things. It is very often equivalent to the
Present, indicating an action completed in the
past, whose effects still continue : thus, ' he is
asleep ' is rendered by an expression meaning :
1 he has lain down ' (and is still lying).

If we bear in mind that both terms are
elastic as to meaning, we can draw a very
clear distinction of form between moods and
tenses in Bantu. The former are distinguished
by suffixes, the latter by prefixes.1

On this showing, the Perfect in -He should
count as a mood, and it appears to me that
there is no good reason against its doing so.
We have seen that some reckon it as a
Derived Form^or Voice.

1 This cannot be taken quite absolutely : for instance,
it does not apply to the Infinitive. (The Imperative,
consisting of the bare stem, might be looked on as the
ground-form whence the others are derived). But this,
in spite of the prefix kit- (which marks neither person
nor time) differs essentially from the tenses proper,

THE VERB 159

Some writers recognise (e.g., in Zulu and
Chvvana) Optative and Potential Moods ; but
these, by their structure, are really tenses,
and, since we cannot adhere to the strict
definition of that word, they may very well
pass fo/ such.

We might reckon in Bantu eight moods,
four of which, the Imperative, Infinitive,
Indicative and Subjunctive, correspond, on
the whole, with the notions expressed by those
terms in European languages. The others
are the Negative, the Perfect in -He, the
Continuative and the Relative.

The Imperative, as we have seen, consists
of the bare verb-stem1 in the singular, and
suffixes -ni (really the pronoun of the second
person) in the plural.

The Infinitive (which, as we have seen, is
identical with the fifteenth noun-class) is
distinguished, as a rule, by the prefix, ku-z This,

1 Perhaps it is better to follow Meinhof in using this
term instead of ' verb-root,' for we cannct tell that these
verbs are not ultimately made up of monosyllabic roots,
going back to, a pre- Bantu stage of speech.

" This prefix has been quite lost in Kongo, except in
the case of the two vowel-stem verbs, /ru'-j'srt and A-u- -ctutti.
Duala shows traces of having had a different infinitive
prefix. (See Meinhof, Grundziiiie, eincrvergleicticnJtn
Crannnatik der Bantuspraclicn, p. 10.)

160 THE VERB

the Indicative and the Imperative all, in the
present state of Bantu speech, end in -a, except
in Herero and some of its cognates, and in the
languages of the extreme north west (Duala,
etc.).1 Bleek seems to have considered this
-a a later accretion, and supposed that the
verb* originally ended in some other vowel.
But this matters little to our present purpose.

The Subjunctive ends in -e. Its uses are
much like those of the European subjunctive,
though more extensive ; they can be better
illustrated from the specimen texts at the end
of the volume, which contain numerous
examples, than by any explanations given
here.2

The Negative, which on our definition we
must reckon as a mood, ends in -/. It is a
feature not found in any European language,
where the addition of some invariable adverb

1 Herero has one present tense which assimilates its
final vowel to that of the stem, as niepiti ' I go out ' (from
pita) ma innnn 'he sees' (from ninn a}. Some of the
Congo languages, such as Ngala, Poto, etc. (not Kongo
itself) seem to possess presents ending in -e and -o,
which are probably to be explained by the same principle
of Vowel -Harmony.

" Ex.: Zulu: ngi-Jiainbe 'let me go'; Swahili : ni-
jenge ' let me build " ; Herero : ngc-tninic ' let me see ' ;
Ganda : a-linie ' let him cultivate,' etc.

THE VERB 161

meaning ' not ' is quite sufficient to negative
any tense of the verb. The only difficulty
that could arise is from the position of the
negative, which, in a compound tense, has to
be inserted between the component parts of
the verb ; and the two particles in French
(ne . . . pas), by doubling this difficulty make
it necessary to learn a negative as well as an
affirmative conjugation. But * not,' nicht, non,
and ne . . . pas do not affect the form of the
verb itself.

It is otherwise in Bantu. There are several
different ways of forming the negative, but
the main principle appears to be that a nega-
tive particle is prefixed and the final vowel
of the verb altered to*'. This is usually
(though not in all languages) the Negative
Present. The Negative Past is formed in a
different way ; and moreover there is not, as
one might expect, a Negative tense corres-
ponding to every Affirmative one. On the
othe* hand, there are some negative tenses
with no affirmative corresponding to them.
This looks as though the Bantu mind con-
ceived of ' not doing ' a thing — just as the
still more primitive mind conceives of ' more
than one thing ' — as a distinct and separate

162

THE VERB

entity.1 And perhaps this is borne out by the
fact that languages 01 relatively advanced
development, like Kongo, have lost the final
inflection, and express the negative merely by
invariable particles. Kongo has one of these
particles before the verb and one after, like
ne . . . pas.

Betonda = ' they love.'
love.'

Ke betonda ko ' they do not

-In Duala, the negative particle si is used for
all tenses, but is placed after the subject
pronoun.

na lotna ' I send ' ; na si loma ' I do not send.'
ba mende jipe ' they will cook ' ; ba si mende jipe
' they will not cook.'

The normal Negative Present is as follows :

Zulu Chwana Swahili Ganda Gisu

a-ngi-hambi

Xa ke reke2

si-pendi3

si-laba

bi-n-teka

' I do not go '

' I do not buy

' I do not love '

1 1 do not see '

' I do not cook '

a-si-hainbi

Xa re reke

ha-tu-pendi

te-ba-laba

hi ba-teka

' We do not
go'

' We do not
buy '

1 We do not
love '

1 They do not
see '

' They do not
cook '

See Language Families, pp. 38, 39.

This e in Chwana is the ' narrow e,' approximating
in sound to i.

The negative particle in Swahili is ha, which is

THE VERB 163

In Nyanja, the negative used throughout is
si (contracting in the second and third persons
singular to sn and sa), and i is sometimes (not
always) suffixed to the verb-stem : as si-nd i
-dziwa-i ' 1 do not know,' but si-ndi-dziwa is also
heard.

We need not enumerate all the different
negative particles in use, e.g., Ila ta, Yao nga,
Zigula nka, etc., but we must say a word in
passing as to the negative in the other tenses
of the Indicative. Swahili has a negative
past formed by means of the infinitive :
si-ku-penda ' I did not love ' ; ha-tu-ku-penda ' we
did not love.' This serves as negative both to
the Past Tense (ni-li-pendd) and the Perfect
Tense (ni-me-ptnda).1 Now, as -me- indicates
that the action is finished, complete, the
sentence ni-me-penda cannot be negatived
merely by the addition of a particle.8 So
another form is used : si-ku-penda is a negation

prefixed to the three plural pronouns, but contracts with
those of the singular : si (originally ha + ni), hu (ha + u)
ha (ha + a). Te, the Ganda negative particle, is in some
dialects, used for all three persons alike as it is in Nyoro.

1 Modern Swahili has disused the Perfect in -tie and
the one which has replaced it is, by its structure, a
tense, not a mood.

2 See Meinhof, Grundzuge (p. 64).

164 THE VERB

of the Infinitive ; literally ' not — I to love.'
(This is different from the form actually in use
as the Negative Infinitive, which is ku-to-penda,
a contraction of ku-toa ku-penda, literally ' to
take away loving.')

But Zulu negatives the Perfect by simply
prefixing the Negative Particle : a-ngi-tandilc
1 1 have not loved'; ka-tandile1 'he has not
loved.' This is what might be called a
mechanical formation ; which means that, the
original force of the inflections having been
more or less forgotten, the prefixes and suffixes
used with some tenses, etc., are applied to
others, without reference to their abstract
congruity.

There is, in Zulu, a Negative Past, made
by prefixing a- as for the Present, and suffixing
nga : a-ngi-hamba-nga l I did not go.'

The Negative Future is, as Professor
Meinhof points out, a recent formation,2 and,
as such, entirely mechanical.

1 Ka sometimes, in Zulu, replaces a, which is never
used, e.g. with the 3rd person singular (if the subject is
of the first class), or with a noun of the 6th class :
itHifanaka-hatnbile ' the boy has not gone1 ; ama-hashi
ka-gijimi ' the horses do not run.'

2 Grundziige, p. 65.

THE VERB 165

Swahili :

ni-ta-pcnda ' I shall love ' ; si-ta-pcndu, ' I shall not
love.'

tn-ta-penda 'we shall love'; ha-tu-ta-penda, we
shall not love.'

This, of course, as it does not change the
final vowel, is indistinguishable from the tenses
we shall have to consider later on. But the
Zulu Negative Future is different. It is
recognizably a compound tense, made up of
the verb ya l go ' and the infinitive ; and the
first part of the compound is negatived in the
same way as the Present.

ngi-ya-ku-tanda ' I shall love ' (lit. ' I go to love ').

a-ngi-yi-ku-tanda ' I shall not love.'

In all these indicative tenses, the negative
particle comes first, but in the Subjunctive,
the Participle, and the Relative Tenses, it
comes after the subject-pronoun.

Subjunctive :

ZULU : ngi-nga-tandi. SWAHILI : ni-si-pcndc ' I may

not love.'
Relative :

dba-nga-yi-ku-tanda a-si-po-setna

' they who will not ' if he does not

love.' —— speak.'1

1 These forms were explained in Chapter VI. As to
the reason for the difference in the position of the
negative particle, see the reference in the last note.

166 THE VERB

It will be noticed that the Negative Particle
here is different from that used with the
Indicative. In Swahili, si is used all through
the Subjunctive, and not with the first
person only.

The Perfect in -He is found in a great many
Bantu languages. Swahili, as remarked
above, has lost it — except in some of the
northern dialects — and it seems to have dis-
appeared altogether from Nyanja, though not
from the neighbouring Yao. It is sometimes
shortened to i or e, and assumes various
modified forms — e.g. it may change the vowel
instead of adding the suffix, as Zulu lele from
lala,1 pete from pata, and in the verbs of the
Reciprocal form, as hlangana ' meet,' perfect
hlangene. Derived forms, especially the Applied,
very frequently shorten the termination to e :
sond-ela ' approach,' perfect sondele.

It is not surprising that missionaries and
others engaged in the reduction of a new
language should sometimes have failed to
recognise this ' Perfect ' when they came
across it, as its use did not correspond with

1 Lalile is also used, but with a somewhat different
meaning.

THE VERB 167

their notion of a tense. Yet that use is not
without parallels nearer home. The Greek
oiSa ' I know,' is really the perfect of the verb
meaning ' to see,' and Latin perfects used in a
present sense, like coepi, memini, odi (which
have lost their presents), and wot;/, (the perfect
of nosco) ' I have come toknow' = ' I know,'
are really exemplifications of the same thing.
The Continuative Mood, with the suffix -ga,
is less frequently met with. It implies that
an action is done habitually, or that it
continues for a long time. It is found in Yao,
Kinga, Konde, Sango, Ganda, Kongo, Benga,
Duala and elsewhere — sometimes in one of
the forms ka, nka, nga, or with other modifica-
tions. It is used in more than one tense, and
is even sometimes added to the Imperative, to
make it more emphatic. This, and the fact
that Kongo suffixes it also (in the form nge) to
the Perfect, might seem to negative its being
counted as a mood ; but, though we do not as
a rule find moods superadded on one another,
after the fashion of the Compound Derived
Forms, there does not seem to be any reason
why we should say it is impossible in these
two cases. Or, again, it is conceivable that
the imperative -ga, at least (which is not exactly

168

THE VERB

continuative, though it might, on occasion, be
so) may not be the same suffix. The follow-
ing are a few examples of this form, which
does not exist in Zulu (unless — which I doubt
— we could count the Negative Past), Nyanja,
Herero or Swahili.

YAO : na-tawa-ga ' I was binding,' or ' I kept on

binding ' (t&wa ' bind ')
ni-ndawa-ga 'if I am binding.'
ni-nga-tawa-ga ' I should be binding ' (if

something else had happened).
GANDA : a-fmnba-nga omupunga bulijo ' she cooks

rice every day.'

a-na-soma-nga ' he will read continually.'
omu-titii eya-kola-nga ebi-bya 'the man

who-used-to-make bowls.'
KONGO : o unu n-tunga-nga e-nzo ame ' to day I am

building my house.'
o unu n-tungidi-nge enzo ame to-day

I-have-been-building my house.'
e lunibu kina ya-tunga-nga enzo ame 'the

other day I-was-building my house."
KlNGA : ndi-tova-ga ' I keep on striking ' (tova,

1 strike.')1
SANGO : vu^a-ga ' go, do ! ' (vu^a ' go.')

The Relative Mood may take, as in Zulu, an

1 O underlined, in Meinhof's notation (used in the
book whence this example is quoted), is the ' broad o,'
like the sound of ou in ' ought.'

THE VERB 169

invariable suffix_for all persons and classes, JOT
as in Swahili, a suffix varying with the class
to which the subject belongs (a-pcnda-ye,
n-anguka-o, li-vwidika-lo, etc.). This applies
to the first and simplest form of the Relative
given in Chapter' VI, which is (in Swahili at
least)1 without distinction of tense. But the
other Swahili forms, if analysed, are found to
follow the same principle : the first part of the
word is an auxiliary (// or 110) in the Relative
Mood, followed by the Infinitive without kit :
a-li-ye-penda, a-na-ye-penda, ki-li-cho-anguka,
vi-li-vyo-vuudika, etc., etc. As the Relative
Pronouns were pretty full/ discussed in
Chapter VI., we need say no more about them
here.

Before passing on to the Tenses, it may be
well to say a few words about Participles.

We have, on previous occasions, referred to
the quasi-participial forms existing in the
Bantu languages: the very common one
formed by prefixing the Possessive Particle to
the Infinitive, and that with a locative termi-

1 Zulu also suffixes -yo to the Perfect. The other
verbal relative formations (those without -yo) are
different in principle from the Swahili ones given in the
text ; but they need not be discussed here.

170 THE VERB

nation, found only (so far as I am aware) in the
various dialects of Chwana (including Sutu).
But Zulu has something like a real Participle,
which ' may be formed for all the Tenses '
(Colenso, First Steps § 232), but, unlike our
participles is preceded by a pronoun and
found in all three persons, singular and plural.
Except in the third person (where the pronouns
are e for the singular and be for the plural,
instead of n and ba), and when agreeing with
a noun of the sixth class (when the pronoun a
is changed to e) the forms are identical with
those of the finite tenses ; and a participle is
often only to be recognised by the difficulty of
construing it as a finite verb in the context.

I shall not attempt anything like a complete
enumeration of tenses. The simple ones are
few and well marked, but there are endless
compound tenses, built up with auxiliaries and
other particles, which are not always easy to
classify. The principle of their structure once
recognized, however, they need present no
great difficulty here.

The tense-particles not immediately recog-
nizable as verbs may have existed as such in
former times — indeed, it is practically certain
that this was the case with most of them.

THE VERB

171

The simplest form of the Indicative Present
(in some languages, as in Swahili, it exists
only in theory) is formed by prefixing the
Inseparable Pronoun directly to the Verb-
Stem.

•See' Zulu Nyanja • Ganda

Si. ls.t pers.

ngi-bona

nJt-ona

ndaba (for

nlab»)

•2.1.1 „

ii-bona

u-on a

o-laba

3id ,,

ii-bona

a-ona

a-laba

PI. 1st „

si-bona

ti-ona

tu-Uba

2nd ,,

ni-bona

n iu ona

niu-laba

3rd „

ba-bona

a-ona

ba-laba

This tense seems, as Junod says of it in
Ronga, not to convey any precise indication
of time. The more usual Present, in Zulu, is
one compounded with the verb ya l go ' :
ngi-ya-bona ' I am seeing ' or, more literally * I
go seeing.'

In some languages, a tense with similar
meaning is formed by means of the prefix
-a-, of which the exact force is uncertain. It
usually contracts with the pronouns. The
Swahili tense given below is that used at
Mombasa ; the corresponding one at Zanzibar
is ni-na-ona, u-na-ona, etc.

Na is one form of the verb 'to be ' ; and in

172

THE VERB

Ronga Cbwana Swahili Zigula

nda-bona (ndi-
a-bona)

ke-a-vona

na-ona (ni-a)

n-a-ona

wa-bona (u-a)

wa-vona

wa-ona

w-a-ona

a-bona (a-a) /

wa-vona

a-oua

a-ona

ha-bona (hi-a)

re-a-vona

twa-ona

ch-a-ona

ina-bona (mi-a)

Iwa-vona

nnva-ona

m w-a-ona

ba-bona (ba-a)

va-a-vona

wa-ona

w a-ona

some languages, instead of the above tense,
we have one compounded with this auxiliary
in one shape or another. Thus, in Nyamwezi,
ndi-wona (for n-li-wona), u-li-woiia, etc. ; in
Nyanja ndi-ri-ku-ona, literally ' I am to see
(= ' I am seeing'), and so on.

Some languages (Zulu, Konde, Ganda and
others) have a Past Tense identical in form
with the Present in -a-. Others use na to
form a Past Tense, e.g., Nyanja ndi-na-ona, etc.

The Future is very often formed with the
auxiliaries meaning ' come 'or 'go ' — Nyanja
ndi-dza-ona, Chwana ke-tla-vona, Zulu ngi-ya-
ku-bona (orngi-za-ku-bona). Swahili has ni-ta-
ona: ta at present means nothing by itself,
but it may be shortened from taka (' wish ' or
' want.') In Ganda, the Near Future is £ormed

THE VERB 173

with the prefix na, and the Far Future with the
prefix ri. Both of them mean ' to be.'

The most peculiar Future is that in the
Likoma dialect of Nyanja, which is identical
with the Negative Past in other dialects — e.g.,
si-ni-fe ' I shall die.' The explanation seems
to be that what one has not yet done is still in
the future, and, therefore, to say one has not
yet died is the same as saying that one will die.

In Swahili, as already stated, there is a
Perfect Tense differing from the form in -He
discussed under the Moods. It is formed
with the particle -me- (which may be connected
with mala ' finish ') — ni-me-ona ' I have seen.'
A similar tense is formed in Pokomo with
-ma-.

Giryama has two Perfects — the Perfect
Mood, which is the older form, ending in -ere
or -ire (ni-onere ' I have seen,' ni-frk-ire ' I have
arrived* — from fika) and the tense, formed
with dza : hu-dza-m'-ona * we have seen him.'

Compound Tenses are very numerous in
Zulu, chiefly built up on the verb 'to be '
(uku-ha) and the particle nga, which mainly
implies potentiality (e.g., nga-ngi-be-ngi-bona,
' I would have been seeing,' etc.)i and
Chwana has a still greater variety', introducing

174 THE VERB

several other verbs. But these, and the
particles which play so great a part in Nyanja
(ma, ta, ka, ngo, etc.), most be left to the
students of the respective languages. It only
remains to say a few words more about
auxiliaries and about monosyllabic verbs.
(The latter are not always auxiliaries, and
there are some auxiliaries of more than one
syllable).

The auxiliaries which we have mentioned
so far are employed as tense-prefixes, and
inserted between the subject pronoun and the
verb. But there are others which are
grammatically separate from it, but necessary
to its meaning. Some of these are defective,
only used in one or two tenses and never
apart from a principal verb ; others are
independent verbs, which have a peculiar use
as auxiliaries. Thus in Zulu ponsa ' throw '
means, as an auxiliary ' to be on the point of
doing ' — ngiponse ukuiva ' I was on the point of
falling.' In Ganda yagala ' like, love, want '
is similarly used to express that something is
about to take place : enyumba eyagala okugwa
'the house is likely to fall.' Va 'go out'
conveys that something has just been done, or
that an act results from something mentioned

THE VERB 175

Before ; in the latter case it is equivalent to
* therefore.'

Tu-va ku-kola ' we have just been working.'
Sometimes the auxiliary is followed by an
infinitive, as in the last examples, and as we
should expect in European languages. But it
is just as often followed by a finite verb, and
this construction gives rise to some of the
most curious and difficult idioms — e.g., in
Zulu : u-buye u-hlangane nabo, ' do thou after
that join with them.' Literally ' do thou
return (buy a) that thou mayest join with
them.' In Ganda, mala ' finish ' is used in
various unexpected ways. It may denote,
with a negative, ' non-completed, though
intended action.' Ya-mala na-ta-kola ' as a
matter of fact, he did not do the work.' Or
we may find it in such sentences as mala ga-lya
1 eat it just as it is' — whether you like it or
not ; (perhaps the idea is ' eat it and have done
with it ! ') — mala ga-genda ' never mind, go ! '

In Ronga, dyuleka neuter-passive of dyula
' seek ' is employed as an auxiliary to express
' it is necessary,' and chuka ' start ' (with
surprise) to convey the notion of ' perhaps ' ' by
any chance,' or to emphasize a negative
imperative.

176 THE VERB

Ku-dyuleka ndi-famba ' I have to go.'
U -ta-mu-khoma loko u-chuka u-mu-bonile ' You
will seize him if by any chance you have
seen him.' U-nga-chnke- u-hlaya ' Don't go to
say. . . / = ' Don't think of saying. . . .*

The use of the verb ti, properly meaning
' say ' is very important. It will be mentioned
again in the next chapter, as it occurs so
frequently in conjunction with the ' descriptive
adverbs ' or ' sound-pictures ' so common in
the Bantu languages. But besides this use,
it enters into a variety of characteristic idioms.

It is found in most languages (except
perhaps those of the Congo), though now
disused in Swahili. In Chwana, it has the
form re1, in Herero ty-a (ti + a) otherwise it
scarcely varies. Its infinitive is often used as
a conjunction, equivalent to 'that' (cf. our
that is to say,') as in Nyanja :

Antu a-ganiza kuti ndi mfiti

' People think that (they) are wizards (who)

zi-sanduka makoswe

1 change-themselves (into) rats.'

Sentences like this, where it is equivalent to
' saying,' show the connection quite clearly :

1 Where e in Chwana corresponds to an i in Zulu, it
is the narrow e — intermediate between French e and i.

THE VERB 177

Tctmbala a-lira kuti ' kukiilukn I '
1 The cock crows saying " kukuluku ! " '

Other tenses, simple and compound, are
used more or less as conjunctions, e.g., Nyanja
nga-ti 'if Ila a-no-ku-ti 'whereas,' Zulu
ku-nga-te, ku-nga-ti-ti. Ila has a-tela l lest,'
antela ' perhaps,' which may be applied forms
of it. A very common idiom is to use it as
an auxiliary at the beginning of a sentence,
with some such meaning as ' when,' ' as soon
as,' or ' it came to pass that. . . .'

Irregular Verbs. Bantu verbs can be
irregular in two ways, neither of which need
cause much difficulty. They may be of one
syllable only, or they may end in some other
sound than a. Ti and a few others are
irregular in both ways at once.

Genuine Bantu verbs of more than one
syllable which do not end in -a are so rare
that we need do no more than mention them.

The monosyllables are more important, but
of these a certain number must be eliminated,
which are not original monosyllables, but have
only become so by attrition. The case of the
Zulu ma or ema ' stand ' (Swahili sim-ama)
was referred to in the last chapter, and there

178 THE VERB

is a whole number of verbs in Zulu, found
either with or without an initial e. Such are
(e)mba ' dig ?1 (e)ba ' steal,'. (e)pa l thin out '2 (as
seedlings) (e)zwa ' hear.'

Some haxve more than one syllable, and
these (like (c)muka l go away ' (e)tula ' take off ')
look like Derived Forms.

It seems clear that these (or most of them,
for some might have been formed later, by

' false analogy ') originally began with a
consonant which was dropped, and then the
initial vowel, when it could not easily be
contracted with the pronoun, was elided also.8
It is interesting to see that Nyanja, while
keeping the initial vowel in ima ' stand ' has
incorporated the infinitive particle with mba
* dig,' which is now ku-kumba. Perhaps this is
to avoid confusion with imba ' sing,' as it has
not been done in the case of ku-ba l steal.'4
Nyanja has no objection to the contact

1 Probably the Swahili j-embe ' a hoe ' comes from
the same root.

a To be distinguished from pa ' give,' which seems to
be an original monosyllable.

'Vowel verbs' are usually reckoned among
' irregular verbs,' on account of this contraction, which
is not always applied in the same way.

4 Cf. Swahili iba (Mombasa) and jepa (Lamu).

THE VERB 179

between tw,o vowels, and, as a rule, sounds
them both distinctly, not often contracting
them into an intermediate sound, as is done in
Zulu ; and perhaps this is why it retains the
original i in ima, which in Zulu is altered to e.
When we come to primitive monosyllables
— or what we may fairly presume to be such—
we find, apart from tense-particles and recog-
nised auxiliaries, several verbs expressing
simple and universal notions (such as ' eat,'
' drink,' ' die,' etc.), in so many Bantu languages
that they are likely enough to have formed
part of the original common stock. The
following table exhibits some, but by no means
all, of these.

The great interest of these primitive verbs
lies in the fact that it may be possible to trace
them in the Sudan languages, as indeed, I
think, has been done with one or two. But
such questions lie outside the scope of this
book.

1 1

2 *

O C

THE VERB 181

1 The Kongo form of this word is spelt as given here in Bentley's
Dictionary, but I have no doubt that it is pronounced dyt, as it is in
Nyanja, where many of the printed books have dia. The same
applies to the spelling nua (for nwii) and kia for kya in Kongo and
Bangi. Bangi is spoken in the district near the junction of the
Kasai with the Congo.

a The old root has been lost in Zulu, probably for hlonifa reasons* ;
the word now used is pitta. Zanzibar Swahili has MJUU, like
Ganda, etc. Note the tendency of Duala verbs to end in-*.

s Gwa, is found in old Swahili : the modern word is anguka. The
usual word in Bangi is kita, but kit is given as an ' indeclinable
adjective ' suggestive of falling. It may be the root of Duala ho.

4 A dissyllabic form of this word is found in Yao (uwa) and
Kikuyu (kiiii). The former must not be confused with Swahili u«
' kill,' which is the same word as Zulu and Kongo bula ' strike.'

* Nyanja Zigula and Swahili have lost this, and use words
meaning 'strike each other' (menya.ua, towana, pigana). Giryama
and Ganda use the reciprocal lu'-ana, and Kongo has nw-antt
evidently another form of the same ro .

6 Ha in Chwana is used in the special sense of " giving food."
The word used in Kongo is vana, reciprocal of va, which could
correspond etymologically with pa.

7 Kia and kya (which should probably be spelt alike) may be the
same sound as that indicated by 'chu and tya. See Noel-Armneld,
General Phonetics, p. 91.

B This is only found in some dialects of Nyanja ; it is not used at
Blantyre, probably to avoid confusion with a similar word, tabcosd

as vulgar.

* This word in Zulu expresses what anthropologists call ' taboo.' People art
said to Monipa a ward, if they avoid it (i) as improper or vulgar, (i) because it
is the name — or part of the name — of a deceased chief, or (in the c*M of
women) the head of the family. Thus, the wives and daughters of • man Mined
u-Langa would have to find some other word when speaking of the tun (t-ianft).

CHAPTER XII

/
ADVERBS AND PARTICLES

I have preferred the term * Particles ' for
the invariable parts of speech — except adverbs,
which are somewhat more clearly defined —
because the words which act as prepositions
and conjunctions may be — and often are —
used in other ways, and, in fact, they usually
prove, on examination, to be different parts of
speech altogether.

Pa, ku, mut which, as sometimes employed,
are genuine prepositions, and treated as such
in all the older grammars, are really pronouns,
as we saw in an earlier chapter. In fact,
Meinhof says there are no such things as
prepositions in Bantu. The Zulu nga ' with '
(in an instrumental sense, as watshaywa
ngomcibitsholo * he was hit with an arrow ') at
one time seemed to me a possible exception,
but its use after the passive1 shows that it is
really identical with the copula, as explained

1 See above, pp. 114, 115.
182

ADVERBS AND PARTICLES 183

above. Na l with,' in the sense of ' along
with,' is really the conjunction ' and,' perhaps
the only undoubted conjunction.

We have already remarked that infinitives
and even finite tenses of verbs may be used
as conjunctions : e.g. (u}kuti in Zulu and
Nyanja ; -xp-hitlha — literally ' to arrive ' (at)
for ' until ' in Chwana ; Swahili kw-amba
1 that ' literally ' to say ' (kuti not being used),
Lala k-umfwa ' and so ' literally ' to hear.'
Nyanja -ngakale 'although,' — used with a
pronoun, a-nga-kale, chi-nga-kale, i-nga-kale,
etc., from kola ' sit,' ' stay, be in a place,' and
so, literally, ' it may be (that. . . .')

There is also an interesting use of nouns as
conjunctions, as, Nyanja, chi-fukwa * because,1
which really means ' fault,' * blame,' etc. ;
Duala onyola na, contracted from o nyolo a na,
' through the body of ' (' the fact that . . .'
also meaning ' because.') In Swina pa musoro
pa ' on the head of ' and pa musana pa ' on the
back of,' are used prepositionally for * because
of,' ' on account of.'

The ease with which these locutions change
places is illustrated by the fact that some
adverbs are. turned into prepositions by the
addition of a particle.

184 ADVERBS AND PARTICLES

Thus pa-nsi (it is found in a great many
languages, even where, as in Zulu, pa by itself
has gone out of use) means * on the ground,'
' down,' but pansi ya is ' under ' ' below.
Tini (chini)''\vhich takes the place of pansi in
Swahili is treated in the same way.

It would serve no useful purpose to attempt
enumerating all the possible words or com-
binations of words which might serve as
prepositions and conjunctions : the above is a
sufficient indication what sort of thing to look
out for in any particular language.

With regard to Adverbs we have several
possibilities to consider. First, there are the
regular adverbs, formed from adjectives with
the prefix ka- ; to which we have already
adverted in the chapter on the Numerals.
These are found in Zulu, Nyanja (only with
numerals), Ila, Nyamwezi (with numerals),
Zigula.1 They do not occur in Swahili, Ganda,
Gisu or Kongo. Kale or kade ' long ago '
found in almost every language, even those
which have no other adverbs in ka, seems to be
the adverbial of the root for ' long.'

1 Besides the numeral adverbs this language has ka-ngi
1 often ' (from -(e) ngi ' many ') and perhaps other words
of the same kind.

ADVERBS AND PARTICLES 185

Then we have nouns preceded by a posses-
sive particle, as in Swahili kufanya kwa uzuri
1 to do beautifully.' Here the possessive par-
ticle agrees with kufanya, but it would also be
used with a finite verb — I suppose still with
reference to the infinitive ; or perhaps because
its real relations had been forgotten, so that
it could be placed indifferently after any form
of the verb.

Another way is to use an adjective with the
prefix of the seventh or the eighth class — as
HereJ) tyi-nene ' very,' from -mm ' great ' ;
Swahili vi-baya ' badly ' ; Duala bu-bi ' badly '
which last might also be classed as a noun.
Other nouns are used, by themselves, or with
a particle prefixed, as Zulu na-muhla ' to-
day,' Nyanja maw a l to-morrow.'

Then we have the locative Adverbs — not
merely those already mentioned, which are
preceded or followed by a locative particle, as
pezulu, pansi (some languages have also kiinsi
and munsi), tint, etc., but such as pano, muno,
kuno, or hapa, pale, or mumona, knkona, (I la)
and momwemo, pomwepo (Nyanja), with other
variations, too many to be enumerated.
These, however, are rather a kind of demon-
strative pronoun. And it should not be for-

186 ADVERBS AND PARTICLES

gotten that the Derived Forms of Verbs, in
many cases, render adverbs unnecessary.

There are some invariable Adverbs, which
do not seem to be derived from other parts of
speech, as lero, leo, lelo ' to-day,' Nyanja
tsopano (Yao sambano) ' now,' kati ' in the
middle ' (sometimes with added prefixes), Yao
soni ' again,' Nyanja -nso (suffixed to almost
any other part of speech), 'also,' 'again' —
and others, which can be found by consulting
the lists in various grammars. It may be
that the etymologies of these are only as yet
untraced, and they may be assigned to their
proper position in time ; but some of them
possibly belong to the class described in the
next paragraph, though they have settled
down to a more assured position in the
language than those we are about to mention.

These are what are sometimes called
' Sound-pictures ' ; other terms for them are
* onomatopoetic vocables' (Stapleton), ' de-
scriptive adjectives ' (Junod), 'onomatopoetic
substantives' (Torrend), 'indeclinable adjec-
tives ' (Whitehead), ' interjectional roots,' etc.

The importance of these has been more
and more recognized of late years. They
occupy a very prominent place in the Sudanic

ADVERBS AND PARTICLES 187

languages, and Westermann has devoted a
good deal of attention to them in his Ewe
Grammar.1 There is also a very interesting
passage dealing with this feature of primitive
language in Levy-Bruhl's Les Foitctions
Mcntales dani les Societes Inferieures? Dr.
Hetherwick (Handbook of the Yao Language,
p. 76) says :

' Certain words onomatopoetic in their
character may be classed as adverbs. They
represent the action or the idea referred to
and may be used either with or without the
descriptive verb ; thus chum signifies the sound
of falling into water, like our English " splash."
Wa-gwile mmesi, chum! " He fell into the
water, splash ! " Myu, with the fingers drawn
across the lips, or accompanied by a peculiar
motion of the hands, one over the other,
signifies completion ; Ngondo jaiche nckumala
wandu myu ! " The war came, and the popula-
tion was completely destroyed." An idiomatic
use of the verb kuti " to say," is used in con-
junction-with such words. To the form -ail
is prefixed the characteristic pronoun of the

1 See The Language-Families of Africa, pp. 43, 66.
a Paris, 1910.

188 ADVERBS AND PARTICLES

object described, and joined with the onoma-
topoetic has the force and application of an
adjective. Ngo jati pyu " red cloth " (literally
the cloth which says pyu or red), Nalet ngope
jakwe jati bi " Look, his face is black" (says
hi — i.e., he is angry).'

Here we see one of the expressions noted
accompanied by a gesture. In fact we may
suppose them to have arisen out of the gestures
which preceded speech — to be, as it were,
gestures translated into sound. To quote
M. Levy-Bruhl (p. 183) :

* It is not even necessary that these
" auxiliaries " of description should be exclu-
sively gestures or movements.' (The previous
paragraph deals with the use of gesture, not
in the absence of speech, but to help it out
and make it more expressive.) * The desire to
describe may also try to find satisfaction by
means of ... a kind of pictures or repro-
ductions of what one wants to express, obtained
by means of the voice. Among the Ewe
tribes, says M. Westermann, the language is
extraordinarily rich in the means of directly
reproducing an impression by sound. This
richness arises from an irresistible tendency
to imitate all that is heard, seen, or generally

ADVERBS AND PARTICLES 189

perceived, and to imitate it by means of one
or -more sounds. . . . What is imitated,
in the first instance, is apt to be movement ;
but we also have these imitations or vocal
reproductions — these " vocal images," for
sounds, odours, tastes, tactile impressions.
There are some which accompany the expres-
sion of colours, abundance, degree, pain,
enjoyment, etc. It is beyond doubt that
many words in the real sense (nouns, verbs,
adjectives) have originated in these vocal
images. They are not, properly speaking,
onomatopoeias, but rather jj&scriptive vocal
gestures.'

Stapleton, therefore, defined them some-
what too narrowly in calling them imitations
of sounds1 — in fact this is contradicted by the
very examples he gives a few lines further
back :

NGALA: inai mabandakani lilili 'the water has

quieted down peacefully.1
butu boindi pi 'the night darkens darkly,

or silently on all the heavens at once, etc.*
LOLO : iitso kwi kwi kwi ' go quickly.'

This writer goes on to say : * These forms
1 Comparative Handbook of Congo Languages, p. 1 30,

190 ADVERBS AND PARTICLES

are used very largely as interjections, and
some are evidently amongst the most primitive
elements of these languages. Some appear to
be the roots from which nouns, etc., are
formed, — sanja abameli bu o mai (Ngala) " the
moon shines on the water brightly " (cf. bo-bu,
11 light "). Ndako foi foi (Kele) "the house is lit
up brightly." '

It is a pity that the author did not illustrate
this point a little more fully, as he does not
tell us what noun is formed from foi foi: by
analogy we should expect bo-foi. In the
cognate Bangi language (which does not seem
to possess the / sound) poipoi1 expresses the
brightness of a shining surface, such as
polished wood or metal. It makes a verb
poibana. A glance through the Dictionary
shows numerous other examples : pioka ' beat
with a stick or whip,' from pio, the sound
made by a switch ; tsakana ' be dispersed '
from tsa ; zonga l surround,' from zo. These
are given in the Dictionary as derived from

As printed in Whitehead's dictionary, this word
has diacritic marks indicating that o is the narrow o
with the ' raised ' tone, and i has the ' lowered ' tone.
This is important, as there are other words quite similar
except for the tones,

ADVERBS AND PARTICLES 191

verbs, but it is not fair to mention this without
quoting the passage from the Grammar (p. 18)
which relates to them :

' For the most part these are derived from
verbs, or the verbs from them. For practical
purposes it is here assumed ' (but why ?) ' that
they are derived from verbs. Those who
maintain that the verbs are derived from
them have the best of the argument, for
these indeclinable adjectives are the most
elementary parts of the language, and many
may be traced to an onomatopoetic origin.
These words are the most graphic in the
language, they are the "colouring" words,
the stories and common speech of the people
are full of them, and often they have such
force that sentence after sentence can be
constructed by means of them, without the use
of a single verb, the verb being indicated by
these indeclinable adjectives. They take the
place of adjectives to a very large extent, and
in the dictionary their meaning will often be
found indicated by an English adverb, yet in
Bobangi they are adjectives.'

These languages of the Middle Congo and
its northern affluents tend to shade off towards
those of the Sudanic family. This would seem

192 ADVERBS AND PARTICLES

to account — I do not say for the abundance of
these roots, for Ronga, Nyanja, Zulu, Yao, etc.,
are very rich in them,1 but for the frequency
of nouns and verbs formed from them, and
the ease with which they can be recognised.

In Zulu (which in many ways seems to be
one of the younger Bantu languages), a number
of verbs are plainly derived from these
particles, though they are more usually intro-
duced byukuti. See § 298 (p. 128) of Colenso's
First Steps in Zulu-Kafir— -a most instructive
passage, though the author did not quite
appreciate the character of these ' vocal
images.' The remark (p. 129) that 'others are
probably imitations of the sounds referred to '
shows, however, that he was on the right track.

Some of the examples given to illustrate
this derivation of verbs incidentally show that
some verbs may seem to be Derived Forms
which are not so in reality ; thus hlepula

1 And probably other languages, where no special
attention has been called to them. In Velten's
Nyamwezi grammar, e.g. (Velten's books are practically
useful, but he is scarcely a safe guide in philology) we
find bn ' abundantly,' and po or pe ' also ' — perhaps
others. It is rather surprising to find no indication of
such ' adverbs ' in Gisu — but the work done on that
language is admittedly very tentative as yet.

ADVERBS AND PARTICLES 193

« break off' looks like the reversive of a (non.
existent) hlepa, whereas it comes straight, so
far as one can see, from nkuti hlepu. So, too,
boboza ' pierce/ which looks like a causative, is
from ukuti bobo * to have a hole in it ' ; and the
same root gives us the nouns im-bobo and
isi-bobo. Perhaps some of us have not left our
childhood too far behind to feel in a dim way
that bobo somehow suggests a hole (and it
does so quite as much as the same word, in
French nurseries, expresses ' something that
hurts') — but even these will not be able to
explain why it is so.

Some of the Zulu examples are so delightful
for their own sake, that I make no apology for
quoting them.

Ngaziti shwangalazi lezo1 zinto zonkc.

' I said shwangalazi to all those things — swept them
away with a swish.'

' He says xafuxafu tl — eats like a dog.

1 It (the sky) said namanama (rained very gently)
this morning."

' He said (or went) gigigi down the slope ' — i.e., ran
down — ' and crossed over to the other side.' (Evidently
getting impetus for the upward effort).

1 The sun said tetete '—was low down in the sky.

1 x indicates the lateral click.

194 ADVERBS AND PARTICLES

Mr. E. W. Smith (Grammar of the I la
Language, p. 66) mentions ' certain particles
suffixed to adjectives which express a. super-
lative or absolute idea. They do not seem to
be used with all verbs.

' NE. Menzhia la tontola-ne, the water is very, very
cold.'

' Bu. Muntu u la tuba-bu, the person is very, or
altogether white.

' NSWA. Menzhi a zuma-nswd, the water is
altogether dried up.'

The acute accent (which is not explained in
the text) may indicate a raising of the tone,
or (more probably) that it is accented
independently of the verb, and does not, as
enclitics in Bantu usually do, draw the accent
forward. In that case, it would surely have
been better not to connect the two by a
hyphen.

It would seem as if Ila had limited the
scope of the Vocal Image to a mere expression
of intensity. Or perhaps the author has to
some extent mistaken its nature ; for it seems
clear — even without an inside knowledge of
the language — that they do not mean ' very '
or anything of the sort ; but ;# is ' cold,

ADVERBS AND PARTICLES 195

bu ' white ' ; nswa ' dry,' piu ' red ' (as in
Yao), and so on. Mr. Smith goes on to say :
' These particles are also used interjectionally,
the verbs being omitted, e.g., Nda kaya ku
menzhl. Nswa! I went to the water.
Quite dry ! ' This could not be explained on
the supposition that nswa simply means
*very.' We should also like to refer back to
the parallel columns of verbs and adjectives
given by Mr. Smith on p. 61 — already ad-
verted to in Chapter VIII. The adjectives
there given look to me like developments
(-biabe and -fwafwi are imperfect reduplications ;
all the others ending in o or u)1 from roots of
this kind, and the verbs as if they had been
formed directly from the roots. Of the
' superlative particles ' I have only been able
to trace one which has given rise to a verb :
-pi, whence pia ( to be hot.' (This, as /
psya, siva, etc., is found in most Bantu lan-
guages with the same, or some closely con-
nected meaning. Meinhof thinks the Proto-
Bantu stem was PIA.) But I have no doubt
that careful search would be able to discover a
great many.

We shall recur to this point in the next chapter.

196 ADVERBS AND PARTICLES

In Swahili these particles are not con-
spicuous, yet I do not know how otherwise to
account for tu * only,' pia-4 also,' ' altogether,'
'entirely,' (watu wote pia) , hima 'quickly.' A
few are hear^i as expletives (' When the doctor
pulled out my tooth, I felt — bu !' — lo-o-o !
expressive of surprise, chub I of impatience, /
etc.), but they are not used otherwise and do not
seem to have given rise to any verbs or nouns.
Perhaps the influence of Arabic, which has
supplied some useful adverbs, prepositions
and ^conjunctions, has favoured the disuse of
the Vocal Image.

The late Revs. D. C. Scott and W. A.
Scott, of Blantyre, collected, in a valuable
little pamphlet — The Manganja Unit of
Thought — some interesting specimens of what
they have somewhat enigmatically called
' Buds or Thorns ? ' I take this title to
imply a doubt whether such particles were
really roots whence speech was developed, or
outgrowths of developed speech — atrophied
' fragments of verbs.' A few of the sentences
may be here given in illustration.

1 The lion did not spring — he just came,
kuputu ! kuputu ! — like a horse.'
' The eagle has swept past — kwa.'

ADVERBS AND PARTICLES 197

' A man with a lame leg goes tintpya, timpya,

1 The soldiers stood nda, uda, nda ' (in line).

'The stars are shining ng'ani, ng'ani, ng'ani.'

' He got into the mud and fell tafrwi ! — he got out

and fell into the water, pabva 1 '

' The guinea-fowl has run away tij'o ! tijo ! njo ! '

Here the verb used is njonjola, clearly formed from the

particle.

Further quotations are unnecessary, and
would take up too much space, but I would
direct the reader's attention to M. Junod's
paragraphs (§§ 378, 379) on Adverbes descriptifs
(pp. 196, 197 of his Grammaire Ronga.)1 He
strongly insists on the importance of these
adverbs ' and on the great number of verbs
derived from them.

One point to notice, in conclusion, is that
Vocal Images frequently contain sounds not
otherwise found in the language, just as we
use clicks not found in any articulate English
words to express surprise, regret or (to a horse)
encouragement. Chum (Yao and Nyanja) and
chub (Swahili) have unwonted final consonants.

1 His Elementary Grammar of the Thonga-Shangaan
Language (in English) is more generally accessible.
The section on ' Descriptive Adverbs ' will be found on
pp. 84-86.

198 ADVERBS AND PARTICLES

In Shambala, quite a number of these words
begin with p — a sound which in that language
is (except when preceded by m) changed to h.
This matter would evidently repay further
study. /

CHAPTER XIII
WORD BUILDING

WE saw, at the outset, that inflexion by
prefixes was a great and striking characteristic
of the Bantu speech-family. We have seen,
also, that suffixes play by no means a negligible
part, as they distinguish both the Derived
Forms and the Moods of Verbs. Further
some languages have the suffixed Locative ;
and we just remarked in passing that a good
many adjectives are formed by suffixes.
What more there is to say about these, and
the other cases not already noticed, can best
find a place here.

Nouns may end in any one of the five
vowels.1 Any one of these may be a suffix.

1 I am -using this expression for convenience sake.
They may, for aught I know, end in any of the fifty or
so vowel-sounds recognised by phoneticians which
exist in Bantu. But the old original five will serve the
purpose of this exposition.

199

200 WORD BUILDING

but is not necessarily so ; e.g., in mbwa ' dog,'
nyati ' buffalo,'1 the final vowel seems to belong
to the stem.

Taking, first, nouns and adjectives formed
from verbs, and going through the classes in
order, we find that one of the commonest
derivatives of this kind is the noun-agent,
where the verb-stem takes the prefix of the
first class, and changes its final -a io-i (in
Herero to -e).

ZULU: um-fiki 'one who arrives,' from fika

1 arrive.'

um-fi ' deceased person,' from fa ' die.'
HERERO : omu-tarere ' overseer,' from tarera,

applied form of tara ' look.'
CHWANA : mo-dihi ' worker,' from diha ' work ' ;

mo-ruti ' teacher ' from rut a ' teach.'
NVANJA : tn-weti ' herdsman ' from weta ; nt-pambi

' robber ' from pamba.
GANDA : omu-zimbi 'builder' from zimba ; omu-

somi ' reader,' from soma.

Other nouns prefix the first-class prefix to
the unchanged verb-stem, as Swahili m-chunga
(m-tunga) 'herdsman,' from chunga (tunga) to
1 herd ' ; m-gema ' one who taps palm trees ' (for

These words are Swahili, but they are found (some-
times in the same form) in many other languages.

^ WORD BUILDING 201

wine), from gttna. But these are really a
species of participle, and their verbal character
is still so far felt in Swahili that they are usually
(not always) followed by an object : mchunga
mbuzi ' a goat-herd ' (' one who herds goats ')
mfanya biashara ' one who makes trade,' i.e.t
1 a merchant.' But both in Swahili and in
other languages we also find nouns of this
kind without an object, which shows that there
is a tendency to lose sight of their verbal
character. E.g., Yao m-langa ' herdsman,'
Nyanja m-londola * one who tracks game,'
from londola l follow up ' and the Swahili
mgema already given.

There are some verbal nouns in -e as
Swahili m-tumel ' messenger,' from tuma ' send.'
We have already pointed out that adjectives
in -e are frequently derived from verbs, and
from these we get names of the first class, like
m-umc ' husband,' the adjective -nine ' male
being derived from an almost obsolete (in this
sense) luma ' cohabit.' Ganda has a set of
nouns ending in -e with a passive significance—
omu-fumite ( wounded man,' from fumita ' stab,
omu-sibe ' prisoner ' from siba, ' bind.'

1 Not often used except in the sense of 'apostle.'

202 WORD BUILDING

Verbal nouns of the first class in o do not
seem to be so common, but are found in Yao,
as m-jiganyo ' teacher,' from jiganya ' teach.'
(Dr. Hetherwick, however, says that ' in
actual use, the relative forms juakwiganya,' etc.
— i.e., the infinitive preceded by the possessive
particles — 'are more frequently employed').
And, in general, it is so easy to make these
forms for oneself that it is well to remember
the warnings of experienced writers, and never
venture on any not ascertained to be used
by the natives themselves. Bishop Colenso
says :

' The above words, however,' (i.e., um-fundi
1 learner ' and um-fundisi l teacher ') ' and most
of the above kind which appear in the printed
books, are formed by Missionaries, not by the
Natives, who employ these derivatives much
more sparingly, but may form them at
pleasure, so that they cannot be entered in
the Dictionaries as standard Zulu words.'
Examples of the latter kind are um-ondhli
1 nourisher ' used in an isibongo1 of Mpande,

1 Isi-boitgo (from bottga ' praise ') is a song (generally a
string of laudatory epithets) composed by the professional
bards or ' praisers ' of the Zulu chiefs, and handed down
by tradition.

WORD BUILDING 203

so that it may be regarded as a kind of
poetical license, and um-hambi ' traveller,'
which occurs in a proverbial expression. Some
such words, however, ' belong to the lan-
guage ' ; and indeed we might add that even
of the others, some (such as um-fundisi) have
been found so useful that they are by this
time fully naturalized.

And the late Dr. Scott, in the Preface to
his Cyclopedic Dictionary of the Mang'anja
Language, says :

' Yet no word can be formed at pleasure :
it must bow to usage and wont. However
clear the formation. ... is ... one
must serve the language, not create it.'

But I cannot refrain from adding to this
a remark I once heard from Professor Meinhof,
to the effect that no one knows a language
really well, until he can play tricks with it.
The application of this, in connection with the
previous quotations, must be left to the indi-
vidual conscience of the linguist.

Adjectives, as we have seen, often end in -u
when derived from verbs (-fu and -vu are
common terminations in this case) and some-
times when their derivation is not so clear.
Meinhof derives -kulu from kula ' grow,' but it

204 WORD BUILDING

is open to question whether the derivation is
not the other way about. Nouns in -u derived
from verbs, do not seem to be so common, —
unless they are verbal adjectives used as
nouns : e.g. m-tulivu l a quiet, peaceable sort
of person,' from tulia.1

There is in Swahili another suffix to
personal nouns, which denotes habitual action :
-_//, as m-sema-ji ' orator ' from sema ' speak ' ;
m-pa-ji ' a generous person ' (but see note
on this word in Madan's Swahili- English
Dictionary], from pa l give.' I do not know
if this particular ending is found anywhere
else.

Nouns of the third class are sometimes
formed from verbs with the ending o ; Nyanja
m-pepo and Herero om-bepo 'wind,' from pepa
' blow,' (this verb and its resulting noun are
found in most Bantu languages, but the latter
is sometimes of a different class) ; Herero
omu-hapo ' shape,' from hapa ' grow ' ; Nyanja
m-kotamiro ' lintel of a door ' from koiamira
1 stoop '• ; m-duliro ' mode of cutting ' (the

1 Chatelain says that, in Mbundu (Angola) -;/ and -o
have, in general, a passive ' or inactive ' force and -a,
-e and -i an active one.

WORD BUILDING 205

hair) ; Swahili m-teremo ' cheerfulness,' from
terema1 rejoice'; m-pako 'plaster' from paka
1 smear,' m-chezo (m-tezo) ' dance,' ' game,' from
cheza (tezci), mw-endo 'journey' from enda' walk'
(the same word is found in Nyanja, meaning
Meg'). Duala has in this class a peculiar
suffix -ko : mu-anga-ko ' roast meat ' from anga
1 roast,' inpoko ' gimlet,' from poa ' bore.'

Verbal Nouns of Class 5 ending in o
are common. They often mean the place
where anything is done, as Ganda e-fumbiro
1 a place to cook in ' from fumba ; Bangi
ebombelo ' hiding place,' from bomba, ' hide.' ,
In Kongo nouns of this kind, end in
u : esumbilu ' place for buying ' from sumba
1 buy.' Sometimes these are only used in the
plural, as Nyanja matero ' limits,' from tera,
applied iorm of ta ' finish ' ; maloicero * place
where the suq sets,' from lowera, Iowa ' go in.'

Another kind of noun in o belonging to this
class indicates the result of an action, or
sometimes the way in which it is done.
Swahili chezo (tezo) ' game,' etc. (used as well
as inchezo 3, but more frequent in the plural,
ma-chezo) ; pcndo 'love,' 'liking' from per,.
Also mapfiizi, not used in the singular: in fact,
many of these nouns only exist in the plural,

206 WORD BUILDING

e.g., Nyanja maganizo l thought ' from ganiza
' think ' ; matyolo ' breaking ' from tyola,
1 break ' ; majebo l notches cut round a stick '
from jeba ; Swahili ma-choro ' carving,'
ma-patano '/agreement ' from ku-patana, reci-
procal of pata l obtain ' : * get (from) each
other ' = ' agree.'

Some have taken these nouns as plurals to
the infinitive (ku) class ; but they should have
been warned by the termination and the
slight, but quite distinct, difference in meaning.
Ku-ckora, e.g., is the act of carving, ma-choro
the carving itself (our English participle is
ambiguous) ku-teza the act of dancing, matezo
the dances (or games) themselves.

I may repeat here that the coining of words
for oneself is apt to be a dangerous experiment.
Because matezo comes from teza, and taka is
1 wish ' it is not safe to conclude that ma-tako
means * wants,' * wishes ' : if you do, the result
will be embarrassing.

The seventh class has a good many nouns
formed from verbs, sometimes without change
in the final vowel, sometimes- with the endings
e, i or o : perhaps the last is the commonest.
The chief meaning is (i) the instrument with
which, or the place or time where, anything is

WORD BUILDING 207

done, but we also find (2) a person who does a
thing habitually, or excels in it. Some,
however, can scarcely be distinguished in
meaning from those mentioned in the last
paragraph, as (Swahili) kitendo * action,'
kicheko l a laugh ' ; Yao chi-nyengo ' deceit '
from nyenga. Ganda : cki-gambo ' wotd,'
from gamba ' speak.'

(l) ZULU : isi-bongo from botiga ' praise,' isi-fo
' disease ' (that by which one dies)?

CHWANA : se-reko ' a thing to buy with,' from
reka ' buy ' ; se-aparo ' clothing*
from apara ' put on.1

YAO: chi-lindo 'watch-hut,' from linda

' guard ' ; chi-gono, ' sleeping-place •
from gona.

SWAHILI : ki-fo ' place of dying ' ; ki-funiko 'lid,'
from funika cover.'

HERERO: otyi-dhgra ' sacred place,' from dhera
' avoid for ceremonial reasons ' .
otyi-kunino ' garden,' from kiina
' plant.'

(2) ZULU : isi-lauli ' habitual jester,' isi-hambi
' traveller.'

RONGA : shi-di ' great eater ' (from da), shi-yaki
1 skilled builder,' from yaka ' build.'

208 WORD BUILDING

Nouns in -e are

NYANJA : clii-ponde ' a mess of pounded food,
from ponda ' pound,' chi-kalidwe
' nature of a thing.'1

Chwana has a large number of verbal nouns
belonging to the ninth class, which in form
(as this class in Chwana has no prefix) and
meaning are much like those of the fifth
already mentioned : tiho ' work,' from diha
1 do ' ; kepo ' digging ' from epa ' dig ' ; picho
1 assembly ' from bitsa ' call.' These usually
have a more and a less- concrete meaning —
e.g., picho may mean the act of calling, or the
assembly which is called ; and there is another
kepo, with a difference in the quantity of the
final vowel, meaning ' a digging-stick.' The
differences in the initial consonant follow the
special laws of sound to which Chwana is
subject, and we may remark that, in kepo, k is
not prefixed to the root, but is a modified
restoration of a lost consonant.

1 From kalidwa, passive of kala ' sit,' ' stay,' ' be '-
verbs which with us cannot have a passive. Chikalidwe
means, I suppose, ' the way in which a thing is,' as
regarded by ^an outsider, and not from its own point of
view — this might make its ' being,' in some sense,
passive.

WORD BUtLDINC 209

In Herero we find some personal verbal
nouns of this class : o-ndodhe ' an artful*
crafty person u from rora ' test,' ' examine,'
o-hodhe t spy,' from hora * spy out.' They are
not so common elsewhere, but Meinhof derives
the Zulu and Chwana word for chief in-kosi,
kxosi, from koka, \oya ' draw,' * lead ' — like dux,
from duco.

We find some nouns of the eleventh and
fourteenth classes derived from verbs, though
the latter oftener come from nouns and
adjectives. Yao, however, has a great many
in «-, which seem to belong to this class.

GANDA: olu-tindo 'bridge' from tinda 'bridge
over ' ; olu-talo ' battle ' from tola ' set
in array'; olu-imba song ' from iinba
' sing ' ; (in other languages this ends in
o: Swahili w-imbo, lla and others Iw -
imbo) ; olu-genilo ' journey' from geiuia
'go' (Nyanja ul~endo)t olu-gero 'a
proverb ' ' story,' from gent ' tell ' (a
story, etc.).

KONGO: lu-keselo 'how the cutting-down came
about,' from kescla, applied form of
kesa, 'cut ' ; lu-vangiln ' the manner of

1 ' One who will take nothing on trust ' is the sense
suggested by the derivation.

210

WORD BUILDING

being made'; 'manufacture,' from vanga
' make.' Of Cl. 14, umbangu ' skill,'
from vanga ; undoki ' witchcraft,' from
loka ' bewitch.'1

I LA : lufuno ' love,' from funa ; lufuko ' dust,'

; from fitka 'rise ' (said of smoke, etc.),

htbeta ' judgment,' no doubt from beta,

but the verb now in use is the derived

form beteka.

SWAHILI : u-fu-ngno ' key,' from fungna ' unlock ' ;
u-jagio and u-peo broom,' from fagia,
pea ' sweep ' ; u-piio, ' passage,' from
pita ' pass ' ; u-puuzi ' nonsense, ' folly,'
from puuza ' talk foolishly.' The last
two are probably of Cl. 14, which is
not now to be distinguished from 11 in
Swahili.

GANDA : Cl. 14, obu-sera 'flour,' from sera, ap-
plied form of sa 'grind,' obn-ganza
' favouritism,' from ganza ' be fond of.'

NYANJA : u-limbo ' bird-lime ' [obu-litnbo in Ganda :
probably from limba ' be firm, hard,
tough,' also ' stick fast,' etc.), u-bvuudo
' decay,' from bvunda ' rot ' ; u-sokedwe
' manner of sewing,' from soka ; «-
endedwe ' manner of walking,' from
enda.

YAO : u-lindi ' watching,' from linda ; uivii

1 Both these words have their initial stem-consonant
modified by contact with the prefix,

WORD BUILDING 211

' theft,' from iwa ' steal ' ; and a number

indicating the way of doing things, like

the last two Nyanja examples: tt-tawe1

plan of building,'1 n-panganyt' 'mode

of making,' from punganya ' make,' etc.

XOSA : nbu-xoki ' falsehood ' from xoka ' tell

lies' (also Zulu) ; nbn-lnntko ' prudence,'

• from Itiinkn; tibu-sika 'winter* (i.e.,

' the cutting time '), from sika ' cut,'

(also Zulu), etc., etc.

In Kele (Congo) we have, e.g., bo-lio * door,'
from lia ' shut.'

The Locative Class sometimes has a kind
of relative form based on a verb-stem in the
applied form; as in Nyanja: po-goncra, nio-
gonera* *' a place to sleep at' or 'in.' The
latter implies that it is an enclosed place, and
can therefore be used as an equivalent for
'bedroom'; so, too, mo-dyer a 'dining-room*
(from dya) and mo-sambira 'bath-room.' from
samba ' wash.'

This does not exhaust the ways of forming

1 Tawa ' tie ' is used to mean ' build ' (as
with the same meaning in Nyanja) because in erei .
the framework of a native hut, a great part of the work
consists in tying the poles, or withes, together — and,
again, in tying on the thatch.

8 From gona ' sleep.'

212 WORD BUILDING

nouns; but these are the principal ones to be
found in most languages. The passage on
the Derivation of Nouns in Bentley's Kongo
Grammar (pp. 528-538) is both interesting and
instructive, but it should be remembered that
the system is not everywhere so elaborate.
See also the section headed ' Formation of
Nouns,' in the Rev. A. J. Wookey's Secwana
Grammar, pp. 10-14.

Before leaving the subject of noun-suffixes,
it is necessary to notice the diminutive in
-ana, which Zulu, Chwana and Ronga have
substituted for the diminutive formed by the
thirteenth prefix (ka-). It is probably like the
suffix -kazi (-kxali)j which forms something
like a rudimentary feminine gender, due to the
influence of the suffix-inflecting ' Hottentot '
language— or languages.

We must say a few words about Denomina-
tive Verbs — i.e., verbs derived either from
nouns or adjectives. They were mentioned
in connection with the Derived Forms, but
they are distinct from these, and probably of
later formation. They are exceptional, in
that they consist of more than two syllables,
without going back to a simple verb ; but they
often look like derived forms ; e.g., Swahili

WORD BUILDING 213

chafu-ka ' be dirty,' which is not the ne^iter-
passive-reversive of a verb chafa, but comes
from chafu ' dirty.' So, too, toroka, ' run
away ' from m-toro ' deserter ' ; pevu-ka ' be
grown-up ' from -pevu ( full-grown.'

Verbs are also formed in this way by the
suffix -pa, as nene-pa • be ' or ' grow stout ' (also
found in Nyanja, where the adjective -nene is
not used). In Herero we have handu-ka ' be
impudent,' from the noun e-handu 5 ; raru-ka
1 be greedy,' from e-raru 5 ' gluttony,' dhandu-
pa * become young ' from -dhandu ; and others
with the suffix para, as potu-para ' be blind,'
re-para ' be long.'

In Zulu there are a few verbs in -pa and
-pala ; de-pa 'grow tall,' kulu-pala ' be fat' (or
' big'). In Yao, verbs are often formed from
Vocal Images by the suffix: -ma, zs'sisima
1 be cold,' from si or sisisi.

Herero has a somewhat peculiar class of
compounds, which we must not leave
unnoticed. We saw, quite early in this book,
that names of trees have practically every-
where the prefix of the third class. In Herero
they are sometimes still further distinguished
by suffixing the root -ti: omit-hama-ti, omu-
tendere-ti, omu-ngwa-ti. (I am unable to say

214

WORD ^BUILDING

what these trees are, except that the last of
the three is called by Brincker ' wild tamarisk.'
Uniu-nga, in Zulu, is a common species of
mimosa.) Again, we have a number of words
compounded with -ndu (the root of omu-ndu,
umu-nhi).

omn-kadhe-ndu

omu-nime-ndu

oniu-hepu-ndu

Also :

on-dutne-wa 9
onui-knrii-kadhc 1
oinu-dhoro-twa 1

' woman ' (-kadhe = ' female ')
' man ' (-nttne — ' male ')
' widow ' (Meinhof suggests a de-
rivation from -hepii
1 discontented ' !)

' male dog ' (on-rume-ombwa)
1 old woman ' (-kuru = ' old ')
' Hill Damara,' from -dhoro ' black '
and onin-twa
1 Bushman.'

Another curious feature is the insertion of
the interrogative particle ke between prefix
and stem, as imu-ke-ndu.1 ' What sort of
person ? ' omu-ke-ti ' What sort of tree ? '

Professor Meinhof says that compound
nouns are unusual in Bantu2 and that perhaps

1 The initial vowel i is sometimes substituted for o
in Herero — as in the Demonstrative Pronoun (Cl. 1)
ingwi.

2 Lautlehre, p. 135.

WORD BUILDING

215

those in Herero are due to the influence of
Nama, in which they are very common. If
he is referring to the compounds enumerated
above, he is probably right ; but he goes on
to give a number of compounds consisting of
a verb and a noun, which could easily be
paralleled elsewhere.
Some of these are :

omu-rara-nganda 1

oni-pit' onganda 11

oniu-dhemba-titua 3

otyi-dhuma-we 7

1 neighbour ' ; that is ' one who
sleeps (rara) in* the same
village ' (onganda).

' vagrancy,' from pita ' go out '
and onganda.

' forgetfulness,' literally ' for-
getting heart ' (dJiemba ' for-
get," omu-tiina ' heart '-
Nyanja mtinia, etc.).

' noise made by the fall of a
meteor ' — strange that this
should be a common enough
occurrence to have a word
to itself — from *dliumal
1 roar ' and e-we 5 ' stone '
(the verb dhuina does not
seem to be in use. Zulu
has duma ' thunder,' but the
word corresponding to the

1 An asterisk prefixed to a root or word means that it
is not actually used in that form.

216

WORD BUILDING

Herero one should be zuma,
which has a different mean-
ing).

It would not be difficult to make a long list
of similar compounds.

ZULU : in-dhlula-niiti 9 ' giraffe,' from dhlula
' pass ' and imiti : it is ' higher than the
trees.'

in-swela-boya ' a horrible portent' (some-
times ' a corpse ') — literally ' a thing
without hair,' from swela ' lack ' and
nb-oya 'hair,' ' wool,' etc.

u-rnahamba-nendhlwane 'caddis-worm,'
etc., from hambet ' go ' and indhlwana
' little house.'

kwa'Mamangalahhva'the Back of Beyond'
— literally : ' At (the place of) " Mother !
I'm lost ! " '

Many Zulu proper names are such com-
pounds, sometimes very curious and
suggestive.

NYANJA : kokalupsya ' early rains ' which ' sweep
away ' (koka) the ' burrft grass ' (lupsya) ;
mpinganjira 3 ' obstacle in the way,'
from pinga ' lie across,' and njira ' road '
fulagombc ' the bee-eater ' (which builds
its nest in a bank, like our sand-martin)
from fula ' dig out ' and gonibe 'river
bank.'

WORD BUILDING 217

ILA : chi-zhinga-lnla 'intestinal fat ' (' that which

surrounds the bowel ')•
ttiu-ilima-ku-bushii ' small -pox ' ('what digs
(holes) on the face ') and numerous
others.

In fact the compounding of words (there
are numerous examples of two nouns connected
by the possessive particle forming an insepar-
able compound1) seems — if not so fully carried
out as in Greek — to be by no means alien to
the genius of the Bantu languages. It is
oftenest found in proper names (as already
remarked) and in the names of animals and
trees.

1 ZULU : inja-yo-mSutu (' dog of the Mosutu ') — a

hairy caterpillar.

iso-le-tikosikazi (' lady's eye ') — a flower (a
kind of jasmine).

' CHAPTER XIV
SOME PHONETIC LAWS

I THINK I have sometimes been asked — by
persons whose philological science was some-
what more imperfect than my own — whether
Grimm's Law was applicable to the Bantu
Languages. Of course, as the law in question
is only a statement of what happens to certain
consonants in the Indo-European languages,
the answer must be no. But the principle on
which it is based, that of the permutatio^ of
consonants, holds good, and seems to work
out with unfailing regularity. That is, if we
meet with any apparent irregularities, they are
probably due, either to imperfect observation
of the sounds, or to the operation of some law
not yet ascertained : in either case, they will
disappear in the light of further knowledge.

' In investigating the relations of any

dialect with its kindred dialects, the first step

218

SOME PHONETIC LAWS 219

is to determine to what sounds in the latter
its own sounds regularly correspond.'1

This was done to a limited extent by Bleek
— with remarkable thoroughness considering
the number of languages at his disposal, and
the small amount of material available for
some of them. But the work can never be
satisfactorily completed till the nature of all
Bantu sounds has been determined with
scientific exactitude. Meinhof, for instance,
after working for years in order to discover
the Bantu sounds regularly corresponding to
the clicks in Zulu and Xosaa was obliged to
give up the task as Hopeless for the present,
chiefly because ' the method followed in these
comparisons was a very rough one. The
question whether the clicks were or were not
aspirated, was never taken into account, and
the tones were not investigated at all. There-
fore, in cases where it seemed as if we had
found two or three instances going to prove a
particular sound-shifting, it is not certain that

1 Whitney, W. D., Language and the Study of
Language (1884), p. 97.

2 Most of these clicks occur in borrowed (Hottentot
and Bushman) words ; but some are found in words
which have parallel forms in other Bantu languages.

220 SOME PHONETIC LAWS

even these were valid, and .... the
results of infinite trouble are worthless from a
philological point of view.'

But, even now, some broad principles of
correspondence can be set down as certain,
though we must await the exacter definitions
of phonetic science before filling in the details.
As this book does not concern itself with
phonetics, I should be straying beyond my
province, if I attempted to do more than
indicate these in the most general way ; but a
few hints on the subject may fitly close our
survey.

The difficult sounds called ' laterals '
(written hi, dhl tl, tlh] are confined to Zulu,
Chwana and Thonga.1 They are absent even
from Herero and the Venda languages of the
North Transvaal. The following table will
illustrate the various sounds in which they
correspond with other languages.

Here we find that hi (which seems to be the

1 Here ' Zulu ' includes Xosa and the various sub-
ordinate dialects spoken by the AmaBaca, Swazis and
others ; Chwana comprises Sutu and Pedi (besides
Rolong, Khatla, etc.) and Thonga (the h is necessary to
prevent confusion with at least three other Tongas), Ronga
and other languages of the Delagoa Bay region.

SOME PHONETIC LAWS

221

Zulu Chwana Herero Vcnda Nyanja SwahW Ganda

Five

-/i/anu

-Hhano

-.'ano

-/anu

-janu

-/ano

•/•no

Python

in-blatu

tlhware

—

n-satu

chatu

Elephant

in-dl.lovu

tlou

on-ilyou

n-dou

n-jobvu

n-dovn

en-jOTu

same sound as the Chwana lateral in the
corresponding word, though written differently),
is represented in the other languages either by
s or (dental) /, except in the case of the
Swahili for ' python,' to which we must return
presently, dhl or //, on the other hand are
represented either by j (dy is probably in some
cases nearer the sound) dental d, or (in
languages not included in the table) z, dz, or s
(Pokomo nzovu, Giryama ndzavu, Konde
i-sofu) .

(The names for * python' used in Herero and
Ganda seem to come from a different root,
and I have not been able to get the Venda
word. Ronga n-hlaru keeps the lateral and,
like Chwana, substitutes Y for /.)

The apparent anomaly of the Swahili -tano
and chain requires a little further explanation.
In the Mombasa dialect there are two kinds of
/, in that of Zanzibar only one is recognised in
practice. They may be distinguished here

222 SOME PHONETIC LAWS

(though this book, on principle, tries as far as
possible to avoid diacritic marks) as t and t.
In Mombasa printed books — the former, as
the more usual sound, is left unmarked, the
latter is underlined, or printed in italics, or
distinguished in some other way. t is pro-
nounced with the tip of the tongue against the
hard palate, t with it touching — or even
between — the front teeth. Get a Mombasa
native to pronounce, first -tatu and then -tano,
and, if you have even a moderately good ear>
you cannot help hearing the difference.

Now many words (but not all) which at
Mombasa have t, as fila ' hide,' teza * dance,'
jito ' eye,' are at Zanzibar pronounced with
what (with apologies to the I. A. P.) I will
write ch : ficha, cheza, jicho. Therefore chain
is the Zanzibar form, which, logically, at
Mombasa, should be tatu. But I am not sure
that it is so, never, to my remembrance,
having heard the latter pronunciation. \ In
fact, the only time I can recall hearing a
python mentioned in Swahili was by a Lamu
man, and I think — but would not swear to it—
that he said chatu. Krapf gives chain, but as
a quotation from Steere : it is possible that
some other word is used 'at Mombasa.

SOME PHONETIC LAWS 223

But, looking again at our table, and taking
it backwards — njl in Nyanja does not always
correspond to a Zulu lateral. We have
njoka ' snake,' njati ' buffalo,' and njuchi ' bee,'
which in Zulu are in-nyoka, in- nyati, in-nyosi.
And sometimes we find Zulu words containing
ny not only in Nyanja, but other languages as
well, such as in-nyama ' animal,' or ' meat ' —
which only in one or two cases that I have
come across is nama (in Chwana and in Venda).
Before coming to a conclusion in a case like
this, one would want to be sure whether all the
ny sounds were the same. Some may, per-
haps, be quite properly written ny, while others
require the symbol which for typographical
reasons is banished from these pages — and as
to the sound, see Mr. Noel-Armfield's General
Phonetics, p. 63.

The comparative tables of words given in
the preceding chapters will already have called
attention to some correspondences, such as
that between Zulu / and Chwana r (which t
and which r are important points to be dealt

The presence or absence of a nasal before a
consonant makes some difference as regards sound-
shiftings.

224 SOME PHONETIC LAWS

with by the phonetician), the tendency of
Chwana (which it shares with Makua) to prefer
a voiceless stop to the same sound voiced and
preceded by a nasal (e.g., rata = tandaand podi1
= mbuzi) ; the curious absence of p from a large
group of East African languages, etc.

In most of these latter, the place of p is
taken by h, in Ganda by w and in Pokomo,
by ' bilabial /' — written bf by the German
missionaries who have chiefly cultivated this
language, and who spell the name of the people
' Wabfokomo.' Chwana, curiously enough,
shares this tendency to a certain extent, though
having no objection to the p-sound per se.
The word usually found as pa-nsi is le-hatse
or le-fatse, (the pa- prefix being incorporated
with the noun-root) and p appears (as we saw
just now) where one would not expect it —
instead of mb or mv. There is no v in Chwana,
but * bilabial v ' is a common sound, and — at
any rate in some dialects — takes the place of

1 Here the o is an extra-narrow o (written in Meinhof's
notation o) which approaches u in sound : the word is
sometimes written poll, in which case it must be
remembered the 1 is ' cerebral 1.' If you try to sound 1
by turning the tip of the tongue up against the palate,
you will find that it approaches very nearly to the sound
of d similarly pronounced.

SOME PHONETIC LAWS 225

b also. P, unless nasalised (i.e., preceded by
m) is not common in Kongo : perhaps the
words in wbich it unquestionably occurs
might on examination prove to be borrowed.

Herero has no s or z, — the sounds substituted
being those of th, voiced and voiceless (i.c.t
as in * there ' and ' thin ') for which I have
written dh and th respectively. Makua, too,
and Kikuyu, have no s sounds : the reason in
all three cases is supposed to be the custom
of extracting — or chipping away part of — two
front teeth. On the other hand, Nyanja,
Yao, and some others have no h, while Swahili
seems to have an exceptional preference for
the sound — as in the pronouns. (It should be
noted that it frequently arises, in this language,
from a contraction which one would scarcely
expect to produce it : e.g. ha- for ni-ka — not to
be confused with the negative ha — and hi- for
ni-ki-1 Conversely, si-, in the negative of the
first person singular, is a contraction of
ha-ni.)

There are three main principles3 which we

1 Steere, A Handbook of the Swahili Language,
pp. 134, 137.

2 See Meinhof, Lautlehre, pp. 12-16.

226 SOME PHONETIC LAWS

must keep in mind when examining the struc-
ture of any language and its relation to others
of the same family. These are :

(1) ASSIMILATION.

(2) DISSIMILATION.

(3) TRANSPOSITION.

We might add FALSE ANALOGY, which often
accounts for phenomena otherwise inexplic-
able, as when in Swahili we have julika ' be
knowable,' fromjua 'know,' which never can
have contained /, as we see by the noun mjuvi,
formed from it, and the parallel forms Nyanja
dziwd, Pedi tzeva. But, as most Swahili
verbs in -ua have dropped /, which reappears
in some of the derived forms (e.g. pindua
* turn over,' applied form pindulia, passive
pinduliwa) jua has been made to ' follow the
rule,' like many modern English verbs
('helped,' '-worked,' etc.).

(i) Assimilation may be (a) Incomplete, or
(b) Complete, and is applied both to vowels
and to consonants.

In Complete Assimilation, two different
sounds occurring in succession are made
exactly alike, for greater ease in pronun-

SOME PHONETIC LAWS 227

elation. Sometimes the second is assimilated
to the first, sometifines_the first to the one t
following it. Thus, in Konde, the verbfisa
1 hide ' (Zulu fihla, Swahili ficha, fita) is
sometimes heard as fifa. Shambala, having
borrowed from Swahili the word for ' paper,'
kartasi (itself borrowed from the Arabic)
makes it into talatasi-. the first consonant being
influenced (in spite of the intervening l=r) by
the / in the next syllable but one.

If a sound becomes, not exactly the same,
but only similar to that which precedes or
follows it, we have Incomplete Assimilation.
This is shown in Bantu when the prefix in- is
followed by a labial (/>, 6, /, v) — when the
dental nasal n is changed to the labial nasal
m. N, again (except in a few languages, e.g.,
Kongo and Makua), cannot be followed by /
or r and changes these sounds to d ; this is why
the plural of u-limi ' tongue ' in Swahili is
(zi)n-dimi, instead of (zi)n-linri.1 In the same
way, verbs whose stem contains o or e take the

1 Another case of Assimilation is when the two sounds
unite to form a third, different from either. \Ve have
already met with the union of a and tt to form o, and a
and » to form e. In Chwana, if n is folfbwed by v, the
two together become />. See Lautlchre, p. 13.

228 SOME PHONETIC LAWS

termination -eka instead of -ika, because the
position of the tongue for c is nearer to that
foro than is that for i. But these terminations
are in most (not in all) languages governed by
the Law oft Vowel-Harmony^ which rests
partly on Assimilation, and partly on Dis-
similation and may be stated thus : If the
verb-stem contains a, ?', or u, the termina-
tion has i: if o or c, it has e. So, in
Nyanja ang'-ana 'look' makes ang'an-ira,
ang'an-itsa, lira ' weep,' lir-itsa ; juna ( seek,'
fun-itsa ; but ycra ' be white ' yer-etsa and the
passive yer-etsedwat and omba ' strike ' omb-era,
omb-etsa. In the case of 0, z, and o the sounds
are made quite identical, or at any rate
brought nearer together ; in that of a and u
they are put further apart.

This last process belongs to Dissimilation.
This arises when two similar sounds occur in
close conjunction, and the speaker, to avoid
confusing them, lays special emphasis on one
and tends to slur the other, in order to make a
difference between them. Some Yao verbs,
whose stems contain /, have a perfect in itc-
instead of -He, as lolite, from lola ' look.'

Under this heading, special attention should
be directed to the law discovered by Dahl, a

SOME PHONETIC LAWS 229

missionary in Unyamwezi, and prevailing in
many East African languages — among others
that usually written as 'Kikuyu.' As a matter
of fact, the people call themselves A-gikuyu;
just as the infinitive prefix for certain
verbs is gu, not ku, and the word for a
stool is gi-tij not ki-ti. All these words, and
many more, are illustrations of Dahl's Law,
which may be stated thus :

When a voiceless stop (k, /, p) is followed by
another voiceless stop, it becomes voiced. In
other words, if k is followed by either k, t, or
p, it becomes g ; t becomes d, and p becomes b.

This principle, if it had been clearly recog-
nised by those who have dealt with Kikuyu,
would have saved them a good deal of trouble.
In Mr. Barlow's Tentative Studies (p. 5) it is
mentioned as the ' Euphonic Change of k ,'
and no doubt the fact that Kikuyu has neither
p nor (except nasalised) d helped to obscure the
real bearings of the case. But the matter
stands exactly as it does in Shambala, Bondei
and Nyamwezi — probably also in Yao, where
we have nguku ' fowl ' (Nyanja, nkuku) mbeko
' fire-stick,' which elsewhere would be ;;//:

NYAMWEZI : inbclio 'cold ' (ntpepo : the second /> bus
become /»)•

230 SOME PHONETIC LAWS

deka ' cook ' (elsewhere teka).
-datu for -tatu ' three.'

SHAMBALA : tn-gate ' bread ' (Swahili m-kate}.

Transposition may occur in two forms :
syllables may be transposed, as in Venda,
where gidima ( run ' is sometimes heard as
diginia, and Nyanja, where ' cough ' is either
sokomola or kosomola. Or a vowel in one
syllable may intrude into another, as in the
Konde perfect of -elupha1 ' be white,' which is
-elwiphe, for -eluphile ; the / being dropped and
the i taken into the previous syllable. Other
interesting examples of this and similar
changes may be found in the sections of
Meinhof's Lautlehre already referred to.

A study of General Phonetics is indispens-
able to anyone taking up an African language.
As already stated, this is a subject with which
I have not attempted to deal, my object being
merely to give an outline of such grammatical
features as are common to the Bantu speech-
family. A list of the most useful grammars,
dictionaries, and other helps towards the
acquisition of particular languages will be
found in the Bibliography.

1 This ph is an aspirated p.

SOME PHONETIC LAWS 231

After embarking on the study ot some one
language, it will be well (though I am aware
that, till the English edition is published, this
is more or less a counsel of perfection) to go
systematically through Meinhof's Lautlehreder
Bantusprachen or, at any rate, Chapters I . to 1 1 1 .
and the one dealing with the language nearest
to that on which the student may be engaged.
But it is a book that cannot be used to much
profit, unless one has some little notion of at
least one Bantu language to begin with. •

As the readers which a book of this kind
can hope to meet with are necessarily limited
in number, and (in one way or another) some-
what specialised in outlook, we may be excused
if, feeling a sort of personal interest and
parting from them not without regret, we
remind them in the words of Brother Hyacinth
that:

1 If any be desirous of learning ....
beyond what is asserted in the preceding
pages for the more easy understanding of
beginners, and their careful recollection, he
ought also sedulously to study and labour in
learning what follows .... and the
preludes of other matters worthy to be known.'

APPENDIX I.

TEXTS

I. ZULU

ta) Why the Rock-Rabbit has no Tail*

Ku-tiwa,1 im-bila1 ya-swela* um-sila4

It is said, rock-rabbit was-without tail

ngo-ku-yalezelas ezi-nye6. Ngokuba7

with-giving-a-message (to) others. Because

na-mhla8 kw-abiwa3 imi-sila la-li-

on the day (when) there were distributed tails it had

buyis-ile10 i-zulu ; za-puma-ke11 ezi-nye

clouded-over the sky ; they went out so (the) others

uku-ya-'u-tata1* imisila lapa13 i-tatwa kona13;

to take tails where they were taken ;

y-ahlul-eka14 e-nye uku-ba i-hambe15 na-zo16,

he was prevented another that he might go with them,

ya-yaleza ezi-lwan-eni17 zonke ezi-ne-misila18

he sent-a-message to-animals all who with tails,

ya-ti, ' 0, nina ba-kwiti19, a-no-ngi-patelaao

he said, ' O, ye our (people), do ye get-for me

owami" umsila; ngi-kohl-we*2 uku-puma

that which is mine tail ; I cannot come-out

em-godi-ni, ngokuba izulu li-ya-na.'

from-hole, because sky is-raining.'

* From Callaway's Nurstry Tales, Traditions and Histories oj the
Zulus (1868), p. 255.

232

APPENDIX I. 233

Za-buya-ke ezi-nye nemisila; leyo yona
They returned, so, others with tails ; that-one he
a i-ba nga i-sa ba" na msila ngo ku enqena
he was not he still being with tail with being-disinclined
uku-puma, izulu li buyisile. Ya-lahla konke
to come-out, sky it has clouded-over. He lost all

oku-hle ngomsila; ngokuba umsila u ya siza
good with a tail ; because tail it helps

eku zi-pung eniM ; ngaloko-ke

in-driving-away-from-oneself (flies) ; and so in this way,

imbila a-i-sa zi pungi nga'

rock-rabbit does not now drive away from himself

luto."

with (any) thing.

Se-kir* izwi eli-kulu loko 'ku-libala kwe-mbila

Now it is word great that loitering of rock-rabbit

ku'bantu" aba-mnyama ; ba-kuluma ngaloko

to people black ; they say with that

'ku tsho kwembila, ku tiwa kw aba nga zi-

saying of rock-rabbit, it is said to those-not-tiring-them-

katazi ngaloko oku-tandwa-yo aba-nye, n aba tsho

selves with that which is liked (by) others, and those who

yo kwabanye, ku tiwa, 'Bani,90

say to others (to act for them) it is said, ' So-and-so,

a w-azi" ukuba loko 'kutsho kwako kw okuti1*,

do you not know that that saying your of saying,

" A-no-ngi-patela " '- a-w azi na ukuba umu-ntu

" Bring for me " — do you not know that a person

ka pat el wa omu-nye, uma into leyo

not is-carried-for (by) another, when thing tluvt

i-lingene35 abakona ? 0 1

it is-enough (only for) those (who are) there ? O !

imbila ya-swela umsila ngokuyalezela.

rock rabbit went without tail by sendiug-message.

234 APPENDIX I.

Nawe, musa ukw-enza"6 njenge-mbila37 ;

And you do not -^^^do^^^' like (the) rock-rabbit ;

ku-yi-ku-zuza88 'lu-to ngokuyalezela ; zi-hamb-eles9

you will not get anything by sending-word; go for yourself

ngokwako.' I-njalo-ke in-daba ye-

as to what ^s yours.' It (is) thus, then, story of

mbila. A-i-kuluma-nga yona ngo-mlomo, ukuti,..

rock-rabbit. He did not speak he with mouth to say,

' A-no-ngi-patela ' ; kwa-Ysla40 izwi kodwa41

' Bring for me ' ; there came forth word only

ngokuba izi-lwane zi ne-raisila,42 kepa yona

because animals they (are) with tails, but he

a-i-na 'msila42; kwa-nga<s ya-swela umsila

not-(is)-with tail ; it (was) as if he went without tail

ngokuyaleza, na ngokuba izulu imbila

by sending-word, and because sky rock-rabbit

i-ya-1-esaba" uma li-buyisile; a-i-

he is fearing it if it has clouded-over ; he does not

pumi emgodini uma li-ng-enzi45 izi-kau

come out from hole if it not making gleams

zoku-sa.

of sunshine.

NOTES

1 Tiwa, passive of ti ' say '; the prefix is that of Class 15,
which is used when there is no definite subject, like our ' it ,
or ' there.'

2 A noun of Class 9, the prefix in- becomes im- before a
labial."

8 Swela ' want,' ' lack.' Ya prefix of the past tense, agreeing
with imbila.

4 Umsila, a noun of the third class; pi. imi-sila, found in
next line but one.

8 Yalezela, applied form of yaleza, ultimately from yala
1 direct,' ' order ' ; yalezu means ' give a message,' and takes the

APPENDIX I. ' 235

thing, not the person, as its direct ohject, whereas yilezel*
takes a direct object of the person by whom, (not to whom)
the message is sent. Ngoku- for iigti-uku- : nga 'by means of,'
prefixed to the infinitive.

8 ezi-nyt agreeing with izi-lwane 8 ' animals,' understood.

7 For nga iikuba, lit. ' with being ' — i.e. ' because.'

8 namhla, for na tiinu hla, often used for ' to-day.'

. 9 abiwa passive of nba ' distribute.' Monosyllabic verbs,
and those beginning with a vowel, make their passive in -iwa
instead of -wa.

10 Agrees with iznlu, which, by a not infrequent exception,
follows its verb. Plupeifect tense; the Perfect having both
the Past (/.£) and present (/»') prefixes before it. Buyisa, lit.
' bring bark,' is the causative of buya ' return' ; the idea being
that the sky ' brings back ' the clouds from below the horizon.

11 ke an enclitic particle, usually rendered ' then,' 'just,' ' so,'
etc. Unlike the interrogative na, it draws forward the accent
of the word to which it is attached, so that we pronounce
zapitmdke, instead of accenting the syllable pu, as would
otherwise be done. It is the subjunctive of the auxiliary ka
(Colenso, First Steps, p. 132).

19 This is a future infinitive, for ukuyakutala : the k of ku is
often dropped.

11 Lapa ' here,' followed by kona, has a relative force, the
two together being equivalent to ' where ' (not the interroga-
tive ' where,' which is-pi). Kona is the pronoun of Class 15 —
or, more properly, of Class 17, which in Zulu has been merged
in 15 — and usually means ' the place,' ' fliere ' (not adverbially,
like lapo, lapaya, but more in the sense of the French y).

14 Neuter-passive of aliliiln ' overpower ' ; the subject is tnyt 9
which seems to refer in a loose way to imbila, properly, it
should be esi-nye 7 agreeing with iti-lu-ane ' animal,' which
seems to be required by the sense.

18 Subjunctive, agreeing with enye. There is no distinction
of tense in this mood.

lc na-zo, agreeing with ezi-nyc (iziltrane) above.

17 Ezi-hvan-eni, locative of izi Iwane. It is not very usual
to have nouns denoting living beings in the locative, though
we sometimes find ebantwini (more often kirf bantu). Here it
is used because yaUza cann.it take a direct object of the
person : it is then-fore equivalent to a dative. Isi-licane for
isi-lo-ane is the diminutive (suffix -aim or -ant) of isi-lo ' wild
animal,' but generally used for ' leopard.' Isilwane means a

236 APPENDIX I.

wild animal in general, — but more especially a carnivorous or
noxious one, whereas in-nyainazane is ' game,' and more
particularly 'buck.'

18 Relative construction (ezi=^a-\-izi ; nemisila=na-}-imisila)
literally ' which they with tails ' — i.e. '• which have tails.'

19 bakwiti or bakiti ' my ' (or ' our ') ' people ' — ki—kn,
followed by the ' prepositional ' form of the personal pronoun
(always in the plural). See Colenso, First S^ps, §91.

*° a-no-ngi-patela, Future Indicative, used authoritatively for
Imperative (First Steps, §222), with a, prefix of Imperative ;
no=niya'u=niyaku : see First Steps, §241.

21 owami, relative form of the possessive^' that which is
mine.' It is generally used for special emphasis, ' my own,'
etc. First Steps, §137.

22 Perf. passive of kohla, properly ' escape,' ' slip (the
memory of) ' — so that the usual meaning of kohlwa is ' forget '
— i.e. 'be escaped' by the thing forgotten. The use here
seems somewhat unusual, but probably means 'coming-out
has escaped me' — i.e. is beyond my power.

28 leyo, demonstrative of Cl. 9 ' that one ' (or rather ' the
aforesaidjjone ') — agreeing with imbila, though the antecedent
is not expressed in this sentence.

24 aibanga negative past of ba ' be ' ; a neg. prefix, i pronoun
agreeing with imbila understood, isaba is best taken as a
participle, sa here=' still '; the two verbs together maybe
taken as a compound tense and translated ' heno longer had.'
Cf. First Steps, §271.

25 For nga-uku-enqcna : itga, instrumental = ' with ' or
' through.'

26 Locative of the infinitive ; zi is the reflexive pronoun.

27 U(lu)-to, usually uncontracted, because of the mono-
syllabic stem. Here nga'luto, not ngolnto, because following
a negative, when the initial vowel is always elided, never
contracted.

26 Verb 'to be ' understood. The subject with which the
pronoun agrees is ukulibala. Concerning se see the chapter
on ' The Particles SA and SE ' in Colenso's First Steps, pp.
112-116, especially §274.

w kiC bantu, more usual than ebaiitu'ini. A little later we
find kivabanyc (u becoming w before ii), not kit' bunye ; but
kit' bantu seems to be preferred, — perhaps because kit-it' bantu
is used with a different shade of meaning — ' at (the house oi)
the people ' — like chez.

APPENDIX I. 237

80 Ubaiti 'who' (interrogative, not relative) is sometimes
used in this way, ' when the name of a person is not known or
not immediately remembered,' or in a familiar and slightly
contemptuous style of address, ' You, sir!' 'you, fellow!' (u
is always dropped in the vocative.)

81 Awazi for a-u-azi, u becoming w before the vowel-stem.
Azi is one of the few verbs which end in i and therefore cannot
change their termination in the negative.

•Ma Kwokuti for Itwa tikuti 'of saying '—the possessive particle
agreeing wilh the preceding loko 'kutsho.

33 The si-ntence breaks off, and the question begins afresh.

84 A'«, not a, is the negative (or the third person singular, or
nouns of the first class ; a being used in the subjunctive
and, in some forms of the relative, as the pronoun of the
third person singular. Patela applied form of pata ' carry '
(in the hand or on the arm — not on the head, which is tu-ula).
This is an instance of the Bantu preference for the passive
when European languages would have the active construction.

84 Lingu ' try,' ' test,' ' strive ' (in some languages, r^.,
Nyanja, it means ' measure ') ; the reciprocal, lingana (perf.
litigene) means ' try or strive with,' 'vie with, ' be as large as,'
and so 'be sufficient for,' as uaibila iilingene labo' bantu 'the
maize is sufficient for those people.' It may take a direct
object, as in tliis sentence, and in the text (abakeua), or be
followed by na. Kulingcne, without a* object, means 'it is
fair, fitting, reasonable.'

88 The Negative Imperative is the Infinitive preceded by
mtisa. I do not know that any satisfactory explanation of this
has been given, butsa is a negative particle in some languages.

8T For njenga (' like') imbila.

* Negative Future (and person sing.) of zuza ' obtain.' Ku
instead of a-u (which would contract into o), probably to avoid
confusion with the same person of the relative.

80 Hantbela here means ' go for,' and not, as most commonly,
1 go to ' '=' visit '). Zt the reflexive pronoun. The Subjunc-
tive is generally used instead of the Imperative, when an
object-pronoun precedes.

40 Past tense (ku-a-vela) with the indefinite subject ku ;
instead of saying izwi la-vela. This construction so exactly
corresponds to our idiom of beginning with ' there ' and letting
the real subject follow the verb, that we can translate quite
literally.

" The adjective -odwa ' only ' (which, like -onkc 'all,' takes

238 APPENDIX I.

the concords of a pronoun, not of an adjective), with the con-
cord of the i5th class — i.e., agreeing with an undefined subject.
It is often used as here, adverbially for ' only,' and thence easily
glides into the very common sense of ' but,' in which it is
synonymous with ktpa.

42 Note the difference between these two words: the first
contracts because the verb (understood) is affirmative ; the
second elides' because it is negative.

43 Nga as an auxiliary is 'used to express a wish or like-
ness ' and is followed by a finite verb. Compare wanga
angawda ' he wished that he might cross.'

44 An unusual order of words, but not unknown.

45 ng' here stands for ngat the negative particle for the
Imperative, Infinitive and Participles: see First Steps, §§259,
263.

Connected Translation

It is said that the rock-rabbit (Hyrax) is without a
tail, because he sent a message through someone else
(instead of going himself). For, on the day when tails
were distributed (to all the animals), the sky clouded
over ; the others went out to get their tails where they
were to be got, but he was prevented from going with
them ; he sent a message to all the animals who had
tails (i.e., all who subsequently received them), saying,
' O ye my people, do ye obtain my tail for me ; I cannot
come out of my hole, because the sky is raining.' So
the others returned with their tails, but he had none,
through being disinclined to come out when it was
cloudy. He lost all the advantage of a tail, for a tail is
useful in driving away flies ; and so the rock-rabbit has
nothing now with which to drive them away.

And so that loitering of the rock-rabbit has become a
great proverb among the black people ; they make use
of that saying of his with regard to those who will not
take any trouble about what other people like, or who
tell other people to do things for them. ' So and so !
As for that saying of yours "just bring it for me,"

APPENDIX I. 239

don't you know that you cannot have anything brought
for you by another person, when there is only just
enough for those who are on the spot ? Oh ! — the rock-
rabbit had to do without a tail because he sent a message.
Do not be like him; you will not get anything by
sending word through another ; go and attend to your
own affairs for yourself.'

This is the story of the rock-rabbit ; he did not
(exactly) say, with his mouth, ' Bring it for me ' ; the
proverb only arose because (the other) animals have
tails, but he has none. It was as though he went without
a tail by sending word and because he feared the
threatening look of the sky ; for he does not come out
of his hole except when the sun shines.

(b) Extracts from Native Letters''

Sa-fika e St. Helena, kwa-t* uba1 ngi pume
We arrived at St. H., it befell when I came out

em-kunj-ini2 aba-kiti ba-jabula kakulu uku ngi

from the ship our people rejoiced greatly to see

bona, nga-puma ngi-nga sa-tandi4 na kancane

me, I came out I no longer wishing even a little

uku-hlala pakati5 kwomkumbi, so-ku ngi-
to stay inside of the ship, it having already

gulisa.8 Yebo-ke, 'Nkosazana ya-kiti

made me ill. Yes, indeed, lady of our (country)

e-tandeka yo,7 ama-kosi lawa akwa'Zulu"

who is worthy to be loved, chiefs these of the Zulus

ay' etanda' kakulu uku finyelela10 England

they are wishing greatly to reach England

a-bone aba-ntu nezwe la-kona

that they may see the people and country of there

* Written from St. Helena, in 1896-7, by a Natal Zulu, employed
as secretary and teacher to the exiled chiefs.

240 APPENDIX I.

nomuzi11 lowo o-dumile yo12 was' e London.18 . .

and town that famous of at London.

Ngezindaba e-zi-vela e-kayau kiti, ngi-zwile

As to news which come from home our, I have heard

ukuti aba-ntwana bami ba-ya-gula15 kakulu ngo-

that children my are ill greatly with

mkuhlane. Nokuti um-kuhlane w-andile1'

fever. And that fever has increased

pakati kwezwe, kodwa-ke kuhle, nje,
in the midst of the land, but yet it is well, indeed,

noma17 ku-njalo,18 ngoba i-kona innyanga

even though it is thus, because there is a doctor

ya-kiti leyo o-y-azi-yo19 nawe ; yena

of our (people) that-one whom you know, you also ; she

u-ya-b-elapa, kambe,20 labo aba-gula-yo,

is treating them, of course, those who are ill,

njengokumiswa21 kwake y'inkosi22

according to the being made-to-stand her by the chief

uyise, aba-ntu ba-ya-m-bonga kakulu,

her father, people are praising her greatly,

ba-ya-jabula nga-ye2* ezwe-ni lonke las'

they are rejoicing on her account in the land all of

Ekukanyeni24. . . .

Ekukanyeni. . . .

Nkosazana — Omunye um-ntwana wenkosi yakwa'

Madam — One child of the chief of

Zulu, u Ndabuko, u-zwile kimi25

The Zulus, Ndabuko, he has heard from me (when)

ngi-m-xoxele28 indaba27 ngawe ya leso' sizwe28

I related to him story from you of that tribe

esi-mnyama o-wa-u-hlezi29 pakati kwa-so,29 na

black which you stayed among them, and

ngezinncwadi lezo o-wa u-zinge u-ngi-tsheleka
from the books those which you used you lending me

APPENDIX I.

zona uku-bangi-funda ngesi kati leso e nga-ngi-gulan

them that I might read at the time that when I was ill
nga-so kiti Ekukanyeni. Nga loko

at it at our (home) at Ekukanyeni. Therefore

u ya-tanda kakulu, u-ya-cela ukuba wena,

he wishes greatly, he asks that you,

Nkosazana, u-m xoxele izin-daba za labo

Madam, you would relate to him affairs of these

'bantu. Ngi-m landisile*J futi uku-ti aba-nye

people. I have narrated to him also that some

baku lezo' zizwe ba-kuluma ngo-limi ' Iwakwa'
of those tribes they speak with tongue of the

Zulu impela, ngitsho1 labo a-ba-biza u Nkulunkulu
Zulus indeed, I mean those who call God

ngokuti ' Mulungu,'3" nabanye aba'lulimi Iwabo*7

by saying ' Mulungu,' and others who tongue their

lu-sondele " kwolwetu ; a nga jabula J u-qobo40
approaches to ours ; he would rejoice in truth

u Ndabuko uma wena, Nkosazana, u-nga-m tumela
Ndabuko if you, Madam, you could send him

in-ncwadi etile" (Book), kumbe" u-m tumele
book some or send him

in-owadi (Note) yo-ku-m xoxela4" indaba ya labo

letter to relate to him story of those

'bantu noku-m-tshela ukuti lezo' zizwe " z'ake"

people and tell him that those tribes they have built

kuyipi46 in-dawo, izi-zwe ezi-ngaki, za-Yela-pi"

in which place, tribes how many? (and) where did they

na?

come from ?

Nkosazana, njengokutembisa kwami ukuti

Madam, as to promising my that

ngi-ya ku-tuma41 ezi-nye izi-bongo zama kosi akwa*

I will send some praises of chiefs of the

242 APPENDIX I.

Zulu kuwe nga lesi 'sitiraela,48 a-ngi-na-wo49

Zulus to you with this steamer, I have not

ama-ndhla ukuba ngi-ku-tumele50 namuhla, kodwa

strength that I might send you to-day, but

ngi-ya-ku-ku-tumela50 ngesitimela esi-za-yo.81

I will send you by the steamer which comes.

Ngi-sa-hamba52 kahle em-zimb-eni wami, uku-gula

I now go will in (as to) body my, illness

loko e-nga-ngi-na-ko53 Ekukanyeni a-ku-ka-ngi-

that which I (was) with it at Ekukanyeni has not again

vuki54 lapa. . . .

arisen (upon) me here. . . .

MAGEMA MAGWAZA.

NOTES

1 Kwati, past tense, like the preceding safika, but with the
indefinite subject kit, 15. An idiomatic use of the verb ti ' say '
(cf. First Steps, §290), which may often be rendered, ' and so,'
or ' once upon a time.' Uba, a contraction of ukuba 'to be,'
is here equivalent to uma ' when.' (First Steps, p. 81.) It is
followed by the subjunctive ngi-pninc.

J tinkunjini, locative of ninkiimbi 3 ' ship.' mb becomes nj in
accordance with the Zulu law that a labial is never followed
by w. There is no apparent w here, such as we find in
endhlw-ini (from indhlu, u becoming w before -ini, but the form
endhl-ini is also in use) ; but it is probably introduced into
the termination by analogy with the u of the stem: nmkumbw-
ini, for umkutnbu-ihi.

3 Abakitilit. 'those of at us,' cf. First Steps, §91, and ante
p. 236, note 19.

4 This is not the Potential Mood (First Steps, §247), but the
Negative Participle (ib., §269). Sa, ' when used with a
negative verb, may be generally expressed by any more, any
longer, at all, etc.' (ib. §271).

6 This is an instance of a word compounded with pa (like
pansi, pezulu, pambili, etc.), though that preposition (or rather
pronoun) is no longer used in Zulu. Um-kati 3 ' space inter-

APPENDIX I. 243

veiling between any two things' preserves the root kati
' between,' which is still so used in Swahili, though not in Zulu.
(There is an adverb kati ' although, ' 'in spite of,' — which may
have the same origin, though possibly a derivative of ti.)

• This is not the infinitive, but a participle, having for its
subject the pronoun ku, which may be the indefinite subject
(' it,' or ' there ') or may agree with nkithlala. ' the staying.1
So=se, the vowel being modified under the influence of u in
the next syllable. Se lias the force of ' now,' ' by this time,'
'already,' etc. (First Stej>s, §§274, ^75.) Giilisa, causative of
gula ' be ill ' ; -ngi- is the object-pronoun, first person,

I Tandeka, neuter-passive of tanda ' love ' (First Steps, §86) ;
for the relative etcnitiekayo, see ib., §132. e=a-}-i is the Relative
Prefix, because agreeing with inbusazana 9.

8 A possessive particle agreeing with amukosi 6 (exceptional
plural of inkosi 9 — see First Steps, §38). Kwti 'Zulu : kwa ' at '
( = French chez) — see First Steps, §92 ; « Zulu (the vowel elided
is probably «=»</«) used for ' the whole Zulu nation ' ; a ku-a'
Zulu is used instead of the locative — as' ezulwuii because the
latter would mean ' in ' (or ' from ') ' the sky ' (ib., §79).

9 A pronoun agreeing with amakosi. A-ya-(lm) ttatuia, lit.
' they are, they loving ' : the verb -ba being understood.

10 finyelela (properly a double applied form of Jinya, but the
latter does not seem to be used in any sense recognisable as
cognate) ' reach, as a traveller a place,' followed by the
locative. (' England,' as it begins wiih e, seems to be treated
as a locative, without further modification.)

II Nomuzi = iu iDituzi 3, ' kraal ; people of a kraal, family ' ;
hence used for ' village ' and ' town.' (Nyanjam/f-t/zi, Swahili
m-ji.)

r- Literally 'that (to.vu) which has thundered ': dum-iU
perf. of duma : the usual expression for ' renowned,' ' glorious,'
etc. Cf. Psalm viii., i : igttma lako lUume kaiigaka tmhlabeni
wonke, lit. ' thy name has thundered how greatly in all the
earth.'

11 s is inserted before the Locative when it follows a
Possessive Particle. See First Steps, §69.

14 Locative of ikaya ' home,' ' dwelling.' Tht word is found
among the ' Nyika ' tribes of Kast Africa (f.g., the Giryama) to
denote the principal (fortified) village of the Tribe. For the
absence of the locative termination, see First Steps, §68.

15 ' Emphatic Present' (Colenso) or ' Present Progressive '
(Bryant) tense.

244 APPENDIX I.

16 For u-andile ; u agreeing with untkuhlant 3, which is
(Bryant) ' a general name for any acute disease accompanied
by fever, such as ague, influenza," etc., etc.

17 noma—na. umci. For lima 'if,' 'when,' etc., see First
Steps, §81.

18 Ku is here (as in the preceding kithle) the indefinite
subject, with the verb ' to be ' understood.

19 Objective Relative (First Steps, §134) y (=yi, for i)
agreeing with innyanga 9. The reference is to Miss Agnes
Colenso.

80 For kambe, see First Steps, p. 75.

21 For njenga ukinniswa. Misa, causative of ma ' stand,'
means, in the first instance ' make to stand,' establish,
' ordain, as a law or custom,' etc., whence the present sense
is easily inferred. The infinitive is here used as a noun of
the i5th class (8th in Zulu grammars), with which the
possessive kivake agrees.

22 The Passive (ukumiswa) is followed by the Copula
denoting the Agent (First Steps, §100 et ?eq., and ante, p. 114).
The ' Chief is Bishop Colenso.

23 ye, pronoun of the third person singular (First Steps,
§104) following, and governed by, uga (ib., §§93-99), and see
ante p. 91.

24 For this locative see First Steps, §69, and ante, p. 243,
note 13.

25 kitni for kuuti, like kiti, etc. — First Steps, §90.

26 X, in Zulu books, stands for the ' lateral click,' made by
pressing the tongue against the side teeth and then with-
drawing it suddenly. Xoxele, perf. of applied form, instead of
xoxel-ile — First Steps, §236.

87 For the various senses of iiulaba, see Colenso's or
Bryant's Dictionary, s.v. Daba (in).

28 isi- xwe ' tribe ' — the root of i(li)-zwe ' country,' -with the
7th prefix — see ante, pp. 45, 55.

29 Relative, with object in an oblique case (First Steps, §134):
viz., pakaii kwa-so (so pronoun agreeing with isi-zive 7). — o-
is the Relative (znd person singular) subject, wa- the subject-
pronoun of the and person combined with (Past) tense-
particle. — Pakati, used prepositionally, is always followed
by kwa.

80 Relative construction similar to owaithlexi. Zinge is a

APPENDIX I. 245

(detective) auxiliary verb, used (First Stef>s, §354) ' to express
"repeatedly," "continually," " habitually," etc.' Zona (agree-
ing with izinncu'iitli 10) is governed by tsheleka, which, like all
verbs of giving, etc. (sse First Steps, §340), takes a doTible
accusative ; but only one objective pronoun can be prefixed
to the verb., viz., here, that of the person, -ngi-.

ni A similar relative, but with the subject in the first person
(prefix <-). The tens» is the Past, which when combined with
a Relative (cf. owauhUzi, above) takes the prefixes both of past
and present (nga-, tigi-). Ngaso agrees with isi-kati 7.

82 Perfect of land-isa, causative of landa ' follow ' : ' make to
follow ' — hence ' narrate.'

83 Literally ' of at those tribes,1 one would have expected
bn lezo'zizive—but the construction is like Iwa kwa'Zulu,
a little further on, and cf. note 8 above. Abanye must be
translated ' some,' or ' others,' according to the context.

84 For nga-t-ii(lit)-linti. The usual word for ' language.'

85 Literally ' I say '

86 Miilnngu is used by the Yaos, Anyanja and other eastern
tribes. It is difficult to believe, with Bleek, that it is the
same word as Unkulitnkitlti, since the latter is plainly derived
from -kitln, a root existing in all the languages where
Mttluiigu is found. Unless, indeed, some other form was
anciently in use among the Zulus, which only became
Unkiiliinkulii through an adaptation of popular etymology.

87 Relative in the Possessive — see First Steps, §133.

88 lu, pronoun agreeing with ulimi n ; sondele, perfect of
sondela (see above, note 26). Sond-ela is properly an applied
form of sonda, which, however, does not seem to be used.

8i) Potential Mood.

40 U(lu)-qobo (q expresses the palatial click), properly ' sub-
stance of a thing,' ' self,' ' person,' ' reality,' is used adverbially
to express 'really and truly' (Coleuso).

41 Tile is an adjective meaning ' certain, when the name or
number is not known ' (Colenso). It takes the prefix o- with
Cl. i, like o-wnyaina, etc. ; hence inncwadi etile, not entile.
The original meanings of inncwadi (Colenso's Dictionary) are :
' mark, made to show whether any one has entered a hut in
the owner's absence ; mark or sign told to a person who
enquires his way by which he will kno,w whether he is going
right or not ; tribal token, as marks ci^t in the skin,' etc., and
hence ' token generally, proof,' and, since the introduction of
writing, ' paper, letter, book.' The writer has bceu compelled

246 APPENDIX I.

to distinguish between the two last-named senses by the
addition of English words. The sense in which he hrre uses
etile seems to be equivalent to ' some ... or other.'

42 kurnbe, ' perhaps, with the idea of hope or expectation '
(Colenso), but also equivalent to the conjunction 'or.' The
latter is often expressed (as in Nyanja) by a word meaning
'perhaps' — the possible alternatives being set before the
mind as conjectures.

43 An example of the quasi-par;iciple mentioned on p. 118,
yoku- =ya uku- : yii referring to ir.iicwadi.

44 ake perf. of c.ka ' build,' which is often used in the sense
of ' live.' Aktlana (reciprocal applied form) means ' to live
near together,' — lit. ' to build for ' (or ' with respect to ') each
other : hence owakelcne ' neighbour.'

46 yipi, interrogative, 'which'? (of two or more), agrees
with indmvo 9 following it. -pi means either ' where ? ' or
' which ? ' (see First Steps, §§169-171). Yipi, as used here is the
object following ku : as subject it would have to be preceded
by the copula (iy'ipi}. -pi 'where?' takes the inseparable
subject-pronoun as prefix : upi ? bapi ? lipi ? ipi? etc.

415 -pi, ' where ? ' is sometimes suffixed to the verb in this
way, and draws the accent forward (zaveldpi).

47 This is the Future, ' I will send,' not the Present, with
object-pronoun, ' I am sending you ' ; tuma in the simple form
cannot take a person as object ; to do this it must be put into
the applied form (tumela), as will be seen a little lower down.

48 An adaptation of the English word ' steamer.' St being
a difficult combination in Zulu, i is inserted between the two
consonants, and the first syllable being taken for the 8th
prefix (isi-), the plural is izi-timela. (There is a genuine Zulu
word isi-timela, meaning ' darkness ' — see Colenso's Dictionary,
s.v., p. 587). The same tendency is observable in Swahiii,
where the Arabic kitab ' book ' becomes ki-tabu, pi. vi-tabn.
Vi-inni has even been heard at Zanzibar, as the plural of
' (lamp-) chimney.'

49 Literally, ' not I with it ' wo ' prepositional form ' of the

pronoun of Cl. 6. Amandhla has no singular.

50 Here tumela takes the direct object of the person, and,
the verb being in the future, ku is repeated, or rather two
different &«-particles follow each other. See note 47 above.

81 esi-, Relative Particle agreeing with isi-timela 8.
M sa may be rendered by ' now,' ' still,' ' already.' See
First Steps, chapter XVI.

APPENDIX I, 247

58 Relative — the construction like engangigula (see note 31),
except that the verb is understood and the whole drawn into
one word. The pronoun -ko refers to uku-gula.

" Vnka ' rise up from a recumbent posture ; . . .
rise in anger, be in a towering passion ' (Colenso). The
personal object -ngi- is unusual with this verb, but may be
used because it is taken in the sense of ' attack,' which is
perhaps not incompatible with the second meaning given
above. In that case, however, one would have expected the
Applied form, vukelu, which is, in fact, so used ; and vuki may
be a mere slip on the writer's part. For the auxiliary ka see
First Steps, §315. .

Connected Translation

We arrived at St. Helena, and when I landed from
the ship, our people were very glad to see me. I also
was very glad to land, having no desire to remain on
board any longer, as I had been very seasick. Yes
indeed ! dear lady coming from our country ! these
chiefs of the Zulus wish very much to come to England,
to see that country and its people, and that famous city
of London. . . .

As to the news which has reached us from home — I
have heard that my children are very ill with fever. In
fact, fever has been very prevalent in the country ; but
there is one very good thing, even though this is the
case — for there is that physician of our people whom
you also know, who is treating the sick according to the
instructions received by her from the Chief, her father.
The people praise her greatly, and they are rejoicing
through her in all the country-side of Ekukanyeni. . . .

Madam, — One of the chiefs of the Zulus, Ndabuko,
has heard from me a story which I related to him,
having heard it from you, of that tribe of black people
among which you (formerly) lived, and (gathered it)
from those books which you used to lend me to read, at
the time when I was ill, at our home, Ekukanyeni.

Therefore he wishes very much to ask that you,
Madam, would relate to him the affairs of those people.

248 APPENDIX I.

I have also told him that some of those tribes speak a
language exactly the same as that of the Zulus — I mean
those who call God by the name of Mulungu — and
others whose language resembles ours (though not quite
the same). He would be very glad if you could send
him some book or other, or perhaps a letter, to give him
an account o£ those people and tell him what place they
live in, and how many tribes there are, and where they
came from. Madam, as to my promise that I would
send you by this mail, some of the traditional songs
praising the Zulu Kings, — I am not able to do so to-day,
but I will send you (some) by the next steamer. I am
now very well in health, as that illness which I had at
Ekukanyeni has not again attacked me here. . . .

2. HERERO.*
Story of the Old Woman -with the Bag

Pa-ri1 omu-kadhe-ndu2 omu-kuru-kadhe,3

There was woman old,

ngu-ya-twa4 ova-natye5 m'ondyatu." E-yuYa7ri-mwe

who put children into bag. Day one

pa rire8 ova-natye, ova-kadhona OY-engi,9 va-ire10

there were children, girls . many, they went

k' oku-nyanda11 k' e-rindi,12 n' a-rire13 tyi14

to play in pool, and it happened that

Ya-hukura otu-Yanda15 n' omi-tombe16 n'

they took off little-skirts and necklaces and

OYi-mbakutu17 n' ou-ndyendye18 n'odho-mbongora19

(see note) and beads and (see note)

n' a-Ye-pundu10 m' omeYa. Kombunda21 omu-atye

and they descended into water. Afterwards child

* Published by C. G. Biittner in Zeitschri/t fiir ajrikanische
Sprachen, Vol. I. (1887).

APPENDIX I. 249

u-mwe wa-tarerew kokure, n' arire ty' a-tara

one she looked far, and it happened that she saw

omu-kadhendu omukurukadhe, ngu n' oka-ti m' eke

woman old, who with little-stick in hand,

oru-horo ti,~' n' ondyatu p' etambo. Nu24 ingwi

a-long-stick, and bag on back. And that

omu-kadhona wa-tyere k' ova-kwawo25 : ' Indyee?1

girl said to (the) others : ' Come

tu-tupukee27 'ka-kurukadhe28 ingwina,

let us run-away-from little old-woman yonder,

ngu-twa ova-natye m' ondyatu.' Indino29 ty 'a-tya30

who puts children into bag. Now when she said

nai, avehe81 arire tyi Ya-piti82 m' e rindi oku-

so all it happened that they came out from pool to

tupuka, nu auhe" wa-torera oru-hira34 r-omu-

run-away, and every she took apron of

kwawo nu omitombe vy-omu-kwawo1" tyinga va-ri

other and necklaces of other as they were

m' oru-haka r-oku-tupuka, ndino arire tyi

in haste of running-away, now it happened that

ya-tupuka k' onganda,R6

they ran to kraal.

N' omuatye umwe wa-dhembire etanda^e-purura'7

And child one she forgot (see note) (see note)

p' ehi.M Ndino 'kakurukadhe arire ty'

on ground. Now little-old-woman it happened that

a-pingene p' epurura, n'a-tora,

she followed (and came) on epurura, and picked-up,

arire ty' a-twa-mo m' ondyatu.89 Nu ing'

it happened that she put-in into bag. And that

omuatye umwe wa-tya: me-yaruka40 me-ka-pura41

child one she said : I return I go to ask

epurura randye k' omu-twa," oka kurukadhe,

epurura my from Bushwoman, little-old-woman,

250 APPENDIX I.

tyiri ! 4S hi n' oku-ri-etha-ko." Irab' ova-kwawo

truly ! not-I with leaving-it-there. Those others

va-tyere : muatye ! arikana,43 omundu eingwi,48

they said : Child ! please (beware of) person that,

ngu, maku-dhu,47 utwa OYanatye m' odhondyatu,

who, it-is- said, she puts children into bags

nil i-ko !4S ; N'e wa-tyere :49 kako ! me-ka-eta

and goes away ! And she said : No ! I go to bring

epurura ra mama oku-kotoka k'omutwa

epurura of my-mother to return from Bushwoman

oka-kuru-kadhe. Nu imb' ovakwawo avehe

little-old-woman. And those others all

Ya-ire k'onganda, n' eye, a- kotoka, a- ende

they went to kraal, and she, she returning, she going

n' a- riri50 n' oma-kono k' otyi-uru. Nu

and she weeping with hands on head. And

ty'a-ri m'ondyira kokure, arire ty'

when she was on road far-away, it happened that

a-ravaere,B1 a-ithana, a-tya : Mu-tyimba,52

she cried-aloud, she called, she said : Pauper,

kakurukadhe, eta nguno epurura ra mama,

little-old-woman, bring here epurura of my mother,

ndi wa-tora, Nu omukadhendu wa-tya:

which you picked-up. And woman she said '

Indyo, kambura. N'e-a-ende, a-me-utuka, arire

Come here, take (it). And she went, she running, and

ty1 a-tumbuka popedhu, n' a-tya : Kakurukadhe,

so she approached near, andshe said: Little old woman,

eta nguno epurura ra mama. N'e wa-tya rukwao :c3

bring here epurura of my mother. And she said again :

Kambura, n' arire ty' a-tumbuka, n' arire

Take (it), and so she approached, and it came to pass

ty' e-mu-tono oru-pyu k' otyi-tama, nu

that she her struck slap on cheek, and

APPENDIX I. 251

okakurukadhe a- hakahana" oku-wira-ko"

little old woman she hastened to fall-upon-her

n' a- petere'* m' ondyatu, n'arire ty' a- kutu

and she doubled (her) up into bag, and so she tied

ondyatu n' omuvia, n' arire ty' a- kutu ondyatu

bag with thong, and so she tied bag

p'etambo, n' a vereka," arire ty'

on back, and she carried (her), it happened that

a i a- tedha ku-kwa-i" ova-natye

she went she followed where-there-went children

k'onganda. Nu m' onganda mwa-vadherwes9

to kraal. And in kraal there-was-arrived,

a-mwa-tu omu-kandi.60 N'e we-ere"

it-there-died feast. And she came (in the)

ongurova n' a- kare kongotwe y- onganda m' okuti.

evening and she sat behind the kraal in the field.

Nu kombunda ova-natye ye-mu-munu,62 arire

And afterwards children they-her-saw, it happened

tyi Ya-raerero ku ihe*3 a-ve-tya : Tate,63

that they cried aloud to their-father they said : Father,

omukadhendu ingwi eingwi" okakurukadhe ngu-

woman this she (is) that little-old-woman who

a-dhepa ova natye nu ngu-a-twa-mo omu-atye wetu

she kills children and who has put in child our

m' ondyatu. Ku va-purire ku ihe a-ve-tya :

into bag. And they-asked from father, saying :

Nga-tu-mu-tyite vi 7s5 Nu ingwi ihe wa-tyere ;

We are to (to) her do what ? And this father he said :

Wererekee6"' ongurova tyi mamu-aruka'7

Catch-with-guile (in) evening •when you (pi.) begin

oku-rara. Nu imb(a) ova-natye ongurova

to sleep. And these children (in the) evening

tyi ma-ve-aruk (a) okurara, ve-ere*8 p'

when they began to sleep, they came to

252 APPENDIX I.

okakurukadhe n' a-ve tya : Mama kakurukadhe,

little-old-woman and they said : Mother, little-old-woman,

mo-Yanga tyike,09 ku-tya tu-ku-pe ? N'e

you want what that we may give you ? And she

wa-tyere : namba70 ami me-vanga tyike ? vanatye

she said : now, for my part, I want what ? children

Yandye, ke-ndyi-pahere71 uri72 orukune (o) ru-nene ;

my, go (for) me look-for just log-of-wood large ;

mba t'7Sombepera. N'owo Ya-ire, arire

I am dead (with) cold. And they, they went, it happened

tyi Ya-ka-paha oru-kune (o)ru-nene, ndu

that they went-to-seek log large, which

Ya-muna rukuru,74 n'arire tyi Ya-tora

they saw long-ago, and it happened that they-lifted

omumbeumbeu,75 n'arire tyi Ya-eta,

all-together and it happened that they brought (it),

a-Ye-tya ; kakurukadhe, twe-ku-etere76

saying: Little-old-woman, we-to you-have-brought

orukune oru-twedhu,™ ndu-rara n' omundu,

log thick, which sleeps with person,

omukadhendu okakurukadhe o-tya ove,78

woman old as you,

n' a-yanyuka oviandonya.79 N'arire

and she stretches (herself) out (on) back. And

tyi Ye-mu-etere orukune. M' ou-tuku

so they (to) her brought log. In night

ty' a-rara, ova-natye arire tyi Ye-kutura80

when she slept/ children and so they untie

ondyatu, n' arire tyi Ya-itha-mo omu-atye

bag, and so they take-out-of-it <5hild

n'oyi-na OYi-tyuma*1 nu m'ondyatu m'

and things vessels and into bag into

otjipurukute82 arire tyi Ya-ongere-mo

dry bag it happened that they collected into (it)

APPENDIX I. 253

ou-puka," ngamwa,84 'kapuka ke-rumata akehe.

animals, all sorts, animal it bites everyone.

Nu ondyatu otyi-purukute, arire tyi va-kutu rukwao,

And bag dry bag, so they tied again,

n' owo a-ye-i k'onganda n' a-ve-twara omu-atye

and they went to kraal and they carried child

n' oYi-na mbf6 va-ithire m' ondyatu, nu

and things which they had taken-out from the bag, and

ve-vi-twarere ku ihe. Ihe

they-then-brought to the father. The father

wa-dhepere omukandi, a-koho nao*7

he killed (a beast for) the feast, he deansed with it

omu-atye. Nu kombunda 'kakurukadhe

the child. And afterwards the old woman

arire ty' a-nununga ondyatu, a-tyangoYathi,^

it happened that she felt the bag, she thought,

omu-atye om' e rf9 arire ty' a-kutura

the child in it she was and so she untied

ondyatu n'e wa-tire omadhenge

the bag and she was (nearly) dead (with) rage

tyinene, kutya™ OYanatye va-ithire-mo m'ondyatu.

truly, that the children bad taken-out from the bag.

Ndino oupuka arire tyi wa-sakumukire

Now the animals it happened that they crawled-out

mu-ye, n'arire tyi wa-hiti m'orutu rwe aruhe,

on her, and so they entered into body her whole,

m' otji-nyo na m' oma-yuru na m' omeho, n'arire

into mouth and into nostrils and into eyes, and so

ty'a-koka. Oputyo.01

she ended. This is all.

254 APPENDIX I.

NOTES

1 Pa locative rprefix ; ri verb 'to be.' Part is the perfect
tense, the one with the suffix corresponding to -He is a ' Plu-
perfect,' or distant past. In the Present, the prefix would be
PC, not pa.

* Herero has a somewhat peculiar way of forming coin-
pounds. Instead of saying oniu-ndn omu-kadht ' female
person,' or using omu-kadhe by itself as a noun, ' woman,' the
root -ndu is suffixed. See ante, p. 215, and note 23, on
oruhoroti ; also Meinhof, Laittlehre dtr Bantusprachen, p.
135. Another curious feature, to some extent analogous to
the above, is the insertion of the interrogative particle ke
between prefix and root, as omu-he-ti, 'what sort of tree?'
omu-kc ndu, ' what sort of person ? '

8 -kadhc is suffixed to omu-kuru, which by itself means ' old
(person),' in order to indicate the feminine, -kazi is similarly
used in Zulu (as in indoda-kiizi ' daughter,' inkosi-kazi
' queen ' : there is no independent word uin-kasi), though less
frequently. The Herero are supposed to have a mixture of
Hamitic blood, or at any rale to have been in contact with
Hamitic tribes (e.g., the Galia or Somali) before their south-
ward migration, and they might have borrowed from them
the notion of a feminine suffix — which is quite foreign to the
genius of the Bantu languages.

4 ngit relative pronoun of the third person. In the Present
it immediately precedes the verb-stem; ngu-twa ' who puts ';
in the Perfect it is followed by -a : ngu-a-tu-a (or ngu-ya-tva ;
the y no doubt introduced to prevent the two syllables from
gliding into ngwa) ' who put.'

5 The singular of this noun is omu-atye, the «, which, as we
know from other languages, belongs to the root, seems to
have dropped out.

6 ondyatu 9 is a leather bag or wallet, carried over the
shoulder by people who go out to collect roots, etc.

7 The 5th prefix is in Herero abbreviated to e (as in Zulu to
i) : its full form is eri, the pronoun ri. Ejuva is the same
word as Sango lidynva, Nyanja dzuwa, Swahili jua, etc.

8 Rira ' become,' ' be,' (rire is the Historic Aorist) ; for its
idiomatic use as an auxiliary, see below.

APPENDIX I. 255

9 ov-cngi for ova-ingi, adjective agreeing with ovanatye. ova-
kmUiona ' girls ' is. though a noun, practically equivalent to an
adjective, being placed in apposition with ovanatye.

10 ire, pluperfect of the defective verb ya ' go.'

11 The frequent use of kit, even where it would seem super-
fluous, as here before the Infinitive, seems a peculiarity of
Herero.

1J Same root as Swahili and Pokomodindi ' hole ' or ' pit ' —
more especially applied to a deep place in the bed of a river
or the sea. It also appears in such place names as Lindi,
Malindi, Kilindini (the harbour at the souih end of Mombasa
Island), etc. The Herero use erindi to mean what is called
in South Africa a ' pan ' — i.e., a depression in which water
collects during the rains, drying up partially or wholly after
they are over.

13 a-rire, followed by tyi is equivalent to ' it happened that,'
' it came to pass that,' etc., — or merely ' and so.' The pronoun
a (instead of «) is prefixed to the 'Historic Aorist ' and the
Subjunctive.

14 tyi 'say,' like Zulu ti, here used as a conjunction (cf.
Zulu

15 PI. of oru-vanda 12 ; the singular is not used in this sense.
The word means a kind of apron or kilt worn by little j^irls
(under 15 or 16) and consisting of a number of hide thongs (in
Cape Dutch rim-pies), hanging from a belt. In front these
reach the feet, — behind they are long enough to sweep the
ground. A more ornamental kind of okavanda is the otjim-
bakutu (pi. ovimbakittu), mentioned a little lower down, which
consists entirely of omitombe (see next note).

18 omitombe 4 are strings of small disks cut from the shells of
ostrich-e^gs, and rounded by rubbing their edges oil a stone.
As the process of preparing these ' beads ' is slow and
tedious, they are highly valued. They are worn, either in
single strings, as necklaces, or the strings are looped together
to forma sort of bodice, called omutombe 3.

17 ovimbakittu, see note 15.

18 oundyendyt 14, (imported) glass beads, usually worn in
strings round the neck.

19 PI. of ombongora, g, a. string of disks 'similar to the
omitombt (see note 16), but made from the shells of snails or
other molluscs.

256 APPENDIX I.

20 3rd pers. pi. ' emphatic aorist ' of punda ' descend ' ; in
this tense the fiual vowel is assimilated to that of the stem.

21 An adverb composed of kit and ombunda g ' the back ' —
therefore ' behind' or ' after.' In Herero, the u of ku is often
elided before another vowel, instead of turning into w.

22 3rd pers. sing., pluperfect of tar a ' look.'

23 oni-lwroti is a compound of omu-ti analogous toomu»kadhe-
ndu. It means ' a long stick,' and is used in apposition with
oka-ti, so that it is practical!}' an adjective=' long.' But
Brincker's Dictionary does not give oruhoro in any sense
which would imply this.

24 Nil is used to join sentences (or, in other words before a
verb) — na nouns.

25 -kuao (=kwawo) is given in the grammars and dictionaries
as an adjective meaning ' other,' but really it is the possessive
pronoun of the 3rd pers. pi. agreeing with Class 15. All three
persons are used with the prefixes of Class i and Class 2 : —
omu-kwctu 'my (our) companion, house-mate, person of the
same village, etc.,' pi. ova-kwettt, omu-kwtnu 'you, etc.,' omti-
kwawo ' his, her, their, etc.' — like Zulu abakiti (see p. 236 ante)
of which, however, there is no singular form corresponding to
omu-kwftit.

26 Imperative plural of ya ' come.'

27 tupnkee, applied form of tiipnka, taking the direct object
(o) kakurukadlie.

28 Diminutive of otnu-kiirukadhe (nole 3). Compare the use
of kizee in Swahili for an old woman, mostly used of a witch
or other uncanny person.

M Indino, properly a demonstrative agreeing with eyuva 5
' day ' (lit. ' sun ') : ' this day,' and so ' now.'

80 Tyi here used in the sense of ' when.' Tya is the form
used as an independent verb, when the meaning is actually
' say.'

81 avehe ' all,' agreeing with Cl. 2. The root is lie which
always prefixes a- followed by the personal or class-pronoun :
a-tu-he ' all of us,' a-mu-he ' all of you, etc.'

82 ' Historic Aorist' (Viehe), one of the tenses which
assimilates its final vowel to that of the stem — cf. pitndu
(note 20).

88 Aiihe : -he 'all ' agreeing with Cl. i — singular of avehe,

APPENDIX I. 257

84 oruhira n, a goat-skin apron worn next the skin, the other
articles mentioned being put on over it. The initial r of the
next word stands for ra, the nth possessive particle, agreeing
with ortihira. This elides the a (instead of combining it, as in
Zulu, with the initial vowel of the noun) — no doubt because
the initial u has already been modified to o (as is also the
case in Ganda).

86 The possessive particle of Class 3 is vya. (not as in most
languages y*), preserving a hint of the original y. Nu before
omttombc seems to contravene the rule given in note 24, but
may be a printer's error.

86 onganda g is the word generally used for ' kraal,' ' village ' ;
the word used in S.W.Africa is wtrft (Cape Dutch, though in
this sense it seems to be peculiar to that district) — see
Pettman, Africandcrisms, p. 550. Tha Zulu timuzi 3
represents the word used in most Bantu languages ; it is
found in in Herero as oru-dhc n, meaning ' principal
village.' Brincker translates onganda by Viehdorj ' cattle-
village,' which among the pastoral Herero would be the
normal type of settlement. A village without cattle is ond.ua
(which, ex hypothesi, appears to be a Nama village) or otjihuro.

87 etanda and cpitrura appear to be more or less synonymous
and consist of strings of iron and copper ' beads,' or hollow
balls, fastened to the lower edge of the omutotnbt.

88 ehi 5 is the word which appears in Swahili as nti 9 ; in
Zulu, Nyanja, etc., as pa-nsi (it is not used by itself) ; just as
-he ' all," corresponds to Nyanja -onsc and Zulu -onke. But it
is something of a puzzle that Zulu should have the ns in pansi,
and so is the elimination of the vowel in Herero).

89 Herero, as we have seen, prefixes pa, ku or mil to a noun,
and does not possess the suffixed locative of Zuhu etc. It
also suffixes the pronouns -po, -ko -mo to the verb accompany-
ing the noun — a usage also found in Nyanja, where the
noun, moreover, frequently takes both prefix and suffix as
m'nyumba-mo ' in the house.'

40 The inseparable pronoun in Herero varies to an extra-
ordinary degree. ' I ' is me- with the Present, e- with the
Aorist, mba- with the Perfect, and with the 'Jussive,' ng'e- or
hi- ; while it also has a distinct object-form nJyi-.

41 The particle ka has a ' directive force ' as mekatona ' I go
(to) strike ' — i.e., ' I am goinj to strike.'

43 Omu-twa, pi. ova-two, (cf. Zulu timu-twa, aba-two) originally
meant ' Bushman,' but seems to be used in a depreciatory

258 APPENDIX I.

sense of any non-Herero, and hence with the meaning of
'slave,' 'bondsman,' etc. Omii-tyimba, applied further on to
the same old woman, is used by the Herero of people who
have no cattle, but pick up a living as they can in the Bush
(and so = ' pauper ') ; but other tribes apply it to the Herero
themselves.

43 tyiri, invariable, is called by Brincker an ' interjection of
assurance.'

44 edha =' leave '; 'ri refers to epurura, ko, locative pronoun
(17).

45 arikana an exclamation, of entreaty, variously rendered
according to the context. Similar expressions are found
elsewhere — e.g., the Yao chonde ! It looks like the imperative
of a reciprocal verb, but none such appears to be now in use.

46 t-ingwi for eye ingwi. Eye is frequently contracted to e.

47 dim from dim ' come from ' means, with the indefinite
subject (kit) ' it is said,' ma-ku-dliu is the tense called by
Brincker the ' Simple Present,' which prefixes ma to all its
pronouns.

48 i is the aorist of the defective verb ya ' go ' ; ko the
locative pronoun, here best rendered by ' away,' but a better
equivalent would be the French en in s'cn aller.

49 e(ye), or eye, separable personal pronoun : tyere,
pluperfect of tya.

50 a-riri: this, like the preceding verbs is a participle, which
in the simple form always assimilates the final vowel. Rira
' weep ' (Zulu lila, Nyanja lira, Swahili Ha, Pokomo ia) must
not be confused with rira ' become.'

51 Ravaera, applied form of rava ' thrust in ' (used, e.g., of
Moses putting his hand into his bosom, etc.). The sense of
' crying aloud ' is derived, according to Brincker, from that of
' thrusting the tongue into the throat ' (stark die Zitnge in die
Kehle sleeken] ; in the applied form ' cry aloud to ' some one,
ithana (cf. Nyanja itana) is properly a reciprocal ; Swahili has
the simple form ita.

51 See note 42. o is elided in this word and the next, because
they are in the vocative.

58 rukwao, used as an invariable adverb, ' again,' but really
an adjective agreeing with oru-vedhc 'time,'=' another time.'

84 Historic Aorist. The rule of vowel-assimilation is not
usually applied to verbs of more than two syllables, but there
are some exceptions.

APPENDIX I. 259

54 Ko locative preposition. According to the usage of most
Bantu languages, one would have expected oku-mu-wira.

88 Pluperfect of peta, ' bend,' Swahili pda ' bend,' ' curve,'
from which comes pete ' ring.'

67 vereka means to carry on th« back, as native women do
babies : bereka is similarly used in Nyanja, and belcka (or
beleta) in Zulu, where im- beleko is the prepared goat-skin used
for tying the child on.

68 Relative of the ku- class in the past. We must under-
stand something like ' to the place,' or ' at the time ' after

a-tedlia.

59 mwa- is the locative pronoun for the past tenses, the
subject of the verb being m'onganda. vadherwe is the passive
of the applied form of vadha ' reach.'

00 Onitt-kandi is a feast of meat, when a bullock is slaughtered
on special occasions. The feast is said to ' die ' because it
was just ending. A -mwa-tu seems to be a mistake for a-mu-tu,
which is the Historical Aorist of ta ' die.' (This verb has a
dental t, which distinguishes it from ta ' be equal with.')

61 ere plup. of ya ' come,' which takes we instead of wa for
the pronoun of the 3rd pers. sing, in the past tenses. This
and the aorist ya (instead of i) distinguisli it from ya ' go
away.' Oku-ti is really the locative (17) which in Herero is
merged into the infinitive class (15) — see ante, p. 85. It
means the open country — in fact is best translated by the
Dutch veld.

62 >nu is here the objective pronoun of the ist class, not the
locative prefix.

ca ihe ' his, her, their father ' — cf. Zulu uyise. ' Your father '
is iho (Zulu uyihlo) ; ' my father' tate (cf. Nyanja tale, atate).
This form is found in a good many languages, while others,
like Zulu and Swahili, prefer (u)baba.

64 e = eye : eye ingwe ' she (is) this ' (or ' that ') one.

65 Subjunctive. In principal sentences (as here) this has
nga- prefixed to it. For tyita 'do,' cf. Nyanja chita. vi is an
invariable interrogative.

86 Brincker translates wercreka by ' do a thing treacherously ;
(verraUrisch etu'as tun).

67 Second person plural of the Present, which prefixes ma
to all the pronouns, though in the three persons of the
singular it is contracted with them into mf, wo, ;;M.

68 See ante, note 61.

260 APPENDIX I.

69 Tyike ' what ? ' stands by itself after a verb, as here.
When the question is asked, ' what is (are) he (it, they) ? '
— -kwatyike is used with the proper class-prefix : omukuia-
tyike, ovakwatyike, otyiku-atyike, etc.

70 In the original n'amba, which, Prof. Meinhof tells me, is
a mistake. ' It should be namba " now," which is derived from
P<*' ('at'). ,

71 Pahere (e) imperative plural of applied form (pahcra) of
paha ' seek.' Ke is the ' directive particle ' ka, which modifies
its vowel when followed by an object-pronoun. ' Go and
seek for me a log.'

72 un an invariable (adverbial) particle, equivalent to ' just,'
' only,' ' so,' etc.

73 For hi, perfect of tal die'; ombepera, a noun of thegth class.

74 Originally an adjective (' old ') agreeing with oru-vedhc, cf.
note 53, on rukwao. A fairly good supply of firewood is to be
obtained from the mimosas and acacias of the Herero country
(Biittner). People are always on the look-out for dead logs
which will burn easily, and, if they see one, note the place so
that they can return for it when wanted. These girls
remembered that they had noticed one in the bush some
time before.

76 Translated by Brincker and Viehe 'zusammen,' ' gemein-
sam ' ; it is evidently a noun of Class 3, but the original
meaning is nowhere given.

76 Applied form of eta (Zulu leta) ' bring,' which enables it
to take the direct object (-ku- object-pronoun 2nd pers. sing.).
Twe, instead of tica, because a always becomes e before the
object-pronoun.

77 Properly ' a log like a bull.'

78 i.e., ' which an old woman like you can sleep with':
meaning that it is large enough to burn through the night, so
that a person can sleep comfortably, without getting up to
put wood on the fire.

79 Properly a plural noun of Class 8, but it only seems to
be used adverbially, ondonya 9 is both noun (' back ') and
adverb (' behind').

80 Reversive of kuta ' tie.'

81 A general word for ' vessels,' ' implements,' ' household
stuff.' Otuma, in Yao, means ' beads,' (applied in Nyanja
to property of any sort) ; in Swahili, ' iron ' — but it is not
certain that the three are the same word.

APPENDIX I. 261

M This seems to be a descriptive epithet applied to the bag
and to mean anything made of hide which is hard and dry
and rattles.

88 PI. of tka-puka (diminutive of otji-pitka), applied to small
animals and insects.

84 ngainwa, indefinite numeral meaning ' all sorts,' ' of any
kind whatever.'

85 ' Every biting animal.' rumata ' bite ' : the simple form,
ruma does not seem to be used in Herero. Ke pronoun of
Cl. 13 with the present tense.

80 Relative Pronoun of Class 8, agreeing with its antecedent
ovina.

87 'with it' seems to refer to omukandi. No doubt some
sort of ceremonial purification is intended, to free the girl
from any evil influences which may have emanated from the
old woman. The sentence seems to mean that the. father
killed an extra beast (wa-dhepere imperfect of the Applied
Form, not Pluperfect) as part of the omukandi, which was not
yet finished.

88 This and some allied forms are derived from tya ' say '
and ndovathi (ndevadhi) ' perhaps,' ' if haply,' and mean ' think,'
' be of opinion that.' ....

89 Oin'eri contracted from omueye uri. (Information kindly
urnished by Professor Meinhof, and see Brincker, IVurterbuch,

P- 83).
00 kutya used synonymously with tyi.

91 Opn, locative adverb, ' there,' ' in that place.' Combined
with a pronoun, as here with -lyo (Cl. 7) it means ' that's all '
(literally ' there (is) that ') — i.e., ' this is the end of the story.'

Connected Translation

There was once an old woman who used to put
children into her bag (and carry them off). One day, a
number of girls went to play in a pool ; they t»ok off
their clothes and ornaments and went clown into the
water. After a time, one of them, looking out to a
distance, happened to see an old woman who had a long

262 APPENDIX i.

stick in her hand and a bag on her back. So the girl
said to her companions : ' Come, let us run away from
the little old woman yonder, who carries off children in
her bag.' When she spoke thus, all of them came out
of the pool, in order to run away, and every one picked
up the apron and the necklaces of her companion, as they
were in sucli haste to run away. So they ran as fast as
they could back to the village. But one child forgot
her epurura and left it lying on the ground; and the old
woman went up to it, picked it up, and put it into her
bag. So the girl said, ' I am going back to ask that
old Bushwoman for my epurura ; I am not going to
leave it there, truly ! ' But the others said, ' Please
don't, dear! — they say that person puts children into
bags and goes away with them ! ' She answered, ' No,
I must get my mother's epurura back from that old
Bushwoman.' So all her companions went hbme, but
she turned back, and walked along, crying, with her
hands on her head. And while she was on the path,
she called out to the old woman, a long way off, saying,
' You horrid old pauper ! bring me my mother's epurura,
which you have picked up.' The old woman said,
' Come here and take it.' The girl ran up to her, and
when she was quite near, said again : ' Old woman, bring
my mother's epurura here ! ' The old woman said
again, 'Take it ! ' and when the girl came close to her,
she slapped her on the cheek." And then the old
woman made haste and seized her and thrust her into
the bag, and then tied up the bag with hide thongs and
fastened it on her back and carried it so, and went on in
the direction which the girls had fe-ken to reach their
village, where a great feast had been going on and was
nearly ended. The old woman arrived there in the
evening and sat down outside the fence in the open
field. When, later on, the girls saw her, they called out

* It is not clear from the t-jxt, as it s'ands, whether it was the
old woman who slapped ths girl, or vice versa. But the women
who related the story to Biittner insisted that the former was the
case.

APPENDIX I. 263

to their father, saying, ' Father, the old woman out there
is the one who kills children, and she has put our child
into her bag.' And they asked their father, saying,
' What shall we do to her ? ' And their father said,
' Wait till the evening, when you are all thinking of
going to sleep, and then you can entrap her.' So, in
the evening, before those girls lay down to sleep, they
came to the old woman (outside the kraal fence), and
said, ' Mother, what would you like us to get for you ? '
And she said, ' For my part, my children, what I should
like is that you should just find me a good big log (to
burn), for I am well-nigh dead with cold.' So they
went to look for a large log which they had marked
down some time before, and lifted it all together and
carried it back and said, ' Little old woman, we have
brought you a regular whopper of a log, such that a
woman can sleep all night beside it, lying comfortably
on her back.' So they brought her the log (and put it
on the fire).:''

But in the night, when she was asleep, the girls went
and untied the bag and took out the child and everything
else that was in it, and they collected and put into it all
kinds of biting insects and reptiles, — every creature that
bites. And they tied up the bag again and went into
the kraal, carrying with them the little girl and the
things which they had taken out of the bag, and brought
them to their father. And he killed another bullock for
the feast, so as to purify his daughter with it. But,
after a time, the old woman got up and felt the bag,
thinking the girl was inside it (but she was not there),
so she untied it and nearly died of rage, indeed, because
the other girls had taken her out. So then all the
animals crawled out on her and got into her mouth and
nose and eyes (and stung her to death), and that was
the end of her. That is all.

* It is to be supposed that the old woman had made a little fire
for herself with such dry sticks as she could find, and only wanted
fuel to keep it up through the night.

264 APPENDIX I.

3. ILA*
The Tortoise and the Hare

Ba-nyama1 bonse' ba ka-fwe2 nyotwa,8

Theanimals all, when they were about todie (of) thirst,
ba-amb,' ati:4 ' A-tu-lukanke lubilo,

they spoke, saying : ' Let us run (with) swiftness,

tu-bone ati a-ka-shike5 ku menzhi.'

(so that) we may see that he may arrive at water.'

Pele, Fulwe ngu a-ka-zhala6 bana

But Tortoise (it is he) who produced children

banjibanji: u-la-ya-bu-zhika7 mwivu;8 umwi

very many : he goes burying (them) in ground ; one

mwana wa-mu zhika, ku-mbadi ku9 menzhi.
child he him buried, by side of water.

Inzho banyama baamb', ati: ' A-tu-tiane,10

So the animals they spoke, saying : ' Let us race,

tu-ka-shike ku mu-longa, tu-ka-nwe menzhi.'

when we arrive at the river, let us drink water.'

Ba-fuma, ba-lukanka, bonse baamb', ati :

They started, they ran, all they spoke, saying :

' Tu-bone11 ati nguni12 u-ka-tanguna13 ku-shika.'

'Let us see that who is it he will be-first to arrive.'

Pele ba-lukanka odimwi,14 ba-fulwe ba-la-ya-

But they ran again, tortoises go along

bu-amb'uati: ' Imbelembele, o-ba-shana-

saying : 'Forward, those who are with

* Ila is spoken in North-Western Rhodesia, by the people
commonly called the Mashnkulumbwe, whose proper name is l3a-ila.
They live on both banks of the Kafue, one of the northern
tributaries of the Zambezi. They are closely allied (at any rate as
far as speech goes) with the Batonga and Basubiya. The above
story is extracted from the Rev. E. W. Smith's Handbook of the
Ila Language (see Bibliography).

- APPENDIX I. 265

sulwe.'"" Odimwi balukanka, odimwi baamb' ati :

Mr. Hare.' Again they ran, again they said :

1 Imbelembele, obashanasulwe ! ' Dimwi izuba"
' Forward, Hare & Co. ! ' Another sun

dia ibila, ba la ya bu-ompolola : ' Dimwi18

it set, they go along shouting : ' Another (day)

kwa shia.1 Imbelembele, obashanasulwe !

it has grown dark. Forward, Hare & Co. ! '

Dimwi banyama bamana kufwa,20 mwana21 fulwe

Next day the animals finished to die, child (of) tortoise

o-wa-ku-di" kumbadi ku menzhi, wo-ompolola28

he who was beside the water, he shouted

ati : ' Imbelembele, obashanasulwe ! ' Wezo

saying: ' Forward, Hare & Co. ! ' That

Sulwe wa-ya ku-fwa,24 wa-bula25 o-ku-shika ku

Hare was going to die, he was- without arriving at

menzhi. Mwana fulwe owakudi kumbadi

water. Child (of) tortoise who was beside

ku menzhi wa-ba-letelela26 menzhi mu-kanwa :

water he brought-for them water in mouth :

ke-ziza27 ku-lapwila28 banyama, Ati :

let him come to spit-out-for animals. He said

, Ndimwe mwa-ku-zumanana, ati, " Fulwe

It is you you were quarrelling, saying, " Tortoise,

tu-la mu-shia29 lubilo." Inzho

we shall him leave (behind) (in) swiftness." So

mwa-ba-nji30 ku-shika ? Mu-di ba-nichi.31

you have become what to arrive ? You are children.

Ndimi mukando, nda-shika ku menzhi.

It is I (who am) a big man, I have arrived at water.

Mudi banichi.' Ngonao"2 wa-ba-lapwila

You are children.' Immediately he spit-out-for them

menzhi a-ku-di mu-kanwa. Ba-bula

water it was in (his) mouth. They were without

266 APPENDIX I.

o-ku-mn-ngula : ba-usa budio. Inzho

answering him : they were-sad only. So

banyama baamb' ati : ' Tu-ka-fumbe84 mu-kalo,

animals they said : ' Let us dig water-hole,

tu-ka-ku-nwa85 u^-mukalo menzhi.' Inzho

that we may drink in water-hole water.' So

ba-fumba. Basulwe ba-kaka kufumba, inzho

they dug. Hares they refused to dig, so

baamb' ati : ' Bus7 mwa-kaka kufumba, inzho

they said : Since you refused to dig, so

ta-mu-ti-mu-nwe38 menzhi. Mu-la-mana

you shall not drink water. You shall finish

ku-fwa nyotwa.' Kwa-shia,

to die (of) thirst.' It grew dark,

bakaka kufumba, ba-ya ku mukalo,

they (who) refused to dig, they went to hole,

ba-kwiba.39 Inzho banyama bamwi baamb' ati:

they stole. So animals other they said :

'A-tu-ba-zube40 basulwe, tu-ba-

'Let us hide (from) them the hares, (that) we may see

bone.' Inzho ba-ba-bona, ba-ba-kwata,

them.' So they saw them, they caught them,

ba-ba-anga. Pele baamb' ati : ' Bu mwa-

they bound them. But they said : ' Since you

tu-anga, inzho twa-beba. A-mu-

(have) tied us, so we (have) repented. Let you

tu-tole41 a bwina, mu-ka-tu yayile42 ngona.'43

carry us to burrow, (that) you may kill us just there.'

APPENDIX I. 267

NOTES

1 Nyama, in most Bantu languages, is a noun of the ninth
class (though, in Swahili, when meaning ' an animal,' it usually
takes the concords of the first). In Ila it has the prefix mu-
(pl. ba-nyama 2) and is thus included in the person-class.
Other names of animals are treated in the same way, c.g.,
ntitnyati ' buffalo' (Z. innyatig ; Ny. njati) mu-sefu 'eland,' mu-
zovu ' elephant.' Some names of animals (also included in
Class i) are compounded with the prefixes s/f«=' father of
and »« = ' mother of (Smith, p. 18). This seems to be dis-
tinct from the use of ska- or shana- as an honorific prefix, e.g.,
shana-sulwe, for sulwe i ' hare.' (Sulwe, ordinarily, has no
prefix in the singular, but, being included in Cl. i on account
of its moaning, its plural is ba-sitlwc 2.) This is very common in
African tales — e.g., in Ganda 'elephant' is enjovu 9, but
when he figures in a story he is called Wanjovu, and in Yao
stories the names of animals have the title Cht prefixed to
them. (Cf. ' Brer Rabbit,' ' Miss Cow,' etc.)

8 Future Subjunctive (Smith, p. 161). This seems, incon-
gruously, to be used as a principal verb, but in reality it is
equivalent to an adverbial clause, with 'when' understood.
' The relation of time is often expressed not by an adverb,
but by moods and tenses of the verb. . . . [e.g.] the
preterite indicative and the subjunctive.' (Smith, p. 240).

8 Nyota means ' thirst ' in some dialects of Swahili, also in
Nyamwezi, Karanga, etc. Cf. also Gauda cnyonta, Yao njota,
Herero onyota, Sutu Itnyora.

4 Ati, properly 3rd pers. sing, of ti ' say,' is used regardless
of number or person, in a way which comes to be equivalent
to the conjunctive ' that.' For other idiomatic uses of ti, see
Smith, p. 185.

6 Lit. ' that he may arrive (first) at the water,' — i.e., ' who
will arrive first.'

8 Past (Preterite) of zhala 'bear,' 'beget': Zulu, sala,
Swahili vyaa (zaa).

T Properly the ' Immediate Future Habitual' tense (Smith,
p. 156) : the narrator goes back to the actual time of the
incidents and treats them as if they were happening before
his eyes.

8 Locative,=w» ivu ' in the ground.' Mr. Smith spells

268 APPENDIX I.

mwivhu, but the sound — see p. 7 of his Grammar — is clearly
that of ' bilabial v.'

9 We should have expected kumbadi kwa menzhi, mbadi 9
being placed in Cl. 17 by prefixing the locative ku, but see
Smith, p. 223. Mbadi is not given in the vocabulary as a
noun, but cf. Nyanja mball g ' edge, side, rim ; ' no doubt the
same word used in Swahili as an adverb ' far,' etc.

10 tia ' be afraid ' (cf. Swahili tisha ' frighten ' — probably a
causative of the verb usually written cha 'fear '), and so ' run
away ' : tiana reciprocal, but apparently with the meaning
' run against ' (or, ' in competition with ') and not ' run away
from ' each other.

11 Present Subjunctive, used in place of Imperative.

12 Nguni, interrogative (Smith, p. 101), lit. : ' it is who ? '. A
relative is understood after it, or rather, it is an example of a
construction very common in the Bantu languages, even
where relative pronouns exist : the demonstrative, or even
the ordinary personal pronoun are often preferred, as though
it were less trouble to make a freslj assertion than to link up
the clause with the preceding one.

13 Second Future (Smith, p. 157) — probably distinguished
from the Preterite by tone. — tanguna, evidently a derived form
of tango, ' begin,' but the force of the termination is not very
clear: it can scarcely be reversive (Smith, p. 130).

14 Dimwi ' another ' (agreeing with izuba 5 ' day,' or some
similar noun, understood), and preceded by the instrumental
preposition o (Smith, p. 224), so meaning ' again.'

15 Same tense as in line 4 — see above, note 7.

16 Obashanasuhvc, a kind of collective pi., including the
person named and those with him — see above, p. 48, and
Smith, p. 18. This, or a similar idiom seems to be universal
in Bantu — e.g. Sumbwa : nge Bandega, ' ce sont des hommes
de Ndega ' (P. Capus) ; Swahili : kina Hamisi, etc. Oha- is the
plural relative prefix (Smith, p. 108).

17 Izuba 5 ' sun ' — here used for ' day.' The same form of
the word is found in Tonga (Zambezi), and cf. Nyoro izoba,
Konde ilisuba, Ganda enjuba, etc. Other forms are lynwa,
dzuwa, riua, jua, iruwa, eyuva, loba, etc. The pronoun for
Class 5 is in Ila di. Di;t is the form with past tenses.

18 Dimwi agreeing with izuba understood.

19 Kwa 17 (ku- with the past tense) is the ' indefinite subject,"
equivalent to ' it ' or ; there ' (' there was darkness ').

APPENDIX I. 269

10 Idiomatic use of inana ' finish ' (see Smith, p. 187), mean-
ing ' they all died together.' The sequel shows that ' died ' is
not to be taken literally.

31 Wa omitted after mwana.

M Past tense of verb ' to be ' (d i, Smith, p. zoo), preceded
by relative particle o.

" wo for wa : a becoming o before the verb ompolola (see
Smith, pp. 12-13). The tense is the ' aorist ' (Smith, p. 150).

u Aorist of yu ' go,' followed by infinitive ku-fwa. Wezo is
th« demonstrative pronoun of the first class meaning ' that '
(already referred to).

u bitla 'lack,' 'be witho*ut,' sometimes followed by a noun,
as ndabiila shidyo ' I have no food,' sometimes, as here, by the
infinitive preceded by a relative particle. ' He was without
arriving ' = ' he failed to arrive.'

20 letelela, ' Double Relative ' (Applied) form of leta : let-ela
' bring to," let-el-ela ' bring to ' a person ' for ' his use.

*7 3rd pers. sing. Imperative of the irregular verb kweza
(=kti eza or ku iza) 'come' — see Smith, pp. 182, 183.

K lapu'ila, Applied (' Relative '). form of lapula ' spit ' — see
Smith, p. 120. We should have expected ktt-ba-lapwila.

23 Immedfate Future Tense (Smith 155).

80 ba is the verb ' to be,' also used in the sense of ' become '
(Smith, p. 184). Translate ' What has become of you that
you did not arrive ? ' » .

81 PI. of. mwanichi (or mwaniche) 'youngster.' The ordinary
word for ' child ' is mwana.

M =ngon'*u'o : ' substantive locative pronoun ' = (' the place)
just there' combined with the demonstrative awo 16 (Smith,
p. 216). Here used as an adverb of time, 'just then,' or
' immediately.'

** budio ' merely.' Smith suggests that it may be a noun of
Class 14 meaning 'nothingness.' But the fourteenth prefix is
' the basis of a number of adverbs of manner' (Smith, p. 217)
and possibly -dio might be explained as a pronominal stem
agreeing with some 5th class noun understood.

84 Future Subjunctive used as Imperative.

•5 Another form of the Future Subjunctive— (see Smith, p.
162).

36 u is the form assumed by the locatives mu and kit before

270 APPENDIX I.

nouns which already begin with those prefixes — e.g., n-kuboko,
for ku-kuboko or (as here) u-mukalo for mu-miikalo.

31 Bu, used as a conjunction and meaning 'since,' is
probably a pronoun, perhaps agreeing with busena 14 ' place,'
understood.

88 Negative Future (Smith, p. 171).

39 Kwiba=kfiiba ' steal.' Cf. Zulu (e)ba, Nyanja ba, Swahili
iba (in the northern dialect jepa, which may preserve a trace
of the lost initial consonant).

40 ist pers. pi. Imperative (Second Form : Smith, p. 163).
Zuba 'hide' (intr.) as it takes a direct object of the person,
must mean ' hide from ; ' but we should have expected the
Applied Form.

41 Second Augmented Form of the Present Subjunctive,
with the particle a prefixed (Smith, p. 163).

42 Future Subjunctive of yayila, which is the applied form
of yaya ' kill.' The force of the Applied Form is not obvious
here, as tuyayile would ordinarily mean ' kill .... for
us,' and no second object is expressed— or, indeed, required
by the sense.

43 Locative demonstrative emphasised : ' just on that
place).' (Smith, p. 91.)

Connected Translation

[This story is not very clear as it stands, but it seems
to be a confusion of two different tales, both of which
are widely distributed in Africa : that of the race
between the Tortoise and the Hare, in which the former
wins by planting out his family along the track (cf.
Uncle Remus, ' Mr. Rabbit finds his match at last ').
and one where all the animals join together to dig a
well : the Hare refuses, and is not allowed to draw
water, but does so by a trick, which is finally detected
and frustrated by the Tortoise. This latter story is
found in Jacottet's Contes Populaires des Bassoutos
(Le Chacal et la Source), in Theal's Kaffir Folk-Lore,
in the Swahili collection entitled Kibaraka (Sungurana

APPENDIX I. 271

Mgomba and Hadithi ya Vinyama), in Mrs. Dewar's
Chinamwanga Stories (' The Rabbit and all the other
Animals ') and elsewhere. All details of the trick by
which the water was stolen«and that by which the thief
was captured are here omitted.]

Once upon a time, when all the animals were dying
with thirst, they said to each other : ' Let us run swiftly
and see who reaches the water first.' But the Tortoise,
who had borne very many children, went on burying
them in the ground (along the course), and one child she
buried beside the water. So the animals said, ' Let us
race each other, and when we reach the river, we shall
drink the water.' They started, they all ran, they said,
' Let us see who will be the first to arrive.' They ran
again, and the Tortoises went on saying, ' Forward !
forward ! Mr. Hare and his friends ! ' Again they
ran, again they said, ' Forward ! Mr. Hare and his
friends ! ' The sun set once more, and they went on
shouting, ' Once more it has become dark. Forward !
Mr. Hare and his friends ! ' Next day the animals all
died together,* and the young tortoise who was beside the
water shouted, saying, ' Forward ! ' as before. The
Hare was just about to die, without reaching the water.
The young Tortoise who was beside the water brought
them some water in his inou'ih, in order to spit it out for
the animals. He said: 'It is you who were spiteful,
saying, " As for the Tortoise, we have outrun him."
Now what has become of you that you did not arrive ?
You are only children ! I am a grown man — I have
reached the water — but you are children ! ' Thereupon
he spat out the water which was in his mouth. They
could not answer him ; they remained sad and silent.
Afterwards the animals said : ' Let us dig a water-hole,
and then we shall be able to drink.' So they dug. But
the hares refused to dig, and so the others said : ' Since
you have refused to dig, you shall not drink any water.

* Probably this is to be understood in the sense of being ' kilt
entirely.'

272 APPENDIX I.

You shall all of you die of thirst.' When it was dark,
those who had refused to dig went to the water-hole
and stole water. So the other animals said, ' Let us
lie in wait for the hares, so that we may see them.' So
they saw them and caught them and bound them. But
they (the hares) said, ' Since you have tied us, now we
repent. Carry us to our burro\v, that you may kill us
just there.'"

4. NYANJA
The Story of the Cock and the Swallow\

Tambala ndi namzeze a-na-palana

Cock and swallow they made-with-each-other
chi-bwenzi,1 ndipo namzeze a-na-ti,2 'Koma3

friendship, and swallow he said, ' But

u-dze4 kwatu.'5 Ndipo tambala

you (must) come to our (house). And Cock

a-na-muka, a-na-ka-peza6 namzeze, a-li pa nsanja.7

he went, he found swallow, he is on nsanja.

Ndipo mkazi wa namzeze, a-na-pula8 ma-ungu,

And wife of swallow, she took-off pumpkins,

ndipo namzeze a-na-lengalenga,9 a-na-tenga

and swallow he-flew-up-aloft, he took

* This seems inconclusive, but no doubt the sequel is omitted as
too obvious : viz., that the too credulous animals did as they were
asked. Brer Rabbit more subtly entreated Brer Fox not to ' fling
me in dat briar-patch.1

| MS. collected at Blantyre, from a boy whose home was in the
neighbourhood of the Murchison Falls, on the Shire River. Mr.
R. S. Rattray has published a longer version of the same story
(from Central Angoniland) in Some Folk-Lore, Stories and Songs.

APPENDIX I. 273

maungu, na patsa10 tambala, ndipo tambala a-na-ti,

pumpkins, and-gave (to) cock, and cock said,

' Udze kwatu.'5 Ndipo tambala a-na-nka

' You (must) come to us.' And cock went

kwao,6 na-uza mkazache11 kuti,12 'u-ndi-ika18 ine

home, and told his-wife saying, ' you-me-put me

mu mpika14 wa mponda,1' 'ndipo a-na-m-pika
into pot of gourds,' and she-him-cooked

pa moto. Ndipo namzeze a-na-dza, na-peza
on fire. And swallow , he came, and found

tambala a-li mu mpika,16 ndipo namzeze a-na-ti,

cock he is in pot, and " swallow he said,

1 Pulani17 mponda, ndi-funa ku-nka kwatu.' Ndipo
' Take-off gourds, I want to go home.' And

mkazi wa tambala a-na-pula mponda, ndipo

wife of cock she took-off gourds, and

a-na-peza tambala, a-ta-fa,18 namzeze

she found cock, he-was-already-dead, swallow

a-na-bwerera19 kwao wo-sa-dya20 mponda.

he-returned home not-having-eaten gourds.

NOTES

1 Chi-bwiiizi 7, from bwenzi 5 ' friend ' (see p. 55). Ndipo,
properly the copula joined with the pronoun of (he ioth

locative) class, is very commonly used in Nyanja for ' and,'
'and so,' ' and then.' In Swahili it is more often found in its
original sense of ' that is where,' ' that is how,' etc.

2 Nyanja has the verb ti 'say ' conjugated in all, or most,
of its tenses, while keeping its original force, unlike Zulu,
where it is apt to pass into adverbial, etc., senses. This is
the past tense in -mi-. St;e Hethenvick, Manual, p. 50.

3 Koma ' but,' often begins a sentence in this way, where
there seems to be no adversative meaning ; but perhaps a
kind of polite deprecation is implied.

274 APPENDIX I.

4 u-dxe, subjunctive (used for imperative) of dza ' come.'

5 Kivatu possessive pronoun of Class 15, ist person plurab
Chez nous is a closer parallel than any we have in English.
But it should be noticed that it isalways ku'atu, never hwanga,
even when the speaker is referring to himself only. It is the
same in the second and third persons — cf. kivao (cltez liti),
lower down, where I have translated it simply by ' home.'
Some nouns ''of relationship are always used with a plural
possessive in the Bantu languages — cf. udade wetu ' my sister,'
umne wabo 'his elder brother,' in Zulu.

6 For this tense see Hetherwick, Manual, pp. 150, 156. It
here seems to indicate the interval between the act of starting
from home (muka) and 'finding' the Swallow — as though we
had to understand ' and, when he arrived, he found.' . . .
Peza, as here used, involves a sort of bull; it is not meant
that the Cock saw the Swallow sitting on the nsanja (see next
note), — but that he did not see him : he found him not there,
he being on the nsanja. It is very common for Africans to
say, ' I saw him not there,' or the like — which, after all, is not
very different from ' I found him already gone,' as we often
say — illogically, perhaps, but not irrationally.

I Nsanja is a kind of staging erected over the central fire-
place in a hut, on which meat is placed to be smoke-dried,
and seed-corn, beans, etc., to protect them from the attacks
of mice and insects. It forms a little loft under the point of
the conical roof.

8 Pula ' to take a cooking-pot off the fire,' is often used, by
an extension of meaning for ' dishing-up food,' and in
European households generally means ' bring in dinner.' —
Ma-ungu, plural of dz-ungu 5.

9 He would, as a matter of fact, have come down from the
nsanja, but having descended (probably at the back) under
cover of the smoke, he would then fly up, as if emerging from
the boiling pot.

10 ' And ' in Nyanja, is ni or ndi, not, as in some other
languages na. This na- is made up of ni and the pronoun a,
and is often found in continuous narrative, prefixed to the
second of two consecutive verbs. Cf. below, na-uza, na-peza,
etc.

II mkazache. Nouns expressive of relationship are often
combined with the possessive in this way : Nyanja : aiuako
' thy mother,' amachc ' his mother ' (but in some dialects mai
wake) ; Swahili mkeo ' thy wife,' mkewc ' his wife,' etc. The

APPENDIX I. 275

rule does not apply uniformly, for Nyanja has atate wako ' thy
father,' while Swahili has babangu, babako, babake (often
further abbreviated into bake) as well as inaniangu, etc. (See
Hetherwick, p. 87).

11 Literally ' to say ' ; equivalent to ' that,' but often used
(as here) where ' that' would be superfluous in English.

18 We should have expected u-ndi-ike, and possibly the MS.
is wrong. Ine ' me ' follows for emphasis : ' But, as for me, you
must put rne.' . . In Mr. Rattray's version, the correspond-
ing sentence runs: ' Mawa (to-morrow) it-pike nntungu, ndi-ka-
itana (and then I will call) bwenzi langa, ndipo int tt-ndi-ike «»'
mpika inomo.' Here the position of ine is varied for still
greater emphasis. Momo is a strengthened form of tno ' in it.'

14 This is written without a hyphen, because the mu really
has prepositional force. Had it still been felt as the locative
prefix (mu-inpika, or m'pika ' the-inside-of-the-pot ') it would
have been followed by mwa, not wa, as the possessive particle.

15 Mponda are a small kind of gourd, not unlike the species
known to cultivation as ' custard-marrows,!— very delicate if
properly cooked.

18 Here, again, it is not meant that the Swallow saw his
friend in the pot, otherwise what follows would lose all point.
Of course the meaning is ' He did not find him, for he was in
the pot.'

17 In Nyanja and some other languages, such as Makuaand
Venda, the plural of the second person is used instead of the
singular where special politeness is intended. But this idiom
does not seem to be very general in Bantu.

18 The verb ta ' finish ' is used as an auxiliary, particle to
indicate ' complete action.' See Hetherwick, p. 161.

19 Bwer-cra, applied form of bwera, ' return,' appears to
imply a return front, though only the place to which he
returned is expressed in words.

10 Wo-sa-dya for wakn-sa-dya, literally ' of to-not-eat ' is a
kind of negative participle in very common use. In the Likoma
dialect, where the ordinary negative serves to express the
future (si-ni-chite ' I shall do ') this participial form is almost
the only one, and is used without reference to person or time :
wo-sa-chitit,p\.o-sii-chita 'not doing,' wo-sa-liina 'not cultivating,'
etc., etc., see pp. 118, 169.

276 APPENDIX I.

Connected Translation

The cock and the swallow made friendship with each
other, and the swallow said, ' But you must come to my
house.' And the cock went, and did not see the
swallow, for he was sitting on the staging over the
fireplace. And when the swallow's wife took the
pumpkins off the fire, the swallow flew up (through the
smoke, as if he had come out of the pot) and took of
the pumpkins and gave them to the cock, who said
' you must come (in return) to my house.' And the
cock went home and told his wife to put him into the
pot of mponda gourds (which she was going to cook for
the guest). So she cooked him over the fire, and when
the swallow came, he did not see the cock, who was
inside the pot. (After waiting for some time), the
swallow said, ' Please dish up the gourds, for I want to
go home.' She did so and found the cock already dead
(in the pot). So the swallow went home, without
having eaten any of the gourds.

5. SWAHILI

(a) Lamu Dialect (Kiainu).
Stories about the People of Shela'"' .

I. Pa-li-kuwa na1 mtu wa Shela,2 hu-amkuliwa3

There was a man of Shela, he was called

Bwana Mgumi, a-ka-twaa4 kibarua5

Mr. Mgumi, and he took a day-labourer

ku-m-tilia6 mai7 katika8 kasiki9, na kasiki

to pour for him water into a jar, and jar

* Pictated by Muhamadi bin Abubakari (Kijuma)

APPENDIX I. 277

hiyo10 hu-ngia mi-tungi11 esherini. Uleia

that one there go in jars twenty. That

kiba-rua a-ka-tia mai, hatta kasiki i-ka-yaa,1"
labourer and he poured water, till the jar it was full,

a-ka-mwambia, 'Bwana, kasiki i-me-ziye

and he said to him, ' Master, the jar it has exceeded

ku-yaa.'M Ka mwa-mbia, ' Shindilia15 mai,
to be full. And he said to him, ' Press down the water,

twaa mti16 huu, u-ka-shi-ndilie.'17

take pestle this, that you may press down.'

A-ka-m-pa mti, ule akapiga,

And he gave him the pestle, and that (man) struck,

kasiki i-ka-vundika18 tini" (kasiki ile

the jar it was broken below (jar that

i-me-zikwa tiati20 nusu), mai

it was buried (in) the ground half), the water

ya-ka-shuka; akamwambia, ' U me-ona,

it went down ; and he said to him, ' You have seen,

ongeza21 tena basi23 mai!' Hatta23 mai

add again then water ! ' Until the water

yakashuka kwa tini,21 ndipo a-lo-po-yua25

and it went down from below, that is where he knew

kasiki imeYundika. Ndiye akatoa28 habari

the jar it is broken. It is he (who) put forth the news

kibarua.

the labourer.

II. Mngwana27 wa Shela mmoya, a-li-weka

Gentleman of Shela one, he put away

baruti, i-ka-ngiwa28 mai, akamwambia

gunpowder, and it was entered (by) water, and he told him,

mtumwake,29 'Twaa kikaango,80 weke moto-ni,

. his slave, ' Take frying-pan, put on fire,

na-taka ku-kaanga baruti yangu,

I want to roast (dry) powder my,

278 APPENDIX I.

i-me-ngiwa mai, na-taka kukaanga mimi

it has been entered by water, I want to dry (it) I

mwenyewe,81 wewe hu-to-yua32 kwa uzuri.'33

myself, you will not know properly.'

Mtwmwake a-ka-twaa kikaango,

His slave / and he took the frying-pan,

kaweka motoni kamwambia, 'Bwana,

and put (it) on the fire and said to him, ' Master,
tayari ! ' Kenda84 bwana, akatia

ready ! ' And he went, the master, and he poured

baruti kikaangoni, baruti

the powder into the frying-pan, the powder

i-ka-m-teketeza uso35 na ndeYu zake.86 Hini37

and it burnt him the face and beard his. This

habari ya-tendeka Shela.

affair is done (at) Shela.

III. Mwinda kungu38 a-li-pata kungu,

A hunter (of) bush-buck, he got (caught) a bush-buck,

a-ka-m-funga kisu na ukumbuu89

and he tied (on) him a knife with (his) girdle

ka-m-wambia, ' Enda kwa Mwana40

and said to him, ' Go to the Mistress, (tell her)

a-ku-tinde,41 nso na ini a-ni-wekee

she is to kill you, kidneys and liver let her put by for

mimi, kiya, nitwelee42 mkakambe.'43

me, when I come, that I may add (them to my) porridge.'

A-ka-mw-eta44 kenenda ule, kisa43

And he sent him and went that one, afterwards

a-ka-rudi, mwinda akenda nyumbani

and he returned, the hunter and he went to-house

kwake46 ; mke wake ka-mu-pa mkakambe mtupu/7

to his ; wife his and gave him porridge bare,

kamwambia, ' Nso na ini li48 wapi ? '

and he said to her, ' Kidneys and liver is where ? '

APPENDIX I. 279

Akamwambia, ' Hu-ku-eta,'

And she said to him, ' You did not send ' (them),

kamwambia, ' Ni-me-m-tuma Bwa' Kungu,"

and he said to her, ' I have sent him, Mr. Bush-buck,

ni-me m-funga ukumbuu na kisu ' ;

I have tied (on) him a girdle and a knife ' ;

kamwambia, ' Ha-ku-ya-' ;

and she said to him, ' He has not come ' ;

a-si-le mkakambe, katoka

and he did not eat the porridge, and he went out

kenda ku-m-zengea,51 a-si-mu-one. Basi,

and went to seek him, and did not find him. So,

hatta sasa watu wa Shela u-ki w-amkua ' Bwa'
until now the people of Shela if you call them ' Mr.

Kungu, hu-teta.

Bush-buck,' they quarrel (with you).

NOTES

1 Pa is the pronoun of Class 16 (locative), and it is quite
easy to translate pa-li-kuwa ' there was,' but the na which
follows seems superfluous. We must remember, however,
that the pronoun represents some noun meaning ' place ' (no
doubt the obsolete pantu, which has been replaced by pahali
or mahali) and that the construction is, literally, ' The place it
was with ' — i.e., 'it had ' — cf. the French use of avoir in il y
avail.

* Shela is a town within a half-hour's walk of Lamu, but
the people consider themselves quite distinct, and the Lamu
men affect to look down on them as stupid and ignorant, and
tell numerous tales against them, of which the three given in
the text are specimens. They resemble those of the exploits
attributed to the men of Gotham, or the mutual taunts of
' Hampshire hogs ' and ' Wiltshire moon-rakers.'

8 Amkua, elsewhere meaning 'salute' is used at Lamu for
' call.' A Mombasa man would have said huitwa or aliitwa,
The ' habitual tense ' in hit- (see Steere, Handbook of the

280 APPENDIX I.

Swahili Language, p. 126), which has no distinction of number
or person and may refer either to present or past, is more
freely used in the Lamu dialect than in the more southerly
ones.

4 Twaa ' take,' here means ' hire,' ' engage.' It is the same
word as the Zulu twain, which, however, means ' carry' (on
the head) — a good illustration of how identical roots may
diverge in meaning.

6 Kibarua, literally ' little letter,' has come to mean, first,
the ' ticket' given to people hired by the day and handed in
when their wages are paid, and, then the person so hired.

6 Tilia, applied form of tia, ' put,' ' pour ' — Zulu tela.

I Mai (mayi), Kiamu for inaji. See yaa, yua, mmoya, etc.,
later on.

8 Originally kati ka 'the middle of (perhaps a trace of the
ka- class which has disappeared from Swahili). Used like a
preposition in the sense of ' in,' ' on,' etc.

9 Kasiki, a large earthen jar, three feet or more in height,
sometimes seen at the door of a small village mosque, instead
of the usual tank (birika), which holds the water for ablutions.

10 Hiyo, demonstrative, Cl. g; here=' the aforesaid.'

II Plural of m-tungi 3 ' water-jar ' ; the usual size holds
about a gallon. Esherini (ishrin, ishinni) is Arabic, like the
other words generally used for the higher numerals.

12 Ule, Kiamu for yule, Distant Demonstrative of Class i.

13 i- pronoun, agreeing with kasiki 9; -ka-, sign of the
Narrative tense (Steere, Handbook, p. 134). Yaa—jaa in
Mombasa and Zanzibar Swahili : cf. Nyanja dzala and Zulu
xala ' be full ' (dist. from zala ' bring forth ').

14 An idiom implying 'not only full but overflowing.' Ziye,
for the more usual zitli (Arabic) ' be abundant,' ' exceed,' etc. —
Kamwambia ; the initial pronoun of the ka- tense is sometimes
omitted.

15 Shindilia (doubly applied form of shimla ' conquer,' of
which the original meaning was, probably, ' beat down,') —
used for pressing down grain into a basket or measure, to
make it hold more.

16 Mti, with dental t (Zanzibar mchi) seems to be a distinct
word from mti ' tree,' which has the cerebral t and does not
change in the Zanzibar dialect. The ' pestle ' used for
pounding grain is a pole of some heavy wood, about four or
five feet long and of a thickness to be easily grasped in the

APPENDIX I. 281

hand. The pestle used in Nyasaland (inunchi, intinsi) is much
thicker and is raised between the open bands which do not
meet round it.

17 For the subjunctive with -ka-, see Steere, Handbook,
p. 141.

18 Neuter-passive of vunda (Zanzibar, vunja). The implica-
tion is that ' it was in a state of being broken,' — ' it was found
to be broken ' ; it' the man's agency had been emphasised, i-ka-
vundwa would have been used.

19 Zanzibar, chini : it is really the locative of nti (nchi)
'earth,' 'ground.' Cf. the Zulu adverb pa-nsi, which has
survived the introduction of the locative in -ni and the loss of
the noun -itsi.

90 Tiati ' earth,' only found in Lamu and other northern
dialects. I have been unable to arrive at its derivation.

21 Causative of ongea, the intransitive verb meaning
'increase,' no doubt the applied form of onga, not in use.

21 Basi, sometimes spelt bassi (but it is better to avoid
double consonants in writing Bantu words), is the Hindustani
bass ' enough ! ' — constantly used in a variety of ways, e.g.,
' that's all ! ' — ' well 1 ' — ' and so ' — ' so then ' — etc. The
position here is unusual.

28 Arabic for ' until,' but often used for ' even,' or (in
narrative) as a mere connective.

" See note 19. K wa might be taken here as having some-
thing like an instrumental force, indicating the way by which
the water disappeared.

K a-lo-po-, Kiamu form of the relative (alipo), Yua for jua —
see note j. Ndipo is the i6th pronoun combined with the
copula, to form the kind of demonstrative (see Steere, Hand-
book, pp. 116-117), which means 'This is he, it, etc.'; in this
instance ' this is that (place) where ' — i.e., ' the time when.'

88 For the various meanings of toa, primarily ' put out,'
' take out,' see Madan's Sivahili-Engiish Dictionary, s.v.

a7 Mngwana ' a free man ' (not a slave) and therefore often
used to denote an educated or civilised person, — or a man of
good position. Also mungwana, and, on the southern Mrima,
or among inland tribes nntlungwana, though it seems doubtful
whether a derivation from Mulungu could be made out. The
word is not in Krapf. Mnwya, see note 7.

28 ngia, sometimes heard as ingia, but in the north at any
rate, the former seems to have better authority. The con-

282 APPENDIX I.

struction illustrates the Bantu use of the passive in cases where
it would be unexpected, or even impossible in a European
language : cf. also amcfiwa ni muute for ' her husband has
died,' and amekivenda kwitwa ' he has gone to be called ' — i.e.,
1 some one has gone to call him.'

29 For mtumwa wake. Such contractions are mostly con-
fined to words denoting relationship, e.g., babake, mamakt,
mumeo, nditguze, etc.

80 Kikaangt (from kaanga, ' roast,' ' fry ') is used for a
European frying-pan, llut in native households denotes a
shallow earthen pipkin, which serves a similar purpose.

31 When following a pronoun this word means ' myself,'
' yourself,' etc. : it is really a contraction of mwenye wakt ' the
owner of it ' (i.e., it is to be supposed, of the identity expressed
by the pronoun).

82 Hu-to- negative future prefix of the 2nd pers. sing., instead
of ha-u-ta-. See Steere, Handbook, p. 149, where this form is
only recognised as used with the Infinitive and is derived
from toa ' take out.' Comparison with other Bantu languages
suggests that it may have had a different origin.

83 Vizuri is often used in the sense of ' rightly,' etc. ; at
Lamu, the abstract noun (uzuri) preceded by the instrumental
kwa, is preferred.

84 For a-ka-enda. The subject, by a not unusual exception,
follows the predicate.

35 The idiom here is more easily parallelled in French than
in English : (la poudre) lui brida le visage. See Steere, Swahili
Exercises, p. 20 : the possessive, in a similar sentence, is seldom,
if ever, used in Swahili.

36 zake agrees with the second noun only. Ndevu 10 is
really the plural of udevu n, which means ' one hair of the
beard.'

81 Kiamu form of the ninth pronoun (hit). Tendeka,
perhaps because no agent is mentioned ; otherwise one would
have expected the passive. But perhaps the meaning is
' Such things are (only) possible to be done at Shela ! '

88 Usually the noun-agent formed by prefixing m- to the
unaltered verb-root is followed (as here) by a noun as object
— so that it might almost be called a participle. Occasionally,
however, a noun of this kind is found standing alone, as
mgema (not tngemi) 'palm-tapper). '

89 ' A girdle made of a narrow cloth' (Steere). The con-

APPENDIX I. 283

struction ' he bound his knife on him ' is similar to that
mentioned in note 35.

40 Mwana, used for ' mistress,' ' lady,1 and, with a woman's
name, as the equivalent for 'Mrs.' or 'Miss' — e.g., Mwana
Somoye, Mwana Esha, etc. This is sometimes called the
feminine of Bwana and is practically employed so to a certain
extent, though bibi is more usual at Mombasa and nana
(originally 'grandmother' at Lamu). Krapf enters this
mwana as a different word from mwana ' child ' (which, in
Swahili, seems to be confined to the meaning 'son'). It is
possible that they may be either (i) different words which by
phonetic change have become identical in form, or (2) the
same word which has become differentiated in meaning. But
Burton's suggested derivation from the Arabic ummand ' our
mother' (see Taylor, African Aphorisms, p. 31), seems very
doubtful.

41 Tinda, Zanzibar chinja ' slaughter ' — especially in the
correct Moslem fashion. Probably it was the time required
for this ceremony that made the hunter unwilling to stop.

" Twelea (spelt by Krapf toelea) is to add the fish, chicken,
or other kitoweo to the rice or porridge.

48 Old word for ' porridge ' (sima or ugali).

44 eta, Kiamu for Ida, which means ' bring ' or ' send '
(a tkittg} according to circumstances (tuma is used of sending
a person), kenenda for a-ka-enenda : enda and enenda are
synonymous. Ule, of course, is the kungu.

46 Mombasa kisha, for a-ki-isha ; ' when he had finished,"
but now practically an adverb. Good Swahili speakers at
Lamu prefer it to the Arabic kltalafu (halafu).

46 Locative concord: in Nyanja it would be ku nyumba.
kwache ; the kit, implying motion towards, has been replaced
by the locative -ni in Swahili.

4T -tupu ' bare,' ' naked ' is often used thus (' porridge and
nothing more,' — whereas it is always eaten with some ' relish '
— kitoweo or intuzi). Cf. the line from a popular song:

1 Wanipa inaji malupu kunisonga moyo.'

' You have given me mere water (the barest minimum of
hospitality), to twist my heart.'

48 Pronoun agreeing with ini 5, the last subject. (Nso 10,
cf. Zulu izi-nso.)

49 A common shortening of Bwana in the Ski and
neighbouring dialects.

284 APPENDIX I.

50 Subjunctive, because the action follows and is in a sense
dependent on what goes before : he did not eat because he
had been told that the buck had not come. Similarly, in
next line asiinuont 'without finding him ' ('so that he did not
find him '). Ha-ku-ya (Kiamu for lia-ku-ja) : Negative Past
which can be used either for ' did not ' or • has not.'

81 Zeiger., used at Lamu instead of tafuta ' seek.'

Connected Translation

I. There was a man of Shela who was named Bwana
Mgumi, and he engaged a day-labourer to fill a large
jar, holding about twenty gallons, with water. The
man poured in water till the jar was full and then said,
' Master the jar is full and running over.' The master
answered, ' Take this pestle and press it down ' — giving
him a pestle, with which he pounded the bottom of the
jar (which was buried in the ground for half its height),
till it cracked, and the water began to go down. So the
master said, ' Do you see ? — now pour in some more
water ! ' And he did so, and it was only when the
water kept running away at the bottom that he knew
the jar was broken. It was through the labourer that
this story got about.

II. A certain gentleman of Shela had put away some
gunpowder, and (after a time, found that) the damp
had got into it, so he said to his slave, •' Take a frying-
pan and put it on the fire ; I want to dry my powder,
which has got damp ; but I want to do it myself, as you
will not know how to do it properly.' So the slave
took the frying-pan and put it on the fire and said,
' Master, it is ready.' Then the master went and
poured the powder into the pan and it (blazed up and)
scorched his face and his beard. This is the sort of
thing that happens at Shela.

III. A hunter caught a bush-buck and tied his knife
round it with his girdle and said to it : ' Go to my wife
and ask her to kill you and put by the liver and kidneys,

APPENDIX I. 285

so that I can eat them with my porridge in the evening.'
So he let him go, and the buck disappeared. In the
evening, when the hunter returned home, his wife gave
him nothing but porridge for supper, so he said to her,
' Where are those kidneys and that liver.' She said to
him, ' You did not bring any ' ; and he said, ' I sent
Mr. Buck and tied a knife round him with my sash'* —
but his wife said, ' lie has not come.' He would not
eat his porridge, but went out to look for the bush-buck
and could not find him. And so, to this day, the Shela
people get angry with you if you address them as 'Bwa'
Kniigu.'

(b) Kimvita (Mombasa Dialect)

•

Story of the Man who did not know when he was
well off.\

Alikuwako mtu mmoja maskini sana, akaketi

There was man one poor very, and he sat

siku hiyo,1 akasema, ' Ni-ta-kwenda kwa

day that, and he said, ' I will go to

Mwenyiezi Muungu, ni-ka-ombe ni-pawe2

Almighty God, that I may pray I may be given

riziki3 yangu, kwani n-na4 dhiki sana.'

living my, for I have trouble greatly.'

A-ka-ondoka a-k-enda zakwe6, akafika

And he rose up and he went his (ways), and he arrived

mbali sana, akaona ziwa li-na6 maji mangi.

far very, and he saw a lake it has water much.

Akaoga, kisha akenda zakwe.

And he bathed, afterwards he went his (way).

* As a rule, in telling this story, the narrator repeats the speech
in full. In dictation, it was given more concisely.

t Dictated by Muhammad bin Macalim '1 Betawi, at Mombasa.

286 APPENDIX I.

Akenda, akamwona7 simba, ha-oni,
And he went, and he saw him a lion, he does not see,

a-na-konda" na ndaa, akamwambia,

and he was thin with hunger, and he said to him,

' Mwana Adamu, wenda wapi ? ' Akamwambia,

' Child (of) Adam, you go where ? ' And he said-to him,

' Nenda kwa Mwenyiezi Muungu,9 nenda omba10

' I go to Almighty God, I go pray

nipawe nami riziki.'

I may be given I too a living.'

Akamwambia, ' Ukenda11 ni-ombea

And he (the lion) said to him, ' When you go pray for me

na mimi, ni-funuke mato yangu,

also me, I may be opened (as to) eyes my,

ni-pate na chakula ni-le.' Akamwambia,

I may get also food (that) I may eat.' And he said-to him,

'Yyema.' Akenda zakwe. Akamwona

'Good.' And he went his (way). And there saw him

nyoka, a-ka-mw-uliza, ' Wenda wapi ? '

a snake, and he asked him, ' You go where ? '

Akamwambia 'Nenda kwa Muungu, nenda

And he said to-him ' I go to God, I go (to)

omba riziki yangu.' Akamwambia,

pray (for) living my.' And he said-to him,

' Ukenda, niombea na mimi ; jua

' Where you go, pray-for me also me ; the sun

li-na-zidi,12 sipati chakula; basi, nataka

has exceeded, I do not get food ; well, I wish

mYua i-nye tu-pate chakula.'

rain may fall, (so that) we may get food.'

Akenda zakwe. A-ka-tokea13 mji

And he went his (way). And he appeared (at) town

mkubwa, akamwona surutani mwanamke,
large, and he saw the sultan, a woman,

APPENDIX I. 287

akamwambia, ' Wenda wapi?' Akamwambia

and she said to him, ' You go where ? ' And he said to her

' Nenda kwa Muungu, nenda omba riziki yangu.'
' I go to God, I go (to) pray (for) living my.'

Akamwambia ' Ukenda niombea na mimi
And he said-tohim ' When you go pray-for me also me,

mimi surutani, mwanamke, raia zangu
I (am) a sultan, a woman, subjects me

ha-wa-ni-sikizi, na mji ha-w-ishi" vita. Basi,
do not obey me, and town does not finish war. Well,

nataka ya-ondoke haya.'16 Akenda

I wish they may go-away, these (matters).' And he went

zakwe.

his way.

Akafika, akamwambia Mwenyiezi

And he arrived, and he said-to Almighty

Muungu, ' Mimi, n-na-ku-ja, na-ona dhiki sana

God, As for me, I have come, I see trouble much

kwa umaskini, nnakuja omba

through poverty, I have come (to) pray

u-ni-wasii18 kwa riziki.' Akamwambia,

you (to) me assign for (my) living.' And he said to him,

' Riziki17 zako sasa nyingi sana,

' (means of) living thy now many very,

zamwayika.' Akamwambia, ' Ni-li-po-ku-ja

they are being wasted.' He said to him, 'When I came

huku, na-li-mw-ona mwanamke, a me-ni-ambia,

hither, I saw a woman, she said to me,

" na mimi, niombea kwa Mwenyiezimgu, mimi

" and me (too), pray for me to Almighty God, I

ni mfalme, raia zangu hawanisikizi, na

am a queen, subjects my do not obey me, and

mji wetu hawishi vita.'" Akamwambia,

town our does not end war."' And he said to him,

288

APPENDIX I.

1 Mwambie, " wewe ni mwanamke, pata mume

' Tell her, " you are a woman, get a husband

a-ku-oe,18 ya-ta-ondoka,

that he may marry you, they will go-away,

yote hayo." ' Akamwambia : ' Kisha

And he said to him : ' Afterwards

ameniambia, jua ni lingi,

he has told me, the sun is much,

chakula, ataka mvua.'

food, he wants rain.'

' Na-tue19 johari

' Let him put down the jewel

all these (th,ings).'"

n-na-ona nyoka,

I saw a snake,

hawapati

they do not get

Akamwambia,

And he said-to him,

iliyo kitwani,

i-ta-shuka myua

nyingi

which is in (his) head, it will come-down rain much

sana.' Akamwambia, 'Kisha na-li-ona simba

very.' And he said-to him, ' Then I saw a lion

ha-oni,

akanambia,

" niombea kwa

he does not see, and he said-to me, " Pray for me to

Mwenyiezimgu

Almighty God

kisha

(and) then (that)

Akamwambia,

He said to him,

yakwe mat::ii,

his

nipate

mato yangu,

that I may get

eyes, my,

nipate

chakula." '

I may get

food." '

' Mwambie

apake mate

' Tell him he is to smear spittle
yatafunuka mato, na

ndio

that-is

on his eyes, they will be opened, the eyes, and

a-taka-cho-ona20 mbele na-le,

that which he sees before (him) let him eat,

Akamwambia, ' Hewallah ! '
And he said-to him, ' Hewallah \ '2

Akamwambia yule surutani

riziki

living

akenda zakwe.

yakwe.'

his.'

and he went his (way). And he said-to her that queen
kama a-li-yyo-ambiwa.22 Mwanamke

like (that) which he had been told.

The woman

APPENDIX I. 289

akamwambia, 'N-oa23 wewe, utapata mali
said-to him, ' Marry me you, you shall get property

mangi ; kisha24 u-ta ku-wa mfalme wewe.'

much ; (and) then you shall be king you.'

Yule akasema, ' Sitaki mali wala

That-one and he said, ' I do not want property nor

sitaki ufalme, mali mimi nayo mangi

I do not want kingdom, property I with it much

sana, na-pawa, ni Muungu.'25

very, I am being given (it), it-is God.'

Akenda zakwe. Akamwona nyoka,

And he went his (way). And he saw him the snake,

akamwambia kana" alivyoambiwa.

and he said to him like (that) which he had been told.

Nyoka akamwambia, 'Twaa wewe hii

The snake said to him, ' Take yourself this

johari.' Yule maskini akasema, 'Si-i-taki,

jewel.' That poor man he said, ' I do not want it,

mimi, mimi mali yangu ni mangi

for my part, I property my is much

sana, ni-li-o-pawa, sitaki tena.'

very, which I was given (it), I do not want again.'

Akenda zakwe. Akamwona simba,

And he went his (ways). And he saw him the lion,

akamwambia, ' Mwenyiezimgu akwambia,

and said to him, ' God Almighty says to you,

" Paka mate yako, mato yako ya-ta-funuka.

" Smear spittle your, eyes your will be opened.

Chakula chako u-taka-cho-ona mbele.'"

Food your (is) thatwhichyou will see in front (of you).'"

Akapaka mato yakwe, yakafunuka.

And he smeared eyes his, and they were opened.

Aka-mwona yeye anasimama,"

And he saw him him (where) he stood,

290 APPENDIX I.

akapeleka mkono28 akamshika, akasema,

and he stretched out paw and seized him, and said,

1 Nitapata wapi chakula chengine, mimi? ' Yule

4 I shall get where food other, I ? ' That

mwana Adamu aka-sema, 'Mimi na-ku-ombea,
son (of) Adam, he said, ' As for me, I pray for you,

yanafunuka mato yako, sasa wataka

they have been opened eyes your, now you want

nila ? ' Akamwambia, ' Sijui,

eat me ? ' And he said to him, I do not know (but),

mimi, nitakula, nitimize maneno

as for me, I will eat you, that I may fulfil words

yako.' Akamla. Hadithi inafika

your.' And he ate him. The story it has arrived

hapa.29

here.

NOTES

1 Hiyo is the demonstrative implying ' that previously
referred to '— here meaning the day to which the story relates.

1 Pawa, Kimvita passive of pa (La.mu powa, Zanzibar pewa)
Omba is used for ' pra)^' in the sense of making a definite
request (also used for ' beg ') -ku-sali means to repeat the
ritual prayers.

8 From the Arabic razaqa ' provide,' (hence er-Razzaq, one
of the names of God) ; — it is used for ' subsistence,' ' daily
bread,' ' rations,' etc.

4 For ulna : the contraction of this pronoun is specially
common at Mombasa. Cf. Stigand, pp. 29-30.

5 -akwt for -akf, the Kimvita form of the possessive 3rd
pers. sing. Stigand unaccountably says (p. 29), ' this is not
often heard.' The expression end a zako (zake, etc.) has njia
(ndia) (pi.) understood after it, and corresponds exactly to the
Scottish provincial ' go your ways.'

8 li is the pronoun agreeing with ziwa 5 (Nyanja dsiwc 5,

APPENDIX I. 291

Ila i-zhiba 5). Zulu and Ganda have the same word with the
prefix of Class 7 : isi-xiba, eki-diba. Mangi, Kimvita — more
southern dialects mengi.

I The construction leaves it uncertain whether the lion saw
the man or vice versa; the insertion of the objective pronoun
would favour the former view, as this usually indicates that
some definite person or thing is meant (performing to a
certain extent the office of the definite article), while, on the
other hand, it is obvious that, if the lion was blind he could
not see any one. However, ona is often used for ' meet ' or
1 find,' as well as ' see,' and this rendering seems to suit the
context best.

8 A-na-konda, not, as in the Zanzibar dialect, a present, but
a perfect, cf. li-na-zidi later on.

9 Kwa only used in this sense before nouns denoting
persons, like the French chez. Really the possessive particle
of the locative ku- class (17) with a noun understood ; (nyum-
bani=ku-iiyu)iiba) kiva. The instrumental kwa (as in kukata
kwa kisu), though the same in origin, is, in usage quite
distinct. Mwenyeezi, compounded of mwenye ' owner ' and ezi,
(enzi, Arabic lizz), ' power,' ' authority,' is never used unless
followed by Muungu (Mungu, Mngu or Mgu as below). The
name Allah is not often used by Swahili Moslems, except when
speaking Arabic ; it seems to be confined (o expressions like
Hcwallah ! — which is now nothing more than a form of assent,
and Allah Allah ! originally an adjuration — ' for God's sake,'
— but generally used to mean ' be sure you don't forget,' ' be
quick, whatever you do,' etc.

10 Nenda oniba. The infinitive following a finite verb
(especially after enda, ja and isha) often drops the ku, thus
constituting a seeming exception to the rule that the verb-
stem is never found without a prefix, except in the Imperative.

II Both -ka- and -ki- are frequently contracted before enda.
Here the sense requires -Art-.

18 This looks like the present in -««-, but that tense is not
used at Mombasa, where the -««- tense has a perfect force
—i.e., it implies an action which has taken place in the past,
but whose effects are still continuing. Cf. anakonda (note 8),
which means ' he became thin and is (or was at the time when
the events occurred) thin.' Cf. also nnakuja, lower down.
The perfect in -me-, however, is also used at Mombasa.

18 tokca, applied form of toka, ' come out,' properly means to
' come out to or for some one ' — i.e., appear to him (it is there-
fore used of ghosts, etc.) and always implies a spectator or

292 APPENDIX I.

spectators. Here the meaning is ' he came in sight of (the
inhabitants of) a town.'

14 Hawishi for ha-n-ishi: u agreeing with the subject mji 3.
Isha, is here used actively.

, u Haya demonstrative 6 agreeing with mambo understood.

*P More usually wasia (from the Arabic wast)— properly,
'make a will;' 'give testamentary directions,' and thence
' appoint,' ' assign.' Some word like viiu or malt is understood
after it.

17 riziki is here treated as a plural. Mwaya (also mwaga
' spill,' ' empty.' The neuter-passive, mwayika, is best
rendered by 'are going to waste,' or ' are lying unused.'

18 »a 'marry' (only used of the man, ohwa is the word
applied to the woman and oza, ' give in marriage ' is said of
the parents or guardian), is the same word as the Zulu lobola,
and illustrates the degree of attrition steins may undergo in
Swahili through the loss of medial consonants.

19 tua 'set down,' as a load off the her\d, also (Madan)' settle
down,' ' rest,' etc. ; hence the applied form tulia 'be calm,'
' be quiet.' Cf. Zulu tula ' be silent,' etula ' take off' (as a hat,
or a pot off the fire), which are probably the same word, the t
having been introduced to differentiate the latter. (Of the
fairly numerous Zulu verbs in e some have lost an initial
consonant and are in process of shedding the vowel, as emba,
or mba ' dig,' in others the e seems to be an accretion (as
above). — N* tue, less usual than katite (Sfeere, Handbook, p.
140). — The jewel in the snake's head seems to be taken for
granted as if possessed by all snakes, but it may be less
summarily treated in the original story.

20 For the construction of the Objective Relative, see Steere,
Handbook, p. 119. A- is the pronoun of the 3rd person agree-
ing with the subject (sitnbet), -taka- the sign of the future, -cho-
the relative pronoun, object, agreeing with kitu understood.
We should, however, have expected the object-pronoun to be
inserted as well as the relative : a-taka-cho-ki-ona. For nale,
see last note.

21 See above, note 9.

22 The full form would be vitu alivyovittmbiura ' the things
which he had been told them,' but the pronouns of the 8th
class are often used without reference to a subject — cf.
the adverbial use of viziiri.

28 It is more usual to substitute the Subjunctive for the
Imperative when there is an object-prononn (e.g., mfc ' give
him '), but we also find nipa ' give me.'

APPENDIX I. 293

" Kisha, in this case, ' moreover,' ' besides ' — not ' after-
wards.'

25 If a connective particle is expressed after the Passive
(there is sometimes none) it is oftener ni than na which
would be the natural word to use, according to European
ideas. The literal rendering of the ni construction is — e.g.,
in this passage : ' I am being given — it is God (who gives).'

28 Kana, equivalent to kama, but not so common.

27 anasimama, Perfect in -na- : so, too, yanafunttka and
inafika, lower down.

18 mkono, properly used of human beings, but also of
quadrupeds when the paw is —as here — used like a hand.

59 Meaning, ' The story having arriveJ at this point, it ends
here.'

Connected Translation

There was once a very poor man, who, on a certain
day, said to himself, ' I will go to (the house of)
Almighty God and pray to him to give me enough to
subsist on, for (as it is) I am in great distress. So he
rose up and went his way, and when he had reached a
place a long distance off, he saw a lake containing much
water. He bathed and then went his way. As he
went, there met him a lion who was blind and very thin
with hunger and said to him ' Son of Adam, where are you
going ? ' So he said, ' I am going to the abode of God,
to pray that I may be given enough to live on.1 And
the lion said, ' When you go, pray for me also that I
may have my eyes opened and get food to eat.' The
man replied, ' Very well ' and went his way. Then he
saw a snake who asked him, ' Where are you going ? '
and he said, ' lam going to God to pray for sustenance ' ;
and the snake said, ' When you go, pray for me also ;
there has been such a drought that I cannot find
anything to eat, so I want the rain to come that we may
get food.' The man went on till he reached a large
town, where the Sultan, who was a woman, saw him

294 APPENDIX I.

and said to him ' Where are you going ? ' [He
answered as before.] She said, ' When you go, make a
request for me also. Though I am the Sultan, I am a
woman, and my subjects do not obey me, so that the
quarrelling and fighting in my town never cease. My
prayer is that this state of things may come to an end.'
So he went on.

And he arrived and said to the Almighty, ' I have
come, because I am in great trouble through poverty — I
have come to beg you to assign me sufficient means of
living.' And He said to him, ' You have abundance to
live on now, but it is being wu; ted.' The man then
said, ' On the way here I saw a woman who said to me,
Pray for me also to the Almighty : I am a queen, but
my subjects do not obey me and war never ends in our
town." He said ' Tell her ; " You are a woman, you
had better get married, then all these troubles will
cease." The man then said, ' After that, I saw a snake
who told me that, because of the drought, his people
cannot get^food— he would like it to rain.' The Lord
answered, ' If he lays aside the jewel which is in his
head, the rain will fall abundantly.' The man went on,
After that I saw a lion who was blind and who asked
me to pray that he might recover his sight and also be
provided with food.' The Lord said, 'Tell him to
smear his eyes with his spittle, and they will be opened,
and then let him eat whatever he sees before him, — that
is (assigned him for) his subsistence.' So the man said
' Hewallah ! ' and went his way. When he came to
the queen, he gave her the message with which he had
been charged, and she said, ' Marry me yourself, you
will acquire much wealth, and you shall be king.' But
the man answered, ' I do not want your wealth, nor do
I want the kingdom, I for my part have very much
wealth of my own, which is being given me by God.'
So he went on. When he came to the snake, he
likewise gave him his message, and the snake offered
him the jewel out of his head, which the man refused,
saying ' I have just had a large property given me, I do
not want any more.' Then he went on and came to the

APPENDIX I. 295

lion, and delivered th« message as it was given him.
The lion did as directed and recovered his sight, and,
seeing the man standing before him, stretched out his
paw and seized him, saying, ' Where shall I get any
other food than this ? ' The man said to him, ' Why, \<<
prayed for you, and your eyes have been opened, arid
now are you going to eat me ? ' And the lion said, ' I
don't know about that, but I have to eat you in order
to carry out your directions.' So he ate him. And the
story ends here.

6. GANDA*

The Story of Ndyakubi and Ndalakubi

Awo1 o-lwa-tukaj omu-saja

Well then, which arrived (there was a) man

eri-nya-lye8 Ndya-kubi,4 ne ba-tas

name his Ndya-kubi, and they made

omu-kago ni Ndalakubi. Awo Ndalakubi

blood-brotherhood with Ndalakubi. So Ndalakubi

n-a gamba Ndyakubi, nti," ' 0-ja-nga >T

and he said (to) Ndyakubi, saying, ' Come '

n-o-n-daba ' ;8 awo Ndyakubi n a genda

and me see ' ; so Ndyakubi and he went

n-a-tuka ewa" Ndalakubi. Ndalakubi

and he arrived at (the house of) Ndalakubi. Ndalakubi

n-a-gamba mu-kazi-we,10 nti, ' Genda

and he said (to) wife his, saying, ' Go (that)

o fumbire11 omu genyi emere.'13 Awo omu-kazi

you may cook for guest plantains.' So wife

* From Engsro za Bagtmia, p. 38. A slightly different version
is given in Manutl de l*u°ut Luganda, p. 237.

296 APPENDIX I.

n-a-genda a-fumbals emere n-e-gya14

and she went she cooked plantains and they were done

n-a-gi-reta,15 omugenyi n-a-lya

and she them brought (to) guest and he ate

na-ta-kuta.18 Nagamba Ndalakubi

and he was not satisfied. And he said (to) Ndalakubi

nti, 'Muna-nge17 sikuse.'18 Ndalakubi

saying, ' Friend my I am not satisfied/ Ndalakubi

nagamba omukazi nti ' Genda ofumbe

and he said (to) wife saying ' Go that you may cook

emere, omugenyi ta-kuse, ofumbe

plantains, guest is not satisfied, (see) that you cook

nyingi.' Awo omukazi nagenda afumba

many.' So wife and she went she cooks

emi-wumbo19 gy-emere e-tano, na-gyo20

bundles of plantains five, and those

n-a-gi-reta Ndyakubi, n-a lya

and she brought them (to) Ndyakubi, and he ate

n-a-gi-mala-wo,31 era12 natakuta.

and when he had finished them still he was not satisfied.

Nagamba Ndalakubi, nti, 'Munange,

And he spoke (to) Ndalakubi, saying, ' My friend,

sikuse.' Ndalakubi nagamba

I am not satisfied.' Ndalakubi and he spoke (to)

omukazi nti, ' Genda ewa munange

wife saying, ' Go to (the house of) my friend

gundi,23 o-n-sabire-yo" emere, nzeM

so and so, that you may beg for me there plantains, I

e-mpwede-ko.28 Omukazi nagenda asaba

they are finished for me. Wife and she went she begs

emere, nagireta, nafumba

plantains, and she brought them and she cooked

emiwumbo kikumi,27 nagireta ; Ndyakubi

bundles 100, and she brought them ; Ndyakubi

APPENDIX I. 297

nalya emere nagimalawo

and he ate plantains and when he had finished them

natakuta. Nagamba, Ndalakubi

he was not satisfied. And he spoke, Ndalakubi

nti, ' Munange, sikuse.' Ndalakubi

saying, ' My friend, I am not satisfied." Ndalakubi

n-a damu,28 nti ' Emere empwedoko.'

and he answered, saying 'Plantains are finished for me.'

Ndyakubi nagamba nti, ' Kalew

Ndyakubi and he spoke saying, ' All right,

ka n gende enjala, munywanyi80 wange,

let me go (with) hunger, dearest friend my,

n-flre ku kubosl . enjala.'

that I may die on road (with) hunger.'

Na-da-yo ewu-we. Na

And he returned there to his (own house.) And

Ndalakubi ya-laba82 a genze,8a naye nagenda

Ndalakubi he saw he has gone, he too and he went

oku mu kyalira, « natuka ewa

to visit him, and he arrived at (the house of)

Ndyakubi. Ndyakubi nagamba omukazi

Ndyakubi. Ndyakubi and he spoke (to) wife

nti, ' Genda ofumbe emere

saying, ' Go that you may cook plantains

y-omugenyi.' Omukazi nagenda

of guest.1 Wife and she went

nagifumba, negya

and she cooked th^i, and they were done,

nagireta ; Ndalakubi n-a-lya-ko™

and she brought them ; Ndalakubi and he ate of them

katono. Awo obude16 te bwa-lwa

a little. So time of day it did not delay

ne bu ziba. Ndalakuki nagamba

and it is stopped up. Ndalakubi and he spoke

298

APPENDIX I.

Ndyakubi nti, ' Nasula wa ? "'

(to Ndyakubi saying, ' I shall pass the night where ? '

Ndyakubi nagamba nti, ' Na-ku-segulira37

Ndyakubi and he spoke saying, ' I will remove for you

ku-kitanda38 kwange kw-o-no-sula.'

from bedstead my where you will pass the night.'

nagamba nti, ' Si-gya-wo.'39

and he spoke saying, ' I do not get-in there.'

nawangulawo empagi,

and he pulled out there a post,

yebaka/0 Nendyakubi

he slept. Ndyakubi too

yebase, Ndalakubi

he is asleep, Ndalakubi

nti, ' Munange,

saying, ' My friend,

si gya-wo ebi-gere,

where I have put up I do not get in (as to) my feet,

bi ri, bweru.' Ndyakubi nagamba

they are outside.' Ndyakubi and he spoke (to)

nti, ' Genda ewa gundi,

' Go to (the house of) so and so,

emuli." Omukazi nagenda

reeds.' Wife and she went

Ndalakubi

Ndyakubi

Ndalakubi nagenda

Ndalakubi and he went

neyebaka. Yali

and he slept. He was

na-mu ita,
and he called him,

we-n-suze41

mukaziwe
his wife

saying,

a-m-pole42

that he may lend me

na-zi-reta. Ndyakubi na-kokera41

and she brought them. Ndyakubi and he pushed out

enyumba ekiro ekyo. Ndalakubi nagenda

night

bi-gere bya-gukira

feet they projected

Ndyakubi

1 Munange,

' My friend,

house night that.

yebaka, bwe-yebaka45

he sleeps, when he slept

bweru. Naita

outside. And he called

nti,

saying,

bwe-wa-ja ewange wa-lya

when you came to my ( House) you ate

APPENDIX I. 299

emere nyingi, laba nze kakano, ebigere biri,
plantains many, see me now, feet are

bweru, ebi-solo bi ja ku-n-dya bigere.'

outside, animals they are going to eat me the feet.'
Ndyakubi nagamba nti, ' Si-ri-ko48

Ndyakubi and he said saying, ' I have not

we-na-gya muli zimpwedeko,

where I shall takeout reeds they are finished for me,

nemiti si-ri-na.' Ndalakubi nagamba

and poles I have not.' Ndalakubi and he spoke

nti, ' Bv/e-wa ja ewange ba-ku-fumbira
saying ' When you came to my (house) they cooked for you

emere nyingi, nolya nogana

plantains many, and you ate and you refused

okukuta emere, ne-zi-gwa-ko

to be satisfied (with) plantains, and they were finished

n-o-ng'amba nti, 'Munywanyi wange,

and you spoke to me saying, ' Dearest friend my,

kang'ende enjala' nange kale! leka

let me go (with) hunger ' and I — well ! let (be) —

ebisolo bindire47 ebweru, munywanyi

animals they may eat me outside, dearest friend

wange.' Ndyakubi nagamba nti,

my.1 Ndyakubi and he spoke saying,

' Munange, wefunye/8 leka kulanama/3

' My friend, draw up (your legs), cease to stretch out,

nange bwe-na-ja-nga50 ewuwo na-lya-nga-ko

and I whenever I come to your (house) I will eat

katono, m onerede.'61 Ndalakubi nagamba

a little 1 have repented." Ndalakubi and he spoke

nti, ' To-kola-ngaM bwotyo, nze

saying, ' Never do like that, I

bwe-na-kw-etonderavt nti, " Emere

when I admitted to you saying, " Plantains

300

APPENDIX I.

empwedeko," wa-yomba buyombi,54

are finished for me," you quarrelled a quarrelling,

nange no55 leka nefunye, nawe
and I — just let me draw up (my legs), and you

bwojanga ewange, emere

whenever you come to my (house) plantains
o-gi-rya-nga bulling!.' M

you shall eat them decently.'

NOTES

1 Aivo is here a mere connective, equivalent to 'and so,' or
the like. It seems to be a distinct word from the locative
awo.

2 The subject of olwatnka is olu-naku n ' day ' understood :
(on) the day which arrived ' being equivalent to ' once upon
a time.'

3 -lye, possessive 5 : in GanJa the possessives of the and
and 3rd persons singular are usually suffixed to the noun.

i Ndyakubi means ' I eat badly,' (the / of ly* becoming d
afier n) and Ndalakubi (from lala) ' I sleep badly.' Lala does
not seem to be used iu this sense now, having been replaced
by ebaka.

5 This is the 'narrative tense' (Pilkington, p. 18) of the
verb 'ta, which properly means ' kill,' but is used idiomatically.
Oku'ta omukago appears to be the technical term for 'making
blood-brotherhood.' See Roscoe, The Bagamfa, p. 19. The
3rd person plural is here (as often) used impersonally, like
the French on.

6 nti seems to be the only trace left in Luganda of the
verb ti 'say,' unless the adverb otyo is connected with it, as
suggested in Elements of Luganda Grammar, p. 206.

7 The Imperative with -nga suffixed is called in Elements
(p. 68) the ' Far Future Imperative,' but it is doubtful whether
it can be restricted to distant time. With a Negative
Imperative, -nga. has the force of ' never.'

8 2nd person singular, narrative tense of luba ' see,' for
na-u-n-laba. Na-{-u con tracts into no, and / becomes d after
-;«-, which is the object-pronoun of the first person.

APPENDIX I. 301

9 Ewa is the locative particle, corresponding to pa and kwa
in other Bantu languages and equivalent to the French chcz.
Kwa does not seem to be thus used in Ganda, though we do
find it as the possessive particle of the rsth class: okufa kwa
kabuka ' the death of the king.' Ewa is a double locative :
wa=pa, while e is a separate prefix meaning ' at ' or ' to ' (see
Elements, p. 97) and possibly connected with the Zulu
locative prefix (ante, p. 83). It is often found with suffixed
possessive — twa-ngt, ewu-wo, etc.

10 -we suffixed possessive, 3rd person. Gamba is the vero
found in Swahili as a:nba (generally used in the applied form
ambia ' say to,' ' tell'). From it we get eki-gambo 7 ' word,' cf.
Yao magambo 6 ' discussion.'

11 2nd person singular subjunctive of the applied form of
fumbn ' cook.' Contrary to what we find in Zulu, Swahili,
Nyanja, etc., it is accented fumbire. This difference in
accentuation and an apparent preference for short vowels
make the sound of spoken Ganda very puzzling to one
accustomed, e.g. to Swahili or Nyanja.

12 emere, g, properly ' steamed and mashed plantains,' but
used for ' food' in general, this being the staple dish of the
Baganda. (See Roscoe, pp. 435-6).

18 3rd person singular, present tense (used for past).

" gya ' be cooked,' ' be done,' etc ; Nyanja psya (/>.}'«),
Herero pya, etc., originally, had the sense of ' burn,' like the
Zulu tslta. It must be distinguished from two other verbs
both of which occur further on in this extract : gya ' take out '
and gya ' get into ' (a space).

15 reta=Zu\u (and other languages) leta, Herero eta, etc.
(r and I are to a certain extent interchangeable in Ganda,
the former being heard before a, o and it, and the latter before
e and i). -gi- is the object-pronoun of cl., 9 the subject-pronoun
being c- or y-. It is very rare to find, except in Class i, the
object-pronoun differing from the subject; its position seems
to have preserved the initial consonant, which has been worn
away at the beginning of a word.

16 The k in huta ' be satisfied ' is the ' exploded ' or 'long '
consonant (see Elements, pp. 14, 15) indicated in C.M.S.
books by a prefixed apostrophe ('ku(a), and in those of the
French Fathers by doubling the consonant (natakkuta) : the
former method seems preferable. The sound is really a
combination of a consonant with the glottal stop, which is
very common in Hamitic languages (e.g, Galla). These
' exploded ' consonants are not marked in the text from which
our extract is taken and it has not been thought necessary

302 APPENDIX I.

to distinguish them here. Natakuta. is negative narrative
tense.

17 mv.no, means ' one of and is therefore never used without
a possessive pronoun or a noun following : muna-ft ' one of
us,' inuna-Budu 'a mail of Budu.' Properly it should not be
used with a singular pronoun, but it has acquired the sense of
' frieud,' ' companion,' etc.

18 si-kuse, Negative Perfect of kuta, 2nd person tu-kuse, 3rd
ta-kusc. ta is the negative particle corresponding with the
Swahili ha.

19 omu-wumbo 3 (from wnmba ' wrap up for cooking in a
leaf ') is a bundle of plantains, which are always prepared in
this way. In the Manuel de Languf Luganda, it is translated
' marmitcs,' but this is evidently a mistake — gyemere for gya
emere— note gya 4, agreeing with emiwumbo.

20 (e)-gy°i demonstrative, agreeing with emi-wumbo : the -gi-
in the next word has the same agreement, and is consequently
4 not 9.

21 Narrative tense, followed by the locative relative -wo
(here=' when,' like -po in Swahili). main ' finish,' with its
derived forms malira, maliza, is found in Swahili (though here
almost ousted by isha), Nyanja, Yao, etc.

22 era seems to be used either as an adverb or as a conjunc-
tion. It may also mean ' and,' 'besides.'

M gnndi, used like goa in Poknmo and fttlani (Arabic) in
Swahili, to designate some one whom one cannot or will not
name.

24 2nd person singular subjunctive of sabira, applied form
ofsafca'ask': — o- (before a vowel w-) subject-pronoun, 2nd
person singular; -«-, object-pronoun, ist person singular; yo-
locative suffix, equivalent to ' in that place ' (Elements, p.*fo) :
the whole word meaning ' where you may request food for
me.'

25 nze, separable pronoun, ist person singular — here used
for emphasis=' as for me.'

26 Perfect (wede) of giva ' come to an end,' agreeing with
enure g and followed by the locative pronoun -ko. The -mp-
represents the object-pronoun of the first person, this being
the ionn assumed by n before u>. The construction suggests
a common Irish idiom, e.g., ' He's lost it on me ' (Jane
Barlow, Irish Idylls.)

37 Note the difference between ekumi and kikumi. Liikumi
is 1,000 and kakuwi 10,000.

APPENDIX I. 303

48 da ' return ' seems in da-tnu ' answer ' to be compounded
with mu in a way not quite easy to explain, but probably
growing out of the usage by which va-mu (e.g.) means ' get out
from inside.' (Elements, p. 71).

211 kale, interjection of 'exhortation,' here equivalent to ' Oh !
very well 1 ' or the like, kang'tnde subjunctive, preceded by
ka, which is generally added to the ist person singular and
plural (Eltments, p. 69). Note ng'ende, not ngtndc — n and g
combining into ng\

30 munywanyi a term of endearment, sometimes equivalent
to ' light of my eyes,' ' darling,' etc.

81 (e) kubo 5 is ' a path trodden down,' possibly connected
with kuba ' beat.'

82 ya-laba, 3rd person singular (note the difference in the
pronoun) of the ' Far Past ' Tense (Elements, p. 27).

88 -genzf, perfect of gtnda.

84 This is the ' Partitive ' use of -ko (like French en) — see
Elements, p. 70.

85 obude 14, constantly used in indications of time. Obudt
buziba is, literally ' the time of day becomes stopped ' (as a
bottle with a cork: the Manuel de Langue Luganda renders,
' le moment se boiiclte' tt-bit'a-lwa), Negative ^Far Past, agreeing
with obude. The whole phrase means ' it was not long before
it got dark.'

86 wa, adverb, indicative of ' place generally,' used interro-
gatively for ' where ? ' (Elements, pp. 46, 51). It is the
locative pronoun pa, a primitive /> becoming w in Ganda.

87 Segulira should properly ba the applied form of a
reversive verb derived from sega, which, however, does not
occur in the vocabularies in any meaning that would be
appropriate. Seguka, intr. is ' move one's position.' Na-ku-
segulira here means, not ' I will make room for you ' on my
bed, but ' I will give up my bgd to you.'

88 Ku-kitanda 17, treated as one word and therefore followed
by the possessive kwange. Kwonosula — ku o-no-sula : and. pers.
sing, of the Near Future, preceded by kn=on (which). Note
the distinction between sula ' pass the night ' and ebaka, used
of actual sleep.

89 Tin's is gya ' get into a space ' — see Note 14 above.

40 ebaka, properly a reflexive verb, e being the reflexive
pronoun (Zuli zi, Swabili ji, Nyamwezi i, etc.). But many
such verbs have acquired distinct meanings of their own.
(Elements, p. 117.) As it begins with a vowel, the Past is
y -ebaka, for ya-ebaka.

304' APPENDIX I.

dl suze, perfect of sula ; we- adverb corresponding to the
ocarive wrt 16 (Element?, p. 96), as in wano we-n-tambula,
here where I walk.'

4S For a-n-wole, from wola ' lend." The nasal n preserves
the p sound elsewhere lost in Ganda and is itself modified by
the influence of the p into m.

43 emuli 10 plural of olu-tnuli n. Reeds are used in
thatching a house, the thatch reaching down to the ground.

44 Ndyakubi, having previously pulled out one of the
supports in order to give his friend more room, now makes
an extension 'to the thatch with the borrowed reeds. Huts
being round, the foot of the bedstead (placed so that, in
a square room it would be parallel with the wall), would
necessarily come in contact with the thatch.

45 bwe, relative=' when ' agrees with obude 14 understood.

46 na understood after ri, as often in the negative. Gya
' take out,' as from a store.

47 For bi-n-lire, applied form of lya.

48 efunya, reflexive of funya ' clench ' (the fist), 'fold' etc. —
used here of drawing up the knees. Wefunyc=o-cfunyt, 2nd
person singular of the Subjunctive.

49 lanama, by its form and sense is evidently a stative, but
no verb lana appears to be in use.

60 The suffix-7Jgtf may denote either present, past or future
action, so long as it is repeated or habitual. (Elements, p. 91.)
Here it is equivalent to ' whenever* I come.'

51 For n-bonerede, perfect of bonera, ' repent ' ; b, after changing
n to m, disappears.

52 Negative Imperative : -nga suffixed to this mood implies
a general prohibition (Elements, p. 35) ; -otyo is an adverb
meaning ' just so ' and bw (e) 'how, 'depends on some implied
I4th class noun meaning ' state ' etc. (Elements, pp. 94-106.)

63 etonda ' confess a fault and be sorry for it ' (Blackledge) ;
ttond-era ' confess to (any one) ' here used of the regretful
admission that his provisions are exhausted.

54 That is ' merely quarrel,' ' do nothing but quarrel ' ; — for
this peculiar use of the i4th prefix, see Elements, p. 107.

55 no is an ' intensive interjection '; leka, ' let, " allow,' often
used before the subjunctive, like our 'let ' though its primary
meaning seems to be ' leave ' ('let alone.') Some languages
have it with the meaning ' stop ' (intr.)

56 Abstract noun, (from -lungi ' good ') used as an adverb
(Pilkington, p. 69.)

APPENDIX I. 305

Connected Translation

Once upon a time there was a man whose name was
Ndyakubi, and he made brotherhood with Ndalakubi.
And Ndalakubi said to him, ' Come and see me (some
day.)' So Ndyakubi went, and arrived at Ndalakubi's
house, and the latter said to his wife, ' Go and cook
food for the guest.' So the wife went and cooked food,
and, when it was done, she brought it, and the guest
ate, but he was not satisfied, and he said to Ndalakubi,
' My friend, I have not had enough.' Ndalakubi said
to his wife ' Go (again) and cook a great deal of food,
for our guest is still hungry.' She went and cooked
five bundles of food and brought them also to Ndyakubi,
and he ate it, and when he had finished he still
had not had enough and said to Ndalakubi, ' My
friend, I am not satisfied.' Ndalakubi said to his wife,
' Go to my friend so and so and ask him for some
plaintains, for mine are all finished.' The wife went
and asked for plantains and brought them and cooked
a hundred bundles. Ndyakubi ate the food and when
he had finished it, he was still unsatisfied and said
. . . [as before.] Ndalakubi answered, ' I have
no food left,' so Ndyakubi said, ' Never mind, I
will go away hungry, my dear friend, and die by the
roadside (if I must).' So he returned home. Ndalakubi
saw that he had gone, and (some time afterwards), he,
too went to pay a visit to him, and when he arrived at
Ndyakubi's house, the latter said to his wife ' Go and
cook food for the guest.' His wife -went and cooked it
and brought it when it was done, and Ndalakubi ate a
little of it. Soon after this it grew dark. Ndalakubi
said to Ndyakubi, ' Where am I to sleep ? ' Ndyakubi
answered, ' I will give you my bedstead so that
you can sleep.' Ndalakubi said ' There will not
be room for me,' so Ndyakubi pulled out one of the
posts of the house, and Ndalakubi went and lay down to
sleep. Ndyakubi also slept. When he was asleep,

306 APPENDIX I.

Ndalakubi called him and said, ' My friend, in the place
you have given me to sleep in, there is no room for my
feet, they are outside.' Ndyakubi said to his wife, ' Go
to so and so and borrow some reeds,' and the woman went
and brought them, and Ndyakubi made an extension to
the house that night. Ndalakubi went and slept, but
when he was asleep (he thrust) his feet (through the
thatch and awoke and found that they) were projecting
outside. So he called out to Ndyakubi, ' My friend,
when you came to my house you ate large quantities of
plantains — now, see how my feet are outside, and the
wild animals will come and eat them.' Ndyakubi said,
' There is no place where I can get any more reeds
they are all done and I have no poles.' Ndalakubi
said, ' When you came to my house, they cooked for
you an enormous amount of food, and you ate, and yet
you kept on saying you had not had enough, and when
the food was all finished, you said to me, " My beloved
friend ! let me go away hungry ! " — and so I say,
" Never mind, dearest friend — let the wild beasts eat me
outside your house ! " So Ndyakubi said, ' Oh ! my
friend, just draw up your legs and don't stretch them
out; and I, too, next time I come to your house, I will
only eat a little ; I am truly sorry for my behaviour.'
Ndalakubi answered, ' Never act again as you did when
I told you, very much to my regret, that there was no
more food in the house, and you did nothing but quarrel
with me. Well, let me just draw up my knees (till the
morning), and when you come to my house again,
remember to eat like a decent human being.'

A free version of this tale is to be found in Roscoe,
The Baganda, p. 482. The point of it lies in the
mutual obligations of blood-brothers, on which Ndyakubi
presumes beyond all permission.

APPENDIX II: BIBLIOGRAPHY

This Bibliography makes no attempt at completeness,
being intended merely as a guide to the books available
for the study of the more important Bantu Languages.
Continental works not easily accessible have only, as a
rule, been indicated where no English ones appeared to
exist.

Languages marked * are those into which the whole
Bibls has been translated ; those marked t possess
complete versions of the New Testament. These
versions are of unequal linguistic value, but as a.
rule are welcome aids to the student. (Most of them,
though not all, are published by the British and
Foreign T>ible Society.) Many others, besides those
marked, have translations of separate parts of the
Scriptures, and school reading-books, etc., which will
often be found useful.

Books marked * are to be found in the library of the
African Society (open to members) at 64, Victoria
Street, S.W.

I. GENERAL

"'Anthropos. Revue Internationale d'Ethnologie et de

Linguistique, Salzburg (Zaunrith).
From 19J36 onwards. Separate items under Congo
(Kiyombe, Kanyoka), Fang, Fipa.

\V. II. J. Bleek. Comparative Grammar of the South
African Languages. Part I. (Phonology), 1862,
3«7

308 APPENDIX II.

Part II. The Concord. Section I. The Noun.
1869. (No more published.) London (Trubner
and Co.)

C. G. Buttner. Zeitschrift fur Afrikanische Sprachen.

Berlin (A. Ascher and Co.), 1887-90.
Contains many valuable contributions, some of which
are entered as separate items in the Bibliography. The
periodical was discontinued on Dr. Buttner's death in
1890.'

*R. N. Cust. A Sketch of the Modern Languages of
Africa, 2 vols. London (Trubner and Co.), 1883.

Journal of the African Society. (Quarterly.) London
(Macmillan and Co.) From 1902 onwards. Contains
some valuable linguistic articles.

J. T. Last. Polyglotta Africana Orientalis. London

(S.P.C.K.), 1885.

Contains vocabularies of over fifty East African
Languages (including a few non-Bantu). They are not
very full, but form useful starting points for languages
not already studied.

*C. Meinhof. Die moderne Sprachforschung in Africa.
1910, Berlin : Berlin Evangelical Missionary Society
(Georgenkirchstrasse).

An Introduction to the Study of
African Languages. (Translated by A. Werner.)
London and Toronto (Dent), 1915.
Being the English edition of the preceding.

Grundriss einer Lautlehre der Bantu-
sprachen (Second edition). Berlin (Dietrich
Reimer) (Ernst Vohsen), 1910.

Grundziige einer vergleichenden
Grammatik der Bantusprachen. Same publisher.
Berlin, 1906.

Das Dahlsche Gesetz. Zeitschrift der
deutschen morgenlandischen Gesellschaft Bd,
LVIL, p. 302.

APPENDIX II. 309

An exposition of the important law of Dissimilation
referred to on p. 229.
C. Meinhof. Linguistische Studien in Ostafrika.

Berlin, 1904-8. Mitteilungen des Sem. fur orient.

Sprachen. Bd. VII-XI.

Phonetic Studies of Swahili, Shambala, Nyamwezi,
Sukuma, Digo, ' Nika,' Pokomo, Bondei, Zigula, Mbugu,
D/a!;i:nj>, Makua, Yao.

'"Mitteilungen des Seminars fur orientalische Sprachen
an der koniglichen Friedrich-Wilhelms-Universitat
zu Berlin, etc. Berlin (W. Spemann, afterwards
G. Reimer), 1898, etc.

These ' Transactions' appear annually in three sections,
of which the third is devoted to Africa, under the title
of Afrikanische Studien. Some of the more important
items are entered separately, under the several languages.
Referred to as Mitt. B. Sem. Or. The series of hand-
books (Lehrbiicher des Seminars fur orientalische
Sprachen} issued by the same institution contains a
number of valuable works, entered under the separate
languages.

Another series of which several volumes will be found
entered under various languages is the Archiv filr das
Studmm deutscher Kolonialsprachen (same publisher),
1895.

:>:B. Struck. Collections towards a Bibliography of the
Bantu Languages of British E. Africa. Journal of
the African Society, London, 1907.

J. Torrend, S. J. A Comparative Grammar of the
South African Bantu Languages, comprising those
of Zanzibar, Mozambique, the Zambezi, etc., etc.
London (Kegan Paul, Trench, Triibner and Co.,
Ltd.), 1891.

:':Zeitschrift fur afrikanische und oceanische Sprachen.

Berlin (Dietrich Reimer), 1895-1903.
Edited by A. Seidel; 5 vols. square royal 8vo., appeared
between January, 1895 and January, 1900. Publication
was then suspended, but resumed (in a smaller fo

310' APPENDIX II.

1902, and ceased with the first issue for 1903. Some
important contributions are entered under special
languages, e.g., Sumbwa, Tabwa. 4 vols. in African
Society's Library.

Zeitschrift fur Kolonialsprachen. (Quarterly.) Berlin
(D. Reimer) and Hamburg (Boysen). From 1910
onwards.

II. SPECIAL LANGUAGES

Adnma (Duma). Spoken along the Ogowe River in
the northern part of French Congo.

R. P. Dahin: Vocabulaire Adouma-Frangais. Parti.,
French-Aduma, pp. 72. Part II., Aduma-Frencb,
pp. 72. Kempten (Bavaria), (Jos. Kosel), 1895.

Angola. See Mbundu.

Bangala. See under Congo Languages.

Bangi. See under Congo Languages.

*Bemba. Between the Lualaba and Lake Nyasa.

*W. G. Robertson. An Introductory Handbook to the
Language of the Bemba People. London (L.M.S.),
1904.

*Father Schoeffer. Grammar as spoken in North-east
Rhodesia. Edited by J. H. West Sheane. Arranged
with Preface by (the late) A. C. Madan. Oxford
(Clarendon Press), 1907.

Benga. (Corisco Bay, West Africa.)

*J. L. Mackey. Grammar of the Benga Bantu
Language, revised by R. H. Nassau. New York
(American Tract Society), 1892.
The original edition of Mackey's Grammar was

published at New York (Mission House, 23, Centre

Street), in 1855.

APPENDIX II. 311

C. Meinhof. Das Zeitwort in der Benga-Sprache.
Berlin, 1890. Zeitschrift fur Afr. Sprachen, Vol.
III., pp. 265-284. Benga und Duala, ib. II., pp.
190-208.

Bondei. Spoken inland from Tanga in East Africa.
G. Dale. Bondei Exercises. Holy Cross, Magila, 1892.

*H. W. Woodward. Collections for a Handbook of
the Bondei Language. London (S.P.C.K.), 1882.

Stories in the Bondei Language
with some Enigmas and Proverbs. Written by
Native Students and edited by the Rev. H. W.
Woodward. (S.P.C.K.)

Bube. (Fernando Po.)

John Clarke. Introduction to the Fernandian Tongue.
Berwick-on-Tweed (Daniel Cameron), 1848.

R. P. Joaquin Juanola. Primer Paso a la Lengua
Bube, pp. 190. This seems to be the most
complete grammar hitherto published. Madrid
(A. Perez Dubrull).

"Sir H. H. Johnston. George Grenfell and the Congo,

Vol. II., Appendix I., p. 882. London (Hutchin-

son and Co.), 1908.

This work contains specimen vocabularies of a great
many other West African Languages, and a discussion
of the various Bantu migrations. The greatest amount
of space is devoted to a comparison of numerals.
Bunda. See Mbundu.
Chaga. (Caga, Dsagga, Djaga, etc.) The Wachaga

live on Kilimanjaro.
*J. Raum. Versuch einer Grammatik der Dschagga-

Sprache (Moschi-Dialekt). Archiv. fur d. Stud.

deutscher Kolonialsprachen, Vol. XI. Berlin

(Georg Reimer), 1909.
H. A. Fokken. Das Kisiha. Mitteil. des Sem. fur

orient. Sprachen. Jahrg. VIII., Abt. 3. Berlin,

pp. 44-93, 1905.
Siha is a dialect of Chaga.

312' APPENDIX II.

Chasu (also called Pare).

*E. Kotz. Grammatik des Chasu in Deutsch, Ostafrika.
Berlin (G. Reimer), 1909. Archiv fur das Studium
deutscher Kolonialsprachen, Vol. X.
Spoken in the mountain? south of Kilimanjaro.

Chinamwanga. See Namwanga.
Chinyanja. See Nyanja.
Chiswina and Chizwina. See Karanga.
*Chwana (Sechwana, Secwana).

It is practically identical with Sutu (Sotho, Sesuto),
and works relating to both are entered under this
heading.

James Archbell. A Grammar of the Bechuana
Language. Graham's Town (Meurant and
Godlonton), 1837.

J. Brown. L.M.S. Secwana Dictionary. Frome
(Butler and Tanner), 1895.

E. Casalis. Etudes sur la Langue Sechuana. Paris

(Imprimerie Royale), 1841.

This is really Sutu. The book is interesting as being
one of the earliest on the subject, and the Introduction
gives a valuable account of the establishment of the
French Mission in Basutoland and its relations with
Moshesh.

*W. Crisp. Notes towards a Secwana Grammar, 1900,

reprinted 1905. (S.P.C.K.)

A useful book, though not very well arranged. The
dialect is that of the Barolong.

K. Endemann. Versuch einer Grammatik des Sotho.
Berlin (Wilhelm Hertz), 1876.

Worterbuch der Sotho-Sprache, Vol.
VII. of Abhandlungen des Hamburgischen Koloni-
alinstituts. Hamburg (L. Friedrichsen and Co.),
1911.

These are really Chwana rather than Sutu, which is
noticed as a dialect under the name of ' Sud-Sotho.'

APPENDIX II. 313

E. Jacottet. Practical Method to learn Sesuto. Morija
(Sesuto Book Depot), 1906.

Elementary Sketch of Sesuto
Grammar, 1893. Published with Mabille's
Vocabulary, which see.

E. Jacottet. Treasury of Basuto Lore, Vol. I. (Sesuto

Book Depot), Morija, Basutoland, 1908. London
(Kegan Paul).

A valuable collection of Native Tales. Subsequent
volumes were intended to contain historical traditions,
songs, accounts of customs, etc., but no more have yet
been issued.

D. Jones and S. T. Plaatje. Sechuana Reader.
University of London Press (Hodder and
Stoughton), 1916.

F. H. Kruger. Steps to Learn the Sesuto Language

(Fourth edition). Morija (Sesuto Book Depot),
1905.

A. Mabille. Sesuto-English and English-Sesuto
Vocabulary. (Preceded by Jacottet's Grammar,
which see.) (P. E. Mission Press), Morija, 1893.

A. Mabille and H. Dieterlen. Sesuto-English Dictionary.
Revised and considerably enlarged. (Sesuto Book
Depot), Morija.

*S. T. Plaatje. Sechuana Proverbs with Literal
Translation. London (Kegan Paul), 1916.

Puisano ea se-Sotho le se-English. Phrase-Book.
Sesuto-English. Morija (Sesuto Book Depot), 1908.

A. J. Wookey, L.M.S. Secwana Grammar, with
Exercises. Frome (Butler and Tanner), 1905.

Congo (languages of). The languages included under
this heading are :

\Bangi, (Bobangi, Kibangi, Kiyanzi). On both
banks of the Congo, from the confluence of the
Sankuru to that of the Lulongo.

314 APPENDIX II.

Kanyoka, between Lulua and Upper Sankuru.
Kele, below Stanley Falls.1
*Kongo, (Congo, Fiote.)

tLo/o, (Mongo, Lunkundu) — on the Equator,
within the great northern bend of the Congo.

Lulua, on one of the Kasai tributaries.

Ngala, (Bangala, Lingala). Middle Congo,
between the confluences of the Mubangi and
the Mongala.

Ng'ornbe, west of the Ba-ngala.

Poto, at and near Bopoto (Upoto), at the top of
the Congo bend.

Soko, near the mouth of the Aruhwimi.

Teke, north of Stanley Pool (also called Ifumu).

Yombe, (Kiombe) in the Mayombe country, North
of the Lower Congo, and inland from the Ba-
vili.

An excellent bibliography of all publications dealing
with the Congo languages up to 1906 (the work of
Professor Starr), was issued by the University of
Chicago2 in 1908.

J. Barfield. Concords of the Congo Language, as
spoken at Palaballa. (East London Mission
Institute), Harley House, Bow, 1884.

W. Holman Bentley. Dictionary and Grammar of the
Kongo Language. London (Triibner and Co.),
1887.

Appendix to the Dictionary, etc.
(Same publishers), 1895.

De Boeck. Grammaire et Vocabulaire du Lingala ou
Langue du Haut Congo, 1904.

1 Not to be confused with Di-kele, the language of a different
tribe of Ba-kele, living near the Gabun estuary.

* Department of Anthropology, Bulletin V.

APPENDIX II. 315

Fra Giacinto Brusciotto di Vetralla. Regulae quaedam
pro difficillimi Congensium idiomatis faciliori captu
ad grammaticae normam redactae. Romas, 1659.

*Brusciotto di Vetralla. Grammar of the Congo
Language, as spoken 200 years ago, translated from
the Latin, and edited, with a preface, by H.
G rattan Guinness. London (Hodder and
Stoughton), 1882.

*R. P. J. Calloch. Vocabulaire Francais-Ifumu
(Bateke), precede d'elernents de Grammaire, 1911.

R. P. Cambier. Essai sur la langue Congolaise-
Brussels (Imprimerie Polleunis and Ceuterick),
1891 (Boko dialect of Ngala).

*H<- Craven andj. Barfield. English-Congo and Congo-
English Dictionary. London (Harley House),
Bow, 1883.

A. Courboin. ' Bangala,' Langue Commerciale du Haut-
Congo, Elements, Manuel de Conversation, Lexique.
Paris (A. Challand), 1908.

R. P. A. Declercq de la Congregation du C. I. de Marie,
Missionnaire au Congo beige. Elements de la
langue Kanioka (Kanyoka.) Vanves pres. Paris
(Imprimerie Franciscaine Missionnaire), 1900.

Vocabulaire Fran9ais - Kanioka.

(Same publishers), 1901.

Vocabulaire Kanioka - Fran9ais,
(Same publishers), 1901.

Grammaire du Kiyombe. Anthropos,
Vol. II., pp. 449-466, 761-794. 1907.

R. P. A. Declercq. Grammaire de la Langue des Bena
Lulua. Brussels (Polleunis and Ceuterick), 1897.

Legendes des Bena Kanioka (Text,
with interlined French translation). Anthropos,
Vol. IV., pp. 71-86, 449-456, 1909.

*L. M. Hailes. Kilolo-English Vocabulary. (East

316 APPENDIX II.

London Institute for Home and Foreign Missions),
Harley House, Bow, 1891.

H. H. Johnston. The River Congo. (Sampson Low),

1884, Second ed., 1895.

Contains vocabularies of Kongo, Teke, Buma and
Bangi (Yanzi).

J. and F. T. McKittrick. Guide to the Lunkundu
Language. (A dialect of Lolo.) (East London
Institute for Home and Foreign Missions), 1897.

*A. T. Ruskin. Proverbs and Similes of the Bamongo
(Mongo is a dialect of Lolo.) (East London
Institute for Home and Foreign Missions), 1897.

Outlines of the Grammar of the
Lomongo Language.

A. Sims. Vocabulary English- Kibangi (Bangi).
London (East London Institute) : Boston (American
Baptist Mission Union), 1886.

A. Sims. Vocabulary English- Kiteke and Kiteke-
English (Teke), 1886.

*W. H. Stapleton. Comparative Handbook of Congo
Languages, being a Comparative Grammar of the
eight principal languages, with appendices on six
other Dialects (Baptist Mission Press), Yakusu,
Stanley Falls, 1903.
The eight languages included in this book are : Kongo,

Bangi, Lolo, Ngala, Poto, Ng'ombe, Soko, Kele.

The six noticed in the Appendix are : Teke, Sakani

(a dialect of Lolo), Lomongo (Mongo, also a dialect of

Lolo), Boko (a dialect of Ngala), Lulua, and Mpombo,

which is not Bantu.

*W. H. Stapleton. Suggestions for a Grammar of
Bangala (the Lingua Franca of the Upper Congo),
with 2,000 words and many useful phrases Yakusu
(Baptist Missionary Society), 1903.

R. P. Ussel. Petite Grammaire de la Langue Fiote,
Dialecte du Loango, pp. 85. Loango (Mission
Press), 1888.

APPENDIX II. 317

(This is spoken by the Ba-vili, whose country is
somewhat to the north of the Congo estuary. The
author is a missionary of the Congregation du St. Esprit.)

R. P. Alexandre Visseq. Dictionnaire Fiot (French-
Kongo), 1889.

Dictionnaire Fiot (dialecte Sorongo),
1890.

Dictionnaire Fiot (dialecte du Kakongo),
1890.

Grammaire (Sorongo dialect spoken at
St. Antonio). Paris (Mission of the Congregation
of the Holy Ghost), 30 Rue Lhomond, 1889.

J. Whitehead. Grammar and Dictionary of the Bobangi
Language. London (Kegan Paul, Trench, Triibner
and Co.), 1899.

*Dnala.

Th. Christaller. Handbuch der Duala-Sprache.
Grammar and exercises ; story with literal interlined
translation; dialogues; vocabulary. Basle, 1892.

C. Meinhof. Die Sprache der Duala in Kamerun, Vol.
III. of Deutsche Kolonialsprachen, 1912.

W. Lederbogen. Daala-Marchen. Mitt. B. Sem.Or. IV,

V, VI, Abt. 3, 1901-3.

A large collection of tales, with German translation
in parallel columns.

A. Saker. Grammatical Elements of the Dualla
Language, with vocabulary. Cameroons (Mission
Press), 1853.

A. Seidel. Leitfaden zur Erlernung der Dualla-Sprache
(with readings and vocabulary). Berlin (Carl
Heymann), 1892.

Die Duala-Sprache in Kamerun. System-

atisches Worterverzeichnis und Einfiihrung in die
Grammatik. , Julius Groos1 Verlag. Heidelberg,
Paris, London, Rome, Petersburg, 1904,

318 APPENDIX II.

Duma. See Aduma.

Dzalatno (Zaramo, Zalamo). East Coast, South of
Zanzibar.

A. Worms. Grundzuge der Grammatik des Ki-Zaramo
in Deutsch-Ost-Afrika. Zeitschr. fur afr. u.
ocean. Sprachen III, p. 289, 1897.

C. Meinhof. Linguistische Studien in Ostafrika, No.
XII., Mitt. B. Sem. Or. X., Abt. 3, pp. 90-110, 1907.

Ediya, see Bube.

Fan (Fang, Fanwe, Pahouin, Pamwe, etc.). West
Equatorial Africa, North of the Ogowe.

*Rev. H. M. Adams. Fanwe Primer and Vocabulary.
Compiled by the Rev. R. H. Nassau, M.D.,
Gaboon and Corisco Mission [from the MSS. of
the Rev. H. M. Adams.] New York (printed by
E. G. Jenkins), 1881.

V. Largeau. Encyclopedic Pahouine, 1901.

Includes Grammar and French-Fan Dictionary,
containing many valuable anthropological notes and
also texts with translation.

R. P. L. Lejeune. Dictionnaire fransais-fang. Paris
(Favre and Teillard), 1892. With a Grammatical
Sketch.

*A. Osorio Zabala. Vocabulary of the Fan Language

(Fan-Spanish.) London (S.P.C.K.), 1887.
A number of stories, with French translation were
published by P. Trilles in Anthropos. Vol. IV., pp.
945-971, Vol. V., pp. 163-180.

Fernandian (See Bube).

Fiote (Kongo. See under Congo).

Fipa (East side of S. part of L. Tanganyika).

B. Struck. Die Fipa-Sprache (Deutsch-Ostafrika).
Anthropos, Vol. VI., 1911, pp. 951-993. Gram-
matical Sketch.

Vocabulary of the Fipa Language. Journal
of the African Society, October, 1908 (Vol. VIII.).

APPENDIX II. 319

*Ganda (Luganda).

*G. R. Blackledge. Luganda-English and English-
Luganda Vocabulary (S.P.C.K.), 1904.

*H. Wright Duta. Engero za Baganda (Proverbs in
the Luganda Language.) (S.P.C.K.), 1902.

*Elements of Luganda Grammar (Exercises and
Vocabulary.) By a Missionary of the Church
Missionary Society in Uganda. (S.P.C.K.), 1902.

*C. W. Hattersley and H. W. Duta (eds.). Luganda
Phrases and Idioms. (S.P.C.K.), 1904.

*Sir H. H. Johnston. The Uganda Protectorate, 2 vols.

London (Hutchinson), 1904.

Vol. II. contains vocabularies of a number of languages
besides Ganda.

Apolo Kagwa. Engero zabaGanda (Folk-stories).
Mengo (C.M.S. Press), 1901.

*Ekitabo kyo Bakabaka beBuganda (Book of the
Kings of Uganda). London (Headley Brothers),
printed, n.d. [1900 ?]

:|;G. L. Pilkington. Handbook of Luganda. London
(S.P.C.K.), 1892 (last edition, 1911).

White Fathers. Manuel de Langue Luganda, par L.L.
et C. D. des Peres Blancs (Grammar and Tales).
Einsiedeln, Switzerland (Benziger and Co.), 1894.

*Giryama.

W. E. Taylor. Grammar of the Giryama Language
(out of print).

Giryama Vocabulary and Collections
(Grammatical notes, and two tales, with translation).
London (S.P.C.K.), 1891.

Gisu (Masaba). Spoken on and near Mount Elgon.

*Rev. J. B. Purvis. A Manual of Lumasaba Grammar
(S.P.C.K.), 1897.

tGogo. Ugogo (the country of the Wagogo) is about
half-way between Zanzibar and Tabora and is
traversed by the Dar-es-Salaam railway.

320 APPENDIX II.

G. J. Clark (C.M.S.) Vocabulary of the Chigogo
Language. London (Gilbert and Rivington), 1877,
pp. 58.

Zimbazi ze Zifumbo, Nhandaguzi ne Zisimo ze Cigogo
(Gogo Reading Book— Native Proverbs, Riddles
and Fables). London (S.P.C.K.), 1901.
Collected and written out by Andereya and Nhonya.
C.M.S. native teachers at Mpwapwa. Preface in
English, signed J. E. B [everley] .

Gwatnba. A dialect of Thonga, spoken in N. E.
Transvaal.

P. Berthoud. Grammatical Note on the Gwamba
Language [1885]. Journal of the Royal Asiatic
Society, Vol. XVI., Part I.

P. Berthoud. Le?ons de Si-Gwamba. (Imp. J.
Chappins), Lausanne. 46 pp., lithographed, 1883.

Hehe. About 300 miles north of Lake Nyasa and to
the south of the Gogo country.

C. Velten. Die Sprache der Wahehe. Mitt. B. Sem.
Or. Vol. II. Vol. III. contains a Hehe-German
and German- Hehe Vocabulary, by P. Cassian
*Spiss, O.S.B.

\Herero.

P. Brincker. Worterbuch und kurzgefasste Grammatik
der Otji-Herero Sprache. Leipzig (T. O. Weigel),
1886.

The Appendix contains some tales, with literal and
free translation into German. Some additional tales,
collected by Biittner, are published in Ztschr. fur afr.
Sprachen.

P. H. Brincker. Deutscher Wortfiihrer fur die Bantu-

Diaiekte, Otji-Herero, Oshindonga, und Oshi-

Kuanjama in S. W. Afrika. Elberfeld (R. L.

Friderichs & Co.), 1897.

A very full German-Herero, etc., dictionary, in four

columns.

APPENDIX II. 321

C. Hugo Hahn. Grundzuge einer Grammatik des
Herero, pp. X+197. Berlin (W. Hertz); London
(Williams and Norgate), 1857.

F. W. Kolbe (L.M.S.) English-Herero Dictionary,

with an Introduction to the Study of Herero and
Bantu in general. Cape Town (Juta), pp. LV. +
569, 1883.

*C. Meinhof. Die Sprache der Herero in Deutsch
Siidwest-Afrika (Deutsche Kolonialsprachen, Bd.
I.). Berlin (Dietrich Reimer), 1909.

*A. Seidel. Praktische Grammatik der Hnupt-sprachen
Deutsch Siidwestafrikas (Nama, Otji-Herero,
Oshindonga), Vienna, Leipzig (Hartleben), 1892.

G. Viehe. Grammatik des Otjiherero (with Vocabulary).

Vol. XVI. of Lehrbiicher des Seminars fur

orientalische Sprachen.

Stuttgart, Berlin (W. Spemann), 1897.

Ifumu (Teke). See under Congo,
\Ila (Seshukulumbwe).

*E. W. Smith. Handbook of the Ila Language.
Oxford (University Press), 1907.

Isubu. Bimbia Peninsula, Cameroons, north of the
Duala.

C. Meinhof. Das Verbum in der Isubu-Sprache.
Zeitschrift fur Afr. Sprachen. Vol. III., pp. 206-
234, Berlin, 1889-90.

Joseph Merrick. A Dictionary of the Isubu Tongue.
(No publisher's name given in the British Museum
copy), 1854. Part I. Isubu-English, only completed
as far as I.

A Grammar of the Isubu Tongue. Unfinished.

Thts is out of print and no doubt rare. The British
Museum copy (Press mark 12907 bb. 22) is bound up
in a volume of ' Philological Tracts.'
Kafir. See Xosa.

322 APPENDIX II.

Kaguru (Kimegi). One of the dialects of Usagara,
lying east of Ugogo.

J. T. Last. Grammar of the Kaguru Language.
London (S.P.C.K.), 1886.

Kamba. Spoken in Ukambani, E. Africa — the district
in which Nairobi is situated.

E. Brutzer. Handbuch der Kamba-Sprache. Berlin,
1906. ;Mitt. B. Abt. Sem. Or. IX., 3, pp. 1-100.

*H. Hinde. Vocabularies of the Kamba and Kikuyu
Languages. (Cambridge University Press), 1904.

*J. T. Last. Grammar of the Kamba Language, pp.
40. London (S.P.C.K.), 1885.

*A. D. Shaw. Vocabulary of Four East African
Languages. See under Swahili.

C. G. Biittner, Deutsch-Kikamba Worterbuch.Ztschr.
f. afr. Spr. Vol. I., pp. 81-123, 1888.

Kami. Spoken in the Ukami country, of which
Mrogoro, on the Dar-es-Salaam railway, is the
centre.

A. Seidel, in Ztschr-fur afr. u. oc. Spr. II., 1, p. 20.
(Grammatical sketch and short vocabulary).

C. Velten. Die Sprache der Wakami in Deutsch-
Ostafrika. Mitt. B. Sem. Or. III., Abt. 3,pp. 1-56,
1899. Grammatical Sketch and Vocabulary.
A few words and phrases are to be found in Last's

Polyglotta Africana Orientalis, pp. 69-72.

Kanyoka. See under ' Congo.'
^Karanga(Chino,Ch\s\vma, Chizwina, Mashona, Shuna).

E. Biehler (S.J.) English-Chiswina Dictionary, with
Outline Grammar. Roermond (J. J. Romer and
Sons), 1906.

•» Four Methods of Teaching English
to the Maswina. Roermond (same publishers), 1906.

Testamente. (BibleStories). Roermond

(same publishers), 1906.

APPENDIX II. 323

Rev. H. Buck. A Dictionary with Notes on the
Grammar of the Mashona Language, commonly
called Chiswina (Compiled at St. Augustine's
Mission, Penhalonga.) (S.P.C.K.), 1911.

W. A. Elliott. Dictionary of the Tebele and Shuna

Languages, 1897.

The second edition of this book (1911) which omits
the ' Shuna ' edition is entered under ' Zulu.'

Louw, Mrs. C. S. A Manual of the Chikaranga
Language (Grammar, Exercises, Useful Phrases
and Vocabulary), p. 397. Bulawayo (Philpot &
Collins), 1915.

Rev. A. M. Hartmann (S. J.). Outline of a Grammar of
the Mashona Language. Cape Town, 1893.

Kelt, (Lokele). See under ' Congo.'

Kele (Dikele). Spoken near the Gabun Estuary.

Missionaries of the A. B.C. P.M. A Grammar of the
Ba-kele Language. New York, 1854.

Kikuyu.

*A. R. Barlow. Tentative Studies in Kikuyu Grammar
and Idiom. Edinburgh (Blackwood), 1914.

Rev. Father A. Hemery. English-Kikuyu Handbook.
Zanzibar-Nairobi (Roman Catholic Mission), 1903.

*A. W. McGregor (C. M.S.) English-Kikuyu Vocabulary.
(S.P.C.K.), 1904.

A Grammar of the Kikuyu Language. London

(Clay and Sons, printed), 1905.

H. Hinde. Vocabulary (See Kamba).
Kinga.

*R. Wolff. Grammatik der Kinga-Sprache (Deutsch-
Ostafrika, Nyassagebiet), nebst Texten u. Worter-
verzeichniss. Berlin, 1905. Archiv fur das
Studium deutscher Kolonialsprachen, Vol. 3.

Kiniassa. See Nyanja.

Kiyanzi ( = Bangi, Kibangi). See under ' Congo.'

324 APPENDIX II.

Konde. North end of Lake Nyasa.

C. Schumann. Grundriss einer Grammatik der
Kondesprache. Berlin, 1899.

Kongo (See under ' Congo.')

Kwanyama. Spoken by a branch of the people called
Ovambo, in S. W. Africa.

P. H. Brincker. Lehrbuch des Oshikuanjama in
Verbindung mit Oshindonga. (Stuttgart) Berlin,
1891.

Deutscher Wortfiihrer fur . . .
Otji-herero, Oshi-ndonga, u. Oshi-kuanjama. (See
also under Herero.)

*H. Tonjes. Lehrbuch der Ovambo-Sprache Osikuan-
jama. Lehrb. d. Sem. f. or. Spr., Vol. 24.

Worterbuch der Ovambo-Sprache. Ib.,
Vol. 25. Berlin (G. Reimer), 1910.

Lala-Lamba. Spoken to the south of Lake Bangweolo.

A. C. Madan. Lala-Lamba Handbook. Oxford
(Clarendon Press), 1908.

Lala-Lamba- Wisa-English, and
English-Lala-Lamba-Wisa Dictionary. Oxford
(Clarendon Press), 1913.

*Lenge. Also called Chopi and Tswa. Spoken in
Portuguese S. E. Africa, between Inhambane and
the Limpopo.

*Bp. Smyth and J. Matthews. A Vocabulary with a

short Grammar of Xilenge. London (S.P.C.K.),

1902, 1912.
Lenje. North-Western Rhodesia ; allied to I la,

which adjoins it on the west, and Tonga, spoken to

the north.

*A. C. Madan. Lenje Handbook. (Oxford University
Press), 1908.

Lolo. See under ' Congo.'

Lotnongo (Mongo) = Lolo. See under ' Congo.'

APPENDIX II. 325

Luba.

W. M. Morrison. Grammar of the Buluba-Lulua

Language, and Dictionary. Privately printed, 1907.

' The Buluba and the Lulua people . . . together
occupy a large area . . . extending, roughly
speaking, from the junction of the Lulua and Kasai
rivers in a general south-easterly direction into
Garenganze, where the language is called Ciluba'
[Chiluba — elsewhere Bu-luba]. (Preface).

P. A. Declercq. Grammaire do la Langue Luba, with
Vocabulary, pp. 504. Louvain (Istas), 1903.

Grammaire pratique de la Langue
Luba. Brussels (Polleunis and Ceuterick), 1911.

Lulua. See under ' Congo.'

Lunda. An important language spoken on the water-
shed between the Congo and Zambesi, near the
sources of the Kasai^pd to the south of the Luba
country.

H. A. Bias de Carvalho. Methodo Pratico parafallar a
lingua de Lunda. Lisbon (Imprensa Nacional), 1890.

Lityi (Rotse). Spoken by the people of Barotseland
(Lewanika's country on the Upper Zambezi.)

E. Jacottet. Etudes sur les langues du Haut-Zambeze.
lr.e Partie. Grammaire Soubiya et Louyi, 1896.
3'oe Partie. Textes Louyi, Contes, Legendes.
Superstitions et Vocabulaires. Paris (Ernest
Leroux), 1901.

Machanie. A dialect of Chaga spoken by about 16,000
people living on the western side of Kilimanjaro.

*Julius Augustiny. Kurzer Abriss des Madschame-
dialekts. Berlin, 1914. Archiv. fur d. Stud,
deutscher Kolonialsprachen. Vol. 16.

Makonde. Spoken in the country north of the Rovuma.
(E. Africa), about as far as Lindi.

E« Steere. Collections for a Handbook of the Makonde
Language. (U.M.C.A.), Zanzibar, 1876.

326 , APPENDIX II.

Makua. In Mozambique

*Chauncy Maples. Collections for a Handbook of the

Makua Language. London (S.P.C.K.), 1879.
Archdeacon Woodward is preparing a new and
revised edition of this little wdfk.

D. J. Rankin. Arabian Tales, translated from Swahili
to Makua. (Tugulu dialect), London, 1887.

Mang'anja. See Nyanja.
Masaba. See Gisu.
Matumbi.

B. Krumm. Grundriss einer Grammatik des Kimatumbi,

1912. Mitt. B. Sem. Or. XV., Abt. 3, pp. 1-63.
Spoken by the inhabitants of the Matumbi hills, inland
from Kilwa. Vocabulary, ib. XVI. Abt. 3, pp. 1-57.

Mbundu (Bunda, Kimbundu, Umbundu, Angola.)
Spoken in Portuguese W._Africa, south of the Congo.

B. M. de Cannecattim. Diccionario da Lingua Bunda

ou Angolense. Lisbon (Impressao Regia.), 1804.
Three parallel columns : Portuguese, Latin, Mbundu.

Colleccao de Observances grammaticaes
sobre a Lingua Bunda ou Angolense, 1805.
Appended to this is a brief Dictionary in four columns.
Portuguese, Latin, Mbundu, Kongo. Second
edition, 1859.

*H. Chatelain. Kimbundu Grammar (Grammatica
elementar do Kimbundu ou lengua de Angola.)
(Port, and English), Geneva, 18S9.

Grundziige der Kimbundu-Sprache, 1890.
Published in Ztschr.filr afr. Sprachen, avowedly as an
abstract of the preceding, though the author says he
has introduced some new points.

Folk-tales of Angola. Boston, New
York, 1894.

Published by the American Folk-Lore Society.
Mbundu text, with English translation on opposite page.

APPENDIX II. 327

W. H. Sanders, W. E. Fay and others. Vocabulary

of the Umbundu Langur«ge, comprising, Umbundu-

English and English-Umbundu. Boston (Beacon

Press), A. B.C. P.M., 1885.

Contains 3,000 words of the dialect spoken inland in

Benguela.

W. M. Stover. Observations upon the Grammatical
Structure and use of the Umbundu. Boston, 1885.

Mongo ( = Lolo). See under ' Congo.'
Mpongw€. See Pongwe.

Naimvanga. Spoken by the Winamwanga, N. W. of
Lake Nyasa.

E. H. Dewar. Chinamwanga Stories (with English
translation.) (Livingstonia Mission Press), 1900.

^Ndonga. The language of one of the tribes known
collectively as Ovambo, in the northern part of

' Damaraland.'

*P. H. Brincker. Lehrbuch des Oshikuanjama in
Verbindung mit Oshindonga. Lehrbiicher des
Seminars fur orient. Sprachen, Vol. VIII. Berlin,
1891. See also Kwanyama.

P. H. Brincker. Deutscher Wortfiihrer fur die Bantu-
dialekte . . . Oshindonga, etc. See under
Herero and Kwanyama.

*A. Seidel. Grammatik des Oshindonga, etc. Also
entered under Herero.

Ngotnbe. See under ' Congo.'

Nika (more correctly Nyika).

There is no language properly called by this name,
which is applied to the Rabai, Giryama, Duruma, Digo
and five smaller tribes.

*J. L. Krapf and J. Rebmann. A Nika-English
Dictionary. Edited by T. H. Sparshott. London,
1887.
The words in this book are chiefly Rabai.

328' APPENDIX II.

A. D. Shaw. See Vocabulary of four E. African
Languages v. Swahili.

\Nyamwezi. Spoken over a large area to the south of
Lake Victoria. Sukuma and Sumbwa are dialects
of it.

*E. Steere. Collections for a Handbook of the
Nyamwezi Language. London (S.P.C.K.), n.d.

R. Stern. Erne Kinyamwesi Grammatik, Berlin, 1906.
Mitt. B,< Sem. Or. IX. 3, pp. 129-258.

*C. Velten. Grammatik des Kinyamuesi (with
Vocabulary). Gottingen. (Vandenhoeck and
Ruprecht), 1901.

*Nyanja (Chinyanja, Mang'anja, Nyasa, Chinyasa).

Is also virtually identical with Sena, and very
similar to Nyungwe (Tete).

*Rev. H. C. R. Barnes. Nyanja-English Vocabulary.

London (S.P.C.K.), 1902.

This is an enlarged edition of Miss Woodward's
Vocabulary of 1892, 1895, q.v.

V. J. Courtois, S.J. Elements de Grammaire Tetense
(Lingua Chi-Nyungue). Coimbra (University
Press), 1900.

G. Henry. A Grammar of Chinyanja. Aberdeen,

(G. and W. Fraser), 1891. Second edition, 1904.
A. Hetherwick. A Practical Manual of the Nyanja

Language. , London (S. P. C.K.), second edition, 1912.
R. Laws. English-Nyanja Dictionary. Edinburgh

(James Thin), 1894.
*R. S. Rattray. Some Folklore, Stories and Songs,

with English translation and notes. London

(S.P.C.K.), 1907.
*J. Rebmann. Dictionary of the Kiniassa Language.

Edited by L. Krapf. St. Chrischona, near Basle,

1877.

(Ki-nyasa=Chi-nyanja. The Anyanja are called
Anyasa by the Yaos and Swahili, Rebmann obtained
his materials from released slaves in East Africa.)

APPENDIX II. 329

*Rev. D. C. Scott. Cyclopaedic Dictionary of the
Mang'anja Language. Edinburgh (Foreign
Missions of Church of Scotland), 1892.

*M. E. Woodward. English-Chinyanjaand Chinyanja-
English Vocabulary, 1892, reprinted 1895
(S.P.C.K.)
Another edition, revised and enlarged by the Rev.

H. Barnes, appeared in 1913.

Exercise-book (S.P.C.K.),..1898, 1909.
Nyika. See Nika.
*Nyoro (Uganda Protectorate).

*H. E. Maddox. Elementary Lunyoro Grammar, with

Lunyoro-English Vocabulary. London (S.P.C.K.),

1902.
Nyungwe (Tele). Spoken in the country about Teteon

the Zambezi. Very similar to, if not identical

with Nyanja.

V. J. Courtois. Diccionario Cafre-Tetense-Portuguez,
1900.

Diccionario Portuguez-Cafre-Tetense,
1900.

Elementes de Grammatica, 1909.

A. v. d. Mohl, S. J. Grammatik der Bantusprache von
Tete. Mitt. B. Sera. Or. VII. Abt. 3, pp. 32-85,
1904. Vol. VIII. 3 (1905), contains a collection of
tales with German translation.

Pahouin. See Fan.

Pogoro. Spoken in E. Africa, somewhat east of the

north end of Lake Nyasa, and north of the Rovuma.
*J. Hendle (O.S.B.). Die Sprache der Wapogoro.

Berlin (G. Reimer), 1907. Archiv fiir deutsche

Kolonialsprachen, Vol. VI.
iPokomo. Tana River, British East Africa.
*C. Meinhof. Linguistische Studien in Ostafrika. No.

VII. Mitt. B. Sem. Or. Jahrg. XIV., Abt. 3.

Berlin, 1911.

330 APPENDIX II.

Pokomo-Grammatik mit Uebungsstiicken. (The work
of one or more of the Neukirchen missionaries, but
no author's name appears.) Neukirchen, Missions-
buchhandlung (Stursberg und Cie), 1908.

*F. Wiirtz. Worterbuch des Ki-tikuu und des Kipo-
komo. Published in Zeitschrift fur afrikanische
und oceanische Sprachen, Vol. I., p. 193.
This is a German-Tikuu and Pokomo Dictionary.
(Tikuu is a Swahili dialect. See under Swahili).

Grammatik des Pokomo, ib. Vol. II., pp.
62, 168.

Vol. 1. of the same periodical contains some Pokomo
songs, and Vol. II. some traditions (all with German
translation).

Some grammatical notes (1889), and a Pokomo-
German vocabulary had previously been published by
F. Wiirtz in Biittner's Zeitschrift fur Afrikanische
Sprachen.

Pongwe (Mpongwe). Spoken in the country adjoining
the Gabun estuary, French Congo.

J. R. Wilson (a late Missionary). Heads of the Mpongwe
Grammar, containing most of the principles needed
by a learner. New York (Mission House, 23,
Centre Street), 1879.

R. P. Le Berre. Grammaire de la Langue Pongouee.
Paris (Maisonneuve et Cie), 1873.

Missionaires de la Congregation de Saint Esprit.
Dictionnaire Fransais-Pongue, 1877-81. Diction-
naire Pongue-Fran9ais, 1881. Paris (Maisonneuve
et Cie).

Missionaries of the A.B.C.F.M. Gaboon Mission. A
Grammar of the Mpongwe Language, with
Vocabularies. New York (Snowden and Prall),
1847.

Poto. See under ' Congo.'

\Ronga (Shironga). A branch of the Thonga language,
spoken in the neighbourhood of Delagoa Bay.

APPENDIX II. 331

H. A. Junod. Grammaire (with Ronga-French-English-
Portuguese Vocabulary and Dialogues). Lausanne
(Bridel), 1896.

Nouveaux Contes Ronga. Neuchatel
Imprimerie (Paul Attinger), 1898. See also Thonga.

Rotse. See Luyi.

Ruanda. N. end of L. Tanganyika.

*P. Eugene Hurel. Manuel de Langue Kinyaruanda.
Mitt. B. Sem. Or. XIV., Abt. 3, pp. 1-159, 1911.

Ruttdi. Between Tanganyika and Lake Kivu, on the
north. Very similar to Ruanda.

R. P. J. M. van der Burgt, des Peres Blancs. Diction-
naire Fransais-Kirundi. Bois le Due, 1900-1903.

Elements d'une Grammaire Kirundi.
Mitt. B. Sem. Or. V., Abt. 3.

R. P. F. Menard, des Peres Blancs. Dictionnaire
Fran9ais- Kirundi et Kirundi-Fran9ais. Paris
(Guilmoto), 1909.

Grammaire Kirundi. Same publisher,
1908.

Sena (Lower Zambezi, virtually identical with Nyanja).

*W. G. Anderson. Introductory Grammar of the Sena
Language (S.P.C.K.), 1897.

J. Torrend, S.J. Grammatica do Chisena. Grammar
of the Language of the Lower Zambe/.i. Chipanga,
Zambezia. (Mission Press), 1900.
In parallel columns, Portuguese and English.

Senga (Middle Zambezi).

*A. C. Madan. Senga Handbook. Oxford (Clarendon
Press), 1905.

\Shambala. Usambara, East Africa, inland from Tanga.

P. E. Homer. Kleiner Leitfaden zur Erlernung des
Kischambala. Mariannhill (Natal), 1900.

*K. Roehl. Versuch einer systematischen Grammatik
der Schambalasprache. Hamburg, 1911.

332 APPENDIX II.

Frau Rosier and F. Gleiss. Schambala-Grammatik
und Worterbuch. Berlin, 1912. Vol. XIII. of
Archiv fur das Stadium deutscher Kolonialsprachen.

E. Steere. Collections for Handbook of the Shambala
Language, 1867. Revised by Archdeacon
Woodward. (U.M.C.A.) Msalabani, 1905.

Shangaan. See Thonga.
Shuna. See Mashona.
Siha (Kisiha). See Chaga.
Soko. See under Congo."
Subiya (Upper Zambezi).

E. Jacottet. Grammaires Soubiya et Louyi, 1896.
Textes Soubiya, 1899. See also Luyi.

Sukuma. (On the south-eastern side of the Victoria
Nyanza. A dialect of Nyamwezi.)

*Capt. Herrmann. Kissukuma, die Sprache der
Wassukuma. Mitt. B. Sem. Or. I., Abt. 3, pp. 146-
1 98, 1898. Gram. Sketch, with Vocabulary and Texts.

A. Seidel. Das Kisukuma. Grammatische Skizze
(with Vocabulary), 1894.

Sumbwa. Spoken in a district of the N.W. part of
Unyamwezi, between Usukuma and Uha, south
of Lake Victoria, but separated from it by Uzinja.

*A. Capus (of the White Fathers). Grammaire de
Shisumbwa (Ztschr. fur afr. u. oc. Spr. Vol., IV.,
pp. 1-123), 1898.
The preceding volume of the same periodical contains

(pp. 358-381) ten stories and some songs and proverbs,

with literal French translation.

Dictionnaire Shisumbwa-Frangais, pp. 147.
Saint-Cloud (Impr. Belin Freres), 1901.

Swahili.

H. W. M. Beech. Studies in Ki-Swahili London
(Kegan Paul, Trench, Trubner & Co.), 1918.

APPENDIX II. 333

E. Brutel. Vocabulaire Fran9ais-Kiswahili et Kiswahili-
Fran9ais, 2 ed. Brussels, 1913.

Mrs. F. Burl. Grammar and Vocabulary (Mombasa
dialect). (S.P.C.K.), 1910.

C. G. Biittner. Worterbuch der Suaheli-Sprache.
2 pts. Stuttgart (Berlin), 1890.

Suaheli-Schriftstiicke in arabischer
Schrift. Vol. X. of Lehrbiicher des Seminars fur
orientalische Sprachen. (\V. Spemann), Stuttgart
and Berlin, 1892.

Anthologie aus der Suaheli-Litteratur.
(E. Felber), Berlin, 1894. Texts (prose and poetry)
with translation into German.

Habari za Wakilindi. Pt. I., n.d. Pt. II. 1904.
Pt. III., 1907. (U.M.C.A.), Msalabani. Traditions
of the Washambala written in Swahili by Abdallah
bin Hemed bin AH Liajjem.

R. P. A. Hemery de la Congregation du St. Esprit
et du Saint-Cceur de Marie. Vocabulaire Fran9ais-
Swahili-Teita. Zanzibar (Mission Catholique).
Paris (30 Rue Lhomond), 1901.

W. K [isbey] . Notes and Corrections of Swahili. I. -IV.
Zanzibar (U.M.C.A.), 1898-1899.

Kibaraka. Zanzibar. (Univ. Mission Press), 1896.
Stories written or dictated by natives.

J. L-. Krapf. Outlines of the Elements of the Kisuaheli
Language, with special reference to the Kinika
Dialect. Tubingen (Friedr. Fues), 1850.

A Dictionary of the Suahili Language,
with Introduction, containing an outline of a Suahili
Grammar. London (Triibner & Co.), 1882.

*A. C. Madan. English-Swahili Dictionary. Oxford
(University Press), 1894. Second edition, 1902.

Swahili -English Dictionary. Oxford

(University Press), 1903.

334 APPENDIX II.

* A. C. Madan. Swahili Grammar. Oxford (University
Press), 1905.

C. Meinhof. Die Sprache der Suaheli. Berlin (Dietrich
Reimer (Ernst Vohsen) ), 1910. Deutsche Kolonial-
sprachen Bd. 2.

W. Planert. Die syntaktischen Verhaltnisse des Suaheli.
Berlin (W. Susserott), 1907.

Ch. Sacleux.; Dictionnaire Fran9ais-Swahili. Zanzibar
(Mission des P. P. du St. Esprit.) Paris (30 Rue
Lhomond), 1891.

Grammaire des Dialectes Swahilis. Paris
(Procure des Peres du S. Esprit), 1909.
This book obtained the Prix Volney from the Institut
de France.

*A. Downes Shaw. Pocket Vocabulary of Four E.
African Languages. (Ki-Swahili, Ki-Nyika, Ki-
Taita and Ki-Kamba; with vocabulary of Kibwyo
dialect). London (S.P.C.K.), [1885].

*E. Steere. A Handbook of the Swahili Language, as
spoken at Zanzibar. London (S.P.C.K.) First
edition, 1871 ; second edition, 1875 ; third edition
revised and enlarged by A. C. Madan, 1884 ; fourth
edition, 1913.

Swahili Exercises. (S.P.C.K.), 1894-
1908.

Swahili Tales, 1889. Reprinted, 1906 and

1917. (S.P.C.K.), London.

Practical Guide to Use of the Arabic
Alphabet in writing Swahili, 1892. (Out of print.)

Capt. C. H. Stigand. Grammar of Dialects in the
Kiswahili Language (with Introduction by the Rev.
W. E. Taylor). Cambridge (University Press),
1915.

W. E. Taylor. Groundwork of the Swahili Language
Tabulated. London (S.P.C.K.), 1898.

APPENDIX II. 335

W. E. Taylor. African Aphorisms, or Saws from

Swahililand. London (S.P.C.K.), 1891.
Swahili Proverbs, translated and annotated. Some
Rabai and Giryama proverbs are appended.

C. Velten. Suaheli -Worterbuch (Part I. Swahili-
German). Berlin. Published by the Author, 1910.

Praktische Grammatik der Suaheli-
Sprache. Berlin (W. Baensch), 1905.

Praktische Anleitung zur Erlernung der
Schrift der Suaheli. Gottingen (Vandenhoeck und
Ruprecht), 1901-1910.

A useful guide to the reading and writing of Swahili
in the Arabic character.

Safari za Wasnaheli. Gottingen, 190L
Narratives of journeys into the Interior (and in two
cases to Europe), written or dictated by natives.

Dtsturi za Wasnaheli. Gottingen, 1903.
A very full account of native customs, written by
natives, in Swahili.

Marchen und Erzahlungen. Stuttgart
(Spemann), Berlin, 1898. Vol. 18 of Lehrb. d.
Sem. fur orient. Sprachen.

Prosa und Poesie der Suaheli. Berlin
(Published by the Author), 1907.
Contains tales, proverbs, dialogues, poems (mashairi)
and popular songs.

Tabwa. Spoken in the Marungu country, between
Tanganyika and the Lualaba.

*G. De Beerst. Essai de Grammaire Tabwa. Berlin,
1896. Published in Ztschr. f. afr. u. oc. Spr. Vol.
II., Nos. 3 and 4.

Taita (less correctly, Teita). Spoken in the Taita Hills,
120 miles N.W. of Mombasa.

*J. A. Wray. Elementary Introduction to the Taiia

Language. London (S.P.C.K.), 1894.
A Taita Vocabulary is included in A. D. Shaw, Pocket

336 APPENDIX II.

Vocabulary. See under Swahili. Also in Hemery,
Voc. Fran$ais-Swahili-Teita. See under Swahili.

Tebele. See under Zulu.
Teke. See under Congo.

*Thonga. Spoken over a large area between St. Lucia
Bay and the Sabi River and including among its
branches Ronga, Hlanganu, Gwamba (now isolated
in the Transvaal), Jonga, etc. Not to be confused
with Tonga, q.v.

*C. W. Chatelain and H. A. Junod. Pocket Dictionary
Thonga-(Shangaan)-English and English-Thonga.
Preceded by an Elementary Grammar. Lausanne
(Georges Bidel et Cie), 1904.

Shangaan (properly Hlanganu) is the name by which
the Delagoa Bay natives in general are known at the
Johannesburg mines. This book, while not ' limiting
itself to any particular dialect ' of the Thonga language,
applies more especially to that spoken in the Spelonken
and Leydsdorp district of E. Transvaal.

Tonga (Zambezi). The Tonga (Gitonga) of Inhambane,
identical with Lenge (q.v.) or Chopi, is distinct
from this. So is the Tonga found on the W. side
of L. Nyasa.

J. R. Fell (of the Baila-Batonga Mission). A Tonga
Grammar. London (S.P.C.K.), 1918.

A. W. Griffin. Chitonga Vocabulary of the Zambezi

Valley. Oxford (University Press).
This Tonga language has been very fully studied by
Father Torrend, who gives some annotated texts in the
Appendix of his Comparative Grammar. The people
are also called (by the Bechwana) Batoka.

Tugulu. See Makua.

Tumbuka. Spoken W. of Lake Nyasa.

W. A. Elmslie. Notes on the Tumbuka Language.
Aberdeen (G. and W. Fraser), ' Belmont ' Works,
1891.

APPENDIX II. 337

W. A. Elmslie. Table of Concords and Paradigms
of Verb. Aberdeen (Fraser), ' Belmont ' Works,
1891.

Venda. Spoken in N. Transvaal, within the bend of
the Limpopo. Sometimes spelt Wenda ; the people
are variously called Vavenda, Bavenda, Wawenda,
etc.

C. Meinhof. DasTsivenda. Leipzig, 1901. Reprinted
from Z.D.M.G.

Th. and P. Schwellnus. Die Verba des Tsivenda. Mitt.
B. Sem. Or. VII. Abt. 3, pp. 12-31. Berlin, 1904.

Vili. Spoken on the Luango (Loango) coast, north of
the Congo. See also " Congo."

P. C. Marichelle. Dictionnaire Vili-Fransais, 1902.

Methode Pratique pour 1'Etude du
Dialecte Vili, 1907.

Xilenge (Shilenge, or Chopi). See Lenge.

*Xosa (' Kaiir ')• Spoken in the eastern part of the
Cape Province, and closely allied to (though not
quite identical with) Zulu.

W. Appleyard. The Kafir Language, comprising a
Sketch of its History . . . Remarks upon its
Nature and a Grammar, pp. 390. King William's
Town, London. (J. Mason), 1850. - (Printed for
the Wesleyan Missionary Society.)

J. Ayliff. A Vocabulary of the Kafir Language.
London (Wesleyan Mission House), 1846.

W. B. Boyce. A Grammar of the Kaffir Language.
London (Wesleyan Missionary Society and J.
Mason). First edition, 1834 ; second edition
(augmented and improved), 1844 ; third edition,
(augmented and improved with Exercises), 1863.
The Exercises were added to the third edition by
W. J. Davis.

C. J. Crawshaw. A first Kafir Course, pp. 133.
Lovedale, Cape Town (Juta). Third edition, 1897;

338 APPENDIX II.

fourth edition, 1901 ; fifth edition, 1903. Grammar,
Exercises and Vocabularies. (These are appended
to each exercise, but can be easily consulted by
means of an index at the end.)

Wm. J. Davis. A Dictionary of the Kaffir Language ;
including the Xosa and Zulu Dialects. Part I.,
Kaffir-English. London (Wesleyan Mission House),
1872.

/ An English and Kaffir Dictionary,

principally of the Xosa-Kaffir, but including also
many words of the Zulu- Kaffir Dialect. London
(Wesleyan Missionary Society), 1877.

I. Bud-Mbelle, Interpreter to the High Court of
Griqualand. Kaffir Scholar's Companion. (Love-
dale Missionary Press), 1903
Contains lists of words, idioms, proverbial expressions,

and a variety of miscellaneous information not always

easy to find elsewhere.

A. Kropf. A Kaffir-English Dictionary, pp. iv., 486.
(Lovedale Mission Press), 1899.

J. McLaren. A Concise Kaffir-English Dictionary.

London (Longmans, Green & Co.), 1915.

A Grammar of the Kaffir Language.

London (Longmans, Green & Co.), 1906.
C. Meinhof. Hottentottische Laute und Lehnworte im

Kafir. (Z.D.M.G.), 1905.

Discusses the question of how far Xosa borrows
sounds and words from the Hottentot language, and in
particular, the origin of the clicks.
W. B. Rubusana. Zenk'inkomo Magwalandini.

Second edition. Frome and London (Butler and

Tanner), 1911.

Traditions and songs of the Xosa, Gcaleka, Tembu
and other tribes, collected by a native minister of the
Congregational Church.
J. Stewart. Outlines of Xosa Grammar, with practical

exercises. (Lovedale Mission Press), South Africa,

1901.

APPENDIX II. 339

J. Stewart. Kaffir Phrase Book and Vocabulary.

Third edition. (Lovedale Mission Press), 1901.
The late Dr. Stewart is well-known as the founder
andTirst Principal of the Lovedale Institution.

J. Torrend. Outline of a Xosa Kafir Grammar, with a
few dialogues and a Kaffir Tale. Grahamstown
(T. and G. Sheffield), 1887.

\Yao (Chiyao, Kihiau). Spoken in the mountains S. E.
of Lake Nyasa, and in the Shire Highlands.

*A. Hetherwick. Introductory Handbook and Vocabu-
lary. (S.P.C.K.) Second edition, 1902.
Contains both Yao-English and English-Yao
Vocabulary.

A. F. Pott. Uber die Kihiausprache. (Z.D.M.G.),
VI., pp. 331-348.

*E. Steere. Collections for a Handbook of the Yao
Language. London, 1871.

Yaunde (a branch of Fan).

P. Hermann Nekes. Praktische Grammatik der Jaunde-
Sprache. Vol. XXVI. of Lehrbiicher des Seminars
fur or. Sprachen. Berlin, 1911.

Yombe. See under ' Congo.'
Zaramo (Zalamo). See, Dzalamo.

Zigula. East Africa, near Luvu River, on the mainland
opposite Zanzibar.

*W. H. Kisbey. Zigula-English and English-Zigula
Dictionary. Loudon (S.P.C.K.), 1906.

Zigula Exercises. London (S.P.C.K.),
1906.

Rev. W. G. Webster (ed.) Zigula Tales. London

(S.P.C.K.), 1912.
Twenty-three stories, written down by natives.

H. W. Woodward. Collections for a Handbook of the
Zigula Language. (U.M.C.A.), Msalabani, 1902.

340' APPENDIX II.

Ziba (Lusiba). Spoken in Kiziba and some other
districts adjoining Lake Victoria on the S.W. It
is not very happily named, as the people speaking it
appear to be called Batundu. It is closely related
to Nyoro.

*Capt. Herrmann (formerly .of Bukoba). Lusiba, die
Sprache der Lander Kisiba, Bugabu, etc., 1904.
Mitt. B. Sem. Or. VII., Abt. 3, pp. 150-200.

*Zulu. ,

A. T. Bryant. A Zulu-English Dictionary, with Notes
on Pronunciation, a revised Orthography, etc.
(Mariannhill Mission Press), Pinetown, Natal,
1905.

An important work, somewhat spoilt by its speculative
etymologies which are not based on any sound principle.
The introduction, too, though containing a great deal of
useful information, is of very unequal value, especially
the historical part, which is not free from parti pris.

H. Callaway. Nursery Tales, Traditions and Histories
of the Zulus. (Zulu Text, Translation and Notes).
Springvale, Natal. London (Triibner & Co.), 1868.

Religious System of the Amazulu. (Zulu

Text, with Translations and Notes. (Printed,
Springvale), Natal. London, 1870.

J. W. Colenso (Bishop of Natal). First Steps in Zulu-
Kaffir. Pietermaritzburg (Vause & Slatter), fourth
edition, 1903.

- Zulu-English Dictionary, fourth edition.

Pietermari.tzburg (Vause & Slatter), 1905.

Three Native accounts of a Visit to,

Umpande, King of the Zulus. With Translation,
Vocabulary and Notes. Third edition. Pieter-
maritzburg and Durban (Vause, Slatter & Co.),
1901.

Izindab'ezinhle, etc. New Testament

(reprinted, 1897 for Miss Colenso), London. (Dent).

APPENDIX II. 341

J. W. Colenso. Pilgrim's Progress. Inncwadi ka'
Bunyane okutiwa Ukuhamba Kwesihambi. Pieter-
maritzburg and Durban (Vause, Slatter & Co.),
1901.

J. L. Dohne (Missionary to the American Board, C.F.M.).
A Zulu-Kafir Dictionary. Cape Town, 1857.

*W. A. Elliott. Notes for a Sindebele Dictionary and
Grammar, with illustrative sentences. (Sindebele
Publishing Co.), Bristol, 1$11.
Sindebele (Tebele) is the dialect of Zulu spoken by

the Matabele in Rhodesia.

Lewis Grout. The Isizulu ; a grammar of the Zulu
Language, with historical introduction. Pieter-
maritzburg, Durban, London (Triibner & Co.),
1859.

James Perrin. English-Zulu Dictionary. Pieter-
maritzburg (P. Davis & Sons), new edition, 1901.

Rev. C. Roberts. The Zulu-Kafir Language simplified
for Beginners. London (Kegan Paul). Third
edition, 1909.

An English-Zulu Dictionary with

the Principles of Pronunciation and Classification
fully explained. London (Kegan Paul, Trench,
Triibner & Co.), 1911.

A Zulu Manual or Vade Mecum.
London (Kegan Paul), 1900.

A companion volume to the two preceding works,
containing grammatical notes and illustrations of special
idioms, — medical, zoological and botanical vocabularies,
etc.

P. A. Stuart. Zulu Grammar, with 400 Useful Phrases,
1907.

INDEX

Abstract nouns, 62.
Adjective roots, 120.
Adjectives, 118, 122.

— few r«al in Bantu, 118.

— concord of, 124, 128.

— derived from verbs,

120.

— nouns made to do the

work of, 119.
Nyanja, 125.
verbs derived from,

130, 154-
verbs used for, 119.

— which take shortened

prefixes, 1*6.

— Zulu, 126.

Adverbial demonstratives, 99, 114.
Adverbs, 184.

— invariable, 186.

— locative, 185.
Agglutinative languages, 12.
Alliterative concord, 14, 20.
Angola and Loango languages, 5.
Animals, names of, 47, 58.
Applied verbs, 148.
Arbousset, 7.

Article, 49, 72.
Assimilation, 226.

Incomplete, 227.
Augmentative class, 56, 66-68.
Auxiliaries, 174.

Bantu, 9.

— family, characteristic fea-

tures of, 14.

— languages, number of

known, 2.

— languages, principal fea-

tures of, 2.

— languages, sounds of, 17.

— name, 3.

— verb, 143.
Barlow, 229.
Bentley, 116.

Bleek, 3, 8, 9, 13, 49, 86. 219.
Boyce's Xosa Grammar, 7.
Brusciotto, Giacinto, 6, 13, 31, 32,
Burton, 31. [33, 72.

Casalis, 7.

Causative verbs. 148.

Cerebral t and d, 222.

" Chiswina," 42.

Chwana, 4, 16, 47, 48, 32, 93.

— participle, 118.

— relative particle, 105.

— verbal nouns, 208
Class, augmentative, 56, 68.

— diminutive, Duala, 62.

— lu, 59-

— ^ meaning attached to each,

— three, concords of, 50. [43.
Classes.hints of several other, 68.
Clicks in Zulu and Xosa, 219.
Colenso, Bishop, 170, 192, 202.
Compound tenses, 173.
Compounds, Herero, 213.

Ila, 217.

Nyanja, 216.

Zulu.

Concord of adjectives, 124, 128.
Concords of Class 3, 50.
Congo, Kingdom of, 31, 43.
Conjunctions, 183.
Continuative mood, 167-8.
Copula, 49.

— combined with per«onal

pronouns, 112.

— old demonstrative root,

in.

— sometimes prefixed to
Cust, 6. [adjectives, 114.

Dahl's Law, 228-9.

De Gregorio, 8.

Degrees of Comparison, 131.

Demonstrative, adverbial, 99.

pronouns, 97, 98

343

344

INDEX

Demonstrative, ya, 48.
Demonstratives, adverbial, 99, 114.
Denominative verbs, 212.
Dental t and d, 222.
Derived Forms, r44, 146, 156, 192.
Diminutive class, Duala, 62.

in -anti, 212.
— in ka-t 60.

plural prefix, Nyanja,
Diminutives, 56. [62.

Dinuzulu, 19.
Dissimilation, 226, 228.
Do Couto, P. AntoSnio, 6.
Double Agreement, 97.
Duala diminutive class, 62.
— language, 89.

Ewald, 8.

False Analogy, 226.
Fruits, names of, 52.
Fulfulde language, 44.
Future Tense, 172.

Ganda, 16, 25, 26, 66, 68, 295.

— relative particle, 105.

— verb ' to be,' 116.
Gisu, 42, 48, 49, 53, 66.
Grammatical gender, 10, n.
Grimm's Law, 218.

Hamitic languages, 14.
Herero, 42, 45, 225, 248.

— compounds, 213.

— relative particle, 106. '

— verb ' to be,' n6.
Hetherwick, Dr., 187, 202.
Hottentot language, 9.
Hottentots, 4, 10.

Human beings, names denoting,

[46.

Ila, 264,

— compounds, 217.
Imperative mood, 159.
Incomplete assimilation, 227.
Indicative present tense, 171.
Infinitive mood, 159.
Inflected families of language, 10,
Initial vowel, 48

absent in the voca-
tive, 49.

' Inseparable Pronoun,' 88.
Intensive verbs, 149.
' Interjectional roots,' 186.
International Phonetic Associa-
Intonation, 15, 16. [tion, 18.

Invariable adverbs, 186.
— particle, no.
Irregular verbs, 177.
Isolating languages, 12.

Johnston, Sir Harry, 9.
Jones, D., 19.
Junod, 186, 197.

' Kafir ' (Xosa), 4.

Ki-class, action of a verb, 55.

— collective sense, 55.

— instrumental force, 55.

— 'likeness, fashion, manner,'
Kikuyu, 225, 229. [55.
Kimvita (Mombasa Dialect), 285.
Kinga, 68.

Kongo, 34, 42, 81, 89.

— relative pronoun, 107.

— verb ' to be,' 116.
Krapf, 7, 31.

Lamu Dialect (Kiamu), 276.
Laterals, 220.

Law of Vowel-Harmony, 228.
Lepsius, 8, 14, 19,
Levy-Bruhl, 187, 188.
Lichtenstein, 4, 6, 9.
Liquids, words denoting, 52.
Locative adverbs, 185.

— class, 76-85, 211.

— prefixes, 66.
Locatives, in mu-, 51.

— suffixed, 82.
Luganda, see Gandn.

Marsden, William, 5.

1 Mashona,' 42.

Materials, names of, 62.

Mbundu language, 6.

Meinhof, 8, 16, 19, 35, 37, 45, 48,
195, 203, 214, 219,
226, 230, 231.

Moffat's translation of the Bible
into Sechwana, 7.

Monosyllabic verbs, 143, 177.

INDEX

345

Mood, Continuative, 167,

— Imperative, 159.

— Infinitive, 159.

— Negative, 160.

— Relative, 168.

— Subjunctive, 160.
Moods, 156-169.
Mozambique language, 5.
Miiller, 8.

Negative mood, 160.
Neuter-Passive verbs, 147.
Noel-Armfield, 223.
Noun agent, 200.

— indicating result of an
Nouns, abstract, 62. [action, 205.

— made to do work of adjec-

tives, 119.

— verbal, 200-211.
Numbers, Ordinal, 140.
Numerals, 133.

— distinct words for

' hundred ' and

' thousand,' 137,

table of, 135, 138. [140.

Nyanja, 16, 17, 43, 44, 46, 48, 62,

63, 98, 223, 172.

— adjectives, 125,

— compounds, 216.

— no true relative

— verb ' to be," 116.

Object-Pronoun, 89.

1 Onomatopretic vocables,' 186.

Ordinal numbers, 140.

Pacconio, P. Francisco, 6.
Particle, invariable, no.

— relative, 101.
Participles, 103-6, 118, 169.
Passive verbs, 147.

Past tense, 172.

Perfect in -He 154, 158, 166.

Perfect tense, 173.

Phonetics, General, 230.

Pitch, 16.

Place, word for 79.

Pokomo language, 80.

Possessive, 70-73.

— particle, 74.

pronouns, 74, 92, 93.

Pott, 8.

Prefix, eleventh, /«-, 59.

— fifteenth, AM-, 63,

— fifth, //-, 51.

— ninth, in- or «-, 56.

— ti- in Nyanja, 62.

— ogu-, 66.

— sixth, in Gisu, ktma-, 33.

— tenth, 37

— <«-, attached to thirteenth

— twelfth, 60. [cla«s, 61.
Prefixes, locative, 66, 77.

not identical with Pro-
noun, 86.

Prepositional verbs, see Applied
Prepositions, 72, 84, 182. [Verbs.
Pronominal forms

suffixed to, 91.

Principiation of nouns, 13, 32, 33.
Pronouns, 86, 182.

Demonstrative, 97, 98.

— Inseparable, 88.

— Longer forms of, 91.

— Object, 89.

— Poisessive, 92, 93.

— Prepositional form of,

9i. 92, 93-
Reflexive, 91.
Relative, 99.

— Separable or Substan-

tive, 92, 94.

Rebmann, 31.
Reciprocal verbs, 131.
Reflexive Pronoun, 89-91.
Relative construction, Zulu, 119.

Chwana, 105.

Ganda, 103.

Herero, 106.

Ronga, 106. ,

Swahili, 101.

Zulu, 104.
form of Locative Class,

— mood, 168. [an.

— particle, 101 it seqq.

— pronoun, 99, 108.
Repetitive verbs, 133.
Reversive verbs, 130.
Rhodesia, Southern, main speech
Ronga language, 54. [of, 42.

346

INDEX

Scott, Revs. D. C. and W. A.,

Semi-Bantu, 2. [196, 203.

Separable, or independent, pro-
nouns, 92, 94.

Smith, E. W., 194, 195.

' Sound-pictures,' 186.

Stapleton, 186, 189.

Stative verbs, 152.

Steere, Bishop, 18.

Stress (accent), 15.

Subjunctive mood, 160.

Substantive pronouns, 94.

Sudan languages, }i, 14, 75.

Suffixes, 199, 204.

Swahili language, 7, 15, 16, 18, 28,
42, 44, 47, 67, 101,
119, 225, 276.

— verb ' to b«,' 116.

Tense, 157.

— Compound, 173.

, — Future, 172, 173.

— Indicative Present, 171.

— Past, 172.

— Perfect, 173.
Thonga, 17.
Torrend, 186.
Transposition, 226, 230.
Trees, names of, 51.
Tribes, names of, 50.
Tuckey's expedition to the Congo,

[5-
>Uncle Remus,' 10, 47.

Van der Kemp, Dr., 7.
Velten, 192.

Venda language, 17, 54, 68.
Verb, Bantu» 143.

— stems beginning with a

— ti, use of, 176. [vowel, 145.

— ' to be,' 109, no.

Ganda, 116.
Herero, 116.
Kongo, 116.
Nyanja, 116.
Swahili, 116.
' to have,' 109.

Verbal nouns, 201, 202, 205, 208.
Verbs, Adjectives derived from, 120

— Applied, 148.

— Auxiliary, 174.

— Causative, 148.

— Compounded forms, 155.

— Denominative, 212.

— derived forms, 144, 146, 136.

— derived from adjectives, 130.

— formed from adjective-

— Intensive, 149. [stems, 154.

— Irregular, 177.

— Monosyllabic, 143, 177.

,, primitive, 179.

— Neuter-Passive, 147.

— Passive, 147.

— Perfect in -He, 154, 158, 166.

— Reciprocal, 151.

— Repetitive, 153.

— Reversive, 150.

— Stative, 152.

— used for adjectives, IIQ.

— which do not end in a
'Vocal Images,' 189.
Vocative, 70.

' Voices,' 146.

Von der Gabelentz, 8.

Westermann, 187, 188.
' Whistling s,' 54.
Whitehead, 186, 190.
Whitney, W. D., 219.
Woodward, Archdeacon, 81.

Xosa, clicks in, 219.

— Janguage, 4, 7, 16.

Yao, 1 66, 167, 225.

Zulu, 15, 16, 20, 44-48, 232.

— adjectives, 126.

— clicks in, 17, 19, 219.

— compounds, 216.

— prepositions, 84.

— relative construction, 119.

— ,, particle, 104.

LONDON :

Women's Printing Society, Ltd.,
Brick Street, Piccadilly, W. i.

KEGAN PAUL. TRENCH, TRUBNER & CO.

BEECH (M. W. H.). Aids to the Study of Ki-Swahili.

65. net.

Ki-Swahili Writing (with 22 full-page facsimiles, with Transliterations
en regard) ; Ki-S%vahili Stories (with Translations and Notes) ; Enigmas
and Aphorisms (with explanations) ; Magic in Pemba.

FOULKES (Capt. H. D.). Manual of the Angass Grammar.
With Vocabulary, 8s. 6d. net.

HAYWOOD (Capt. H. W.). EngHsh-Hausa Vocabulary.

Second Edition, containing nearly 1,400 words in
everyday use. 25. 6d. net.

LARAJASSE (F. E. de). Somali-English and English-
Somali Dictionary. 155. net.

- and CYPRIEN de SAMPONT, Practical Grammar
of the Somali Language. With Sentences. 155. net.

MERRICK (Capt. G.). Hausa ProYerbs. 35. 6d. net.

MIGEOD (F. W. H.). The Languages of West Africa : a

Linguistic Survey of about 200 Languages or Dialects.
2 vols., each 125. 6d. net.

Grammar of the Hausa Language. 6s. net.

The Mende Language: including Grammar,

Vocabularies, and Useful Phrases. 75. 6d. net.

- Mende Natural History Yocabulary : a description
of the Fauna and Flora. 45. 6d. net.

ROBINSON (C.). Grammar of the Hausa Language. 6s.net.

STAPLETON (W. H.). Comparative Grammar of the
Congo Languages, and of Swahili. With Vocabulary.

75. 6d. net.

TORREND (J.). Comparative Grammar of the South
African Bantu Languages, comprising those of
Zanzibar, Mozambique, the Zambesi, Kafirland,
Benguela, Angola, The Congo, The Ogowe, The
Cameroons, the Lake Regions, etc. 305. net.

BROADWAY HOUSE, 88-74, CARTER LANE, LONDON, E.G.

BINDING GEC7. FEB 1 0197f

PL Werner, Alice,

8025 1859-1935

. 1 Introductory sketch
W4? of the Bantu languages

K. Paul, Trench,
Trubner (1919)

PLEASE DO NOT REMOVE
CARDS OR SLIPS FROM THIS POCKET

Internet Archive Audio

Featured

Top

Images

Featured

Top

Software

Featured

Top

Books

Featured

Top

Video

Featured

Top

Mobile Apps

Browser Extensions

Archive-It Subscription

Save Page Now

Full text of "Introductory sketch of the Bantu languages"

See other formats