A LINGUISTIC SURVEY OF ADOPTIVES IN VENDA

by

MBULUNGENI RONALD MADIBA

submitted in fulfilment of the requirements for

the degree of

MASTER OF ARTS

in the subject

AFRICAN LANGUAGES

at the

UNIVERSITY OF SOUTH AFRICA

SUPERVISOR: PROF G POULOS

JANUARY 1994
 (i)

DECLARATION

I declare that A LINGUISTIC SURVEY OF ADOPTIVES IN VENDA is my own work

and that all the sources that I have used or quoted have been indicated and
acknowledged by means of complete references .

.~.
SIGNATURE
(Mr M R Madiba) DATE

. -·..:o.... •• .. (~' ; , sA-·. ---~-

818!. :r ·.,-,:; "~ I l ••: :-I,A.RY

1 •nn::. _n')_ t ~

~
s 496.327 S20 05
as MADI
Ct
m\ein ........... .
'lr.JI'-)_.. ~,._ .......- . s

II IIIIII ill/
01565312
I /I~ 11111111111
 (ii)

ACKNOWLEDGEMENTS

I am deeply indebted to the following people who assisted me during the course of this
study:

Professor G. Poulos- My supervisor, whose deep insight into and knowledge of Venda
linguistics facilitated my study and analysis to a large degree.

Mrs L. Pretorius who sacrificed her holiday time to edit this work. Dr M.J. Mafela for
his assistance in proofreading the final draft of this dissertation.

My colleagues in the Department of African Languages who showed an interest in this

work and who provided continuous support, as well as all those people who helped
me during my research. Amongst others, Mr P.A Mulaudzi, who shared some views
on Venda language contact, and Mr M.A Mudau of Mashau, one of my informants,
who helped me during my research. I am also indebted to Prof V.N.M.N Ralushai for
his contribution to some of the historical facts in Chapter 1.

The Subject Librarian for African Languages, Mrs S.G Napaai, who provided me with
relevant sources for this study.

Finally, _my wife, Livhuwani, and my son, Mulalo, whose unfailing support and
encouragement have always been a source of inspiration.

The financial assistance awarded to me by the Human Sciences Research Council

for the STANON programme is hereby also gratefully acknowledged.
 (iii)

ABBREVIATIONS AND SYMBOLS

Afr. Afrikaans
Arab. Arabic
Eng. English
Ve. Venda
NS. Northern Sotho
Tlo. Tlokwa
Lo. Lobedu
Zu. Zulu
Xho. Xhosa
Sho. Shona
Tso. Tsonga
KaL Kalanga
Port. Portuguese
Swah. Swahili
< from
> to, becomes
I I phonemic form
phonetic form
low tone
high tone
stress
 (iv)

SUMMARY

This study deals with the influence of other languages on Venda. It begins by looking
at the various contact situations and then analyses the adaptation of foreign linguistic
forms and their impact on the Venda language.

Chapter 1 gives a historical perspective of Venda. The focus here is on the origin of
the Venda language and the different contacts it has had with other languages during
and after migration.

Chapter 2 analyses the adaptation of foreign linguistic forms to the Venda lexical-
semantic system, while chapter 3 focuses on adoptives in relation to the Venda sound
system.

Chapter 4 focuses on the grammatical (i.e morphological and syntactic) adaptation .

This chapter concludes by looking at the impact of adopted linguistic forms en the
Venda grammatical system.

The final chapter gives a general conclusion on the adjustment of adopted linguistic
forms and their effects on the Venda language.
 ( v)

TABLE OF CONTENTS

INTRODUCTION

(i) Aims of study 1

(ii) Definition of concepts and viewpoints
on adoptives 1
(iii) Research methods 7
(iv) Scope and composition of chapters 9
(v) Conclusion 10

CHAPTER 1

1.0 Historical background of the Venda 11

1.1 Introduction 11
1.2 A brief outline of the history of Venda 11
1.2.1 Possible original migrations of the Venda
and their contact during migration 13
1.2.1.1 Migration from Egypt 13
1.2.1.2 Migrations from Central Africa 14
1.2.1.3 Migration from Malawi 17
1.2.1.4 Migration from Vhukalanga (Zimbabwe) 19
1.2.1.5 The Northern Transvaal as a place of origin 27
1.2.1.6 Summary of the origins of Venda 31
1.2.2 Postmigratory contacts 32
1.2.2.1 Shona 34
1.2.2.2 Northern Sotho 49
1.2.2.2.1 Tlokwa 50
1.2.2.2.2 Lobedu 53
1.2.2.2.3 Other cognates between Venda and Sotho 56
1.2.2.3 Tsonga 62
 (vi)
1.2.2.4 Nguni 65
1.2.2.5 English and Afrikaans 70
1.3 Conclusion 83

CHAPTER 2

2.0 Lexica-Semantic adaptation 86

2.1 Introduction 86
2.2 Lexical adoption 86
2.2.1 Reasons for lexical adoption 87
2.2.1.1 The need-filling motive 87
2.2.1.2 The prestige motive 89
2.2.1.3 A combination of both factors 90
2.2.2 Spheres of influence 90
2.3 Semantic adaptation 90
2.3.1 Semantic adaptation of adoptives 91
2.3.1.1 Meaning correspondence 91
2.3.1.2 Meaning extension 92
2.3.1.3 Narrowing of meaning 98
2.3.1.4 Radical change (shift) of meaning 99
2.3.1.5 Changes in emotive value 101
2.3.2 Semantic adaptation of native words to
adopted concepts 102
2.4 The effects of adoption on the Venda
lexicon and semantics 104
2.4.1 The enlargement of the lexicon of a language 105
2.4.2 Enriching the language with synonyms 105
2.4.3 Loss of native words 108
2.4.4 The problem of homonymy 109
2.4.5 Taboo and euphemism 113
2.5 Conclusion 114
 (vii)
CHAPTER 3

3.0 Adoptives and the sound system of Venda 115

3.1 Introduction 115
3.2 The sound system of Venda 115
3.2.1 Vowels 115
3.2.1.1 Simple vowels 116
3.2.1.2 Diphthongs 116
3.2.2 Consonant phonemes of Venda 117
3.2.2.1 ·The basic inventory 117
3.2.2.2 Venda phonotactics 117
3.2.2.2.1 The complex consonants, compounds and clusters 120
3.2.3 Venda syllable structure 124
3.2.4 Tone 127
3.3 Phonological adjustment on adoptives 128
3.3.1 Attempted explanations of the phonological
changes of adoptives 129
3.3.2 Sound substitution 154
3.3.2.1 Vowel substitution 155
3.3.2.2 Consonant substitution 179
3.3.3 Syllable adjustments 199
3.3.4 Tonal adaptation of adoptives in Venda 216
3.4 Concluding remarks on the effects of
adoptives on Venda phonology 222
3.5 Conclusion 222

CHAPTER 4

4.0 Grammatical adjustment in adoptive forms 224

4.1 Introduction 224
 (viii)

4.2 Morphology and some syntactic aspects of the

adoptive nouns in Venda 224
4.2.1 The morphology of the Venda noun 224
4.2.1.1 Morphological adjustment in the
ad·option of nouns 232
4.2.1.1.1 Affixing of class prefixes to adopted nouns 234
4.2.2 Some syntactic aspects of adoptive nouns in Venda 246
4.3 Adopted words and other word categories 251
4.3.1 Adjectives 251
4.3.2 Adopted verbal roots. 255
4.3.3 Adverbs and conjunctives 258
4.3.3.1 Adverbs 258
4.3.3.2 Conjunctions 261
4.4 The effects of adoptives on Venda grammar 262
4.4.1 The effects on the morphology of the noun 263
4.4.2 The effects on the concordia! agreement system 265
4.4.3 Enrichment to the grammatical system of the language 266
4.5 Conclusion 266

CHAPTER 5

5.0 General Conclusion 268

BIBLIOGRAPHY 277
 1

INTRODUCTION

(i) Aims of study

It is a well-established fact in linguistics that all languages change through time.

According to Thomason and Kaufman (1988:9) change in a language may occur as
a result of three factors, namely:

"drift i.e. tendencies within the language to change in certain ways as a result
of structural imbalances; dialect interference, both between stable, strongly
differentiated dialects and between weakly differentiated dialects through the
differential spread (in 'waves') of a particular change; and foreign interference".

In this study we shall be mainly concerned with the latter change i.e~ change that
occurs in a language as a result of the influence of another language. Therefore, the
primary aim of this study is thus to describe and analyse linguistic changes which
have occurred in the Venda language as a result of the influence of other languages.
The influence of one language on another depends mainly on contact. "The speech
communities must obviously be in fairly regular contact for borrowing to take place"
(Ciuver 1989:267). Although the different contact situations of the Venda language will
be outlined, the main focus will be on the adaptation of foreign linguistic forms which
have been adopted into the Venda language and their impact on the language. In
other words, our approach will be based mainly on linguistic factors, even though
certain social factors will also be taken into account when aspects of the adopted
linguistic forms are explained. The adaptation of these foreign linguistic forms will be
discussed at the phonological, morphological, semantic and syntactic levels.

(ii} Definitions of the concepts and viewpoints on adoptives

In this section the key concepts used in this study are defined and different
approaches to adoptives are briefly sketched in order to situate this study within a
broader perspective.
 2
(a) The tenn ·adoptive'

The term 'adoptive' is used in this study to refer to linguistic forms that have beeri
adopted from other languages. The process of transfer will be called 'adoption'.
These terms have been used. by scholars such as Rassman (1977), Chimhundu
(1982), Khumalo (1984), Cole (1990) and Poulos (1990),. However, other terms have
also been used in studies on adoptives, namely, 'borrowings', 'loan-words', 'foreign
acquisitions' etc. Terms such as 'interference', 'mixing' 'stealing' and 'acculturation'
have also been used to refer to the process by which linguistic forms are transferred
from one language to another. It appears as though there is nO consensus amongst
scholars on the use of any one specific term.

As already mentioned, adoptives have been referred to as 'foreign words' by some

scholars, for example Serjeantson (1968). What is interesting to note is her apt
realisation of the inappropriateness of such a term. She criticises its usage as follows:

"The first thing that is wrong in this book is the title. (The book is entitled: A
history of foreign words in English). Very many of the words dealt with are
certainly not foreign words". (Bracketing - mine)

Therefore it is perhaps a misnomer to refer to the adopted words that have assimilated
into the language structure as 'foreign words'.

The terms 'borrowing' and 'loan words' have been widely used in the literature, mainly
by Haugen (1950), Weinreich (1953) and many other scholars. However, scholars like
Rassman (1977), Chimhundu (1982), Cole (1990) and Haugen himself have expressed
their dissatisfaction with the use of these terms. Cole (1990:345) says:

"Incidentally. such lexical acquisitions are commonly referred to, even in

professional linguistic circles. as 'borrowing' or 'loanwords', but having failed
to find any disposition in languages ever to return such 'loaned' commodities,
I prefer to term them adoptives. In the process of adoption they are adapted
 3

to their new linguistic home and environment, phonologically, morphologically,

and often also semantically, just as an adopted child adapts to the way of life
and behaviour patterns and attitudes of its new family. Having adopted and
assimilated such words from other languages, having thus granted them
linguistic 'citizenship', we relatively quickly forget that they are foreign elements
and believe them to be part of our own linguistic heritage".

Chimhundu (1 982:XIX) finds the terms 'borrowing' and 'loan words' unacceptable for
Shona due to the fact that it gives the general assumption of monolithic and
homogeneous donor languages lending items to equally monolithic and homogeneous
recipient languages. Although some linguists use these terms on the grounds that
they have been widely used in the literature, their use seems to create the impression
that the adopted linguistic forms are in the language temporarily. As long as adoptives
are regarded as 'borrowed' or 'loan' words in the true sense of the-word, they will
never be accepted as standard forms in the recipient language.

Terms such as 'mixing' and 'interference', create the impression of a process which
takes place haphazardly. The adoption of foreign words takes place in an orderly
manner and does not always interfere with the recipient language structure. In most
cases adopted linguistic forms conform to the grammatical structure of the adopting
language. Therefore, it is not always appropriate to refer to adoption as 'mixing' or
'interference'. The use of these terms also expresses undesirability in the purist sense
in that they suggest that there are certain languages which are 'pure' and others which
are 'mixed' (Chimhundu 1982:XIX). The adoption of words from another language is
hence seen as 'interfering with' or 'polluting' the purity of the language. The term
'acculturation' is also used to refer to the process of adoption. Acculturation is the
process by which one group assimilates the cultural traits of another group (Collins
Concise Dictionary). The problem with this term is that it seems to put more emphasis
on culture rather than on language.

In sum the use of terms such as 'borrowings', 'loan words', 'foreign acquisitions'
'language mixing' and 'stealing' may reveal some or other prejudice toward the
 4
process of adoption. This kind of prejudice can be regarded as a purist or
conservative standpoint on language. The use of these terms will thus be avoided in
this study and the term 'adoptive' is preferred. This term is derived from the verb
'adopt' which, according to Webster's New Collegiate Dictionary (1980:16), is defined
as follows:

'to take by choice into a relationship'.

'to take up and practise or use as one's own'.
'to accept formally and put into effect'.

'Adoptives' will therefore be regarded as those foreign linguistic forms that have
received formal acceptance and reflect widespread use in the recipient language. The
advantage of this term is that it lacks the prejudices that are generally implied by the
other terms discussed above. Furthermore, its use represents a somewhat liberal
view of language change. The different viewpoints regarding language change and
adoptives will be discussed in the following section.

(b) Views on adoptives.

The different terms discussed above reflect different viewpoints on adoptives. Here we
shall look at three such viewpoints, namely, the purist, the liberal and the moderate
viewpoints.

The purist viewpoint

This view is maintained by those who, according to Cheng (1985:178), "have

established themselves by going the road of traditional scholarship and those who
believe in purism and superior culture". Such people will try to block any influx of
foreign elements into their language.

According to this view, adoption of linguistic forms from other languages is not
tolerated. The views of purists notwithstanding, the crucial question is whether there
 5
is a language, in this ever shrinking world with all its technological advancement,
which can purport to be totally free from all foreign influences. Knappert (1972:2)
rejects the existence of such a language and says:

" ... pure races do not exist among the human species and I have never been
able to discover a pure language".

Santiago (197 4:328) also vigorously reject~ the notion of a pure language:

"All languages are capable of borrowing from other languages. In fact, no

language in the world is pure or completely free of borrowed words. Somehow
. a language borrows from other languages especially now that the world is
shrinking so fast".

Chinese and Afrikaans are examples of languages where attempts have been made .
to maintain language purity, but to no avail. Afrikaans for example, resisted the
influence of English but could not prevent the adoption of English words into the
language. Nevertheless adoption in this language was restrained to some extent (Van
Wyk 1978:49).

The liberal viewpoint

The purist view is rejected by what is regarded as a more liberal view of language
change. This view unlike the purist one, maintains that languages change with time
since they, cannot be divorced from reality. According to this view, the adoption of
linguistic forms from other languages is seen as a natural development of the
language. In other words, adoption is not seen as a decaying process in a language
but rather as an enrichment of the language. Thus Kunene and Sukumane (1987: 116)
state:
 6

''The process of borrowing should not be looked upon as a negative device of

polluting the language, but it should be considered as one of the devices that
enables the language to expand itself - thus indicating that the language is
complete and alive".

The attitude towards adoption determines the status which adoptives will have in that
language. In languages where the purist view of language predominates, all forms of
adoption are avoided. Kunene and Sukumane (1987:117) t1ave this to say in this
regard:

"The attitude towards borrowed words is one of the main factors that keeps
some languages 'sterile' in this area of terminology that can be used in order
to express new ideas and concepts. Thus the attitude towards borrowed words
might be the cause of lexical sterility in some languages".

In languages such as English and Japanese, adoptives are given equal status to that
of native words. Arlotta (1972:189) describes adoption in the English language as
follows:

"English stands out as being almost unique with regard to the extent to which
it has indifferently borrowed words from any language for whatever purpose.
English speakers do not ordinarily worry ·about the notion that extensive
borrowing causes their language to be corrupted or less 'pure'. ·In English a
native word and a borrowed word may exist comfortably side by side with very
similar or identical meanings".

Each of these two views discussed above has its own merits and demerits, depending
on the situation of a given language. In some languages it is not possible to adopt
one approach only. This then leads us to the third type of approach which is more
moderate in outlook.
 7

The Moderate viewpoint

This viewpoint is described by Okonkwo (1977:322) as one which accepts purism in

some areas and allows adoption in others. According to Santiago (1978) both
extremes ( i.e. purist and Iibera! viewpoints ) are bad and can cripple the natural
growth of a language.

"There must lJe some flexibility, an 'elbow room' for the language by way of
relaxing its inflexible system of borrowing ... On the other hand, stability should
also be maintained by way of putting some constraints on borrowing" · (ibid.
334).

It would seem that most of the African languages, have opted for this kind of
approach. Whilst use is made of the vocabulary expansion devices within the
language itself, certain linguistic forms from other languages Oare also adopted. The
advantage of this approach is that the language is able to make maximum use of both
internal and external resources, and adoptives are usually. accorded their rightful
position when this viewpoint prevails.

(iii) Research methods

(a) Collection of material

Given the fact that very little work has been done on adoptives in Venda, the first task
which faced the researcher in this study was the collection of material. In an effort to
arrive at a representative database a number of techniques were used, namely,
informal observation, formal interviews and the examination of written materials. Each
of these techniques will be briefly discussed below:
 8
lnfonnal observations

The main focus in this study is on the spoken language rather than on the written
form. It is a well supported view that change in a language is first attested in the
spoken language rather than in the written form. As such, collecting data by means
of informal observation is likely to yield valuable lexical material which has not yet
been recorded.

Observations were made during informal meetings, with children at play, during
informal conversations with people in buses, taxis, market transactions and in work
situations. I also participated in some of these interactions since I am a mother-tongue
speaker of the language. Through this method I was able to make some systematic
observations of language change in its social context. Insights into the acceptability
of innovations and frequency of use of the adopted linguistic forms were also gained
in such an investigation.

Fonnal interviews

Formal interviews were also used in selected areas. The main objective of these
interviews was to get more background information of the area under investigation.
People with formal linguistic knowledge were interviewed to provide some clarification
on the linguistic forms existing in their area of expertise.

Examination of written materials.

Various types of literature also served as sources for the collection and analysis of
adoptives. These included, amongst others, literature books, dictionaries, the
handbook of Venda Terminology and Orthography of the Department of Education and
Training, the Thohoyandou newspaper, etc.
 9

(b) Treatment and analysis of the material.

The data was scrutinized to determine the general trend underlying the ways in which
the Venda language is changing. The following steps were followed in the analysis of
the data:

An attempt was made to establish the source languages of the data collected. To this
end, both diachronic and synchronic approaches were used. The diachronic approach
assists in establishing the original forms of the adopted words whereas the synchronic
approach investigates words in their present forms.

(iv) Scope and composition of chapters

This study deals with the influence of foreign languages on the Venda language. For
any two languages to influence each other there must be some form of contact
between them. This study thus begins by looking at the various contact situations and
thereafter the adaptation of foreign linguistic forms and their impact on the Venda
language are analysed.

In Chapter 1, an historical perspective of the Venda language will be given. The focus
here will be on the origin of the Venda language and the different contacts it has had
with other languages. The influence of the foreign languages in these different contact
situations with the Venda language will be assessed.

Chapter 2 will look at the semantic adaptation of foreign words in Venda. An attempt
will also be made to discuss the adaptation of native words to express foreign
concepts and ideas. The chapter will conclude by looking at the impact of the adopted
words on the Venda semantics system.

In Chapter 3, an analysis of the adaptation of the foreign linguistic forms to the Venda
sound system will be made. The effect of the adopted linguistic forms on the Venda
sound system will be assessed.
 10

Chapter 4 deals with the grammatical (i.e. morphological and syntactic) adjustment of
the adopted words in Venda. The morphological modifications of different parts of
speech will be discussed and the syntactic behaviour of some of these parts of speech
will also be considered.

Chapter 5 will form the conclusion. In this chapter the general conclusions regarding
changes to the Venda language due to the influence of other languages will be drawn.

(v) Conclusion.

The purpose of this introductory section was to set out the aims of this study, which
are to analyse and describe the nature and general trends of change in Venda as a
result of its contact with other languages. Different terms which have been used in the
literature to refer to linguistic forms transferred from one language to another were
discussed. The term 'adoptive' appears to be the most appropriate term to refer to
these linguistic forms and this is the term used in this study. The process of transfer
is referred to as 'adoption'. Three different viewpoints on adoptives, namely, the purist,
the liberal and the moderate viewpoints, were briefly discussed and the methodology
and the scope of this research were outlined.
 11

CHAPTER 1

1.0 HISTORICAL BACKGROUND OF THE VENDA

1.1 Introduction

The main focus of this section will be on the history of the Venda language and its
contact with other languages. The historical background of a language is essential in
a study on adoptives since the latter are inextricably bound up with the history of the
speech community (Hanan 1973: 389). Clark (1981 :37) reiterates this view as follows:

"This aspect of vocabulary (adoptives), cannot, therefore, be studied except in

parallel with history, cultural, social and even political".

Although the history of a speech community is important for a study of adoptives, only
cultural-historical aspects which serve as a source of guidelines for analysing
adoptives will be considered in this study. An historical record of a language or
speech community can help one to infer which words were adopted from what
language, approximately when and by whom. This study will therefore take the origin
of the Venda people as its point of departure. Thereafter it will proceed to the
migration of the Venda to their present country. Tlle emphasis here will be on the ~

contacts which Venda had with other languages during this period of migration. Last
but not least, this study will look at the different postmigratory contacts. Postmigratory
contacts arise mainly through geographical proximity and urbanisation.

1.2 A brief outline of the history of Venda.

So far, very little has been written on the history of the Venda. The situation is
exacerbated by the fact that even those few recordings which have been made on the
history of the Venda seem to be superficial and sometimes unreliable (Ralushai.
1977:22). In the absence of a good recorded history it becomes very difficult to
postulate reconstructions of the historical contacts Venda had with other languages,
 12

particularly the earlier contacts. However, it is hoped that those few recordings will go
a long way in reconstructing the history of the Venda. Adoptives can also contribute
in one way or another to the reconstruction of the history of a speech community. In
this regard Knappert (1970:81) has the following to say :

"The total body of adoptives in a language yields a fair picture of the material
and cultural acquisitions of its speakers since they settled in that part of Africa
... the lexicon of a language still gives us exact information on the stage of
acculturation of its users".

This method of reconstructing the history of a speech community by means of

linguistic evidence, adoptives in particular, was pioneered by Ehret. According to him
"historical inferences of much more detailed and more complex kinds can be made
from the diachronic analysis of adopted elements, ~ecifically word-borrowings" (Ehret .,. r'
1976:18). In most of his works dealing with the history of the Bantu languages 1 he
used adoptives to illustrate the development of such languages (cf. Ehret 1967; 1968;
1973). These works are regarded by Phillipson (1977:110) as being of additional
interest because of the light which they may throw on the transmission of the cultural
traits which adoptives describe. For example, Ehret (1973:1) suggests that the names
applied to cattle and sheep by many Bantu speakers were probably derived from the
non-Bantu languages known collectively as Central Sudanic.

In outlining the history of the Venda language, the following aspects will be discussed
and an attempt will be made to identify the role which adoptives can play in the
reconstruction of these different aspects.

i) The original migration of the Venda; .

ii) Contact during migration;
iii) Postmigratory contacts.

1. The word 'Bantu', even though it has become stigmatised in South Africa, will be
used in this study on purely linguistic grounds.
 13

1.2.1 Possible original migrations of the Venda and their contact during migration.

The origin and political history of the Venda have intrigued generations of scholars
ever since missionaries like Beuster published works on the Venda (Loubser 1988:2).
Traditional views differ with regard to the origin of the Venda. According to some
traditions, the Venda are said to have originated from various areas, for example, from
Egypt, Lower Congo, the Great Lakes area, Vhukalanga, and finally the Northern
Transvaal (Ralushai 1977:22). There is no consensus amongst scholars on the place
which could be regarded as the original home· of the Venda. Some scholars feel very
strongly that the Congo or Central Africa is the place ·of origin of the Venda. However,
scholars like Loubser (1988; 1989) reject the Central Africa hypothesis in favour of the
local origin hypothesis of the Venda. In this section an attempt will be made to·
scrutinize the different views on the origin of the Venda so as to establish an area
which could serve as a starting point for our study of adoptives. These traditions will
also be viewed against the background of the origins and migrations of the whole
Bantu language family, of which Venda forms a part.

1.2.1.1 Migration from Egypt

One viewpoint is that the Venda originated in Egypt from whence they left because
of civil wars following succession . disputes. There are no written documents to
substantiate this view. This view, is according to Ralushai (1977:23), only maintained
by a few individuals such as chief Davhana and John Mulaudzi . The latter claims to
have seen old Venda work of art in a big house in Egypt during his stay in North
Africa during World War II..

This viewpoint of Egypt as the place of origin of the Venda seems to have emanated
from the Ancient Egyptian theory of the Bantu origins which is st,Jpported by some
students of Negro linguistics like Hamburger (Bryant 1963:123). This theory, like that
of the origin of the Venda people, seems to be decidedly weak and unconvincing.
Thus Bryant (1963:127) after his intensive comparative study of the Egyptian
languages and the Bantu languages, came to the following conclusion:
 14

"All in all, then granting Egyptian certain minor similarities to Bantu, one feels
that they do not suffice to convince one of even a prima facie case for Bantu
relationship let alone Bantu parenthood or that there are any stronger than
similar evidence producable from many other tongues, ancient or modern ...
Indeed it were difficult to believe that a language so infantile in its simplicity
without even regular tense, mood and voice forms as was the Egyptian, could
ever have borne as its offspring a system of speech so scientifically complete
as is the Bantu: one cannot produce figs from thistles".

Bryant's conclusions also throw some light on our problem of the origin of the Venda.
It is very clear from the above passage that there is little or no linguistic evidence
which can justify the Egyptian theory. Mulaudzi's evidence, (as recorded by Ralushai
1977), based on works of art needs to be further substantiated before it is accepted.
Since there is no linguistic evidence which supports this view it will be- fruitless for our
study of adoptives to pursue this theory further.

1.2.1.2 Migration from Central Africa

Another view on the origin of the Venda claims that the Venda originated from Central
Africa. Some scholars such as Wesseman (1908) specify this place as being that of
the Congo region. Thus Wesseman (1908:9-1 0) describes the place of origin of the
Venda as follows:

"The Bawenda [Vhavenda] are probably only a part of a greater tribe whose
dwellers are looked for on the Congo where according to the oldest Portuguese
maps, a people of the same name is still living ".

This view is also supported by people like Beuster (1879), Grundler (n.d), Thea!
(1916), Mathivha (1 966) and many others. These people base their argument on the
existence of a tribe bearing a Venda name, speaking Venda and practising customs
similar to those of the Venda. The fact that the Central Africa hypothesis of the origin
of the Venda seems to be based on the supposed existence of a tribe with the name
 15

Venda raises a number of questions. In the first place there is no proof of the
existence of such a tribe in Central Africa. Lestrade (1927:489) states this point as
follows:

"There does not in fact exist to the writer's knowledge any Congo language
which is sufficiently similar to thE~ irreducible third of Tshivenda as to place
Tshivenda in the Congo dialect".

Secondly, the origin of the name Venda may also throw some light on this problem.
According to Venda oral traditions the name Venda originated from the Northern
Transvaal. The Singo who immigrated to the area south of the Limpopo, adopted this
title from the people whom they found in the area. However, some scholars like
Mathivha (1966:9) state that the. name Venda . was derived from Vele-e-nnda, which
~

means that "one of their leaders Vele is outside. the original .home which is
Mashonaland" (lit. 'when Vele 'is outside'). I do not intend to delve deeply into these
arguments. What is important here is that the name Venda seems to have originated
at a much later stage in the Venda migration. This suggests that the Venda people
were probably known by another name before they settled in the land south of the
Limpopo. Another name by which the Venda were known was Senzi. The name Senzi
originally derives from · Zendj' (cf. Wentzel 1983: 148) which was used by Arabs to
refer to the Ethiopians or Abyssinians. The Arab geographers use the word 'Zinghi'
to describe the interior of Ethiopia (Von Sicard 1952: 112). The name Senzi seems to
have been first applied to the Venda by the Lemba who according to Wentzel
(1983: 148) had lived among the Venda for some time long before their migration to
their present country. The Lemba referred to the Venda as Senzi (i.e. the
uncircumcized people) because the Venda did not practice circumcision. This name
also seems to have been used for the Venda much later in their migration to the
south. Thus Wentzel (1983:149) says:

"Whatever the origin of the word Senzi may be, by the time it was used by the
Lemba with reference to the Venda, they obviously already had migrated south,
but not yet as far south as the present countrt of the Venda''.
 16

From these discussions on the name Venda it becomes very clear that there is no
justification to link the tribe with a Venda name which is said to exist in the Congo
with the present Venda people. Even if there were a Congolese tribe with such a
name, "it would not be sufficient reason to link up two tribes ethnically, certainly not as
intimately as implied here" (Lestrade 1927:49).

The Central Africa theory of the origin of the Venda may perhaps be more clearly
understood when viewed against the background of the Central Africa hypothesis of
the origin of the Bantu languages. Most scholars seem to agree on the origin of the
Bantu languages as being somewhere in ·central Africa, the precise location having
changed over time (cf.Greenberg 1963, 1972, Guthrie 1967, Ehret 1972 and Phillipson
1977). The original location was previously believed to be the savannah belt lying to
the south of the equatorial forest i.e. the Congo area, but later it was established that
the Bantu languages originated fro in an area on the north western limit of the present
Bantu present distribution i.e. Central Cameroon (see Map no.l, on page 273 taken
from Phillipson (1977: 109) to illustrate the origin and the dispersal of the Bantu
languages over some 2,000 years).

Against the background of the origin of Bantu languages, one may be tempted to
accept the Central Africa theory of the origin of the Venda, more so because most of
the Venda traditions seem to point to the origin of the Venda as being somewhere
around the Great Lakes. Although it is not clear which lakes the different traditions
point ta.,. there is a strong possibility that the Vendas might have stayed around Lake
Victoria or Lake Tanganyika before they migrated further south. Archaeological
evidence also supports the fact that some Bantu tribes stayed around Lake Victoria,
Lake Tanganyika and Lake Albert before migrating to the south (Phillipson 1977). The
Venda might have been part of those Bantu tribes, although they were not yet called
the Venda. It is to be regretted that no record is available on the migration of the
Venda from these areas. As such, no contact between the Venda and other tribes in
the area can be established except the contact which the Bantu tribes had with
Sudanic languages. The contact between Bantu tribes and Sudanic people led to the
adoption of animals such as cattle and sheep and also some pottery wares by the
-=--
 17

former from the latter. Phillipson (1977: 111) has this to say in this regard:

"Along the first part of this route we find an area east of Lake Chad where the
pottery that was made in the first millenium B. C strongly resembles Early !ron
Age wares. The area is adjacent to the territory where Central Sudanic
languages are spoken today. The linguistic evidence indicates that some early
Bantu speaking people acquired domestic cattle and sheep, together with the
words for both animals, from speakers of Central Sudanic. Did they acquire
them during the eastward stage of such a detour around the equatorial forest?
There is further evidence favoring this interpretation: around Lake Victoria,
where the southward turn must have been made, we find the sites that have
yielded Urewe ware".

It is interesting to look at the Venda names and their proto-forms for-these animals.
The word for cattle in Venda is nombe and it has been derived from the starred Bantu
word *gombe. The word for sheep is nngu from the starred Bantu word * gu. These
Venda forms appear to be very close to those of the Proto-language. This may
indicate that Venda got these forms directly from the proto-language.

1.2.1.3 Migration from Malawi

Some Venda traditions point to Malawi as the supposed place of origin of the Venda.
This theory is supported by scholars like Lestrade (1960:v) who describes the place
of origin of the Venda as follows:

"The original nucleus of tribes which forms the distinctive element in the present
Vendaspeaking group had its home probably somewhere in the region now
constituting the Nyasaland Protectorate."

Lestrade supports his view by citing one of the Venda scholars Mud au (1940: 10) who
describes the place of origin of the Venda as follows :
 18

"... a place of great rivers and lakes, a country overflowing with water and many
forests and fruit of bananas growing in many groves and tubes and peanuts in
great variety".

This has been regarded by Lestrade as suggesting an area to the west of Lake Malawi
or Lake Shirwa. However, Mudau never m·ade mention of Malawi as the place of origin
of the Venda. The view that the Venda originated in Malawi is also supported by
scholars like Gottschling (1 905), Stayt (1 931 ), Van Warmelo (1 960) and Wilson (1 969).
Some Venda chiefs such as Sinthumu!e claimed to speak the Malawian languages
well (Ralushai 1977:25). This has been regarded by some scholars as evidence that
the Venda originated in Malawi. According to Mullan ( 1969) the Venda or Senzi, as
they were then called, were displaced from this area by the Malawian invaders who
came into the area from the Congo region around 1600 and occupied the territory on
both sides of Lake Nyasa. From this area the Venda moved southward-into Sena (an
area which lies just across the Zambesi in northeastern Zimbabwe). At Sena the
Venda are said to have come into contact with the Lemba and the Sotho. With regard
to the contact with the Sotho, Mullan (1969:75) says:

"Speaking of Vendaland reminds us that the Bavenda appear to be related to

the Basotho or that they had at least close association with one another over
a lengthy period as would appear to be established by similarity of language.
It is very possible that this close association took place in the Sena area, where
we find not only the people of Mongaz, but also the Bakwena across the
Zambezi from Mongaz".

Ralushai ( 1 977) does not accept the view that the Venda originally migrated from
Malawi. He rejects this view on the grounds that there is no substantial linguistic or
historical evidence that shows that the Venda originated from Malawi. The fact that
some Venda people speak Malawian languages is, according to him, not convincing
enough since the Venda have contact with Malawian people in work places such as
the mines. This language argument is also weakened by the fact that no study has yet
been undertaken to compare Malawian languages or dialects with Venda. Perhaps
 19

such a study may throw some light on this problem.

Archaeological researches made in Malawi (Robins 1971, 1979) have not produced
any evidence to link the Malawian ceramics with those of the Venda. The only
ceramics which resemble those of the Venda were found by Inskeep (1965:34-35) and
they date back to A.D 1800. By that time the Venda were already in the Transvaal.
Loubser (1989) regards any claims about the similarities between Malawian and
Soutpansberg ceramics as superficial at best.

1.2.1.4 Migration from Vhukalanga (Zimbabwe)

Various Venda traditions indicate that the Venda reached their present home acros~

the Limpopo from a vaguelly designated area called Vhukalanga (Lestrade1927:XXI).

Vhukalanga has been taken by certain scholars as referring to the present Zimbabwe.
/ But in Vendi_Vhukalanga or Galang a may be used to refer to a. northerly direction,
which in this case can be taken to mean that the Venda came from north of the
Limpopo river. Recent scholars such as Loubser (1988; 1989; 1991) argue that the
Venda actually originated in Zimbabwe rather than central Africa. Loubser supports his
argument by refering to the names of the earliest Singo 1 chiefs such as Lozwi (Rozvi),
Mambo, Dyambeu and Thoho-ya-ndou which all are titles of Rozwi rulers. The Rozwi
people inhabited the western part of Zimbabwe. The following Venda names also
resemble those of the Rozwi, viz. Nyadenga, Dombo, Madanda, Bvumbi, Makoni,
Nyatsimba, Ngou (Zhou). This similarity of names may be taken as an indication that
the Venda had some connection with the Rozwi. The Rozwi dynasty has been well
recorded in various history and historical linguistics books (e.g.Van Warmelo 1940,
1960; Lestrade 1960; Mullan 1969; eaton-Thompson 1970; Beach 1984).

1
The origin of this name is not known. What is known is that it is a mutupo (totem)
name of the prominent Venda group.
 20

Lestrade collected informative data about the Rozwi and he established that although
the Venda are associated with other Shona groups like the Kalanga, they were
specially connected with the Rozwi and they probably belonged to the same stock. In
this regard he has the following to say:

"The accounts (i.e. reports of Transvaal Native Locations Commission of 1907-

8 & 1910 mainly by Mr.C.N.Manning) mention the names of tribes with whom
the original Vhavenda are said to have been in close contact, and from one of
which they are stated actually to have descended. Two such names are the
Banyai and the Balozwi(BaRozvi), the former allied to the Karanga group, the
latter according to Mr.Manning's informant, the tribe with whom the Vhavenda
. were intimately connected" (Lestrade1960:XXV).

The question that arises here is whether the Venda people are an -offshoot of the
Rozwi or whether they were an independent tribe which later became part of the
Rozwi Changamire dynasty. According to Manning's informant the Rozwi people found
the Venda already in the area and they presented them with a leopard skin saying that
the Vendas were of the same original stock as they. This event has been interpreted
by Lestrade (1960:XXV) as follows:

"The Leopard skin is a form of royal tribute, given by a subject to his chief, and
by an inferior tribe to a superior of the same stock. If, then, we are inclined to
believe that this act indicates similarity of origin between the original Vhavenda
and the Balozwi, we have further evidence for believing that the Vhavenda are
from a stock other than the· Karanga tribes since the BaLozwi are not a
Karanga people".

According to this passage it is clear that the Venda did not originate from the Rozwi.
Probably the Venda had a common origin with the Rozwi, who are said to have
originated from areas around Lake Tanganyika (Marodzi 1924:88). This is quite in line
with the Venda traditions which point to this area as being the origin of the Venda.
 21

This view counters the view that the Venda originated in Zimbabwe. Lestrade echoes
this claim as follows:

"... the ancestral home of the original Vhavenda was not in the Karanga country
but in the region considerably further north ... "( 1960:XXII)

While it is most probable that the Venda did not originate in Zimbabwe, their sojourn
in this area has been supported by historical and archaeological evidence. The Venda
stayed in Zimbabwe for a considerable length of time (Stayt 1931 :258; Wentzel
1983: 152). They are said to have stayed around Matongoni Mountain (Von Sicard
1952:1 0). In written sources this area is often referred to as Dhlodhlo (Wentzel
1983:1.52) and it was also the capital of the prominent Changamire Rozwi dynasty
. (loubser 1990:15). According to Wentzel, this city was built on a site which is now
about 67 kilometres to the north-east of Bulawayo and the remains of its ruins can
still be seen. (see Map II on page 27 4 of this dissertation for the location of
Bulawayo).

The sojourn of the Venda in Shona country has resulted in the Venda absorbing "a
certain proportion of Shona and certainly a considerable amount of Shona culture,
especially as regards language" (Lestrade 1960:1 ). As already mentioned, Venda
appears to have been influenced more by Western Shona than by other Shona groups
such as the Korekore, Manyika and others (Wentzel 1983:170-171 ). By Western
Shona we are referring particularly to the Kalanga and Rozwi dialects. The influence
of Western Shona on Venda can be clearly seen from the following words:
 22
r

Venda Western Central Proto-forms 3

Shona (Kal.) 1 Shona 2

vhomme vhomme madzamai -mayi 'mother'

-elekanya -ala kana -rangarira -gan- 'think'
-ambara -ambala -pfeka -yamb- ' wear clothes'
-vhamba -bambula -rovera --- 'stretch'
-edela -ezela -vata -d6tk6pe 'sleep'
""
-end a -end a -fa mba -yend- '
go'
-kuvhanganya -kubunganya -ungara -bung- '
gather'
-
vhutshilo buchilo upenyu -yumi 'life'
khwidzi mwizi hwai -gu 'sheep'
-kuku -kuku mbuya -kuuku '
grandparent'
malofha malopa ropa -dopa 'blood'
tshinoni nyunyi shiri -nuni/-yoni 'bird'
""
fola fola fodya -kaya ·'tobacco'
inwi imn'ilin'wi imi -nyu 'you'

Wentzel (1983) has written fairly extensively on the relationship between Ka!anga and
Venda. However, no detailed comparative studies have yet been undertaken between
Venda and the Rozwi dialect1 or the influence of the latter on Venda with the
exception of some similarities in the names of people and places pointed out in some
studies. Such a detailed comparison would perhaps provide some evidence of the

1 Kalanga examples were taken from Wentzel (1983) and Webb.er (1988).
2 Central Shona examples are based on the " Standard Shona Dictionary "

by Hannan (1959).
3 Proto-forms listed here, and in other sections of this study, are based on

"Comparative Bantu "by Guthrie (1971).

 23
the connection between the Venda and the Rozwi. However, such a comparison
cannot be easy since the Rozwi dialect has fallen into almost complete disuse. The
Rozwi dialect has only been recorded in an article by Fortune (1956). He also provided
an English translation of this text. A linguistic analysis of this text shows some strong
similarities between Venda and. Rozwi. Our focus here will only be on the features
which occur between Rozwi and Venda and not with other dialects of Western Shona.

Lembeu: Asi Lembeu wakati: "Bge halimhanye asi imni, ana vangu, cimhanyani
zenyu; imi ndofira pano paManyanga. "Paiva pacia:1, paraware gwedzimbabge
paivunganyidziwa nyanga dzemhuka dzaivuya navanhu vaihoshuma
kunaMambo. Akati kuvana vane: "/we, Cikole, cinda neShamba, musatovere-
zira imne"; nokuti wakanga afunga kuti Ntevere wogara vushe, ndiye otevegwa
.nepfumo liciuronda uyu Cikole haacahorondewa ...

Kwakahoti gale lapera, Cikole akahoti: "Ndicatini? Ndinotini kwangu kuci-

kururarama here? Ko, vana vangu zo vopera nenyota nokushaya zoku/a,
ndoburuka zangu." Ndopaakaburuka mugumo akasika paushasha wamaDziti
akati: "Hayi takatumigwa kuyi: 'Mundovuya naye, musaubaya ndihovona
nyanga dzidze, idzodzo dzinoyidziri muulomo mume'." Ndo kunda naye kuna-
Muzilikazi ndo kundosika kukayi: "Mavuya naye here?" Vakati "Tavuya naye".
Kukayi: "Vuyani naye ndimbone." Zino ndo pavakasika wovuza kuti: Ko, za
waiyi unenyanga muulomo, dziriyi?" lye akati: "Maingorevegwa marutu
navanhu." Muzilikazi akati: "lndazo kuusha kuko, vanhu vakangokurevera
marutu; kakale handicakubaya nokuti ndakatorera Mambo mhuri inji, ndilo
pfumo landibata" (Fortune 1956:74).

Literary translation:
Lembeu (Seed-eater):but Lembeu said: "A stone does not· run but you, my
children, flee for your part, I die here at Manyanga". At the council place, at the
flat rock of the court was piled up with the horns of animals which had come
with people who paid tribute in meat to the Mambo. He said to his children:
You, Cikore, go now by Shamva(?), do not follow one path:" because he had
 24
thought that Ntevere inherits the kingship, it is he who is followed by an army
surrounding him, this Cikore will not be surrounded.

When the year was up Cikore said:"What shall I do? I say does my place still
prosper? Tell me, since my children are perishing of thirst and for lack of food,
I'm going down." Then it was he who went down from the hill and reached the
camp of the Nguni. The Nguni said: "No, we were sent with instructions: "You
come with him you must not kill him so that I may see his tusks, the ones
which are said to be in his mouth." And they took him to Mzilikazi and arrived
and it was said: "Have you come with him?." They said: "We have come with
him." It was said: "Come with him that I may see him."Then it was when they
. arrived he questioned, saying: "Tell me, since you were said to have tusks in
the mouth, where are they?" He said: "You were just told lies by people."
Mzilikazi said: "Go you to. your village, the people just told lies about you;
further I shall not attack you because I took many families off the Mambo, it is
the army which held me" (Fortune1956:80-1 ).

The first outstanding similarity is that of names. In this text there are Rozwi names
which are similar to Venda names, e.g. Lembeu, Mambo and Ntevhere (cf.
Ntevheleni). However, some of these names do also occur in other Shona dialects,
Mambo for example, is used to refer to a chief. The similarity between Rozwi and
Venda names has already been mentioned and will not be discussed furthe'A' ·

Another feature which distinguishes Venda and Rozwi from other Western Shona
dialects like Kalanga, Lilima, Nambya, etc., is their use of the r sound. From the text
given above it appears that this sound was used interchangeably with I, and this is
also a characteristic of Venda. Kalanga does not make use of the r sound.

Venda and Rozwi are also distinguished from other dialects of Western Shona by their
use of the formative -chi- in the participial present. This can be seen frorn words such
as liciuronda cf. Ve.li
.... tshi u Ianda 'when it followed you. 'Other dialects of Western
Shona do not make use of this formative.
 25

In this text there are many words which are similar to Venda eg. nyanga (cf. Ve.
nanga
,.. 'horns'). Further examples are listed below:

pfumo cf. Ve. pfumo 'assegai'

ronda cf. Ve. -londa 'follow'
gole cf. Ve. gole 'cloud'
-tumigwa cf. Ve. -rumiwa 'to be sent'
mhuri cf. Ve. phuli 'slaves'
-bay a cf. Ve.-baya 'injure'

These words, with the exception of -tumigwa and mhuri, are the same in other Shona
dialects. Another slight difference is the sound I in gole which in other dialects like
Karanga, Zezuru, Manyika, is written as an r. An attempt was not made here to
distinguish the Rozwi lexicon from that of Kalanga. Such an endeavour would not be
easy since there are not enough records of the Rozwi language. Although Venda
shares some linguistic features with Rozwi it differs from it in other ways. In the Rozwi
text given here, one cannot fail to note this difference. For example, the use of the
formative ci- in Rozwi words like cimhanyani, cinda, etc. is different from Venda. In
Venda imperatives are never formed in this way. Differences are also realised in the
lexicon, phonology and the syntax of these two languages. These differences show
that although Venda has been influenced by Rozwi,the influence is not so great that
Venda can justifiably be regarded as an offshoot of Rozwi.

Traditions indicate that the Singo separated from the Rozwi and went south as a result
of dissension (Motenda 1940:52). Alpers (1970:217) has the following to say with
regard to the departure of the Venda from Zimbabwe:

"The Venda crossed the Limpopo at the end of the seventeenth century, almost
in flight from the warring Changamire Dombo".

It seems that the Venda broke away from the Rozwi around the 1680s (loubser
1989:58). From Matongoni (Dhlodhl
 26

Matongoni after a disaster struck the city, presumably fire. They were led to the south
by Tshilume, son of the God King Mwali. On the way they are said to have been
protected by their sacred drum, Ngoma-Lungundu. Wentzel (1983: 154) describes their
movement to the south as follows:

"Tshilume did however, not guide his people on a straight route to the
Limpopo. They moved via Chibi (Tshivhi)'s country, where they stayed for
about three years. Here they were in close contact with the Karanga (whence
the many borrowings from Karanga). From Chibi's country the Venda moved
via Buxwa (Vhuxwa) in the present Belingwe (Mbelengwa) district".

At Vhuxwa (Mbelengwa) the Venda came into contact with the Lemba (Refer to Map
II on page 27 4 of this disertation for the different places given above). From Vhuxwa
the Venda together with the Lemba who joined them crossed the Limpopo into the
Northern TransvaaL It is interesting to note that the sojourn of the Venda in Southern
Zimbabwe is marked by 'a great many place-names and sib-names in Vendaland
today identical with place names and sib-names in Southern Rhodesia (Zimbabwe)'
(Caton-Thompson 1970:250). This includes place names such as Makonde, Thengwe,
,
Mutale (Mutare) Nwanedi
A
(Nuanetsi), etc.

Concerning the date of arrival of the Venda in this land south of the Limpopo, there
is considerable difference of opinions. Their date of their arrival seems to range from
the end of the 17th century to the beginning of the 18th century. (cf. Stayt 1931;
Wentzel 1983; Loubser 1988; 1989)

In the Transvaal the Venda came into contact with other groups such as the Sotho,
Shona groups such as the Lembethu and Twanamba, and at a much later stage they
also came into contact with languages like Tsonga, Zulu, English and ·Afrikaans. Much
about the Venda contact with these languages witl be discussed later in the section
on postmigratory contacts (cf. section 1.2.2).
 27
1.2.1.5 The Northern Transvaal as a place of origin

Recent research studies on the origin of the Venda have indicated that their origin
may be local rather than Central Africa or any other area north of the Limpopo
mentioned in the previous sections. This view is propounded mainly by Loubser (cf.
Loubser 1988; 989; 1990; 1991 ). Loubser has done an intensive study on the
ethnoarchaeology of the Venda people in South Africa. In his study of the origin of the
Venda he integrated oral traditions, archaeology and ethnography to evaluate the
competing hypotheses about Venda origins and political history. Tt1e need for this kind
of an approach to the study of the history of the Venda has been expressed· by
scholars like Ralushai (1977) and Wentzel (1983). For languages like Venda with no
good recorded history such archaeological findings can be of great assistance.

Perhaps before we discuss Loubser's view on the origin of the Venda it may be
necessary to look at the people he regards as the ·real' Venda. According to him the V
real Venda originated from the Northern Transvaal, while other groups like the Singo
migrated much later into the area. The Singo are the most documented of all Venda
groups. This is mainly because of the very significant role they played in the history
of the Venda. This group seems to be the one that broke away from the Changamire
Rozwi dynasty. The names of their leaders like Dimbanyika, J:hohoyandou, Vefe·la-
Mbeu suggest such connections. These connections were discussed in the previous
sections (cf.section 2.2.1.4).

Venda traditions unambiguously show that the Singo took the Venda title from the
people they found in the area despite the claim· by some Singo informants that this
name was brought into the Northern Transvaal by their ancestors. According to
Motenda (1940) the original inhabitants of this land are the Ngona such as the
Raphulu clan. He has the following to say in this regard:

"according to tradition the true or real Venda are those of Raphulu. They are
the original inhabitants of Venda" (Motenda 1940:154).
 28

Whether or not the Raphulu people are the first inhabitants is not the issue here, the
fact is that when the Singe arrived in this land they found some people already in the
area. Most traditions on Venda point to this fact.

1 Loubser excavated different areas in Venda which were

1 occupied by early Vendas.
These include areas like Mapungubwe Hill, Tshitaka-tsha
A
Makoleni, Vhunye!a, Princes
Hill, Mutamba, Verulam, Tavhatshena,
;..
Tshirululuni, Begwa, Manavhela and Nkgaru.
He compared the ceramic styles, settlement patterns and 'mitsheto' patterns found in
""
these areas. The Mapungubwe ceramic style marks the earlier Shena settlement south
of the Limpopo while the Khami ceramics show the arrival of new Shena dynasties
from Zimbabwe. In the south of the Soutpansberg the earliest ceramic style found was
the Eiland followed by Moloko. Both are Sotho styles and they show the presence of
the Sotho speaking people in the land south of the Limpopo long before the Venda.
Excavations in some areas has shown the overlap between Khariii and Moloko
ceramic styles. This overlapping "indicates the close interaction between Shena and
Sotho speakers, then the development of Tavhatshena and Letaba Venda language~~.

(Loubser 1989:58). This therefore suggests that the Venda language is a result of a
fusion between Sotho and Shena. Loubser (1 991 :396) gives the following schematic
representation of the ceramic sequence which was discussed above :
 29

1600

1500

1400

1300

1200

, 100

After comparing these ceramic styles, Loubser (1991 :339) came to the following
conclusion:

"The culture history of the Soutpansberg thus supports the view that the Venda
had a local origin involving more than one ethnic group 11 •

Although Loubser describes the fusion of the Shona and Sotho ceramic styles so well,
it is to be regreted that he cannot establish how other aspects like language came to
 30

be fused. Perhaps Loubser's view on the Venda language as an amalgamation of

Sotho and Shena can be attested by the affinities which Venda is said to share with
these languages. Lestrade (1932:21) has this to say in this regard:

" ... Venda grammar and phonology are similar to Shena, particularly Western
Shena, while Venda vocabulary has its greatest equivalent in Sotho".

These affinities may be regarded as paralleling those found by archeaologists between

Venda pottery and adjacent Sotho and Shena styles (Schofield 1937, Sinton-Schoetter
1971, Loubser 1988, 1989). The dual character of Venda has created problems to
many linguistic scholars. Scholars such as Janes-Phillipson (1972:201) tried to
account for this problem in the following manner:

"In general, the evidence in respect to terms of domesticity seems to suggest

that there may be some link between Venda and the Shena group on the
female side. The link with the Sotho groups on the other hand, appears to be
on the male since a number of the terms refer to hunting and cattle keeping.
One cannot interpret this kind of fact with certainity, but one way to account
for it would be to suppose that at some period or other, the Venda acquired
Shona-speaking wives and on the other hand, that some Venda-speaking
wives were acquired by the Sotho".

Janes-Phillipson also seems to suggest that the affinities which exist between Venda
and other languages like Sotho and Shena is a result of a fusion of some kind which
happened in the distant past. Her explanation seems to be confirmed by Loubser's
archaeological findings:

"Traditions hint at intermarriage between the Sotho under Raphulu and the
Shena speaking Tshivhula dynasty. Significantly, this interaction is indicated by
the occurrence of Moloko and Khami ceramics on the same
sites ... lntermarriages must have been intensive because Moloko and Khami
elements appear together on vessels" (Loubser 1991 :418).
 31
Much on the affinity between Venda and these languages (Sotho and Shena) will be
discussed later in this chapter.

1.2.1.6. Summary of the origins of Venda

In the aforegoing discussion an attempt was made to establish the place of origin of
the Venda. Different theories on the origin of the Venda were discussed. Some
views, such as the Egyptian hypothesis, were found to be unacceptable. A view which
seems to hold water is the one which states that the Venda originated in Central
Africa. However, the problem with this view is the lack of historical evidence to
support its claim. Even though oral traditions point to Central Africa as the place of
origin .of the Venda, these traditions seem to be very ambiguous. This view is only
acceptable when viewed against the original migration of the Bantu languages.
Archaeological research has also shown that the Bantu speaking people did indeed
stay in Central Africa, particularly around Lake Victoria and Lake Tanganyika.
Linguistic researches also point to the nucleus of all the Bantu languages as
somewhere in Central Africa. This therefore shows that the Venda may have
originated from this place too. Studies on the Venda language have also revealed that
the Venda language has unique linguistic features (i.e. words and sounds) which are
not found in any of the languages spoken in the region. These linguistic features are
regarded by some linguists as an indication that the Venda originated in Central Africa.

Other views such as those of the Venda migration from Malawi and Zimbabwe were
also discussed. If one takes Central Africa as the place of origin of the Venda one
would regard these palces, i.e. Malawi and Zimbabwe, as mere milestones in the
migration of the Venda to the south. If the Venda originated from these places it
would mean that they are a branch of one of the Bantu group which migrated from
Central Africa. So far there is not enough evidence of a language which can be
regarded as a parent of Venda. One last interesting view is the one which states that
the Venda originated from the Northern Transvaal. The proponents of this view accept
that some groups which today form part of the Venda tribe, the Singe for example,
immigrated to the Northern Transvaal from the north. This view is substantiated by
 32
the fact that the interaction of groups like the Sotho-Tswana and the Shona south of
the Limpopo somewhere in the distant past resulted in a new group which spoke a
mixed language of the two groups. This nev.(ly emerged language had a great impact
on the speech of groups such as the Singo who arrived much later in the Transvaal.
Therefore what the Venda language looks like today is not only due to its place of
origin, but also due to the interactions which took place before and after the arrival of
the Singo in the land south of the Limpopo.

1.2.2 Postmigratory contacts.

The settlement of the Venda on the land south of the Limpopo opened a new chapter
in the history of the Venda language. However, the earliest postmigratory contacts are
obscured by the fact that very little is known about the nature of the language which
was spoken when the Singo arrived in the area and also the language which was
spoken by the Singo themselves. According to Mudau (1940:73-7 4)

"The language spoken at this time(i.e. when the Singo arrived) was, they say,
the same as that spoken today viz. the Tshivenda which has absorbed all
various forms of speech of the Lembethu, Twanamba , Senzi and Lemba. It
was of course not quite identical with present day Tshivenda, here were certain
differences. Thus the Ngona used u dida for strike lusese for tribe, mukhaha
for milk where we use other words today".

With regard to the language spoken by the Singo when they migrated to Venda
Motenda ( 1940:53-5 4) has this to say:

"As for their language, when they left Rhodesia the invaders had spoken
Tshikaranga. However, amongst the captives only the men aFld aged persons
were killed, whilst the women and girls and small children were spared and
became slaves of the royal kraal, others being divided amongst the great ones
of the land. From this time onwards these people began to abandon their
 33

language because they had married so many women of the Venda. In course
of time, the Tshikaranga language disappeared and only Tshivenda survived".

However, Motenda 's interpretation of events is rejected by Ralushai (1977:44), who

says:

"As to the argument that the Ramabulanas used to speak Kalanga and not
Venda, I found no confirmation of this. For centuries I was told, the Venda
language has been in existence on both sides of the river Limpopo. The
presence of Kalanga words in Venda does not necessary mean that whoever
lived in Rhodesia must have been a Kalanga".

From the aforegoing discussion it would appear that more research is needed in order
to either confirm or reject Mud au and Motenda's views. However, the- only certainty
that exists is the contact which Venda currently has with the languages spoken in the
area. At present, beside Shona, Venda can be said to be in contact with Northern
Sotho, Tsonga, Zulu, English and Afrikaans.

Venda contact with these languages occurred or is still occurring in different ways. In
the first place the contact can be said to occur through geographical proximity.
Tsonga, Shona and Northern Sotho are good examples of languages which are in
contact with Venda through geographical proximity. Venda contact with English and~
..........
Afrikaans present a very interesting case of contact. This is because of the fact that
although some of the Venda people have direct contact with the speakers of these
languages, other people who have never had any contact with speakers of these
languages seem to have been influenced mainly through the diffusion of cultural
material, concepts or ideas from these languages.

The intensity of contact and the degree of bilingualism play an important role in the
adoption of linguistic forms in a contact situation. The most intense kind of contact
may be said to exist in fully bilingual and/or multilingual communities. In such
1 situations lexical items but also phonological and grammatical rules may come to be
 34

shared by the languages in question. In the following sections the different contacts
which Venda has with other languages during the postmigratory period in this region
will be discussed.

1.2.2.1 Shona

Venda has been considerably influenced by Shona, Kalanga in particular. This is

indicated by Shona words and grammatical forms in Venda (Wentzel 1983, Loubser
1988). The influence of Shona on Venda seems to have occurred in different ways.
Some linguists regard the presence of the Shona component in Venda as a result of
the fact that Venda originated from the fusion of Shona and Sotho. Aspects of this
issue were discussed in the previous section. Other linguists regard the influence of
Shona as the result of language contact. The contact between Shona and Venda can
be seen to have occurred in different ways. Wentzel (1983:169) has th1s to say in this
regard:

"Venda did not only have its major contact with dialects of this cluster of Shona
group in the distant past in. what is now known as Zimbabwe, but this contact
was continued through the years".

In other words, Venda had contact with Shona during migration and even after the
Venda· s settlement in the land south of the Limpopo. The contact between Venda
and Shona during the migration period was discussed at some length in the previous
sections. It was during their stay in Zimbabwe that the Venda accumulated many
Shona forms into their language. In this section, however, the focus will be on the
postmigratory contacts. The contact between Venda and Shona continued in many
different ways after the Venda migration. In the first place it should be borne in mind
that when the Venda (Singo) arrived on the land south of the Limpopo they found
some of the dialects of the Western Shona cluster already in the area. These were the
Lembethu and Twanamba dialects. Although these Shona groups were defeated and
subjected to the rule of their conquerors, the influence of their language on Venda
remained a factor ( Wentzel 1983: 126).
 35
Venda contact with Shona also occurs as a result of geographical proximity. Wentzel
{1983:126) describes this contact as follows:

"The geographical closeness is so much more obvious if we take into

consideration that between 10% and 20% of all Venda speakers still live to the
North of the Limpopo river in Zimbabwe ... Here they live in direct contact with
the Karanga speaking Shona peoples".

· The contact between the Venda north of Limpopo and those to the south did not stop
despite the establishment of the Limpopo river as the colonial border. Thus Mathivha
{1972:19) says that "the Venda always ran to Mashonaland whenever there was
trouble from outside in Vendaland."

It is this interaction between the Vendas on both sides of the Limpopo which keeps
on bringing fresh Shona words into the Venda spoken to the south of the Limpopo.

Some of the influence is due to religious contact. The Mbedzi of Malungudzi

(Zimbabwe) and Vhumbedzi {Venda) attend their religious thanksgiving ceremonies
either at Mianzwi (Venda) or Malungudzi together (Ralushai 1977:17).

The contact between Shona and Venda also occurs as a result of urbanisation. In
areas like Messina one finds a conglomeration of many population groups including
the Shona and the Venda. Venda spoken in this area, which is regarded by some
scholars as the melting pot of different cultures, is also influenced by Shona.

The influence of Shona on Venda, as already mentioned, can be realised at the lexical
and grammatical levels. With regard to lexical items, it is interesting to note that most
of the words which are a clear case of adoption are very often not found in everyday
discourse, instead they are found in proverbs, riddles, praises, folktales, traditional
songs and such like (cf. sections (a) - (f) below). Most of these Shona adoptives are
regarded by the Venda not as foreign forms, as is the case with words from other
languages, but as "the original" Venda. The fact that most of the Shona words are no
 36

longer found in everyday Venda discourse may be regarded as an indication of the

fact that although Venda was at one stage heavily influenced by Shona, this influence
has slowly diminished. This influence can be traced back to the contact of the Venda
with the Shona groups during their migration to the south. The fact that many of these
Shona forms are no longer used in Venda may show a shift in contact to the
languages of the South-Eastern Zone, Sotho in particular.

(a) Proverbs with words of Shona origin

The following proverbs have words of Shona origin. These words have been written
in bold.

i} U tshi dza muro wa muvhuda

A
ranga u ya tshilaloni.
"When you think little of the gravy made from a hare, go and look at the size
of its lair."

ii} Mutshimbili u wana riombe, mavhele a wanwa nga nzula.

"A wanderer will always get cattle and maize from a settler."

iii) Tshe wa da iwe tshilavhi na murundo wa phofu u £o vhuya wa nwa.

"What you set your mind on, you chip, you even drink an eland urine."

iv} Mungona mubikwa na ive, ive Ia vhibva mungona a sala.

"A Mungona boiled with a rock, when the rock is soft he is still as before."

v) Nyavhumbwa wa dagaila wa kanda vho u vhumbaho.

"Creature of another [placed in power by another] when you start getting full of
yourself [ too prosperous ] you trample on those that raised. you."

vi) Muima woga shaka ndi nnyi?

"You who stand alone, who is your relative?"
 37
From the above proverbs, the following Western Shona adoptives are illustrated:

• ·cattle'
nom be cf.-kholomo
-da cf.-funa 'love/like'
ive cf.-tombo 'stone'
-dagaila cf.-tonga 'boast'
/'

-woga cf.-u wothe 'alone'

/'

(b) Shona words found in Venda praises

The following praises are for Mwali or Raluvhimba, the God-King of the Venda who
according to Stayt ( 1931 :230), "is supposed to live somewhere in the heavens and is
to be connected with all astronominal and physical phenomena". The-praises are as
follows:

i) Muvhusa-denga. 'The Ruler of Heaven'

ii) Ngwaniwapo. 'The One who was found there'
iii) Ntangiwakugala. 'The First to settle'
iv) Tshidzatshapo. 'He who was there'
(cf. Sho. Chidzachapo)

The parts of the praises written in bold are of Shona (Western Shona) origin.

(c) Shona fonns found in Venda riddles

Shona forms are also found in Venda riddles, even though there are a few of them,
e.g.:

Tsha nk.unda ndilipo. "Something which I couldn't get even though I am there."
 38

Ndilipo is a Shona construction with typical Western Shona forms, the use of I for
example.

(d) Shona forms in traditional Venda songs

Some Venda traditional songs have some Shona forms. Consider the following
examples:

i) Potilo
"
Hang ala
Hangala
Nda tema
Temiso
Tshinoni
"
Tshagala
Mutanda
Mandule
Gunu-wee

From this song the following Shona forms are illustrated:

-tema (cut), -temiso (cause to cut), -gala (sit)

ii) Funguvhu.
Funguvhu tanzwa mulomo
"
Tanzwa mulomo
A

Ri kane ri tshi Ia rothe

.... "
Ri tshi Ia rothe
"" "
Vhomme vha ka enda pi
Vha ka enda pi
Vho lima davha .,..Ia khombe
Vho lima davha .,..Ia khombe
 39
Vho lima ndima nngana
Ndima nngana
Vho lima ndima ntharu
Ndima ntharu
Ya vhuna
... ndi ya u fhedza
Ndi ya u fhedza

"Crow, wash your mouth

So that we may eat together.
Where have the mothers gone to?
They have gone for compulsory davha
How many acres have they tilled?
They have tilled three acres.
The fourth is the last."

From this song the following forms are illustrated:

-enda, ka, pi (go, past tense marker, where ) and the construction Vha ka enda pi
(Where have they gone to) which is typically Shena.

(e) Shona words in Venda folktales

Shena words are also found in Venda folktales. Here they are only found in the song
which forms the core of the folktale. Consider the following two songs from two
folktales in Ngano dza vha kale.

i) Mukegulu we a vha a sa funi nwana wawe a tshi mala

{The old lady who did not want her son to marry)

In this folktale we find the following song which is a dialogue between the son and his
old mother.

" Wa vhuya-vhuya Lutsindi,

 40
Wa vhuya-vhuya Lutsindi,
Vhaeni vha nga za pano;
Vha na nweli, vha na mapa;
Tshawe ndi tshilili tsho ga

(Come back Lutsindi

Come back Lutsindi
Visitors may come
They have nweli, they have mapa.
Yours is just a lonely cry)

The son then replies by singing as follows:

"Nda te ndi nga ngeno,

Nda farwa nga tshithavhi,
. "
Nda te ndi nga ngeno
Nda wana ku mitundo ... "

(I tried one side, and I felt pain on my side l tried on the other side, it was
urine)

The mother concludes by singing and she says

11
Ewe,ewe, Lutsindij
Zwithu zwangu ndi a tsenga ... "

(Yes, yes Lutsindi

My things I leave)

And then the child says

 41

"Tsengani zwanu
.... maye
Vha do sala vha no thenga zwavho
/'o

(Fine, you may leave mom. They will remain with me)

ii) Mukukulume na cj.ongololo

(The cock and the millipede.)

The song found in this folktale is a dialogue between a cock, a millipede and a
hen.The hen starts the song as follows:

"Nandi vhomukukulumel
Nandi vhodongololo
.A

Nalwo muthu nalwo!"

(Hey Mr Cock, hey Mr Millepede/ Go with it, go with it)

The cock and the millipede reply by singing as follows

"Pembela ndi ku pe riombe ahaa!

Pembela ndi ku pe riombe ahaaj"

(Dance excitedly so that I should give you cattle

Dance excitedly so that I should give you cattle)

The parts written in bold in these songs are Shena forms.

(f) Shona fonns in some Venda speech communities

Some Shena forms are only found in certain sections of the Venda speech community.
These forms are found in areas previously occupied by speakers of Shena dialects
such as Lembethu and Twanamba or areas which are geographically still in contact
 42

with Shona. In areas like Mutele and Tshikundamalema one still finds some remnants
of Shona forms. One can still find words such as the following:

-pip a ct. Ve.vhifha '

ugly'
xwilo ct. Ve.tshikhau 'short-cut'
son a ct. Ve.runga 'neat'
ninzwi ct. Ve.murudi 'kind of vegetable'
-pe ct. Ve.-fha ' give'

-ti ct. Ve.ri 'say/do this'

Sentence constructions which are typically Shona are also found in these areas, e.g.

Nwana a no gala pano. 'A child stays here'

Ndi no do Ia vhuswa. 'I will eat porridge'
.... "'
U no gala zwau henepa 'Stay here'

It is interesting to note the use of the formative no in these constructions. The

formative no in Shona is basically a present marker in the conjugation of the verb. In
Shona, except in Kalanga, it occurs in the present tense only, but in Western Shona/
Kalanga, it occurs in the future form as well (Wentzel 1983:165-166 ). Venda also
makes use of this formative. It appears that of all the languages of the South-Eastern
Zone Venda is the only one which makes use of this formative, and this may show its
connection with Shan a. In Venda the formative no is used in relative constructions and
verbal forms. With regard to the latter it is used in the present tense as well as the
perfect tense but with an aspect meaning.

There are also words in Venda which came from languages like Portuguese, Arabic
and Swahili. It is interesting to note that these words are also .found in Shona.
Consider the following examples:

fola : Sh.fodya < Port. fohla 'tobacco'

pes a : Sh.pesa < Port. peca 'piece of cloth'
 43

mali : Sh.mari < Arab. mal ·money'

mbanzhe : Sh.mbanje <Arab. bhan 'dagga'
khanzu : Sh.hanzu < Arab. hanzu : Swah. kanzu."clothes"
ndalama : Sh.ndarama <Greek.drachma 'money'
ngalavha : Sh.ngarava < Greek.karabos 'ship'
(Knappert 1970:88; Jone-Phi!ipson 1972:167 -8)

These words are undoubtedly adoptives in Venda. If we take words like like ndalama,
mali and ngalavha, we note that these words denote things which were foreign to the
Venda. These specific words, as indicated above, came from Arabic and Greek
respectively. A comparison of these original forms and their adapted form in Venda
suggests that Venda did not adopt them directly from their source language.
Phonetically there is no justification for words like Greek drachma and karabos to be
adopted as ndalama and ngalavha. This suggests that some Intermedi-ate languages
were involved before these words were taken into Venda. Shona seems to be the
language from which Venda adopted these words. There seems to be very close
similarities between these words and their counterparts in Shona. If we compare the
Potuguese and Western Shona words for 'tobacco', namely, fohla and fola
respectively, it is very unlikely that Venda adopted this word directly from Portuguese.
Instead, the Venda seemed to have adopted the word fola from the Rozwi who uses
the same form. We do not intend here to establish the route of each word given
above. We hope the examples we have given here are enough to show how these
words were adopted into Venda via Shona.

A comparison of Shona and Venda vocabulary reveals a number of words which are
common in both languages. Some of these words are not found in languages which
belong to the same zone as Venda. The following words are found in both Shona and
Venda and they are exactly the same in shape and meaning.

Venda Shona Sotho

mutupo mutupo seano 'totem '

madekwana madekwana mant~iboa 'evening '
 44
ngano ngano nonwane 'folktales '
v
ndalama ndarama tshelete 'money'
mangolomela mangoromera ---------'supernatural power'
mvumi mvumi maru ·dark clouds '
gale .gore leru 'cloud'
-gum a -gum a -fela ' .end'

-bika -bika -ape a 'cook'

-pin a -pina ----- 'growl at menacingly
-pond a -panda -bolaya 'kill I

-sum a -sum a -beg a 'report '

-tumbuka -tumbuka -thoma 'originate '
-peta. -pet a -koba 'bend '
-I ala -rara -robala 'sleep'
-gala -gara -dula ·sit'
-bay a -bay a -gobatsa 'injure'
-dza dza -belaela 'doubt'
-tema -tema -sega 'cut'

Words such as these create a problem in distinguishing adoptives from cognates.

According to Knappert (1970:81} "similarity, close resemblance between two words in
two otherwise unrelated languages is not in itself a proof of borrowing."

While this does not dismiss the fact that affinity may be brought by adoption, it helps
to show that there are other factors that can bring about similarities in the languages
in question. In his 90-word core vocabulary list, Ehret (1972:12) has established that
besides adoption, Venda shares several words with Shona which are common
retentions of earlier Bantu words. Therefore it is important to check whether a word
is not a retention of an earlier Bantu form before declaring it an adoptive form.
Consider the following examples:
 45

CB Venda Shena
-nuni/-nyon! ,.
tshinoni nyunyi (W.Sh. 'bird'
-ny6ta dora
A
nyota 'thirst'
-buta mvula mvula (W.Sh.)'rain'
-bua mbwa mbwa 'dog'
-kadi -kadzi -kadzi 'female'
'
-dume -lume -rume male'
-nama nama nyama ' meat'
A

-yati nari nyati 'buffalo'

....
-yoka now a nyoka 'snake'
I'

tang a dang a dang a 'kraal'

-tope . dope dope 'mud'
-kanya fola fodya · tobacco'
'
-budi mbudzi mbudzi goat'
-yaka riwaha mwaka
.year'
-gu nngu gwai 'sheep'
-jada ndala zhara 'hunger'
"'
-jogu ndou zhou 'elephant'
....
-juba duvha zuva ' sun'
"'
-gubu mvuvhu vubu 'hippopotamus'
-juki notshi nyuchi 'bees'
A

-gombe riombe n'ombe 'cattle'

-gudube nguluvhe nguruve 'pig'
-guena ngwena ngwena 'crockodile'
-gue nngwe nngwe ·tiger/leopard'
-p6ng6 thongo nhongo 'he-goat'
-gupa gufha gupa 'tick'

This similarity may be brought about by the fact that the languages in question may
adopt a word from the same source. Names for cattle and sheep are similar in most
of the African languages, as the examples below attest to the fact that the Africans got
cattle from the Sudanic groups ..
 46

cattle from the Sudanic groups.

nngu (Venda) 'sheep'

nku (Sotho-Tswana)
imvu (Zulu)
hwai (Shona)

In trying to resolve the problem of distinguishing adoptives from cognates Knappert

(1970:83) provides three criteria that can be used to determine words which are
adoptives in a language:

" .. the shape of a word (its morphophonemic form), its meaning and the extent
of its distribution i.e. the extent to which simifar forms are found in neighbouring
languages."

First, we shall look at the shape of the words or morphophonemic form criterion.
Words which show unusual morphorphonemic form are very often of foreign origin. In
Venda there are words whose forms are similar to Shona. If one takes words like
ntangiwakugala, ndilipo and muhwadzamai one cannot fail to notice that the shape of
these words is typically Shona rather than Venda. The following are further examples
of words or parts of words with a typically Shona form:

pano cf. fhano 'here'

-pi cf. -fhi where'
-pip a cf. -vhifha 'be ugle'
henepa cf henefha 'here'
-za cf. -da
/'
'come'
maye cf. mme 'mother'
wangu cf. zwanga 'mine'
nyoka cf. now a
.snake'
A
 47

proto-forms instead of adoptives. The reason why these forms may be regarded as
Shona adoptives is due not only to their shape but also to the context or constructions
in which they have been used. Even though these words are no longer found in day
to day speech, they are still found in Venda proverbs, riddles, traditional songs and
folktales, as was shown at the beginning of this section dealing with the influence of
Shona on Venda. Thus although the morphophonemic criterion may help in the
identification of adoptives in a language, it alone is too narrow because the oldest
adoptives which have been completely assimilated into the language may not be
recognizable from form alone. Therefore there could be many Shona adoptives in
Venda which cannot be identified by using only this criterion.

The second method of identifying adoptives is based on semantic criteria. In this

method one should "make classified lists of words within certain categories of meaning
- such as tools, clothes, fruits, spices, weapons, furniture, parts of the -house, metals,
cereals - and to study these words in all the languages of a given area.n (Knappert
1970:80). According to him adoptives will be found in these categories in almost all
the languages. If one takes a word like pfumo (assegai), which is an example of a
traditional weapon, one finds that it only occurs in Venda 's neighbouring Sotho
languages (i.e. Northern Sotho, Tswana and Southern Sotho) and not in Nguni and
Tsonga languages. e.g. :

Venda Shona Sotho Tsonga Zulu Xhosa

pfumo pfumo lerumo tlhari umkhonto umukhonto

Although we have given one example here, it illustrates a very interesting point. This
word is the same in all of Venda's neighbouring languages except Tsonga. Its non-
occurrence in languages like Tsonga could be explained by the fact that Tsonga came
to the area at a much later stage. Nguni languages are geographically far removed
from Venda. This may suggest that the Venda adopted this word from Shona and then
spread it to their immediate neighbours, the Northern Sothos and Tswanas. The form
of this word in the Sotho and Venda languages is more or less the same.
 48

The third criterion which Knappert suggests is the distribution of the suspected words
in the neighbouring languages. In adopting this criterion one should identify the
suspected (on the basis of morphophonemic and semantic criteria) and compare them
with similar words with similar meaning in other neighbouring languages. If these
words are found in a whole chain of neighbouring languages then there is a good
chance that such words are adoptives. This criterion has a number of problems that
renders it inefficient for the identification of adoptives. First, words change their shape
and meaning when they are taken over by other languages. Therefore the adopted
words which have completely assimilated to the adopting language may not
necessarily be identified by means of this criterion. The second problem with this
criterion is that words can be the same in neighbouring languages due to their
common origin. If one takes African languages like Venda, Shona, Sotho, Zulu, Xhosa,
and many others it is evident that these languages have many words in common due
to their common origin. This can be clearly seen from the following examples:

Venda Shona. Sotho Tsonga Zulu CB

v
dang a dang a lesaka shanga isibaya -tanga 'kraal'

mvula mvura pula mpfula imvula -hula 'rain'

mbwa mbwa
v
mpsa mbyana inja -bua 'dog'

,..nari nyati nare nyarhi inyathi -yati 'buffallo'

The words given above are similar due to common origin and should not be confused
with adoptives. Nevertheless, there are certain adoptives that could be identified by
means of this third criterion. In Venda there are many words adopted from English and
Afrikaans which are also found in other African languages spoken in the region:

Afr.winkel cf. NS.Iebenkele : XH.ivenkile : Ve. vhengele 'shop'

Eng.steamer ct. NS.setimela : XH.isitimela : Ve.tshidimela. ' train'
Afr.plaas ct. NS.polase : XH.ipulazi : Ve.bulasi 'farm'

The words given above are similar not because their source languages have
originated from one common language, but simply because of adoption.
 49
The criteria discussed above, however, can obviously not lead to the identification of
all adoptives in a language. Knappert (1970:80) himself concludes by saying

"In spite of all our efforts to trace probable loanwords to their potential originals
in the giver-languages, there is a large number of words the origin of which will
remain undecided or at least unproven .. ".

1.2.2.2 Northern Sotho

Venda shows some significant influence from Sotho. The influence of Sotho on Venda
occurred in different ways as already stated in (cf.1.2.1.5). According to Loubser
(1988,1989, 1991) the Sotho component in Venda may be explained by the fact that
Venda originated from an amalgamation of Sotho and Shona. In this section we shall
be looking at the influence of Sotho on Venda during the postmigratory period. This
covers the time from when the Venda settled south of the Limpopo up to the present
day. Of the Sotho language group Northern Sotho is the one that seems to have had
a transparent influence on Venda. In referring to Northern Sotho one should take
cognizance of the different dialects of this language. These dialects include, amongst
others, Tlokwa, Lobedu, Hananwa and Pedi. Venda has been influenced by some of
[r: these dialects through geographical contact (Van Warmelo 1935; Ralushai 1977).
Kruger (1965:1 0) describes the contact between Venda and Northern Sotho as follows:

"Hierdie (Venda) is 'n buurtaal van Sotho dialekte wat op die noordelike grens
van Sotho-gebied voorkom. Moontlike beinvloeding kon op die Lobedu, Tlokwa,
Hananwa en Birwa dialekte van Noord Sotho plaasgevind het".

In this passage, Kruger brings two important points to light. First, the influence of
Venda on Northern Sotho or vice versa should be seen as a result .of geographical
proximity. (The proximity between Venda and Northern Sotho dialects such as Tlokwa
and Lobedu can be seen from Map Ill on page 275 of this dissertation). Secondly,
different Northern Sotho dialects should be taken into consideration when the influence
of Northern Sotho on Venda is discussed. An attempt will be made here to show how
 50

the Northern Sotho dialects impacted the different sections of the Venda speech
community.

1.2.2.2.1 Tlokwa

The Tlokwa are found to the West and South-West of Venda. (Refer to Map Ill for
Tlokwa). They are or have been in contact with the Western Venda. Van Warmelo
(1935:117) describes their contact with the Venda as follows: " ... the Western Venda
or most of them have not escaped Sotho influence ... " With regard to the Venda to the
South-West of Venda he says:

"In the South-West, too, the Venda have been affected in both language and
culture by prolonged and fairly close contact with northernmost- Sotho tribes,
mostly Tlokwa".

It is an established fact that the Venda have been in contact with the Tlokwa for a
considerable length of time. The Venda contact with the Tlokwa occurred in different
ways. Mokgokong (1966:19) describes the Venda- Tlokwa contact as follows:

"At first the Tlokwa lived peacefully with the Venda who then occupied the
country which now belongs to the Tlokwa. They accepted Venda customs and
traditions, a point which is borne out by the fact that to this day the Tlokwa
have retained Venda names of places, rivers and mountains. In fact Tlokwa
vocabulary abounds in 22 words of Venda extraction which are not found in
other dialects such as Pedi."
 51

The following examples illustrate the impact of Venda on Tlokwa

N.Sotho Tlokwa Venda

'
nag a lesako shango country'
mo~ate mosata musanda 'chiefs place'
belega pep a beba 'give birth' .
thaba thakgala taka Ia 'be happy'
bole Ia apa a mba 'speak'
nnyane thoko thukhu 'small'
A

On the other hand Tlokwa has also influenced Venda, particularly the Venda spoken
to the West or South-West of Venda. The former includes areas like Sinthumule and
Kutame and the latter areas like Mulima. However, it is difficult to locate all the points
of contact between these languages since the boundaries have shifted greatly
following political upheavals. Thus Ralushai (1977: 198} says:

"Before the 1969 mass removals and resettlement of people on ethnic lines,
Ha-Mashamba and Ha-Mulima were highly mixed areas consisting of the
Maguvhu-tswera (Northern-Sotho speaking people who are not regarded as
true Northern Sotho speaking but as a hybrid of Venda and Northern Sotho},
Shangana-Tsonga and the Venda. The contact between the Tlokwa and Venda
i.n areas like Mulima, Kutame and Sinthumule is presently no longer intensive
as before these resettlements. Therefore the use of the modern maps may be
misleading since they do not reflect on the previous contacts".

The influence of Tlokwa on Venda, particularly in those areas which are or have been
in contact with Tlokwa, vary from the lexical to tonal level. In these _!3reas one finds
a lot of words from Tlokwa which are not found in other areas which do not adjoin
Tlokwa. Consider the following examples:
 52
Tlokwa W.Venda Standard Ve.
ipshina phina taka Ia '
enjoy oneself
swab a swavha shona 'be winkled'
tokwane -tukwana -tuku 'small one'
""
basi mana vhasimana vhatukana '
young boys'
nyaka nyaga toda
r~"
'look for'
bomma vhomma vhomme 'mothers'
tia dia rwa 'beat'
tswenya tswenya dina· 'disturb'
ntsha ntsha bvisa/tusa
/'
'take out'
fokola fogo Ia lwala 'be sick'
leba Ia livhala hangwa 'forget'

The influence of Tlokwa on Venda is also realised in cultural aspects such as initiation.
Initiation schools like musevhetho and murundu were introduced to the Venda by the
Sotho, Tlokwa and Lobedu in particular. Thus the songs in these initiation schools are
predominantly Sotho. The Tlokwa forms can be clearly seen from the following Venda
musevhetho song:

Vhasimana vha Vhudogwa

vha fhorwa vho lingenea

Ri a reda, ki redela li nnani tshakha-khale

Ri apara kuvho ka pedi
nnwe mathatha wa mathatha
Ki apere nguvho ka pedi.
Vhana a vho ngo mu laya
Vho mu lukutshedza.
(Van Warmelo 1960:93-94)

(The lads of Vhodogwa

Have been snipped off of length
 53

I put on two dresses

The children did not teach her correctly,
they have intentionally led her astray. )

From this song we find the following Tlokwa words:

vhasimana (NS.basimane) ct. Ve. vhatukana 'boys'

ki a reda (NS.ke a reta ) cf. Ve. ri a renda 'I praise'
apara (NS.apara) cf. Ve. ambara 'clothes'
ka pedi (NS.tSe pedi) cf. Ve.nga mbili 'two'

1.2.2.2.2 Lobedu.

Lobedu is one of the Northern Sotho dialects which is spoken in areas which border
Venda in the South (see Map no Ill on page 275 of this disertation). The Lobedu
migrated from Zimbabwe and it appears they arrived in the land south of the Limpopo
before the Venda (Singo). According to Krige (1937:329) the Lobedu people migrated
from the Kalanga area about the middle of the seventeenth century.

Krige (1937) identifies different groups of the Lobedu viz. the 'kolobe of Mohale' (the
Pigs of Mchale), the 'kolobe of Sephoko' and 'the kolobe Mmamaila'. The identification
of these groups is essential in as far as contact with Venda is concerned. In this
study we shall be looking at the last group, i.e. the 'kolobe of Mmamaila' since this is
the one which had intensive contact with Venda. Krige (1937:330-1) describes the
migration and the contact of this group with the Venda as follows: .

"The 'kolobe' of Mmamaila, who have retained more Lobedu elements of

culture than those of Sekhopo, left Bolobedu, according to reliable traditions,
at an earlier date (c.1750) than Sekhopo and Rakwadu. They moved
northwards into a country then called Thobe Ia, which they say was uninhabited.
 54

They were offered hospitality by the Venda in the Ndze!ele [Nzhelele]

valley ... They appear to have completely dominated the Venda of Moila who still
go to them and not to Modjadji (the Queen of Bolobedu) for rain and with whom
close ties of marriage as well as of tribulation during Zwazi raids(1848-54)
exist".

It is clear from this passage that the Venda had intensive contact with Lobedu after
the middle of the eighteenth century. To date the areas which are close to Lobedu are
those of Magoro and Nngwekhulu.

Contact between Venda and Lobedu also came about through marriages, musical
expeditons, trade, initiation schools, political upheavals and the Venda practice· of
coming to Modjadji to supplicate for rain. This past association is attested by the
notable similarities in their languages, customs and material culture (Ralushai
(1977:155; Davison 1984:45 ).

(a) Trade.

Trade between the Venda and the Lobedu seems to have been mainly in pottery. This
is shown by the fact that a number of Venda pottery terms have equivalents in Lobedu
e.g.

thidelo ".large smooth pebble from river bed used as polishing tool ."
phomo "black stone like graphite which is mixed with water and smeared on spots"
nkho "large beer pot"

These names are exactly the same in, Venda and Lobedu. It is not e.asy to establish
the direction of adoption for these words, but it seems that Venda might have adopted
these names from Lobedu. This is because of the fact that pottery in Lobedu seems
to be more advanced than in Venda. There are more types of pottery in Lobedu than
 55
in Venda. For example, types of pots such as modzheha, phafa, kheritswana, Iebeda,
Iesabelo, khetsikhayo etc. are not found in Venda.

(b) Marriage.

Intermarriages between the Venda and the Lobedu have probably contributed to the
adoption of Lobedu words in Venda. The following kinship words appear to have been
adopted from Lobedu:

Lo.molamo mulamu cf.malume 'brother-in-law'

Lo.ngwet~i nwedzhi cf.muselwa ·newly-wed-lady'
Lo.mohadibo muhalivho cf.muvhuye 'sister-in-law'
Lo.bohweni vhuhweni cf.ha vhomakhulu 'in-law's place'

Although the words given above are said to have been adopted from Lobedu, they are
also found in other Northern Sotho dialects. However, the pronunciation of these
words in Lobedu is very similar to that of Venda. Therefore, these words were
probably adopted from Lobedu.

(c) Miscellaneous examples

The following are further examples of words which are common between Lobedu and
Venda .

khelemo cf.Ve.tshi!imo 'summer'

lethabula cf.Ve.lutavula
A
·spring'
maria cf.Ve.vhuria 'winter'
mosada cf.Ve.musanda ·chiefs place'
ho seda cf.Ve.u sinda
,... 'to grind mElalies'
bomakgolokoko cf.Ve. vhomakhulukuku ·ancestors'
 56
It is interesting to note here that although some of these words are found in other
dialects of Northern Sotho, their sounds are closer to Venda than in those other
dialects. However, it is not easy to establish how these similarities came about,
whether it was through a common source in Zimbabwe or through adoption. The only
certainty that exists relates to the influence of the Venda sound system on the Lobedu
forms.

Although an attempt was made here to show how the different Northern Sotho dialects
have influenced Venda there are certain adoptives which are best assigne~d to
standard Northern Sotho as it is today rather than to a particular dialect:

VENDA N.SOTHO TLOKWA

gwineha < konega koneha 'backslide'
phuthego < phuthego phutheho ·congregation'
lovhedza < kolobetsa kolobetsa 'baptise'
ngosha < kosa kosa 'hymn or song'
..;
phuluso < pholoso pholoso 'salvation'

These words refer to religious activities. This may be taken as an indication that they
were introduced to Venda by the missionaries who are known to have played a very
important role in the development of the language. Missionaries like Beuster worked
with the Northern Sotho people before coming to Venda and they might therefore have
acquired. the above Northern Sotho words during that time. The influence of the
missionaries on the Venda language will be discussed later in this chapter (cf.1.2.3.5).

1.2.2.2.3 Other cognates between Venda and Sotho

The study of adoptives from Northern Sotho is complicated by the many cognates
which exist between the two languages. It therefore becomes very difficult to
distinguish some of these cognates from adoptives. In our discussion of the cognates
between Shona and Venda an attempt was made to establish some criteria which may
 57
help in distinguishing between adoptives and cognates. Some of these criteria will be
used in this section. let's begin by looking at the following examples:

CB Venda N.Sotho
-gombe kholomo NS.kgomo
.cattle'
phulu NS.pholo
.ox'
mboho NS.poo ·bull'
thole NS.sethole 'heifer'
namana NS.namane 'calf
-diba mafhi NS.maswi 'milk'
-kam- ham a NS.gama 'to milk'
-nama nama NS.nama
.meat'
"' .goat'
-budi mbudzi NS.pudi
-bua mmbwa NS.mpsa 'dog'
-kuku khuhu NS.kgogo 'chicken'
-gudube nguluvhe NS.kolobe 'pig'
-gu nngu NS.nku 'sheep'

The words given above are names of domestic animals. It is interesting to note that
most of these words have proto-forms. This suggests that the similarity of these words
is due to a common origin rather than to adoption. But the problem still remains with
those words which have no proto-forms. The question is whether or not to treat these
words as.adoptives and if so, to establish the direction of adoption. First we shall look
at the first question. There is no doubt that words like kholomo, mafhi, phulu, mboho,
etc. have been adopted from Sotho. This becomes clear if one looks at a word like
kholomo. It seems that cattle keeping was originally reserved for royal people among
the Vendas. Consequently, the commoners obtained cattle from the. Sothos, and as
such they call their cattle kholomo, after the source, instead of using the Venda name
for cattle which is nombe. But what is surprising is the dominance of these cattle
names over those that are of Venda origin. The Venda term nombe, for example,
seems to be losing frequent use. Other words which pertain to cattle, e.g mafhi, were
also adopted with the word kholomo and are now used alongside mukhaha, which is
 58

similar to Shona. This might also be the case with words like phulu and mboho.

Cognates are also found in words pertaining to wild animals (including birds and
insects). Consider the following examples:

CB Venda N.Sotho
-cfmba ndau NS.tau 'lion'
-jogu ndou
,... NS.tlou 'elephant'
-tuiga thud,..a NS.thutlwa 'giraffee'
I'

-budu phunguhwe NS.phukubje 'jackal'

-yati nari NS.nare 'buffalo'
I'

-y6ka now a NS.noga ' snake'

"
-guema ngwena NS.kwena 'crockodile'
-gue nngwe NS.nkwe 'leopard'
-yiba ....liivha NS.Ieeba 'dove'
-nungi nungu NS.noko 'porcupine'
-gige nzie NS.tsie 'locust'
-kamba khamba NS.khudu 'tortoise'
ntsa NS.phuthi 'duiker'
phithi NS.phuthi ·red duiker'
phala NS.phala 'impala'
muvhuda A
NS.mmutla 'hare'
mbidi I'
NS.pitsi 'zebra'
pfene
/'o
NS.tshwene 'baboon'

From the examples given above it is very clear that Venda shares a strong affinity with
Northern Sotho in so far as names of wild animals are corncerned. _This affinity can
be ascribed to a common origin since most of these words have proto-forms.
However, words like muvhuda,
,A
ndau, ntsa, khamba, mbidi and pfene present a very
A /"

interesting case for our study. A word like ndau has a proto-form which is not similar
to it. Instead it shows dose similarity to the Sotho form. In this case it can be regarded
as a Sotho adoptive. Words like muvhuda,
1\
mbidi
I'
and pfene
I'
are similar in both Venda
 59
and Northern Sotho and have no proto-forms. This may suggest that the Venda people
adopted these words from the Sothos since the latter are said to have settled in the
area before the Vendas. Venda also seem to have certain names of animals which are
not found in Northern Sotho, e.g. ntsa and khamba. The word khamba resembles its
proto-form. And this illustrate the fact that Venda retained some of the proto-forms for
names of animals which other languages like Northern Sotho do not have.

There are other words which are common between the two languages besides names
of animals. First, we shall consider the words with proto-forms:

CB Venda N.Sotho
.yadim- -adzima -adima 'borrow/lend'
-te 9 -renga reka 'buy'
-badi vhadzi lebadi 'scar'
-buda mvula pula 'rain'
-poku bofu sefofu 'a blind'
-tak6 shahu Iera go 'a buttock'
-p6d- fhola fola 'be healed'
-domo mulomo molomo 'mouth'
-yem- -ima em a 'stand'
-bepo phepho 'cold' phefo 'wind'·
gina dzina leina 'name'
gubo nguvho kobo 'cloth'
-jada ndala tlala 'hunger'
A

-jida ndila
A.
tsela 'path'
-ju nndu ntlo 'house'
"'
-kade kale kg ale 'long ago'
-kemde dzhende lerete 'testis'

From the examples given above it appears Venda retained the proto-forms rather than
adopting these words from the Sotho languages. If one compares the proto-Bantu
forms with those of Venda one realises that the sound shift in some of these words
 60

agrees well with what has been established by Guthrie and other linguists on the
Venda proto-Bantu reflexes. From the examples given above the following shifts have
occurred:

C B. * p > fh e.g * -po'd - > -fhola

*k>f e.g * -poku > bofu
*k>h e.g *-taka > rahu
* k > tsh e.g * -kimde > -tshende)
*k>k e.g * -kade > kal.e
*t>r e.g * -teg > renga

These reflexes attest to the fact that although the words above are similar to Sotho
they are in fact proto-Bantu retentions rather than adoptives. The fact that Venda
retained certain words from proto-Bantu can also be seen from the following examples:

CB Venda N.Sotho
'
-juba duvha
A
letsatsi sun'
-dopa malofha madi 'blood'
-cue khovhe hlapi 'fish'
-yimb- imba -opela 'sing'
yimbo nyimbo kosa '
song'
-dong6 dongo(shubi) mmu 'soil'
-t6nga thong a pat! a 'stick'
-denge mulenzhe leo to 'leg'
-yambo nyambo tab a 'affair'

Even though these examples have nothing to do with the affinity bet~een Venda and
Northern Sotho, they demonstrate that Venda has retained certain proto-Bantu forms
which Northern Sotho does not have. And this in turn adds weight to our argument
that the similarity between some of the words in these languages is due to a common
origin rather than to adoption.
 61
Although we have attempted to explain some of the cognates between Venda and
Northern Sotho there are others which we cannot account for. These words either
have proto-forms which are different or they do not have proto-forms at all. We shall
first look at those words with proto-forms:

CB Venda N.Sotho
-buk- -alafha -ala fa 'cure'
-diad- -dzula -dula 'sit'
-de mad- -huvhala -gobala ·get injured'
badi mulovha maloba
.yesterday or few days ago'
degu halwa bjalwa 'beer'
'
-keb- -lumelisa -dumedisa greet'

To explain some of the words like halwa and -lumelisa one needs to take into account
the social and cultural history of the speech community. In the case of a word like
halwa it is important to check which speakers of the two languages originally used this
product. It seems that this name originated with the Sotho people. This is mainly
because of its occurrence in other languages of the South-Eastern Zone viz. Zulu,
Xhosa, Tswana, and Sotho. The Venda probably adopted this word from the Sotho
--
whom they found already in the area when they arrived from the north. With regard
to the word lumelisa it is important to note that there is another word for this in Venda,
viz. vusa. The word -vusa seems to be the real Venda word. Some of the old people,
particularly those far from the areas which have been influenced by Sotho rarely use
the word lumelisa. This suggests that the word Iumelisa was adopted from Sotho.
There are many other words whose origin can be identified in this way, but this will
need a study which is based on the etymology of words only.

From the aforegoing discussion an attempt was made to show the influence of Sotho,
particularly Northern Sotho, on Venda. Northern Sotho dialects which have influenced
Venda were identified and their influence discussed. The influence of these dialects
is easy to identify, especially at the lexical level. The problem of identification occurs
in words which are common in these two languages yet do not seem to be clear cases
 62

of adoption. Scholars have attempted to explain the affinity between Venda and Sotho
in different ways. Scholars such as Loubser (1 988; 1989) would regard this affinity as
an indication of the fact that Venda originated as a result of a fusion between Shona
and Sotho. However, what we have found from this study suggests a case of influence
rather than that of amalgamation. The fact that Venda has many words in common
with Northern Sotho does not mean that Venda originated from that language. To date
Venda also shows a very great influence from Afrikaans and English and yet this does
not necessarily mean that it originated from these languages. The fact that Venda
originated as an amalgamation of Sotho and Shona can be disputed on the grounds
of the many words we have given in this section which either show close
resemblances with their proto-forms and not with Sotho or those words which occur
in Venda and not in Sotho.

Some scholars account for this similarity by looking at a common origin. This seems
to be the main cause of the affinity which exists between Venda and the Sotho
languages. Many examples given in this section attest to this fact. However, in some
cases the affinity between Venda and Sotho is a result of adoption as was indicated
earlier.

1.2.2.3 Tsonga

The influence of Tsonga on Venda can be attributed to contact between the two
languages. The close contact between Venda and Tsonga can be traced as far back
as 1835 when significant numbers of Tsongas moved to the North-Eastern Transvaal.
However it is not intended here to give a detailed history of the Tsonga migration to
the Northern Transvaal. According to Van Warmelo (1 935:93) the Tsonga came over
the border in small parties and settled down wherever they coul~. Some of them
settled in Venda areas like Hatshivhasa and Mphaphuli. Amongst those who settled
in Mphaphuli's areas are Xikundu, Mhinga and Xigalo (see Map IV on page 276 of this
disertation). As the number of Tsongas increased, particularly in the North-East and
South of Venda (Refer to Map Ill for the contact between the Venda and the Tsonga
in the south), they started to impact on the Venda speech community in many ways.
 63

Van Warmelo (1935: 117) describes the settlement of the Tsonga to the south of
Venda as follows:

"This Venda population to the South of the Zoutpansberg was partly driven out
and partly absorbed by the influx of the vaTonga (Tsonga) who drove a wedge
in between the Venda of the Zoutpansberg and the Sotho of the present
districts of Groot Spelenken and Duiwelskloof. But a number of the Venda
remained though the waves of Tsonga immigration surged all around them "

However, the description of the contact between Venda and Tsonga made by Van
Warmelo in this passage seems to have changed drastically after the 1968/69
removals which saw hundreds of Tsonga removed voluntarily or involuntarily to the
land today called Gazankulu. These removals were an attempt by the South African
government to group people according to ethnic affiliation. At that stage the Vendas
had been living happily with the Tsonga with whom they shared the territory for a
hundred years. Their children attended the same schools in some areas and there was
extensive inter-marrying between them (Desmond 1969: 146). The removal of the
Tsongas from Venda reduced the contact between the two population groups to border
areas. However, in some areas the boundary between Venda and Gazankulu has
never been clearly demarcated. As a result some strips or islands of either the Venda
or Tsonga remained in each other's territory. Tsonga areas like Elim, Valdezia and
Bungeni form a strip of land which juts into Venda country.

In his research on Tsonga dialects, Baumbach (1970:59) established that certain

Tsonga dialects had been influenced by Venda. He describes this influence as
follows:

"In Venda word die dentale konsonante as foneem gebruik en as daar in

aanmerking geneem word dat daar veral in die Suidelike, Suid-Oostelike en
Oostelike gedeeltes van Vendaland, 'n baie groat mate van
ondermekaarw'oning tussen Vendas en Tsongas bestaan, kan 'n beskouing van
taalbeinvloeding (in hierdie geval kan in eensydige beinvloeding van Venda na
 64

Tsonga met sekerheid aanvaar word. Verder as die getal Tsongas wat as
onderdane van sekere Venda kapteins in Vendaland woon, in aanmerking
geneem word, kom die politiese faktor oak in die gedrang en dit speel 'n baie
definitiewe rol by taalinvloeding".

This influence of Venda on Tsonga is found in Xiluleke of Makuleke and Mhinga.

These areas border with Venda on the East or South-East. The Venda spoken in
these border areas is also influenced by Tsonga.

The influence of Tsonga on Venda, as already mentioned, is very intense in the south.
This include areas like Tshimbupfe, Masia, Nngwekhulu, Magoro, Mashau and Muila.
Van Warmelo (1937:81) describes the language situation in these areas as follows:

"Their Venda language and tradition were swamped by numbers. All their
children grew up bi- and trilingual.

The influence of Tsonga on Venda can be illustrated by the following examples:

dzhomela Tso. dzhomela 'kala bash'

tshilibana Tso. xirimbane 'iron-trap'
matshomane Tso. macomane ' a kind of worship'
.tshidzholo Tso. xidzholo 'kala bash-clay'
mathwasane Tso. thwasani '
person undergoing divine training'
mungome Tso. mungoma 'diviner'
manzhozi Tso. mandlozi 'divine-spirits'
thandanda Tso. thandanda 'tambourine'
mungana cf. khonani Tso.mungane 'friend'

It is interesting to note that these examples denote different cultural aspects. The word
tshidzholo shows the influence of Tsonga on Venda music. Words such as
matshomane, mathwasane, mungome, manzhozi, thandanda show Tsonga
 65
influence on the religious sphere of Venda culture. Venda has also adopted a number
of words pertaining to utensils. Words like dzhomela and tshilibana are good examples
of this.

In some areas the contact between Venda and Tsonga is so intense that, besides the
many adopted lexical forms ,Tsonga influence is realised at other grammatical levels
such as semantics and syntax. These aspects will receive attention later, in the
chapters that will follow. Nevertheless we will only give a few examples of some of
these aspects.

Madukana ndo rna vhona mulovha. 'I saw the boys yesterday'
Madi rna khou elela. 'The water is flowing'
""'
Ndo rna AIa haala
A
maswiri. · 'I ate those oranges' _
Halwa vhula
,.... byo tevhuwa. 'That beer has been spilled out'

When one looks at these examples one is fascinated by the use of the ma form which
in this case has a concordia! function. The use of this form is typically Tsonga since
in Venda there is no such concord and instead, the concord a is used. The use of
byo instead of ho also shows the influence of Tsonga on Venda grammar and sounds.

1.2.2.4 Nguni

Venda has also been influenced by Nguni languages such as Zulu and Ndebele.
But what is surprising is that geographically speaking, Venda is far removed from
areas where these languages are spoken. The historical contact with Nguni
languages which can still be recalled by elderly Vendas that of Mzilikazi's raids
during the 1930s. Mzilikazi and his men (called Mabunyu i.e. ·naked' in Venda)
raided Venda on their way to Mashonaland. It was during this time that the Venda
adopted a few Nguni words pertaining to military and political fields. This fact is
also attested by Janes-Phillipson, who after comparing Venda and the Nguni
languages, came to the following conclusion:
 66

"Historically the S.40 (Nguni) group have displayed dominance militarily and
politically over surrounding groups ".

The military dominance of the Nguni groups over Venda can be illustrated by the
following examples:

Ve.tshitangu
,. < Zu. isihlangu 'shield'
Ve.mmbi < Zu.impi 'army'

Some of the Venda names still bear resemblances to Nguni. The name
Mulangaphuma is said to have originated from the Zulu phrase ifangaliphuma
. (When the sun rises) .

The more significant contact between Venda and Nguni, Zulu in particular, came
about through urbanisation. Urbanisation can be divided into two phases. The first
one is the mining period following the discovery of diamonds in Kimberley and gold
at the Witwatersrand during the 1860s and 1880s respectively. The second phase
is that of industrialisation.

The discovery of diamonds and gold attracted many people to the minefields. Thus
Van Jaarsveld (1975:159) says:

"Kimberley (1871 ), Johannesburg and Witwatersrand initially exerted the

biggest attraction. Blacks living on a subsidence economy in the reserves
occasionally went to the diamond and gold fields for two or three months
to supplement their incomes and then went home again".

The Vendas were no an exception to this. According to Van Warmelo (1940:59) by

1895 there was already a significant number of Venda men working in the mines.
When Makhado died in 1895 Mphephu, who was to be his successor was at
Kimberley having gone there to seek work with a group of his age-grade and some
elders. It is therefore assumed that in the mines the Venda came into contact with
 67

Nguni groups such as the Zulus. The Zulu language seems to be the one which
was most dominant in the minefields. Its dominance was not due to large numbers
of Zulus in the minefields but due rather to Fanagalo which was the lingua franca
of the mines. Fanagalo is described by Ngcongwane (1983:62) as "a pidgin
language used extensively in the mines and other industries in South Africa". Cole
(1953:2) describes its role in the mines as follows:

"... Fanagalo soon became established as the means of communication

between European employers or overseers and Bantu labourers ... lndeed
without some lingua franca the gold mines would be faced with an almost
insuperable language problem for they employ Bantu labourers speaking forty
or more different languages from as far afield as Angola, Northern Rhodesia,
Nyasaland and Tanganyika".

Fanagalo was also taught in the mines. Cole (1953:2) has this to say in this regard:

"On many mines new recruits are regularly put through a course of Fanaga!o
before being sent underground and others learn it while undergoing training
in their various duties but learn it they must ·before becoming effective
workers"

It is clear from this passage that all the people who went to work on the mines
were fo(ced to learn Fanagalo. Fanagalo seems to be dominated by Zulu forms.
Cole (1953:1) describes its composition as follows:

"Its vocabulary is roughly 70% Nguni (mainly Zulu), 24% English and 6%
Afrikaans."

The fact that Nguni, Zulu in particular, forms the main component of Fanagalo may
perhaps explain the influence of Nguni on Venda. Venda appears to have
accumulated substantial Nguni vocabulary via Fanagalo. The migrant labourers
continued to take the knowledge of Zulu that they had learnt from Fanagalo back to
 68

their own territories. Of course most of the Africans at that time could not hesitate
to use Fanagalo words outside the workplace since "they were convinced that they
were speaking the language of the white man which is so different from their own ..
(Cole 1953:7).

The following adoptives from Zulu refer mainly to mining and they seem to have
been adopted into Venda via Fanagalo.

mugodi Zu.umgodi 'mine'

ngolovane Zu. ingolovane 'mine-trolly'
ngozi Zu.ingozi 'accident'
tshova Zu.chova 'push'
khwela Zu.khwela 'board'
phakhamisa Zu.phakhamisa 'lift-up'
tshona Zu.chona 'go down'
khathala Zu.khathala 'ired/worry'
kavhanga Zu.cabanga 'think'

The influence of Nguni on Venda also came about through industrialisation. As in

the mines, many Vendas flocked to industries. In most industries Fanagalo was the
medium of communication, particularly in those early days (cf. Ngcongwane
1983:70). Therefore Nguni elements were adopted into Venda through Fanagalo in
industries. Beside the influence of Nguni on Venda via Fanagalo,. the Venda
sometimes had direct contact with Nguni groups such as Zulu. In this way Venda
was influenced directly from interaction with these people. It should also be realized
that in some areas "Zulu is about the widest (sic.) used black language"
(Ngcongwane 1983:71 ). Areas like Johannesburg are a good exaf1!ple of this fact.
Today the interaction between these groups is very common in the townships and
work situations. Intermarriage between the Zulus or any other Nguni group and the
Venda is very common in urban areas these days. Thus the Venda speaks of
mutshato, which is an adoptive from the Zulu umshado (wedding).
 69
The following are further examples of adoptives from Nguni languages

A. mabalane cf. muiiwaleli Zu.umabalane 'clerk'

nduna cf.gota Zu.induna 'foreman'

B. Mugivhela Zu.umgqibelo 'Saturday'

tshikhathi cf. tshifhinga Zu.isikhathi 'time'

C. mutshato cf.mbingano Zu.umshado ' wedding'

D. tshifevhi cf.phombwe Zu.isifebe ' a loose woman'

tshidakwa Zu.isidakwa 'drunkard'
tshigevhenga cf.tshira Zu.isigebengu
.criminal'
tshivhotshwa Zu.isibotshwa ' prisoner'.

E. gokoko Zu.ikokoko 'water-bucket'

F. -zama cf.linga Zu.-zama 'try'

-tshaela cf.laula Zu.-chaela 'drive'
-phumelela cf.wedza Zu.-phumelela ' succeeds'
-hoi a cf.lam bedzwa Zu.-khola · 'get paid'
-thela Zu.-thela 'pay tax'

The examples given above show that Nguni forms have been adopted into different
spheres in Venda. Adoptives under 8 have mainly to do with time. The word
M ugivhela is used to refer to the sixth day of the week and is from the Zulu word
uMgqibelo. In Zulu this word has been derived from the stem -gqibela and it means
the last working day. The word Mugivhela seems to be of recent origin. Before the
advent of westernisation people used to work throughout the week. The
introduction of christianity and the modern labour system led to the choice of the
seventh day of the week as a day of rest, and the sixth day as the last working day
of the week. This word was probably brought to the Venda speech community
 70

either by missionaries or those who went to work in urban areas. The word
tshikhathi was probably also adopted from Zulu. The word for 'time' in Venda is
tshifhinga. The adoptive mutshato in C as already indicated, shows the influence
of Zulu in the sphere of marriage in Venda. The adoptives under D are very
interesting since they show social problems that were common in urban areas and
but not in Venda. The word gokoko under E shows that names of some of the
utensils were adopted from Zulu or Nguni groups. The last group of adoptives has
mainly to do with diverse activities. Some of these have counterparts in Venda e.g.
the Venda verb stem -linga is used for -zama. However, word forms like -hola and-
thela have no equivalents since they denote activities that are foreign to Venda.

1.2.2.5 English and Afrikaans

Venda, like many other African languages of South Africa, has been considerably
influenced by English and Afrikaans. The influence of these languages on Venda is
mainly a result of westernization. This section will therefore attempt to show how
the Venda, after their settlement in the land south of the Limpopo, came to be
influenced by Western culture (civilization) which in turn affected their language.

Whereas some African languages came into contact with European languages as
early as the seventeenth century, Venda only seems to have had significant
contact with these languages during the nineteenth century (Blacking 1967:6). The
Venda are therefore among the last people to have come into contact with
European languages. Nevertheless, Venda has adopted linguistic forms from these
languages very extensively. To provide a better background on how Venda came
to be influenced by these languages, different contact situations ~ill be looked at.
The first contact to be discussed will be the contact with the Voortrekkers, then the
role of the missionaries, labour migration and lastly contact through conquest.

The first contact with European languages such as Afrikaans was a result of the
Great Trek (ibid. 6). Some of the trek movements, the trek movement under Louis
 71

Trichardt for example, went as far as the Zoutpansberg mountains in the Venda
area, hence their contact with the Venda. The Voortrekkers settled on Venda soil
around 1838 and then established towns such as Schoemansdal and Louis
Trichardt. The Venda tried to resist the settlement of these foreigners in their
country but in vain. When the Voortrekers had settled on Venda soil, they used to
interfere with the Vendas in one way or another. Blacking (1 967:6) tells of the
interaction between Louis Trichardt and some Venda chiefs. According to him Louis
Trichardt was very often induced to interfere in a dispute between two factions led
by the sons of a deceased chief. However, these contacts had little influence on
Venda since the Venda ·were antagonistic to European settlement' (Mathivha
1985:11)

The missionaries who came to Venda during the 1870s also played an important
role in the influence of European languages on Venda. In this study we do not
intend to describe all the missionary activities in detail, but to make a brief survey
of their contribution to the influence of English and Afrikaans on Venda. The first ·
missionaries to come to Venda were from the Berlin Missionary Society. Although
they were German speakers they contributed greatly to the influence of English and
Afrikaans on Venda. The influence of German language on Venda seems to be
very minimal. Only a few examples of words in Venda can be traced back to the
German language. It is very difficult to distinguish between words adopted from
Afrikaans and those from German. The word muneri
,...... which is regarded as an
Afrikaans adoptive, might have been adopted from Afrikaans via German.
(Personal communication with Z. Mutsila). This may be supported by the fact that
Venda h;3S the word tshiQeri (hair-style-fashion achieved by cutting away at one
level right round, Van Warmelo 1989:412) which is not found in Afrikaans. This was
the hair-cut style of the Germans. Today this hair-cut is called the. German cut. It
appears that missionaries who came to Venda, Beuster for example, also had a
background of Afrikaans and English from the Sotho areas where they worked
before they came to Venda. The role of the missionaries in the influence of English
and Afrikaans on Venda was in culture. Their activities caused a gross change in
the Venda culture. The missionaries all over Africa are known to have converted
 72
more people to Western civilization than to Christianity. To the missionaries
civilization and Christianity seemed to go together. Thus Mathivha (1985:7 4)
quoting from Livingstone says:

"Neither civilization nor Christianity can be promoted alone. In fact they are
inseparable ... Wherever a missionary lives, traders are sure to come, they
are mutually dependent, and each aids the other".

In their mission of spreading the gospel or converting people to Christianity, the

missionaries seemed to have destroyed much of the cultural material of the Venda
people which they regarded as heathenism. By so doing they left a vacuum in the
culture of the people· and this paved the way for the influence of Western
civilization on the Venda. Mathivha (ibid. 74) levels criticism against the missionary
activities in the following way:

"... it should be stated that the greatest mistake the missionaries made was in
the breaking down of all the organizations and customs of the black people
without waiting to discriminate between good and bad ... As it was, they often
destroyed what they were not able to rebuild and left many of the black men
in a worse state than they had been in before".

Many of the Venda people have, for better or for worse, accepted Western
civilization in the name of Christianity. Many western materials such as clothes
replaced the traditional cultural materials of which the missionaries disapproved.
Through westernization and perhaps the influence of the missionaries, the Venda
people came to use foreign materials such as clothes which in turn brought a
significant number of foreign words to Venda. Consider the following examples:

vhurukhu Afr.broek 'trousers'

hemmbe Afr.hemp 'shirt'
tshikete Eng.skirt
rokho Afr.rok 'dress'
 73
badzhi Afr.baadjie 'jacket'
tshikhipha Eng.skipper

The missionaries became directly involved with the Venda language. In the first
place the need to translate the Bible into Venda compelled them to codify the
language. This was followed by the translation of the Bible. Many words of
Afrikaans and English origin are still found in the Bible or in other Christian
activities. Consider the following examples:

kereke Afr. kerk 'church'

evangeli Eng. evangelical
bivhili Afr. Bybel 'Bible'
muneri
A
Afr. meneer 'Mr '
muporofita Eng. prophet
mufarisei Eng. pharisee
paseka Eng. passover
themmbele Eng. temple
mubishopho Eng. bishop
veine Eng. vine
mudikoni Afr.diaken 'deacon'

The influence of English and Afrikaans on Venda also came as a result of

urbanisation. The discovery of diamonds and gold in Kimberley and on the
Witwatersrand respectively saw the flocking of many people including the Venda
into the minefields. In the mines migrant labourers seemed to have accumulated
English and Afrikaans vocabulary through Fanagalo. It was mentioned in the
previous section that English and Afrikaans constitute about 39% of Fanagalo.
Therefore the migrant labourers can also be regarded as the agents of words from
these two foreign languages into Venda. On their return home they did not only
have a broadened vocabulary but they also had new ideas and commodities of
foreign origin. Cole· (1953:4-5) gives a list of English and Afrikaans words found in
Fanagalo and some of these words have found their way into Venda. He gives the
 74
following examples:

tshikurufu Fan. sikulufu Afr. skroef 'screw'

bongisi Fan. bontshis Afr. boontjies 'beans'
tshikolodo Fan. skwelete Afr. skuld 'debt'
mbitshana Fan. bitshan Afr. bietjie 'little'
phaiphi Fan. payip Eng. pipe
fosholo Fan. foshol Eng. shovel
nyuwane Fan. nyuwan Eng. new one
-badela Fan. batala Afr. betaal 'pay'
-laisa Fan. layisna Afr. laai 'load'
-mosha Fan. mosha Afr. mars ' waste'

-basoba Fan. basop Afr. pas op 'watch'-

-tshipa Fan. tshipisa Eng. cheap
-tshova Fan. shova Eng. shove

It is interesting to note that Venda adoptives from English and Afrikaans seem to
be closer to Fanagalo than the forms in the original source languages. This may
suggets that Venda adopted these words via Fanagalo. In Fanagalo the structure
of most of the words adopted from English and Afrikaans has been modified. For
instance, vowels have been added to some of the words which end with a
consonant, e.g sikulufu, skwelete, -batala, chipisa, and mosha. Details of sound
ditferenc.es will be discussed later in Chapter 3. This made it easier for the Venda
to adopt these modified Fanaga!o words rather ·than from their source language.
Words like fosholo and mosha clearly show the adoption of these words via
Fanagalo. There is no reason for Venda to adopt the forms as fosholo and mosha
instead of shovele and morosa. This shows that Venda got th~se words from
Fanagalo, otherwise their form would have been modified differently.

The influence of English and Afrikaans on Venda became greater after the
conquest of the Venda by the Europeans. The Vendas who for so long resisted the
 75

control of the Europeans were finally defeated during the Mphephu-Boer war of
1899 (Blacking 1967: 16).

The subjection of the Venda to foreign rule affected their language in many ways.
What happened after the conquest could be paralleled with what Serjeantson
(1935:1-2) says of English:

"When one nation subdues another which speaks a different language the
conquerors if their object has been political power rather than settlement may
constitute an authority or ruling class which is in point of view of number
much in the minority compared with the whole body of the conquered people.
!n cases like this it is usually the native language that survives though the
. incoming dialect (language) will very probably transfer to the native language
vocabulary words which expresses its own methods of governments and
other cultural words".

Just as Serjeantson put it , the conquerors of the Venda people established their ·
form of government over the Venda after conquest. The first people to establish a
foreign rule over the Venda were the British who by then had occupied the Boer
Republics following the Anglo-Boer war of 1899. They divided the Venda country
into districts. These districts were ruled by commissioners, hence the adoptive
khomishinari. Policemen were also stationed in •different districts t? help maintain
peace and order. This resulted in adoptives such as pholisa, phikhaphu, etc. It
seems it was during this time that most of the words relating to money e.g
hafukoroni, bonndo, disheleni, !iki, peni, etc. came into frequent use in the Venda
speech community. The influence of Afrikaans seems to have gained momentum
after the Boers regained control of the Transvaal with the Limpopq as its border to
the North. Venda became part of the Union Government. It was under the rule of
the Union government that Venda became greatly influenced by the European
languages. The changes that took place in Venda after it was subjected to
European dominion were not different to those of other areas where the Europeans
established their control. Thus Lestrade (1967: 106) says:
 76
"Native culture in South Africa is everywhere in contact with ever-widening
channels of European civilization which are gradually but effectively modifying
its traditional manifestation. It is interesting to note that English and
Afrikaans,. became the bearer of the new and powerful culture, the key to the
white man'.s supremacy, the secret to his superiority".

In other words, English and Afrikaans were accorded a higher status compared to
Venda. Prestige-wise Venda was reduced to a lower status. Such a position has
put Venda under severe pressure to adjust to the level of these foreign languages
with regard to the communicative needs of its speakers. It is this state of affairs
which has resulted in the adoption of many words from these foreign languages.

The influence of English and Afrikaans has also been accelerated- by schooling.
These languages became the official languages after the Union· government took
over control of schools from the missionaries. The learning of these languages by
pupils makes it easier for them to transfer certain forms from these languages to
their language. Some of the subjects they study, science for example, are full of
concepts which have no equivalents in Venda. Therefore words from the foreign
languages which have no equivalents in Venda are very often adopted into the
language. Consider the following examples that have to do with science:

okisedzheni 'oxygen'
saintsi 'science'
'
pharafeni paraffin'
veini 'vine'
khomphiutha 'computer'
bailodzhi 'biology'
tshipiriti ·spirit'
""
fanele 'funnel'

In a nutshell, the political conquest of the Venda by the Europeans, together with
the missionary effort, education, administrative action and above all the introduction
 77
of new economic systems has inevitably changed the role and the nature of Vemda
and also led to the adoption of many words from English and Afrikaans. The
adoptives from these languages also reflect these different spheres of influence.
Adoptives that have to do with religious activities of the missionaries have already
been given. Therefore only adoptives that has to do with administration, education,
and the new economic system will be given below:

a) Government (administration and judiciary)

phalamennde Eng. parliament

hulumennde Eng. government
m uphurisidennde ~ng. president
khabinete Eng. cabinet
minisi~ Eng. minister
khothe Eng. court
dzhadzhi Eng. judge
axennde Eng. agent
"'
tshipikara Eng. speaker
m adzhisitirata Eng. magistrate
pholisa
A
" Eng. police
burigadia Eng. brigadier
kaputeni Afr. kaptein 'captain'
/'o

These examples show the dominance of English in these fields of administration

and the judiciary.

b) Education.

tshikolo Afr. skoal 'school'

bugu Afr. boek 'book'
maraga Afr.mark 'market'
bammbiri Afr. papier 'paper,
 78

desike Eng. desk

ravha Eng. rubber.
thitshere Eng. teacher
tshoko Eng. chalk
bulupheni Eng. blue-pen
phirisipala Eng. principal
tshipikitere Eng. inspector
"
kholishi Eng. college
yunivesithi Eng. university

From these examples it is clear that both English and Afrikaans have had a great
influence on the the field of education.

c) Economic system

indasitiri Eng. industry

"'
ikonomi Eng. economy
tshelede Afr. geld
.money'
bisimusi Eng. business
bannga Afr. bank 'bank'
vhengele Afr. winkel · 'shop'
shopho Eng. shop
bonndo Afr. pond 'pound'
tiki Afr. tiekie 'ticky'
A

disheleni Eng. ten-shillings

hafukoroni Eng. half-a-crown
rannda Afr. rand ·rand'
bulasi Afr. plaas 'farm'
maini Eng. mine

These examples show the influence of English and Afrikaans on different economic
fields of the Venda. The influence ranges from business institutions, money,
 79

mining, industries to farming. Whereas English seems to dominate in industrial and

mining terms, Afrikaans dominates in farming terms. A very interesting case is that
of the monitory system. Venda shows the influence of both English and Afrikaans
in this sphere. These words for money are used interchangeably even today.

d) Transport and communication

The influence of English and Afrikaans on Venda can be clearly seen in, these two
areas. Most of the vehicles used for transport nowadays, cars, trains, motorbikes,
and aeroplanes for example, were . not available before the Venda came into
contact with the Europeans. The introduction of these vehicles to Venda brought a
number of foreign words to the language:

Iori Eng.lorry
mod oro Afr.motor 'motor-car'
"'
thirakha Eng.truck
"'
baisigira Eng.bicycle
gariki Afr.karretjie ' wagon'

.....thireni Afr.trein 'train'

vulamatshini Eng.flying machine.
bisi Eng.bus
thekhisi Eng.taxi

The following adoptives have mainly to do with modern communication.

thelefoni Eng.telephone
thelegiramu Eng.telegram
'
poswo Afr.pos post'
vhurifhi Afr.brief ·letter'
 80

e) Clothing

This aspect was discussed in the section on the influence of the missionaries in the
Venda culture. However, westernization also played a very important role in this
regard. The introduction of western types of clothes to the Venda led to the
adoption of many foreign words from English and Afrikaans into Venda.

dzhasi Afr.jas 'jackef

tshinorovhagi Afr.onderbaadjie 'waist coat'
sogisi Afr.sokkies 'stockings"
bulausu Eng.blouse
gebisi Eng.cap
vese Eng.vest
tshuniki Eng.tunic
dugu Afr.doek 'handkerchief.

f) Foods and household

The contact between Venda and English and Afrikaans has led to the adoption of
many words from these languages pertaining to food and households:

i) Foods

vhurotho Afr.brood 'bread'

khekhe Eng.cake
tshizi Eng.cheese
Iegere Afr.lekker 'sweet'
raisi Eng.rice
khasitede Eng.custard
bivi Eng.beef
 81

ii) Drinks

tie Afr.tee 'tea'

kofi Afr.koffie 'coffee'
bia Eng.beer
nyamunaithi Eng.lemonade

g} Business, professions and institutions

The contact of the Venda with Europeans has in one way or another transformed
Venda society. New forms of business came into existence in the Venda society.
· The following adoptives designated such businesses:

vhengele Afr.winkel 'shop'

tshitolo Eng.store
butshara Eng. butcher
khefi Eng.cafe
hodela Afr.hotel 'hotel'
holosele Eng.who!esale
makete Eng. market

New professions also came into the Venda society due to their contact with the
Europeans. Let's look at the following examples:

vhunese Eng.nursing
vhukhanikhe Eng.mechanics
vhudokotela Afr.doktor 'doctor'
vhulekhitshara Eng.lectureship
vhupurofesa Eng. professorship
vhuthitshara Eng.teaching
vhuhweta Eng.waiter
 82
Institutions such as schools, colleges, universities, hospitals, etc. were introduced
by the Europeans and have resulted in adoptives such as tshikolo (cf. Afr.skool),
kholishi (cf. Eng. college}, yunivesithi (cf. Eng.university).

h) Sport and social entertainment

The westernization of the Venda can also be seen in the domains of sport and
social entertainment. Most of the things in these domains were introduced by the
Europeans and names of these activities were then adopted into Venda:

V Sports

bola Eng.ball
tshese Eng.chess
the nisi Eng.tennis
hoki Eng.hockey
dzhavelina Eng.javelin

ilj Modem entertainments

radio Afr.radio 'radio'

thelevisheni Eng.te!evision
baisikopo Afr.bioskoop 'bioscope'
ogeni Eng. organ
giramafoni Eng. gramophone

i) Agriculture and husbandry

The influence of English and Afrikaans on Venda can also be found in the domains
of agriculture and husbandry. The contact of the Venda with the Europeans has
transformed Venda· traditional ways of farming. Modern farming methods were
accompanied by many foreign words which denote new farming methods, plants
 83
and equipment for farming. Consider the foll~wing examples:

bulasi Afr.plaas 'farm'

mogeni Eng.morgen
many oro Eng.manure
deretere Afr.trekker 'tractor'
I' "'

New plants includes amongst others, the following:

khavhishi Eng.cabbage
nyala Eng.onion

...tamatisi
,.. Afr. tamaties 'tomato'
tshipinisi Eng. spinach
gorou Afr.koring ' wheat'

'
ndabula
,.... Afr. aartappel potatoes'
muberegisi Afr. perskeboom 'peach tree'
munngo Eng.mango tree
muapula Eng.apple tree

The following domestic animals were introduced to the Venda :

donngi Afr.donkie 'donkey'

bere Afr.perd 'horse'
meila Eng.mule
'
gamela Afr.kameel camel'

1.3 Conclusion

In this chapter the original migrations of the Venda and their contact with other
population groups were discussed:

With regard to the origin of the Venda, different theories were discussed. There are
 84

two main opposing views on the origin of the Venda, namely the Central Africa
origin and the local origin hypotheis. An attempt was made in this study to
reconcile these two opposing views. It appears that some Venda groups, such as
the Singo, originated from Central Africa while the earliest Venda people (the so-
called ·real' Venda) have originated locally as a result of the interactions or
amalgamations between the Sotho-Tswana and the Shona speaking groups wr10
inhabited the area before the arrival of groups such as the Singo.

The contacts of the Venda with other population groups during migration also
discussed. The main focus here was on the Singo group since it forms a big group
and has a well-recorded history compared to other Venda groups. However, the
study of the Singo migration to this part of the country is obscured by the lack of
evidence as to their place of origin. Although some scholars maintain that this
group originated in Central Africa, it is not possible to trace the migration of this
group before it reached the country today called Zimbabwe. It is in this area where
the picture of their migration becomes clearer. Here they came into contact with
the Western Shona groups and the Karangas. Their contact with these Shona
groups is supported by linguistic, archeaological as well as historical evidence.

This chapter concluded by looking at the postmigratory contacts, i.e. the contacts
which the Venda had with other population groups after their settlement in the land
south of the Limpopo. Here it was established that when the Singo arrived in the
area south of the limpopo they found the Venda groups that were already in the
area in contact with other languages spoken in the region, namely, Lembethu,
Twanamba and Sotho. Thus the different Venda groups came into contact with
both Shona and Sotho languages. With regard to the latter the contact continued
mainly with the Northern Sotho groups such as the Tlokwa and Lob~du. The arrival
of the Tsonga in the 1830s added another contact situation to Venda. Through
urbanisation the Venda came into contact with other languages like Zulu, English
and Afrikaans. Contact with the latter two languages was mainly due to
westernization. These languages have influenced Venda in different domains.
Some of these influences were discussed in this chapter. From this discussion it
 85
has become clear that the influence of other African languages, viz. Sotho, Zulu,
Tsonga and Shona on Venda is mainly in cultural aspects such as music, marriage,
initiations and religion. While English and Afrikaans have also influenced the Venda
culture to a certain extent, they made an impact on the language in other semantic
domains such as education, science, technology, government (including
administration), politics, sport, entertainment, telecommunication and others.
 86
CHAPTER 2

2.0 LEXICO-SEMANTIC ADAPTATION

2.1 Introduction

The focus of this chapter will be on lexical and semantic adoption. These
processes are often 'viewed as twin processes since the adoptive elements may
themselves undergo changes of meanings in their adopted language' (Chimhundu
1982:2). The main emphasis in this chapter will be on the semantic adaptation of
foreign words to Venda. However, other aspects of lexical adoption like the factors
which lead to lexical adoption will also be discussed. The discussion on semantic
adaptation will also include meaning changes to native words.

2.2 Lexical adoption

The lexicon of a language is more open to foreign influence than other grammatical
levels. Thus Arlotta (1972: 184) says:

"Words can be taken very freely from one .language into another, with very
little, if any, effect on the rest of the grammar or lexicon of the borrowing
language. The vocabulary or lexicon is the most unstable part of any
language, and words may be picked up or discarded as a given community
feels the need".

Weinreich (1953) regards the lexicon as a domain of adoption par excellence.

Lexical adoption was discussed in the previous chapter at some .length, where it
became clear that Venda has been considerably influenced by other languages in
as far as the lexicon is concerned. In this chapter we shall be looking at the factors
which contribute to the adoption of lexical items in a language. The lexical items
adopted into Venda may, as was indicated in the previous chapter, be divided into
semantic domains.
 87
2.2.1 Reasons for lexical adoption

The first significant question in the study of adoptives is what motivates lexical
adoption. Linguists have in many studies given different reasons for lexical
adoption. In his book, Languages in contact, Weinreich (1953}, suggests at least
seven reasons for the adoption of lexical items in a language. Emslie (1 977:23-24}
summarises these factors into two main motives, viz. the need-filling motive and
. the prestige motive. These factors will be discussed in view of the different contacts
that Venda has had with other languages. Some of these factors have .. already
been discussed in the previous chapter, albeit scantily.

2.2.1.1 The need-filling motive

Words may be adopted into a language because there is a need for them.
According to Robins (1971:133} the need to adopt foreign words into a language
may arise in many ways:

"The conditions of life of individuals in society, their artefacts, customs, forms

of organisation and the like are constantly changing, and consequently the
referents of many words in languages and the situations in which they are
employed are equally liable to change in the course of time. New products
often require new designations, and some words pass out of current
vocabulary as the particular sorts of objects or ways of behaviours to which
they refer become obsolete" .

Robins shows in this passage that words may be adopted because of certain
needs in the receiving language. These words are adopted to _designate new
concrete objects or abstract ideas and other cultural material that have come into
the speech community. It seems that all languages have, at one stage or another,
a need for new words. There is no language which can claim to be self-sufficient.
Partridge (1973:306) expresses this as follows:
 88
"No standard language exists on its own capital - no language can thus exist
if it is to continue to be a language and not become a mausoleum."

The need to adopt new words in African languages is described by Lestrade (1967
as follows:

"The indigenous forms of speech, rich as they are in vocabulary, flexible as

they may be in the expression of ideas, admitting as many shades of
accuracy and delicate nuances of meaning, could not in their unaltered state,
serve as an instrument of expression when it came to designating all the new
things, all the new ideas, all the new content of the white man's civilization.
Progressively as the Bantu came into contact with this civilization, there arose
the need for new words to name the new things, new forms oLexpression, to
convey the new ideas, a new speech habit, as it were, to meet the needs of
the new enviroment in which they found themselves".

From this passage it becomes clear that a language adopts words from other
languages to fill the communication gaps introduced into it, mainly through contact.
In Venda a considerable number of foreign words seem to have been adopted for
this purpose. From what Lestrade says, it appears that the need for African
languages to adopt foreign words was increased by their contact with European
languages like English and Afrikaans. Our findings in the previous chapter seem to
attest to this fact. Venda appears to have been influenced more by these
languages than by other African languages. English and Afrikaans are according to
Lestrade (1967:106 ) " ... the bearers of the new and powerful culture". Therefore
many words were/ are adopted into African languages due to westernisation.
However, in some instances words from these languages we_re adopted not
because of their function, but because of the status of these languages and this
brings us to the second motive, that of prestige.
 89

2.2.1.2 The prestige motive

While some words are adopted to fill the communicative needs in a language,
others are adopted for prestige. Prestige here has to do with the status of the
source language. The prestige of a language is usually determined by factors
outside the realm of language. Political and social factors play an important role in
this regard. When two or more languages meet, it often happens that one has a
higher status than the other. This view is supported by Burling (1970:169) who
states that languages seldom come into contact on genuinely equal terms. Certain
languages when they are in contact with others, become dominant over. There can
be two types of dominance. According to Coetsem (1988:13) dominance can be
linguistic or social. He describes linguistic dominance as that which occurs in a
situation where the speakers of the adopting language are in the majority and have
greater proficiency in the latter language than the speakers of the source language.
Social dominance "refers to the social status of the r1 (receipient language) as
opposed to the sl (source language). Therefore a language can be dominant not
because of the number of its speakers, but because of the social status which is
accorded to it.

Therefore the dominant language may be that of a conquering tribe or nation or it

may be the language of an established majority among whom socially subordinate
immigrants filter (Burling 1970:170). The prestige motive seems to account for the
many foreign forms adopted into Venda from languages such as English and
Afrikaans. As was indicated in the previous chapter, these languages enjoy a
higher, status than their African languages counterparts. Because of their prestige,
words which do not fulfil any function in the adopting language are adopted merely
for speakers to identify themselves with the source languages. Th~s, 'the prestige
factor might be looked upon primarily as a cause or a reason for linguistic
borrowing whether across language or across dialect boundaries' (Arlotta
1972:204).
 90
2.2.1.3 A combination of both motives

Although in some instances adopted words could be accounted for by either of the
factors we have discussed above, in others there is a combination of both factors.
Coetsem (1988:14) has this to say in this regard

"A speaker using 'borrowed' words may be motivated to do so by need but

nonetheless derive prestige from this situation. On the other hand a speaker
using borrowed words may be motivated to do so by prestige while such a
usage is rejected as pretentious by another speaker".

It seems that some words from English and Afrikaans are adopted for both the
prestige as well as the need-filling motive. It is interesting to note tt"!at words from
other African languages, with the exception of Shona are never adopted for
prestige. The words from these languages (African languages) are adopted only
when there is a need for them to express certain meaning significance. However,
some words from these languages, though they fulfil certain functions in Venda,
are often rejected because of the fact that their source languages are not highly
regarded by the Venda.

2.2.2 Spheres of influence

Adoptives may be divided into different semantic domains. Some of these semantic
domains were discussed in the previous chapter, where an attempt was made to
show how the different semantic areas of Venda have been influenced by the
languages it came into contact with. These semantic domains will therefore not be
discussed again in this section.

2.3 Semantic adaptation

Studies on adoptives have indicated that when words are adopted from one
language to another they undergo certain adjustments with regard to their meaning.
 91

In this section, therefore, we shall be looking at how adoptives from different

languages are adapted semantically to the Venda language. An attempt will also be
made here to show how the native words extend their meaning to accommodate
the foreign concepts.

2.3.1 Semantic adaptation of adoptives

In discussing the semantic adaptation of adoptives in. Venda we shall be looking at,
amongst ott1er things, meaning correspondence, meaning extension, narrowing of
meaning, radical change (shift) of meaning and change in emotive value
(cf.UIImann 1957, Ullmann 1970, Mojela 1991, Louwrens 1 993) 1

2.3.1.1 Meaning correspondence

Some adoptives are taken over without any change of meaning. Their meaning
remains exactly the same as in their source language. Consider the the following
examples:

Afr.brood > Ve.vhurotho 'bread'

Eng.dish > Ve.ndishi
Eng.gravel > Ve.giravhulo
Eng.pi!ls > Ve.philisi
Eng.watch > Ve.watshi
Afr.kart > Ve.garata card'

1.Mojela (1991) and Louwrens (1993) have carried out an intensive study of the
semantic changes of adoptives in Northern Sotho. Their framewo~k of analysis with
regard to some aspects of meaning change ·will be used in this study. Some
examples included here have a/so been extracted from the above mentioned works
because of their appropriateness with regard to the present analysis on Venda
adoptives.
 92

Afr.botel > Ve.bodelo

A
'bottle'
Afr.suiker > Ve.swigiri 'sugar'

Most of the adoptives that fall into this group are names of foreign objects ideas
which have entered the speech community.

2.3.1.2 Meaning extension

Meaning extension according to Arlotto(1972:177) has to do with a change in which

a word refers to more items than was the case with its original occurrence in the
source language. In other words, the adopted word widens its meaning range. 'By
meaning range or area of meaning of a word we refer to the totality of meaning it
can express' (Mokgokong 1975:26). Although the meaning is extenged, the basic
meaning is retained and is comparable. There are various ways in which the
· meaning of an adopted word can be extended. First, an adopted word can acquire
additional meaning in the adopting language irrespective of the context in which it
is used. Secondly, semantic meaning of an adopted word may be extended by
means of figurative meaning. In this case the adopted words are used in
metaphoric and idiomatic expressions.

(a) Additional meaning irrespective of context

We shall. begin by looking at instances where the meaning of an adopted word is

extended irrespective of the context in which it can be used. Let us consider the
following examples:

ENGLISH: STEAMER VENDA: TSHIDIMELA

Steam engine train Steam engine train

Diesel/electric train
 93

ENGLISH: LEMONADE VENDA: NYAMUNAITHI

A cooldrink made from lemon juice A cooldrink made from lemon juice
Any kind of cooldrink

ENGLISH: GAS VENDA:GESE

A type of air substance A type of air substance

Electricity
Lights

ENGLISH: CHECKERS VENDA: TSHEKASI

Name of a chain store Name of chain store

Any plastic bag

SOTHO: BAGATLA VENDA:VHAKHADA

A

A type of Tswana group A type of Tswana group

Any Sotho speaking person
(particularly women)

An examination of the examples given above indicates that the adopted words
have extended their meaning to cover a semantic field wider than that of the
source language. If we take a word like nyamunaithi < Eng. 'lemonade' we realise
that its original meaning has been extended. In its source language this word is
used to refer to a specific type of cooldrink which is 'a still drink made from lemon
juice' (The Concise Oxford Dictionary 1990:677). In Venda this word is used to
refer to cold drinks in general. Meaning extension can also be seen from the
 94

adoptive gese < Eng.'gas'. In English this word refers to 'any airlike substance
which moves freely to fill any space available irrespective of its quantity' (The
Concise Oxford dictionary 1990:487), but in Venda this word, besides this meaning,
can be used to refer to electricity or light. Sometimes it is used figuratively to refer
to a beautiful person with a very light complexion. Figurative meaning will be
discussed in greater detail the section that follows.

(b) Mean.ing extension through figurative meaning.

This aspect of meaning change, as indicated above, is still part of meaning

extension even though it is discussed separately. The meaning extension in this
section differs from the one discussed under (a) in the sense that the one
discussed here has to do with the addition of figurative meaning to an adopted
word. This process of adding figurative meaning to an ordinary word is quite a
familiar process in Venda. Adoptives are also used in metaphors and idioms to
express figurative meaning. We shall begin by looking at the figurative meaning of
adoptives as they occurs in metaphoric expressions.

(i) Meaning extension through metaphoric expressions

"Metaphm changes the meaning of words and creates new expressions on spur
the of the moment. It is born from the instanteneous glimpse of similarity between
two objects or two acts" (Breal 1964: 122). Breal's discription helps us to
understand the origin of metaphors in a language. Speakers of a language
compare two objects or two events which they see and this comparison results in a
change in the meaning of the original word. Thus Waldron (1967:169} says:

"Metaphors enlarge the semantic range of words, momentarily or more

permanently changing the frontiers of our lexical categories".

What Waldron says above, is also true of adopted words in Venda which when
used in metaphoric expressions, may acquire new figurative meaning. This change
 95
of meaning can be clearly seen from the following examples:

Ve.limindidi < Eng.limited

VE. : Vele ndi limindidi.
LIT. :'Vele is a limited (company)'
FIG. :'Vele is a person with unlimited resources (riches).

In this example the word 'limited' is used as a metaphor of wealth. Limited

companies are thought to be rich. Vele is therefore equated with such companies
to show that he is rich, particularly with regard to money. In this example the
meaning of the word 'limited' which in the source language only denotes a
company, has extended its meaning to refer to 'wealth or riches'. The extension of
meaning through metaphors can also realised in the following examQ_Ies:

VE.: Mulalo o tou vha baisikopo namusi.(Eng.bioscope)

Lit. :' Mulalo is a bioscope today'
· FIG.:'Mulalo is a laughing-stock today'

VE.: Mutukana u tou vha sheleni (Eng.shilling)

LIT.: 'The boy is shilling'
FIG.: 'The boy is naked'

In the examples given above, the English words 'bioscope' and 'shilling' have been
used in metaphor expressions and have therefore acquired figurative meaning. A
'bioscope' is a kind of film projector which was used in the past before cinemas
were established. It was watched by many people and people often laughed at
some of the scenes that were shown. Therefore, .if a person does things that
attracts peoples attention and becomes a laughing-stock, he is referred to as a
'bioscope'. The word 'shilling' on the other hand, refers to a person who is naked
since a shilling has no covering. As such one can see everything that is written on
it.
 96

(ii) Meaning extension through idiomatic expressions.

An idiom is defined by Langacker (1968:83) as "a phrase whose meanings cannot

be predicted from the individual meanings of the morphemes it comprises". It is
also important to note that although idioms are made up of more than one linguistic
item they function semantically as a single lexical item (Mokgokong 1975:55). If we
take an idiom like "U swika bagasititshi" (to arrive at Park station) we find that the
"'
word bagas~hi is an adoptive from Eng. 'Park station'. In Venda this word has
acquired the figurative meaning of ·a final or last point', particularly in a journey or
in· doing something: This figurative meaning has been derived from the fact that
Park Station is the biggest train station for mainline trains in the Transvaal. When
the mainline trains arrive at this station they proceed no further. Therefore the
idiom U swika bagasi;!_itshi (to arrive at Park Station), means to arriv~ at a final or
last point in what one is doing. This therefore shows that the meaning of the name
·Park station' has been extended to an additional figurative meaning. This kind of
meaning extension is also realised in the following examples:

Mashudu u vho sokou

.
amba vhudaridari
/" A
nge a wanedzwa a tshi khou tswa.
Hone thala ha Bonyongo o vha a tshi khou tou tambela tshiporoni. Arali li mu
A ~

wane zwavhudi o vha a tshi do vha o raha bakhethe.

"' ,.....

Lit. (Mashudu was just saying that and that because he was found stealing.
But .there at Bonyongo he was playing on the railway lines. If he (Bonyongo)
had found him he would have kicked a bucket)

From the above paragraph , we find the following idioms:

IDIOM : Mashudu u vho sokou amba vhudaridari nge a wanedzwa a tshi khou
X I'

tswa (Afr.daar,daar)
LIT. : 'Mashudu was just saying that and that because he was found stealing'
FIG. : 'Mashudu was confused because he was found stealing'
 97
In this example the Afrikaans words 'daar, daar' have acquired the figurative
meaning of 'confusion'.

IDIOM : Hone fhaJ.a ha Bonyongo o vha a tshi khou tambela tshiporoni (Afr.spoor).
LIT. : But there at Bonyongo he was playing in the railway lines.
FIG. : But there at Bonyongo 's place he was playing at a dangerous place.

In the idiom given above the Afrikaans word 'spoor' has acquired the figurative
meaning of 'danger'. The following idiom is the last one in the paragraph above:

IDIOM : Arali li mu wane zwavhudi

~ A
o vha a tshi do vha o raha bakhethe
~

(Eng. bucket).
LIT. : 'If he had found him he would have kicked the bucket'
FIG. : 'If he had found him he would have been dead'.

In this example the English expression 'to kick the bucket' is used to refer· to
'dying', hence this meaning extension has been adopted into Venda as well.

The following are other examples of idioms with adoptives which have acquired
figurative meaning in Venda:

IDIOM: U maketa (Eng.market)

Lit. : To .sell something
Fig. : To look for a job

IDIOM : U tshina thandanda (Tso.thandanda)

Lit. : To undergo thandanda divine practice.
Fig. : To be in trouble.

IDIOM : U dzhena mabanndani (Afr.band)

Lit. : To get in belts.
Fig. : To be arrested
 98

The adopted words, u maketa, thandanda, and mabanndani have acquired

figurative meanings which do not occur in their source languages. When a person
is selling something, he moves from place to place. The same thing applies to a
person who is looking for a job, moving from one place to another until he finds it.
Thandanda is a kind of dance perfomed by people who are undergoing divine rite
in Tsonga. The dance is very fast, following the fast rhythm of the drum. The
dancers never rest until the song is over. The fact that it is not an easy dance is
likened with a person who is in trouble. Such a person never settles down until the
'
trouble is over. The 'belts' in the last example refer to handcuffings. When one is
handcuffed one seems to be bound up by belts.

2.3.1.3 Narrowing of meaning

Whereas some words extended their meaning when taken over into Venda, others
narrowed their meaning. 'What is involved here is that the number of things a
word refers to is reduced, or we might say that a word becomes more specialized'
(Arlotta 1977:178). If we take a word like mubomo < Afr.boom (tree), in its source
language this word refers to ·a tree', but in Venda it refers to ·a bluegum' tree only.
Likewise, the word muneri < Afr.meneer (male church official) refers to 'any male
. "
person in its source language. The narrowing of meaning of this word can be
presented in a diaaram as follows:

A. Mubomo

B. Muneri
"

The following tables also show instances of the narrowing of meaning of adoptives:
 99

ENGLISH: BEER VENDA: BIA

Alcoholic drinks made from hops or Brewery beer only (Traditional beer
malt. not included)

ENGLISH : RAILWAY VENDA: RALIWEI

""

Railways Railway company (The meaning of

Railway company railways is not included)

The last example is quite interesting. The meaning of ·railway lines' is not included
. -
in the adopted word raliwei.
,., Instead, the adoptive word tshiporo from Afrikaans
word 'spoor' (railway line) is used in this regard.

2.3.1.4 Radical change (shift) of meaning.

louwrens (1993:14) says the following with regard to radical change of meaning to
adoptives:

"It may happen that the semantic content of a loan-word shifts so radically
from what the word originally meant, that only a meagre semantic relationship
between the original word and the loan-word can be observed".

The difference between radical shift and other changes of meaning such as
meaning extension and narrowing of meaning is a matter of degree. Radical shift of
meaning can be clearly seen from the adoptive J!dzhagane. This word has been
probably adopted from the Afrikaans word 'diaken' (deacon). A deacon is ·a lay
official who assists the minister, esp. in secular affairs' (Collins Concise Dictionary).
The adopted form of this word in Venda refers to ·any person who is a christian'.
The word mudikoni is now used in Venda for ·a deacon'. Another interesting
 100

example of radical shift of meaning is the adoptive noun tshigidi from the Afrikaans
word 'skiet'. In Afrikaans the word 'skiet' means to shoot, but when adopted into
Venda it means 'a gun'. Although the meaning between these two words is still
related the radical shift is in the word category. The word 'skiet' is a verb whereas

tshigidi is a noun. Both words have related meanings despite the shift. Louwrens
( 1993:14) explains this point as follows:

" .. whatever the extent of the shift, there will always remain some semantic
connection between the borrowed word and its counterpart in the language
from which it was taken over"

These changes, as well as others, are illustrated in the following tables:

AFRIKAANS: DIAKEN VENDA: LIDZHAGANE

....
Church official who attends to secular Any person who is a Christian
affairs of the church

AFRIKAANS: SKIET VENDA: TSHIGIDI

Shoot A gun

ENGLISH: FOX VENDA: LIFOGISI

A

A person who is cunning and sty A detective

ENGLISH: MARCHING IN THE LINE VENDA:MATSHINGILANE

People who march in a line Nightwatchers

 101

AFRIKAANS: VANMELEWE VENDA: MULIVHE

"'
Long ago Forever

ENGLISH: LIMITED VENDA: LJMINOIOI

""
A type of a company A wealthy person/ Person with
unlimited resources

2.3.1.5 Changes in emotive value

The emotive value of words can be described as the words "capacity to produce a
certain emotional effect upon the hearer or listener" (Lyons 1977:175). When words
are taken over from one language to another, they may change their emotive
value. Louwrens (1993:14) has this to say in this regard

" ... borrowing often goes hand in hand with changes in the emotive value of
loan-words, i.e. it causes a shift in the emotional effect a word has in its
language of origin on the one hand, and in the borrowing language, on the
other".

The emotive value of words can change to express either appreciative connotation
or pejorative connotation. The first type of change is also referred to as
amelioration. This type of change can be shown by the adoptive mesitirese
..... from
the English word 'mistress'. In English this word originally referred to 'a woman at
the head of the household or family I woman in authority who gives orders to the
servants'( Hornby 197 4:562). In Venda the word mesitirese
,... refers to 'a lady
teacher'. The connotation of 'a housewife' in this case is completely excluded in
Venda. Instead, a specific adoptive for a domestic servant is used viz. gele
/'\
from
English word 'girl'. In English the word 'girl' is used to refer to 'female child;
daughter; young woman; woman working in a shop, office etc' (Hornby 1974:363).
 102
The fact that the adopted word gele refers to a female domestic servant only
A

shows that a pejorative connotation has been added to the word. Another example
of pejorative shift is found in the word 'boyi' (an adult labourer) from the English
word 'boy'. The word ·boy' refers to a 'male child up to the age of 17 or 18'
(Hornby 1974:98). The English word 'boy' has therefore acquired some pejorative
connotation when taken over to Venda. The change in the emotive value of the
words discussed above can be clearly illustrated in the following tables:

ENGLISH: MISTRESS VENDA: MESITIRESE A

Woman associated with domestic A lady teacher (Meaning of a domestic

services. servant is excluded)

ENGLISH: GIRL VENDA:GELE A

A young female person (under 16 An adult domestic female worker

years of age) (particularly one who works for whites)

ENGLISH: BOY VENDA: BOYI

Any male young person ( under the An adult male labourer ( particularly
age of 17 or 18) for whites)

2.3.2 Semantic adaptation of native words to adopted concepts.

Native words can undergo some semantic adjustments due to the influence of
foreign languages. In such cases the changes here are mainly brought about by
the new concepts introduced into the language. According to Ullmann (1959: 170)
 103

·a semantic change will occur whenever a new name becomes attached to a sense
and I or a new sense to a new name.' Here we shall be looking at the way in which
the meaning of native words is modified to accommodate new foreign concepts
which have entered the speech community. The most common way in which native
words are modified to . accommodate foreign concepts is by extending their
meaning range or area of meaning (The meaning extension of words has already
been discussed, 4.2.3.1.2). Native words in Venda, like the adopted words, may
have their meaning range extended to accommodate new senses or concepts
brought into the language. Just like adopted words, native words while accepting
the new meaning, retain their original meaning. The word mudi
......
which means ·a
house or traditional village' can be used to refer to ·a town or city'. In this case the
meaning range of the word mu~i has been extended to accommodate a new
meaning. Another good example is the word maine which means ·a traditional
healer'. The meaning of this word can be extended to mean 'a medical doctor'.
Another example is the word riwedzi. This word refers to 'the moon'. The meaning
of this word has now been extended to mean ·a month'. What happened here is
that when the modern calendar was introduced with a certain number of days
constituting a month, the Venda people who relied on the appearance or shape of
the moon as their calendar, then extended the meaning of the word nwedzi to also
rt:;lfer to 'a month'. The word musanda is used in Venda to refer to ·a house of a
chief. Today the meaning of this word has been extended to refer to ·a capital city'.
For instance, _,
Thohoyandou and Pretoria may be regarded as the capitals or
· misanda' of Venda and South Africa respectively. The meaning of the word
musanda has in this instance been extended to include that of a modern capital
city. However, the modern capital city does not necessarily mean the dwelling
place of a king or chief, but a centre of administration. The following are other
examples of native words whose meaning range have been extended:

Native :Original and adopted meaning

vhurala :-traditional bed made of wood

-modern bed
 104

khoro :-traditional court at the chiefs place

-modern court
halwa :-traditional beer
-modern beer in cans or bottles
sosa :-traditional place where beer is made
-modern brewery
mutshaeli :-a person who controls spanned oxen
-a person who drives a vehicle
muhatuli
,... :-a person who judges cases in a traditional court
-a person who judges cases in a modern court

In each of the examples given above there are two meanings. The first meaning in
each case represents the original meaning of the word. The second one is the
additional meaning to the native word. The meaning of the native word has been
extended to cover the new meaning introduced into the speech community.
Although in some instances the adopted words may be used instead of the native
words, the purists often prefer to use the native word with extended meaning. For
instance, the purists may prefer to use the word mutshaeli ·a person who controls
yoked oxen' to refer to a 'driver' instead of an adoptive .....diraiva .

2.4 The effects of adoption on the Venda lexicon and semantics.

Studies on adoptives (e.g.Weinreich 1953) have shown that adoptives can affect a
language in various ways. The influence of adoption on a language can be either
negative or positive. With regard to the latter, adoption may result in the
enlargement of the lexicon of a language, enriching the language with synonyms
and other words that have more expressive meaning. On the other hand adoption
may cause the obsolescence of native words, and the problem of homonymy in a
language. These different effects of adoptives on the Venda language will be
discussed in detail in the sections below.
 105

2.4.1 The enlargement of the lexicon of a language

When a new object, activity or idea enters a culture, the word or words which
express it may be adopted (Hoffer 1980: 12). The introduction of these new lexical
items enlarges the lexicon of the language. With time, some of these words
become completely assimilated into the adopting language. Hundreds of new words
have been added to Venda through. this process of adoption.

2.4.2 Enriching the language with synonyms

The adoption of foreign words into Venda has resulted in many synonyms in the
language. "The term synonymy is used in semantics to refer to the phenomenon
whereby words are believed to have the same meaning" (Poulos 1986:30). Nida
(1969:73) defines synonyms as

"words which share several (but not all) essential components and thus can
be used to substitute for one another in some (but not all) contexts without
any appreciable difference of meaning in these contexts".

This definition brings us to the questi.on of whether real synonyms do occur in a

language. Real synonymy can be regarded as an instance where two words which
appear in the same language have exactly the same meaning and can be used
interchangeably in all contexts without the slightest alteration either in cognitive or
emotive import (Ullmann 1963:108-109). From Nida 's definition it ls clear that real
synonyms are very rare in a language. This view is supported by Palmer (1976:60)
who says:

"It can, however, be maintained that there are not real synonyms, that no two
words have exactly the same meaning. Indeed it would seem unlikely that two
words with exactly the same meaning would both survive in a language".
 106

Mokgokong (1 975:45) also rejects the view of synonyms as words which are
exactly the same in all respects:

"Synonyms are therefore words of similar significance in the main, but with a
certain dissimilarity as well; with very much in common, but also with
something private and particular, which they do not share with one another.
They are not on the one hand words absolutely identical in meaning; but
neither on the other hand only remotely related to one another".

For our study we shall be mainly concerned with the synonymous relationship
between adopted and native words and also the synonymous relationship between
adopted words themselves. We shall begin by looking at those instances where
adopted words enter into synonymous relationship with native words. Let us
consider the following examples:

ADOPTIVES SOURCE LANGUAGE NATIVE WORDS

sibadela Afr. hospitaal vhuongelo 'hospital'

nese Eng.nurse muongi
modoro
,... Afr.motor tshiendedzi 'vehicle'
thiibii Eng.T.B lufhiha 'tuberculosis'
thoilethe Eng.toilet bung a
mungane Tso.mungana khonani 'friend'
sivhara Afr.swaer malume 'brother-in-law'
muhalivho NS.mogadibo muvhuye 'sister-in-law'
mungome Tso.mungome nanga 'diviner'
dokotela Afr.dokter nanga 'doctor'
tshifevhi Zu.isifebe phombwe ··adulteress'

,....lisuhana NS.Iesogana muthannga 'young man'

A

phutheo NS.phutego tshivhidzo"congregation"

morodza Afr.more vusa 'greet'
 107

The examples given above illustrate the synonymous relationship between adop-
tives and native words. The adoptive word sibadela (hospital) has the native word
vhuongelo as its synonym. However, the meaning between these words, though
related, is not exactly the same. The word sibadela refers to 'a hospital' whereas
the word vhuongelo may refer to 'any place where a sick person is looked after'.
Another example is the adoptive nese (nurse) which has muongi as its synonym.
Although both of these words may refer to 'a nurse' the latter example may be
used to refer to 'any person, not necessarily a trained person, who looks after a
sick person.' A synonymous relationship is also realised between the adoptive
thoilethe and the native word bunga. Although these two words refer to a place
where people excrete, there are some differences in their meaning. The adopted
word thoilethe is used to refer to the type of toilet that contains water whereas
bung a refers to ·a toilet made of a hole dug into the ground and walls .built around
it perhaps with a roof. This type of toilet, because it does not use water, smells
and attracts flies, hence the word bunga, which is derived from the verb stem -
vhunga (to wipe away flies, in this context). The word, ndilo and phuleithi, although
they both refer to ·a container used for porridge' have slight differences in their
meaning. Ndilo is used to refer to ·a wooden food container' whereas the word
phuleithi refers to 'a metal food container'. Another example is that of the words
muhalivho and muvhuye. The word muvhuye refers to 'a sister-in-law ' who has
been married by money used as the 'lobolla' of a sister to the brother. The word
muhalivho has no such implication. It merely shows the relationship between a
sister to the brother and the wife. Although we have not discussed all the examples
given in the table, the above discussion illustrates the kind of synonymous
relationship which adoptives may enter into with native words.

As has already been mentioned, an adopted word may also enter into a
synonymous relationship with another adopted word. These synonyms may be from
the same source language or different source languages. The adoptive vhengele
(Afr.winkel) may enter into a synonymous relationship with the adoptive shopho
(Eng. shop). Another example is the Afrikaans adoptive · kamara' which is
 108

synonymous with the English adoptive · rumu.' The meaning in the two examples of
synonyms given above is axactly the same. Therefore these synonyms constitute
real synonyms in Venda. The following are other examples of real synonyms from
different source languages:

silaha (Afr.slaghuis) : butshara(Eng.butcher)

swigiri (Afr.suiker) tshugela (Eng.sugar)
. tshitolo (Eng. store) shopho (Eng.shop)
shopho (Eng.shop) vhengele (Afr.winkel)

The meaning of the words of each pair given above is exactly the same and so
therefore these words may be regarded as true or real synonyms.

2.4.3 Loss of native words

Although the adoption of foreign words enriches a language, it may also lead to the
loss of some native words. This loss of words in a language is commonly known as
obsolescence. According to Arlotta (1972:169) obsolescence

"takes place when the speech community ceases to use the item in question
and as a result the word falls out of everyday usage ... For all practical
purposes, these words have left the language, and their current usage is
essentially restricted to antiquarians or historians of science".

Although there are various factors which lead to the obsolescence of native words,
adoption seems to play a very important role in this regard. The native words
become obsolete because of the introduction of better words into the speech
community through adoption. Sometimes words become obsolete because the
object or action they used to refer to has ceased to be of use to the speech
community and has been replaced by a new one. Nkabinde (1968:19) has this to
say in this regard (with reference to the Zulu language):
 109

" .. the borrowings found in the language are a record or register of concrete
objects or abstract ideas that have been incorporated into Zulu life due to
contact with foreign cultures".

The fact that there are many foreign objects that have been incorporated into
Venda society at the expense of native ones explains why many native words have
become obsolete in the language. If we take food, for example, we note that most
of the food products used in the olden days have been replaced by modern ones.
Food such as tshitagada,
.... ,.. tshigume, mufumbu, munamba, mutumbula etc. have
been replaced or are in the process of being replaced by modern food. The
replacement of these food products also leads to the replacement of words which
designate these products.

Lastly, a word may become obsolete because of its unfavorable connotations. The
replacement of words with negative connotations by adopted ones will be
discussed later in this chapter. Taboo words may also be replaced by adopted
words. This may lead to their permanent loss.

The prestige factor may play a role in the loss of native words. Speakers of a
language, particularly the educated ones, prefer the adopted words above the
native ones and this causes the corresponding native word to lose frequent use. As
a result of this, the native word becomes obsolete.

2.4.4 The problem of homonymy

"Homonymy is a term used to describe pairs or groups of words with the same
phonetic form but totally different meanings" (Mokgokong 1975:40). According to
Yule (1985:96-97) such words "have accidently come to have exactly the same
form". Lyons (1977:21) indicates that it is not just the difference in meaning which
makes two words to be identified as homonyms, but also their unrelatedness of
meaning. Although homonymy has been clearly defined here, several problems are
encountered when a distinction is drawn between homonymy and other sense
 110

relations such as polysemy. Polysemy is used to refer to a situation where the

same word may have a set of different but related meanings. The main problem
here is "how does one draw the line between homonymy (roughly 'two or more
words having the same pronunciation· and/or spelling') and polysemy ('one word
having two or more senses')" (Leech 1974:228). Ullmann (1957:127-128) illustrates
the problem of the distinction between homonymy and polysemy by the English
words 'flower ' and 'flour. These words were originally one word; 'flour' was
'flower the finest part of wheat. Therefore synchronically, these words may be
regarded as homonyms, and diachronically they are in fact polysemous. Lyons
(1977:550) also recommends that the historical derivation of words is important in
distinguishing between homonymy and polysemy. However, he regrets that this
criterion may not be always possible due to the large number of words which may
be under investigation. Sometimes the historical information needed may not be so
relevant to the synchronic study of the language.

Although some linguists like Lyons ( 1977b:552) feel that the theoretical status of
the distinction between homonymy and polysemy should be left unsolved, we hope
that the study of adoptives will throw some light on this problem.

In Venda adoption has resulted in many homonyms. These homonyms may be

categorised according to their similarity of shape with regard to tone. Some
adopted words have exactly "the same form (i.e. the same speech sounds and
tonal pattern) but with different meanings" (Mokgokong 1975:40). This type of
homonym is called homotonym. Other homonyms have corresponding segments
but differ in meaning and tonal pattern. This type of homonym is referred to as
heteronym. We shall begin here by looking at examples of homotonyms.

Homotonyms

p6sa - throw something

p6sa - post a letter (cf.Afr.pos)
 111
sala - saddle of a horse (cf.Afr.saal)
sala - remains of burned wood or coal

samba -curse (cf.Eng.samba)

samba - part of a car engine

d6r6bo' -town (cf.Afr,dorp)

/'.

d6r6b6 -venereal diseas (cf.Eng.drop)

A

tshidina -type of a fern plant

tshldina -brick (cf.Afr.steen)

fUru - lining of a cloth.

furu - food for cattle (cf.Afr. voer)

tshlkapa - type of a blanket

tshlkapa - sheep (cf.Afr.skaap)

This type of homonymy creates ambiguity in communication. For instance, the

meanings of the following sentences are ambiguous

(a) --
,... ha sala.
Mulalo o dzula ntha

(i) Mulalo sat on a saddle

(ii) Mulalo sat on the remains of burned coal

This sentence may have two meanings. The word sala in this sentence may be
used to refer to 'the saddle of a horse' or ·remains of burned coal or wood'. As the
above sentence shows, the meaning of this word can be ambiguous in some

contexts.

The problem of ambiguity in the usage of homonyms in tonal languages like Venda
 112
can be partially solved by means of tone. Tone may play an important role in
distinguishing words which have the same form but different meanings. Let us look
at the following types of homonyms which have different tones (i.e heteronyms):

Heteronyms

....libugu -big book (cf.Afr.boek)

Jibugu -big worm
"
The words in this pair of words constitute heteronyms in Venda.These words differ
in their tonal patterns. The first word begins with a low tone followed by two high
tones. The second word begins with two low tones and ends with a high tone.

gulu - intestine
gulu - bullet (cf.Afr.koel)

The words in the examples above are also heteronyms since they differ only in
tone. The first word has a low and a high tone, ·whereas the second word consists
of high tones only. The following are further examples of heterotonyms:

mub6m6 - bluegum tree (Afr.boom)

mubomo - the edge of a basket.

makete - market (cf.Eng.market)

makete - big skirts (derogatory).

dzhesi -jersey (cf.Eng.jersey)

dzhesi - type of grass

mugodi -mine (cf.Zu.umgodi)

mug6di (~~A)-gatherer (goods)
 113

2.4.5 Taboo and euphemism

Adoption may assist the language in replacing some of the taboo words in the
speech community. The term taboo, has according to Ullmann (1970:465)
originated, from the Polynesian language and was introduced to English by Captain
Cook. According to Ullman (ibid:465) this term 'has a very comprehensive meaning
but, in general, signifies that a thing is forbidden'. To avoid the usage of taboo
words in a language, adoptives may be employed. The verbal stem -nya (to pass
faeces} is taboo in Venda. Therefore an adoptive stem -kaka (from Afrikaans 'kak'
} is used in place of -nya. The word/stem -kaka is regarded as less offensive. The
use of a less offensive word in the place of the offensive one is referred to as
euphemism. Stern (1931 :331) describes this substitution of an offensive word by a
less offensive one as follows:

"A foreign word is substituted, which being less definite in meaning, is the
same degree less offensive. Beckman explains this by the circumstances that
the foreign word is more or less a blank, which obediently conforms to the
use we make of it, while the native word has numerous undesirable
associations with the offensive thing, and means exactly what it means".

The following are other examples of euphemisms in Venda

Munna o binya musadzi thavhani mulovha.

Munna o reipa musadzi thavhani mulovha.
(The man raped a woman yesterday at the mountain}

Luvhengo u na thumbu.
Luvhengo u na phurege.
(Luvhengo is pregnant).

In the examples given above the words binya and thumbu have been replaced by
 114

the adoptives reipa and phurege respectively. These adoptives are regarded as
less offensive compared to their native counterparts.

2.5 Conclusion

In this chapter we discussed the factors which have led or which lead to the
adoption of foreign words in Venda. In .discussing these factors, we specifically
referred to the need-filling motive, the prestige motive or a combination of both. An
attempt was also made to show the different ways in which foreign words were and
are adapted semantically into Venda. Here it was established that certain words
have meanings which correspond with that of the adopting language; it was noted
that certain words extend their meaning range whereas others narrow it. Some
words extend their meaning range figuratively. Some adopted wor~s also change
their emotive value when they are taken over into another language. Some words
acquire pejorative connotations, whereas others acquire ameliorative connotations.
It is interesting to note that some native words, too, extend their meaning to
accommodate new concepts which have been introduced into the speech
community. Lastly, the effects of adoption to the Venda lexicon were assessed.
 115

CHAPTER 3

3.0 ADOPTIVES AND THE SOUND SYSTEM OF VENDA

3.1 Introduction

When words are adopted from one language to another they appear to be
subjected to different kinds of adjustments, amongst others, phonological
adjustment. This chapter will attempt to describe and illustrate how adoptives are
adapted to the sound system of Venda i.e. to show the changes that occur and to
account for the factors that underlie such changes. To start with, a general
overview of Venda phonology will be presented. This will be followed by an
analysis of. the phonological adaptation of adoptives to Venda p~onology. The
focus here will be at adaptation on the segmental level, adaptation to the syllable
structure and tone assignment on adoptives. This chapter will conclude by looking
at the implications of adoptives to the Venda sound system.

3.2 The sound system of Venda

In this section we shall be looking at the sound system of the Venda language.
However, the intention here is not to make an indepth study of all the aspects of
the Venda sound system since much on the sound system of Venda has been
written by linguists such as Westphal (1946), Doke (1954), Van Warmelo (1989)
and Poulos (1990). Instead only those aspects which may help to throw some light
on our study of adoptives will be outlined in this section. We shall begin by looking
at the vowels.

3.2.1 Vowels

Vowels are defined by Poulos (1990:500) as "a category of sounds which are
produced with a relatively unobstructed air passage ... when vowels are produced,
the articulators are fairly wide apart ... " Vowels can be divided into simple and
 116
compound vowels or diphthongs.

3.2.1.1. Simple vowels

Venda, unlike other languages. such as Northern Sotho, English, Afrikaans, etc.,
only has five basic vowels. These vowels are represented phonetically and
phonemically as follows

1 a I : [a] as in [-amba ]
.speak'
I e I : [e] as in [ k'erek'e] 'church'
I i I : [i ] as in [-ima 'stand'
I o I : [:>] as in [-:>ra ] 'draw'
I u I : [u] as in [-dura ' expensive'
J
""
Two of these vowel phonemes, viz. I e I and I o I, have variants which are realised
phonetically as [ e J and [ o ] respectively. These realizations occur where such
phoneme segments are followed by a high vowel or a syllable with a high vowel.
Venda vowels can be clearly illustrated in the following chart

[ i] - - - - - - - , - - - - - - - - - - - - - - . , [u]

[e ]

[a ]

3.2.1.2 Diphthongs

"A diphthong is a two-part vowel sound which, as a matter of convenience, can be

 117

described as a sound which involves a movement of the tongue from one vowel
quality to another". (Poulos 1990:505). According to Ward (1948:111) the vowels
which combine should be so pronounced as to form one syllable if they are to be
regarded as diphthongs. The occurrence of diphthongs in Venda and in other
African languages such as Tsonga and Zulu, is still a matter of controversy.
However, Poulos argues for the occurrence of such sounds in Venda :

"The occurrence of diphthongs in Venda is particularly noticeable in a number

of aspect prefixes and, as far as can be ascertained, consist of a combination
of two vowel qualities, namely o and u " (Poulos 1990:505).

Diphthongs are according to him more noticeable in quick speech in aspect

prefixes like to9_ou, tou, konou ~tc. However, in normal or slow_ speech these
vowels do not come out clearly as diphthongs but rather as separate vowels. The
fact that diphthongs rarely occur in the language may in a way show that
diphthongs are not a common feature of the Venda language system. It is hoped
that a study of adoptives will throw some light on this question of the occurrence or
non-occurrence of diphthongs in Venda. It will be interesting to see how foreign
words with diphthongs are incorporated into the Venda sound system.

3.2.2 Consonant phonemes of Venda

In this section we shall be looking at the Venda consonant phonemes. Here we

shall be looking at the Venda phonemic inventory and its phonotactics. Let us
begin by looking at the phonemic inventory of the language.

3.2.2.1 The basic inventory

Every language has its own set of phonemes. The concept 'phoneme' will be used
here to refer to the sound segments of a language which have the capacity to
distinguish between the meanings of words. Lass (1987:357) has defined the
concept phoneme as follows:
 118

"A phonological unit serving a contrastive function; a minimal segment-sized

unit capable of distinguishing meaning."

This definition somehow answers the question of how to identify the sound
segments which constitute the phonemic inventory of a language. ·The distinctive
sounds or phonemes of a language can be identified by comparing pairs of words
which are phonetically identical in every respect except for one sound segment.
Such pairs of words constitute what is called minimai pairs (Poulos 1990:529).
From each pair at least two phonemes can be identified. Consider the following
examples:

u tala 'to swim' Ill

u tata 'to be sensitive' I tI

u mala 'to marry' I a I

u meta ·to germinate' I eI

In the examples given above, four phonemes have been identified, viz. /1/, It!, Ia/,
and lei by means of minimal pairs. This is attested by the fact that the substitution
of one phoneme for another in each pair results in a change of meaning of the
respective words. The phonemic inventory of the Venda language in accordance
with Poulos (1990:534) will be presented below.
 119

Bi DL D AL. Pa. Ve. AlB. LabP

ph pf th th ty kh tsw phw
"" t dy dzw pw
p bv t
I'
k
b f d d tsh g sw bw
I'

fh v I
A
ts dzh X zw nw
vh n
I'
dz sh h
m zh
r ny
s
z
n

The consonant phonemes in the chart above have been grouped according to their
place of articulation. These phonemes may be further classified according to the
manner in which they are articulated. However, 'it might, in fact, be argued that
phonemes do not have any phonetic properties because they are not units of the
phonetic level of representation: phonemes one might say, have nothing to do with
phonetics because they are abstract non-phonetic units" {Giegerich 1992:39).
Giegerich, although he maintains the validity of this statement also emphasises the
need for its qualification. In· order to classify phonemes or describe them, their
phonetic characteristics should be considered. This is supported by Ladefoged
(1982:245) who says that "each feature (phonetic feature) can be used to classify
the phonological oppositions - the phonemes - that occur in a language."

Some phonemes have different phonetic realisations. The phonetic realisation of a

phoneme is commonly known as an allophone. If we take the phoneme /m/ for
example, we will find that it is realised as the sound [m] in certain enviroments and
as [11J] in other enviroments. Therefore, these sounds are mere variants of one and
the same phoneme /m/ and as such they do not function distinctively.
 120
We will not pursue the discussion of the phonetic features of phonemes or their
phonetic realisation any further. This brief discussion suffices for the purposes of
this study.

3.2.2.2 Venda phonotactics

Here we shall be looking at the possibilities of combinations of distinctive sound

segments i.e. permissible strings of phonemes and the rules which govern such
combinations. The focus here will be on consonant combinations such as
consonants clusters. However, other aspects of phonotactics !such as distribution
restriction will also be discussed.

3.2.2.2.1 The complex consonants, compounds and clusters .

.Consonants in a language may be arranged in a way that they succeed each other
without having a vowel in between. "A language imposes certain restrictions on the
kinds of combinations into which speech sounds can be put, and every speaker
knows, albeit subconsciously perhaps, which sounds and which sound sequences
belong to his own particular language" (Poulos 1990:521 ). Sound segment
combination may result in complex consonants, consonant compounds and
consonant clusters. We shall begin by looking at the complex sound segments:

{a) Complex consonant sounds

Certain stop sounds in Venda are complex. These consonant sounds are mainly
affricates. In the articulation of these sounds there is "a sharp plosive character,
followed by an element of audible friction" (Poulos 1990:489). Examples of these
consonant sounds are:

tsh as in tshanda
A
'hand'
pf as in pfene 'baboon'
/'-

dzh as in u dzhena 'to get in'

 121
tsw as in tswio 'kidney'

Each of the sound segments given above appears to be comprised of more than
one sound segment. The sound segment tsh can be regarded as comprising
consonants, viz. t and sh. The question which is often raised by linguists is
whether the sound segment tsh should be regarded as a single phoneme or as a
compound or cluster segment. In Venda, these sound segments are articulated as
single segments. Therefore they can rather be considered as complex consonants
rather than compound or cluster consonants.

(b) Consonant compounds

Some consonant sounds in Venda form sequences which constitute compounds.

Consonant compounds may be formed by nasals or by a nasal and an oral s.ound.
It is interesting to note that when a nasal combines with another consonant it
always comes first. When some nasals occur before certain consonants they
constitute independent syllables. Nasals that constitute syllables are referred to as
syllabic nasals. These nasals may be influenced by the consonants which they
precede. Because of these influences they may have different phonetic
realisations. Consider the following examples:

nkunda ·defeat me' [IJ]

mmala ·marry me' [m]
nnyala ·dislike me' ijt]

In each of these examples the homorganic nasal N- constitutes an independent

syllable and is realised differently in each case. In the first example it is realised
as a velar nasal, in the second example as a bilabial consonant and in the last
example as a palatal nasal. It is also interesting to note that the homorganic nasal
N- has, in the examples above combined with oral stops and with other nasals.
The following are further examples of nasal compounds:
 122
i) Nasals and plosive consonants

Plosive consonants which can combine with nasal consonants include ejectives,
aspirated plosives and voiced plosives.

Ejectives: : ntota 'pinch me'

p : mpima 'measure me'
: ntama 'admire me'

Aspirated : th ntho 'wound'

kh nkho 'clay-pot'
ph mpho 'gift'

Voiced plosive: d ndowela

/'
'get used to me'
d ndina 'worry me'
b mbeba 'put me on the back'
g ngavha 'grab me'

ii) Nasals and affricates

The affricate sounds can also combine with nasals to form nasal compounds

tsh ntshipa 'strangle me'

pf mpfuna 'love me'
bv mbvula 'strip me'
dz ndzea ·marry me'
dzh ndzhena ·get into me'

iii) Combination with other nasals

n nne a ·give me'

"'
m mmona ·go around me'
 123

n nriala ·give up on me'

The homorganic nasal N-, however, cannot form compounds when used with
certain consonants. Some of these consonants are as follows:

I th, vh, f, v, I, s, z, I, r, sw, zw, sh, zh, h I

Where the homorganic nasal preceded such sounds there is always some form of
sound change. Hence the homorganic nasal is known to cause quite a number of
sound changes in Venda. Consider the following examples:

N + fhola :pholo 'healing' [ <P > [ ph ]

N + ramba : tharribo ·invitation' [r > [ th ]
N + swaya : tswayo 'signs' [ sw] > [ tswh]
' winnowing
N + sela : tselo basket' [s > [ tsh]
N + shaya : tshayo 'lack' [J > [ tJh ]

(c) Consonant clusters

Consonant sounds may form a sequence which constitutes a cluster. Consonant

clusters differ from the complex consonant sounds in that in their articulation, unlike
the complex sounds, there is more than one utterances. It appears that consonant
clusters . rarely occur in Venda. However, languages like English and Afrikaans
allow several consonant clusters, e.g. :

Eng.stamp
Eng.truck
Afr. knoop 'button'
Afr. straat 'street'

It is interesting to see how foreign words with consonant clusters are treated in
Venda. This will be discussed in detail later in this chapter (cf.section 3.3.1.3).
 124

3.2.3 Venda syllable structure

Linguists have defined a syllable in various ways. Dake (1954:207) defines a

syllable "as a sound or combination of sounds uttered with a single impulse of the
voice". According to same linguists such as Janes a syllable consists of a
sequence of sounds which contain one peak prominence. Paulos (1990:522)
regards these peaks of sonority as having same or ather correlation with the whale
question of tone. According to him every syllable in Venda is associated with tone.
In this section our main focus will be an the structure of the syllable in Venda.
However, the syllable structure of other languages, particularly those languages
which serve as source languages far Venda adoptives, will also briefly be
examined.

In Venda the syllable can consist of any of the following:

i) a vowel (V)
ii) a consonant followed by a vowel (CV)
iii) a nasal (syllabic nasal) (C)

i) A vowel

A vowel may constitute a syllable in Venda e.g

a-mba V- CV 'sing'
e-nd a V - CV ·go'
i-nga V- CV 'make beer'

ii) Consonant and a vowel

Some syllables in Venda consist of a consonant and a vowel. In same languages ,

syllables, particularly in ward final position, nd with a consonant only. Such
syllables are called closed syllables. In Venda, syllables generally end in a vowel
with the exception of same ideophanes:
 125

khu-hu :CV- CV 'chicken'

mu- n- na :CV- C-CV 'man'
kho- lo-mo :CV- CV-CV 'cattle'

Examples of ideophones are as follows:

tswo-rrr CV-CCC
bvu-mm CV-CC
nzi-rr CV-CC
thwa-la-111 CV-CV-CCC

iii) A nasal consonant

It was indicated in the previous discussion that some nasals constitute a syllable on
their own. Consider the following examples

m- pho C-CV 'gift'

m- mvi C-CV ·grey-hair'
n- ngu C - CV 'sheep'

It has already been mentioned that nasal consonants which function in this manner
are referred to as syllabic nasals.

In recent years the syllable has received a considerable amount of attention from
phonologists (Katamba 1989: 153). Many different approaches to the study of the
syllable have come to light. The branching or hierarchical approach was adopted
by many phonologists to represent and explain the syllable structure of a language.
Some phonologists such as Kiparsky (1979), Halle and Vergnaud (1980). Harris
(1983) and many others used the hierarchical branching approach in the framework
of multi-tiered phonological theory. According to this approach "phonological
representations are viewed as consisting of a number of independent levels that
are linked to each other" (Katamba 1989: 154). Much on this approach will be
 126

discussed in more detail later in this chapter (cf.section 3.3.1 ). According to the
traditional model of the branching approach, a syllable (6) is considered as
consisting "of an onset (0) and a rhyme (R); and the rhyme of a peak (P) and a
coda (Co" (Lass 1984:252). According to Lass the syllable constituency can be
represented in a BRANCHING TREE as follows:

1\
0 R

A
P Co

The branching model which represents the syllable structure so well is the CV-
model of phonology by Clements and Keyser (1983). This model has added a
great impetus to the the study of the syllable structure. This model has also been
used in the study of the African languages syllable structure {e.g. Khumalo 1982).
According to this theory a syllable is made of syllable trees. The syllable trees
consist of:

"three-tiered representations, in which each tier has a certain vocabulary

associated with it. The vocabulary of the first, or o-tier, consists of a single
element o. The vocabulary of the second or CV-tier, consists of two elements
C, V; and the vocabulary of the third, or segmental tier consists of single-
column phonetic matrices characterising consonants and vowels in the usual
manner" ( Clements and Keyser 1983:25) .

The version of the CV-model proposed by Clements and Keyser "has the
advantage of being conceptually simpler than the alternatives. Intervening between
the syllable node and segmental tier there is a 'flat' CV-tier, lacking internal
constituent structure. Contrast with the more complex syllable models with onsets
 127

and rhymes ... " (Katamba 1989: 158). The different tiers in CV~phonology are linked
to one another. Katamba (1989:158) describes the linking between tiers as follows:

"In CV-phonology, the linking is done using ASSOCIATION LINES which are
subject to a WELL-FORMEDNESS CONDITION. To relate the CV-tier to the
segmental tier, association lines are drawn following certain univeral rules.
Normally, these rules link V elements to [-cons] segments (vowels) and C
elements to [+cons] segments ... "

He then illustrates this representation with the following example:

. d

/\ /~
c v c c v

m
II u n t u ' person in Luganda'

According to this model, any segment dominated by a C-element of the CV-tier is

nonsyllabic while any segment dominated by a V-element is syllabic. However, as
has already been mentioned, a nasal on its own can constitute a syllable. The way
in which foreign words with syllables consisting of a consonant only and those
which end with a consonant are treated in Venda will be discussed later in this
chapter (cf· section 3.3.2.2).

3.2.4 Tone

Venda is a tone language. Poulos (1 990:564) defines a tone language "as a

language in which pitch variation plays a role in changing the meaning of a word
i.e. the meaning of a word can be changed by merely changing the pitch of the
word... He refers to the distinctive pitches or pitch levels as tones or tonemes
 128

Studies done on Venda tones (Westphal 1946, Mathivha 1966, Van Warmelo 1989 ,
Poulos 1990) have shown that Venda has two basic tone levels viz. high (H) and
low (L). Two other relative tone levels are a falling tone and a rising tone. Tone is
assigned to each syllable. Where a syllable consists of more than one element i.e.
a consonant and a vowel, tone is assigned to the latter.

The high and low tones of a word may be marked by acute and grave accents i.e
['J and ['] respectively.
Another way of representing tones in a word is by using
.
symbols such as H and L after a word (Poulos 1990:564). Tones may occur in
patterns in some words (Poulos 1990:564-565). According to him certain general
statements can be made about the tones which occur on formatives or words of a
language. Noun prefixes, with the exception of a few classes, are known to be
marked with a low tone. In this section we shall not make an in depth study of the
tonal patterns of different word categories in Venda. ·Enough work has been written
in this regard (cf Westphal 1946, Mathivha 1966, Van Warmelo 1989, Poulos 1990,
Cassimjee 1992). What we has been said here forms the basis for our study of
tonal adjustment on the adopted words in Venda.

3.3 Phonological adjustment on adoptives.

As already mentioned, adoptives seem to be subjected to different phonological

treatment when they are incorporated into the adopting language, in this case
Venda ... In this section we shall be looking at the different ways in which adoptives
in Venda are modified phonologically. However, before we discuss the phonological
changes which occur to adoptives in Venda, we shall present a brief outline of the
different explanations that have already been given by other linguists in this regard.
Most of these explanations are within the phonological theory framework. In
discussing the phonological adaptation of foreign words to Venda ·we shall first look
at the changes which take place at the segmental level. The focus here will be on
sound substitution. Secondly, we shall be looking at the adjustment which occurs
on the syllable structure of adopted words. The main focus here will be on the
 129

constraints against clusters and words with consonant-final syllables. Lastly we

shall look at the tonal adaptation of the adopted words to the Venda tonal system.

3.3.1 Attempted explanations of the phonological changes of adoptives.

"The analysis of borrowed lexicon has been one of the undecided areas in the field
of phonology" (Yavas 1982:123). As a result, there are as many explanations of
this problem as there are different phonological theories. The problem of the
phonological adaptation of adoptives can be traced back to the days of pre-
structuralism. Although pre-structuralists partly addressed the problem with their
phonetic approximation hypothesis, they did not solve it. The problem once again
captured the attention of Structuralists like Haugen, Bloomfield and Weinreich.
Nevertheless their explanations never satisfied the Generative phonologists such
as Hyman. Hyman's explanation was followed by a revised version, still within the
framework of the Generative approach, by Kaye and Nykiel (1979). The emergence
of the Natural phonology theory was also of great significance to the study of the
phonology of adoptives. Phonologists such as Lovins (1975), Bjarkman (1976) and
many others took pains to show how their theory could resolve the problems which
exist in adopted words. However, as we shall see later in this section, some of the
problems regarding adoptive phonology remained unresolved. Some linguists,
realizing the shortcoming of working within a phonological theory to address the
problems of adoptive phonology, decided to find solutions in terms of the adopting
language constraints, i.e. the adopting language is regarded as having certain
constraints which determine the nature of phonological change an adoptive word
will undergo.

In the following sections we shall be considering six different approaches in greater

detail under the following subheadings: phonetic approximation, phonemic
approximation, generative phonology, natural phonology, autosegmental phonology
and the recipient language constraints approach.
 130

(a) Phonetic approximation

This explanation is found in pre-structuralist literature. Hyman (1970:8) describes

this approach as follows:

"The form this argument usually takes is that speakers of a language, in

hearing a foreign sound replaces that sound with the most closely related
phonetic (or perhaps phonemic) unit in its inventory".

From this passage it is clear that the substitution of one sound segment for another
is explained in terms of physical phonetics. The speaker of the receiving language
will as a rule replace a foreign sound by the one which is the most closely related
to it in his language. This therefore, presupposes that some form_ of contrastive
analysis takes place in the mind of a speaker before deciding on the substituting
sound segment. This approach may lead to some generalisations on the nature of
substitution. Hyman (1970:9) has this to say in this regard: (NB.L 1 in the following
passage refers to the source. language and L2 to the adopting language).

"A contrastive analysis of the occurring phones of L1 and L2 should then

suffice in such a 'theory' to explain how (and demonstrate why) the phones of
L 1 will be realized by speakers of L2, and the phones of L2 will be realized
by speakers of L1".

The phonetic approximation theory seems to provide a satisfactory explanation for

some cases of sound substitution in Venda, particularly the substitution of sound
segments which do not occur in the language. There are a number of sound
segments in words adopted from English, Afrikaans and other African languages
which do not occur in Venda. Some of these sound segments, as it will be shown
later in this chapter, are replaced by native sounds segments on the basis of their
phonetic approximation.

Although this approach can satisfactorily account for the substitution of some sound
 131

segments, it is inadequate when it comes to others. Hyman (1970:9) demonstrates

the inadequacy of this approach by us.ing the example of the English sound [8]
which is realised in French as [s], and an English [0] as [z]. The substitution of
these sound segments in these two languages has a physical phonetic explanation,
[s] is the closest sound French has to English [8] and [z] to English [OJ. The
problem with this theory arises when another language, L3 comes into
consideration. Hyman (1970: 10) then compares how the English sounds [8] and [0]
. are adapted to the Serbo-Croatian language. In this language the English [8] is
realised as [tJ (and not [s]) and [0] as [d] (not [z]). Although substitution here seems
to be based on phonetic approximation, Hyman seems to be dissatisfied with the
fact that these sounds are not replaced by the same sounds in both French and
Serbo-Croatian. This in itself shows that there are other factors which are at work
beside physical phonetic properties. He therefore comes to _the following
conclusion:

"It appears more realistic to say that a language adopts that sound that is
'felt' to be closest to the prototype. Thus a Frenchman feels that English [8]
most closely resembles his own [s], while a speaker of Serbo-Croatian feels
that [6] most closely resembles his [t]. As soon as we accept this notion of
Sprachgefuhl we can no longer adhere to a physical phonetic interpretation of
foreign sound assimilation. The unequivocal conclusion that we are led to is
that foreign sound adaptation is mental in nature. The only way in which we
can explain why a sound X from L2 is realized as sound Y (and not as sound
Z) in borrowing language L 1 is by having recourse to the phonological facts of
L 1 and the phonetic data of L2. Differing phonological properties are then
responsible (at least in part) for different nativization processes." (Hyman
1970:11-12)

It is clear from this passage that Hyman rejects the phonetic approximation
approach in favour of the phonological explanation. The phonetic approximation
approach also suffers another setback in cases where foreign sounds that have
 132

equivalents in the recipient language are replaced by other different sounds. Thus
Lovins (1975:14) says:

"The phonetic approximation theory most obviously fails when there are
members of L 1 phonetic. inventory essentially identical to the Ls (source
language) sound in question, and demonstrably much more like it than
anything else in LT (recipient language), and yet an entirely different
substitution is made. This means that some more abstract aspect of LT
phonological structure is getting in the way ... "

Whilst it is true that this approach is not sufficient to account for all the
phonological changes (including substitution) which adoptives undergo, it cannot be
totally abandoned since "there are well-documented cases for which !10 satisfactory
explanation has been offered in terms of anything more abstruse than low-level
phonetic detail" ( Lovins 1975:12). This will become much clearer later in this
chapter.

(b) Phonemic approximation

The advent of structuralism, with its emphasis on the sound system of a language
rather than on the study of isolated speech sounds, resulted in a reinterpretation of
the phonological changes which occur on adoptives. According to this view,
"foreign sounds are not reanalysed as isolated phenomena, but instead fit into the
phonological system (as opposed to the phonetic system) of the borrowing
language" (Hyman 1970: 12). Haugen (1950:217), one of the phonemicists who
supports this approach, illustrates it with a Spanish adoptive estufa 'stove' which is
reproduced in Yaqui Indian as [ehtupa]. In this case the sound [h] has substituted
[s] and this is because the normal allophone of Is/ before It! and /kl in Yaqui, is [h].
Hyman describes Haugen's explanation as follows:

"Haugen should have been led to postulate that sounds are borrowed on the
basis of phonemic approximation, that is, a ·language first finds the closest
 133

phoneme that encompasses the phonetic quality of the foreign sound, and
then this foreign sound is appropriately phonemicised and subject to the
phonological constraints of that phoneme 's members ... Thus the lexicalized
form of Spanish estufa is /estupa/ with a phonetic realization of [ehtupa]. This
would account for why [h] was used, when I assume there was some other
element appearing in that context that would have more closely approximated
[s] phonetically".

The advantage of this approach over a solely phonetic one ''lies in distinguishing
underlying from surface levels of phonological structure (phonemes vs. allophones)
and in advocating the description of at least some of the distributional restrictions
on these different sets of units" {Lovins 1975:15). However, Lovins does not fail to
recognize the shortcoming of this approach

"But as a purely descriptive theory, taxonomic phonemics says little about

what should happen in practice in a language contact situation: what it is that
is perceived in Ls as significant, and how it is classified and produced 'in
terms of the units of LT'.

Hyman illustrates the shortcoming of this theory, especially with regard to its failure
to give general explanations to the phonological processes of a language, by giving
the following examples from Nupe:
v
1 [s .. ] [s ]
[z]
[ts ] ·
.·> [~ J
[c]
I 1 e

l
..;

[dz] [j J e

These examples illustrate the distribution of stridents in Nupe. A palatalization rule

which derives strident palatal sounds from strident alveolars results in the
appropriate allophones before front vowels as shown above. According to
phonemic approximation one would predict that foreign sequence [si] will be
analysed as [si] in Nupe. This, according to Hyman {1970:18) is not the case. The
 134

foreign sequence [sij is realised as [~i] or [syi]. This leads Hyman to postulate that
foreign sounds are perceived in tenns of undertying fonns, and thereafter, they are
subjected to the phonological constraints of the system of the recipient language.

Bjarkman (1976:314) describes the shortcoming of this approach as follows:

"In the final analysis, however, Structuralist approaches to borrowing (and to

most other questions in linguistic theory) fail essentially because they can't
provide a comprehensive enough model capturing the inevitable interplay
between phonemic (underlying) and phonetic (surface ) factors and
accounting for constraints in both the target and source systems".

This theory, therefore only describes the phonological changes_ that occur to
adoptives without showing how a speaker would actually map one linguistic system
onto another.

(c) Generative phonology

Another attempted explanation of the phonological · changes which occur to

adoptives is the one which operates within the framework of Generative phonology.
In this study we do not intend to make an in-depth study of this theory.
Nevertheless, we shall give a very brief background of the theory and then show
how it has been applied to the study of adoptives by some linguists.

Generative phonology (or standard theory as it was later called) was developed by
Chomsky and his collaborators. This theory was given its first full and authoritative
statement in their book 'The Sound Pattern of English' (1968). Katamba (1989:X)
explains the basic objectives of this theory as follows:

"The basic goal of generative grammar is to explore and understand the

nature of linguistic knowledge. It seeks answers to questions like: what does
knowing a language entail? Are there any properties of language that are
 135

universal i.e. is there such a thing as 'Universal Grammar'?"

From its conception Generative phonology was mainly focused on 'phonological

representation' and 'rules'. With regard to the former, two levels of analysis are
recognized, viz.the systematic phonemic level and the systematic phonetic level.
This causes the Generative approach to differ from that of Structuralism which
instead recognizes three levels of analysis which are the phonetic, phonemic and
morphophonemic levels. Anderson (1985:325) describes the differences with regard
to rules between the generative approach and its predecessors in the ·following
manner:

·"the emphasis of generative phonology broke with previous work from the
beginning by emphasizing the centrality of rules in a theory of language".

The phonemic levels of Generative phonology, though similar in some respects to

those of. structuralism, also have marked differences. The systematic phonemic
level corresponds to the morphophonemic level of Structuralism. The systematic
phonetic level is more abstract than that of Structuralism. In generative phonology,
allophonic variations fall within the systematic phonetic level. Rules play an
important role in deriving surface ·forms from underlying forms. This short
background on the Generative approach suffices for our study of· adoptives.
Whether or not, this theory is suitable for explaining phonol~gical changes which
occur to adoptives, is an issue that will be taken up later in this section. The first
significant attempt to explain phonological changes of adoptives by this theory was
made by Hyman. He uses adoptives to justify generative phonology.

As his point of departure, Hyman (1970) rejects both the phonetic and phonemic
approximation theory' whether fairly or unfairly' in favour of an explanation based
on Generative phonology. This approach, as we have already mentioned, differs
from others in the sense that it recognizes two levels, viz. systematic phonemic
(underlying) and systematic phonetic (surface) levels. According to him " foreign
sounds are perceived in tenns of underlying fonns" (1970:19). The source word
 136

would be subject to modification by the synchronic phonological rules of the

borrowing language (Steinbergs 1984: 119). Hyman justifies his analysis by
examples from Nupe. After analysing the phonological changes of adoptives in this
language he makes the following observations

"a word or formative when borrowed becomes part of the lexicon with a
phonological representation which is subject to the morpheme structure
conditions and then all of the phonological rules. These new words then
consist of phonological strings which are composed of underlying
phonological segments already in the language, and not new phonological
segments or segments which may be closer to the phonetic output in all
cases" (1970:21).

In this passage, Hyman emphasises the necessity of recognizing an independent

level of abstract phonological representations and an ordered set of Phonological
rules relating these abstractions to their concrete phonetic realizations. Thus he
further says:

"It is by recognizing a deeper level than the autonomous phonemic level and
by conceptualizing the phonological component as a system of rules relating
abstract underlying forms to surface phonetic realizations that borrowing can
be coherently handled" (1970:21).

Hyman illutrates this operation with the Spanish word estufa which was lexicalized
with a phonetic realization [ehtupa] in Yaqui. The underlying sound /s/ had been
transformed to a phonetic sound [h] by a phonological rule which changes the Is/
before /tl and /kl in Yaqui. This rule can be formalised as follows:

s -> h '---
 137

Hyman (1970) also demonstrates, with reference to Nupe adoptives from Yoruba,
how speakers of a native language perceive foreign sounds in terms of their
underlying forms. Here it is important to note the vowel systems of both Nupe and
Yoruba. Yoruba has a seven-vowel system, viz.:

e 0

while Nupe has a five-vowel surface system, viz.:

e 0

Hyman then introduces a context-free rule

[ +low ] --> [+back,-round]

which will account for the absolute neutralization of I e/ and /o/ as low surface
vowel [a] (Bjarkmann 1976:324). Bjarkman illustrates the application of this rule by
the following Nupe adoptives from Yoruba:

Yoruba Nupe
[keke] [kyakya] 'bicycle'
[egbe] [egbya] 'Yoruba town name'
[tore] [twarya] 'to give a gift'
[kobe] (kwabwa] 'penny'
 138

In this case Nupe speakers can be assumed to make the following analysis. Since
the vowels in Yoruba utterance [keke] can be recognized as acceptable underlying
forms, they are texicalized as such according to Hyman's Principle 1 which states
that foreign sounds ·are perceived in terms of underlying forms in the native
phonological system.

Although Hyman's generative approach accounts for some of the phonological

changes which occur in adoptives, it is not without some flaws. The first
shortcoming lies in the theory itself. The question of abstractness has caused much
criticism of the Generative approach. The main problem here is the discrepancy
which sometimes exists between the underlying and the surface representation.
Kataniba (1989:134), though an adherent of Generative phonology, raises some
thought-provoking questions with regard to abstractness in phonologi_9al theory:

"The controversy centres around the extent to which phonetic realisations can
differ from underlying representations. Is there a principled way of determining
when underlying representations are too far removed from surface
manifestations to be credible? ... Phonological theory would become empty if
any sound was allowed to change arbitrarily into any other sound at the whim
of the analyst" (1989:134-135).

He further explores the possibility of finding the mechanism which may be used to
determine the degree of relatedness between the underlying and surface
representations:

"This begs the question of the degree of relatedness that is sufficient to allow
the derivation of phonetic representations from the same base form. Can
effective mechanical procedures that determine which forms are derivable
from the same synchronic source ever be devised? Unfortunately, the answer
is 'no' (ibid. 135).
 139

According to him "abstractness is a scale rather than a dichotomy. It is not possible

to establish the absolute cut off point beyond which the degree of abstractness
becomes intolerable" (ibid.150).

The fact that Generative theory has no mechanisms to constrain abstractness

makes it difficult to reject whatever an analyst deems as the underlying form of a
surface representation of an adoptive form in a language. In some studies of
adoptives the situation is exacerbated when the underlying form has been
deepened to account for the maximum possible surface representations. Bjarkmann
(1976:324) criticizes Hyman 's treatment of Nupe with regard to this fact:

·"Hyman's treatment of Nupe has received recent notoreity (especially as

developed in Hyman 197Gb) to the extent that he has employed his loanword
data in the attempt to justify an abstract solution to Nupe phonology".

This positing of abstract underlying forms which may sometimes be remotely

related to the derived surface form weakens the Generative explanation to the
problem of adoptive phonology.

Lovins (1975:19) also criticises Hyman's approach on the ground that it deals only
with foreign words which have become completely nativized. It does not account for
the words which have not yet been completely assimilated. In such cases Hyman
takes refuge in 'rule exception without describing what type of exceptions entail
what sort of foreignness' (1970:19}. Although Steinbergs (1984} found Hyman's
explanation useful for some cases in Tshiluba, an African language spoken in
Zaire, he takes issue with it with regard to other examples:

Fr. (mustike:r] > Tsh. [m usitekele] ·mosquito net'

Ki. [polisi] > Tsh. [mpoluse] 'policeman'
Fr. (soset ] > Tsh. (nsesete] socks'
Fr. (muzet] > Tsh. [muse:te} 'box, bag'
 140

Normally Tshiluba alveolars are palatalized before [i] not [e], yet in these examples
. alveolars are only palatalized before (e]. This makes it difficult to apply Hyman's
explanation since there is no synchronic phonological rule in this language which
accounts for this change.

This failure on the part of Hyman's hypotheses has led some phonologists such as
Kaye and Nykiel (1979) to refine generative phonological theory of adoption by
stressing the importance of deep phonotactic constraints in a language. "Certainly it
is evident that phonotactic constraints (whether deep or surface) must play a large
part in modifying the shape of some loanwords" (Stein bergs 1984: 120).

(d) Natural Phonology

The failure of standard generative phonology to provide satisfactory explanations to

the phonological changes which occur in adoptives has made some linguists take
recourse to Natural phonology. Natural phonology arises from the need to find
natural or more transparent explanations of the language processes instead of
using abstract rules. This theory was founded by Stampe. The basis of the theory
is the claim that our innate phonetic capacity can be represented in the form of a
set of very general natural processes. The trademark of this phonological theory is
the distinction between rules and natural processes. According to this theory,
natural processes and not learned rules, are the ones which govern our phonetic
behaviour. Since phonological processes form the basis of this theory, it would
perhaps be expedient to cite what are processes in Stampe's view. Stampe
( 19 72: 1) defines a phonological process as

"... a mental operation that applies in speech to substitute for a class of

sounds or sound sequences presenting a specific common difficulty to the
speech capacity ... an alternative class identical but lacking the difficult
property".

Stampe argues further that although the substitutions are mental in nature, they are
 141

motivated by the physical character of speech. In other words processes are

defined according to their function which is "to maximize the perceptual
characteristics of speech and minimize its articulatory difficulties" (Stampe 1972:9).

As already mentioned this theory distinguishes between processes and rules:

"Natural processes are constraints which the speaker brings to the language,
rules, on the other hand, are constraints which t11e language brings to the
speaker" (Yavas 1982: 123).

Yavas further explains the distinction between natural rules and natural processes
as follows:

"Learned rules need not express phonetically transparent alternations or

minimal substitutions. They are always context-sensitive, are not
synchronically productive, and do not apply to nativize loanwords. Natural
processes, on the other hand always have phonetic motivation, they may be
context-free, they can make only minimal substitutions, and do apply to
loanwords. As to the order of application, all natural processes apply after
learned rules" (Yavas 1982: 124) .

From the above discussion it becomes clear that .natural processes apply to
adoptives. These natural procecces differ from rules in that they are productive.
One of the objectives of Natural Phonology is to limit the global status of rules
which in Generative phonology resulted in so many exceptions, especially with
regard to adoptives. Global rules are known to add too much power to grammars
(ibid.128). Natural Phonology differs from the Generative approach on the question
of perception of foreign sounds. According to Bjarkman (1976:353), "loanwords are
taken to be perceived in terms of 'natural phonemic' and not 'systematic phonemic'
representations". Thus it is natural processes, not rules, that determine the
nativization of the adopted words. It is interesting to note how these processes
function according to Natural phonology. This can be illustrated by the following
 142

examples which show the changes that occur to Arabic adoptives in Turkish (these
examples have been taken from Yavas 1982:125):

Arabic Turkish
/kita:b/ [kitap) 'book'
/istibda:d/ [istibdat] 'despotism'
/imda:d/ [imdat] 'help'
/intixa:b/ [intihap] 'choice'
/muha:rib/ [muharip] ·'warrior'
/mutana:sib/ [mutenasip] 'proportional'

These examples show a phonological process in Turkish which devoices stops in

word final position. This process takes place without exception. The phonological
processes in this approach apply sequentially (i.e. apply one after another) and not
simultaneously (i.e. they do not apply at the same time) (Steinbergs 1985:113).

Although Natural Phonology theory might be useful in explaining some problems in

adoptive phonology, it is also not without flaws. Even Bjarkman, who did an
intensive study of this theory and its application to adoptives, admits that "certain
inherent flaws remain unrectified in Stampe's pioneer model". There are a few
inherent problems with regard to this theory, viz. the distinction between rules and
processes, ordering relationships between processes, and its exclusiveness (i.e.
the exclusion of unmotivated and morphological alternatives).

Let's begin by looking at the problem of the distinction between rules and
processes. Although the distinction between rules and processes is the basis of the
Natural Phonology theory, such a distinction does not appear to be clear-cut,
especially when it comes to adoptives. The problem here is the fact that "certain
constraints do not lend themselves easily to such classification" (Yavas 1982: 130).
According to him "the rule/process distinction appears to be too simplistic to
account for certain facts of borrowing" (lbid.130). Yavas illustrates the problem of
the distinction between rules and processes by citing the case of vowel harmony in
 143

Turkish. Although in some languages vowel harmony is a natural process, in

Turkish it is a rule according to 'Natural Phonology' since it does not apply
consistently. Yet to adoptives it applies very consistently with phonetic motivations.
In this case vowel harmony suits the· definition of it being a process. There are
many such instances, as we will see later in this chapter.

"The rule/process identification would run into other problems when we consider
the ordering relationships of the constraints in accounting for the borrowing
assimilations" (Yavas 1982: 127). As we have already indicated, processes occur
sequentially and they occur after the learned rules. Yavas uses the same case of
vowel harmony in Turkish which, according to the theory, is a learned rule. Vowel
harmony, as can be seen from the following examples, takes place after vowel
epenthesis, which is a natural process.

Arabic Turkish
/jism/ Oisim] 'substance'
/fikr/ [fikir] 'idea'
/xayr/ [hayir] 'good'
/qahr/ [kahir] 'anxiety'
/kibr/ [kibir] 'pride'
/sabr/ [sabir] 'patience'
/tawr/ [tavir] 'manner'
(Yavas 1982:126)

In these examples vowel harmony (which is a learned rule) takes place after vowel
epenthesis (which is a natural process}. This, according to Yavas, undermines
natural phonology's claim that rules only occur before processes. In the examples
given above the learned rule appears to come after the natural process. This
hypothesis therefore still needs some modifications with regard to the rule/process
ordering.
 144

The third problem relates the exclusiveness of the theory. As already discussed,
only processes are considered for the nativization of adoptives. The issue of
unmotivated and morphologically motivated alternations falls outside the scope of
this theory. Anderson (1985:345) expresses his disatisfaction with regard to this
point as follows:

"If we attempt to limit the explanatory domain of phonology to the set of

natural processes in Stampe's sense, the result is that any alternation which
is not phonetically motivated or shows phonetically arbitrary properties,
immediately falls into the category of learned rules and thus fall outside of the
theory. In that case however, a great deal of the descriptive content of the
sound systems of natural languages, indeed, nearly all of it is not describable
as 'phonology' in this sense at all".

From this passage it becomes obvious that not all the problems of the nativization
of adoptives can be fully accounted for by a theory such as natural phonology,
which restricts its explanation to phonetic motivated processes only. linguists like
Anderson (1985), Bach and Harms (1972) regard the search for naturalness purely
on phonetic grounds as a futile excercise. Other domains may contribute
significantly towards solving the problem of the phonological adoption of foreign
words in a language. According to Steinbergs (1985:90) social, cultural, and
historical factors just as purely linguistic ones, also play a role in the modifications
of adoptives.

(e) Autosegmental phonology

Autosegmental phonology was an attempt by some phonologists to break away

from "the monolithic nature of uniformly segment-based theo.ries" (Anderson
1985:347). Some of these theories have been discussed in this section. Katamba
(1989:196-197) describes the differences between these theories and
autosegmental phonology as follows:
 145

"In autosegmental phonology the focus shifts to ways in which phonological

rules can change the organisation of phonological representation. In
autosegmental phonology, phonological representations are no longer seen
as simple rows of segments, with all phonological processes taking place at
one single level. Rather, they are regarded as complex arrays (in principle of
independent) elements arranged on different levels or TIERS. The different
levels though interconnected, are in principle autonomous".

The autosegmental approach brought about innovation with regard to phonological

represenation. In Anderson's words it is "the elaboration of richer notions of
phonological representation, going beyond the traditional picture of forms as
composed at all levels of sequences of segment-sized units" (1985:342). Giegerich
(1992) also emphasises the need for such phonological representations which are
more highly structured than those used in linear approaches. Thus he says the
following in favour of the autosegmental approach:

"Indeed, recent phonological theory (under headings such as Dependency

Phonology and Autosegmental Phonology) is notable for such an interest in
(and proliferation of) representational devices, and in particular concerning the
· structure of 'segments'-whatever those are" (1992:323).

Lass (1992:1) illustrates the phonological representation in Autosegmental

phonology with the word 'cattle'
 146
H

I
Tonal Tier

I
s w Foot

I Syllable

0
;\ 6""'A R
Syllable-constituents

~ N Co

C
I I I I
V C ¢ ,V C CV-Tier

I I
/k a
I
t
I. \
. a, I] Segment Tier (Phonemic)

I
[kh re
I I
t
I \
a · :!:] Segment Tier (Phonetic)

(Lass 1992:1)

In linear phonology (i.e. where· phonological representations are seen as simple

rows of segments) it is not possible to show all the phonological aspects of a word
in one representation as in the example given above. In this example the different
elements on each TIER associate with each other by means of association lines.
However, there are certain rules that govern such association. As already
mentioned, phonological processes take place at different levels or tiers instead of
taking place at one single level or row of segments. In this way it is easier to show
the different phonological changes that occur in a word.

This approach has been used by some linguists to explain phonological processes
such as vowel and consonant harmony, palatalization and many others (cf.
Katamba 1989; Khumalo 1987) in language. Khumalo (1987) employs this theory to
explain some of the phonological changes which occur on Zulu adoptives. Here he
established that the substitution of certain foreign sound segments in Zulu is
 147

determined by the harmony process which can be vividly explained by the

autosegmental approach. Another innovation brought about by this theory concerns
tone. Katamba (1989:201-202) says the following in this regard:

"In autosegmental phonology, given the starting assumption of the autonomy

of tone, the existence of such free, unassociated tones is not surprising. But
the same facts cause severe embarrassment in the theory of linear phonology
where tone features are assumed to be an integral part of the segmental
representation of tone-bearing units such as vowels because the prediction
that tone is the segmental feature is not borne out by the evidence. The
existence of floating tones was one of the original arguments used to justify
autosegmental representations".

Although this study will not make a detailed study of tone, it is hoped that this
approach will provide some explanation of the tonal adaptation of foreign words in
Venda.

(f) Recipient language contraints approach

From the aforegoing discussion it appears that the phonological adjustments of

adoptives have created problems for most of the phonological theories proposed so
far. In some instances, as has been discussed above, it is the shortcoming of the
theory that complicates the analysis of the changes rather than the processes of
adoption themselves. It is for this reason that some linguists like Holden {1976),
Yavas (1982), Steinbergs (1984, 1985), and Kaye and Nykiel (1972) decided to
seek explanations in the constraints of the recipient language. Yavas (1982: 130)
describes this approach as follows:

"What seems to be a better approach is to look at the native language

constraints as having different degrees of strength, and such a strength
hierarchy among the native constraints is reflected in loan adaptation" .
 148

He further mentions that

" ... an approach which considers the native constraints as a continuum

according to their synchronic strength (productivity) would be better equipped
to account for the facts than the one which commits itself to a binary
rule/process distinction based on phonetic naturalness" (ibid.131 ).

This kind of an approach has been used by Steinbergs (1984, 1985) in the analysis
of adoptives in Tshiluba, a language spoken in Zaire, and Oshikwanyama a
language spoken in Angola and Namibia. It seems that most linguists accept the
fact that there are constraints in a language which determine the nativization of
adoptives. Holden (1 976) gives a lengthy discussion on how such constraints in the
recipient language modify the newly. incorporated words. He discusses these
constraints under a hypothesis which he labels the 'The Magnetic Attraction
Hypothesis'. This analogy according to him implies

"that each phonological constraint of the target system exerts a 'magnetic'

pull on the appropriate segments of the borrowings, in order to assimilate
those segments to the native system. Different· constraints have differing
strengths of attraction, and the 'magnetic force' of each constraint acts
differently on different segments which· satisfy the structural conditions of that
constraint (as though different metals were being attracted by the magnet). I
further hypothesize that these varying strengths are a direct measure of the
relative 'productivity' or viability of phonological rules or constraints in the
native system, and provide us with a glimpse of how such rules begin to
decay 'from within', so to speak" (Holden 1976:133).

Holden raises a very interesting point with regard to the way in· which language
constraints work in the modification of adoptives. He shows that language
constraints operate in the form of a continuum. These constraints operate in
different degrees depending on different factors. Although this approach, may lead
to some generalisations on the nature of change, it avoids such general rules or
 149

processes advanced in theories such as Generative grammar. In opposition to

approaches which propagate explanations based on ordered phonological rules,
Holden favours one which regards adoptives as being "incorporated into a system
expressing constraints on the concatenation of surface phones" (Holden 1976:144).
Steinbergs's research on the Tshiluba and Oshikwanyama languages (1984 and
1985 respectively) shows how different constraints operate in a language. With
regard to the latter, he established the following hierarchy of constraints: The first is
the constraint against consonant final words; the second is the constraint which
results in substitution and the third and last constraint relates to processes which
derive from constraints against clusters. These constraints can be illustrated by the
following examples:

Source language Tshiluba

Eng. [farm) [ofalama) 'farm'
Germ.[kart:;)] [okalita] 'map'
Eng. [krismgs] [okilisitimisa) 'Christmas'
Germ.[brot] [ombolota) 'bread'
(Steinbergs 1985:95)

In these examples there is a constraint against the occurrence of foreign segments

in Oshikwanyama. The sound r and the 'schwa' vowel ~ do not form part of this
language's sound inventory. They are therefore replaced by I and a or i
respectively. These examples also illustrate a constraint against consonant
clusters. Consonant clusters -nn, -rt-, kr- and br- have been broken up by
epenthetic vowels. The first and the last examples also illustrate a constraint
against consonant final words in Oshikwanyama. As a result of this constraint
vowels are added onto the final positions of these words. Steinbergs's research
also established that some modifications to adoptives derive from deep structure
constraints rather than surface ones (1985:100).

Whilst this approach ·may appear to be very comprehendable since it operates

within the physical or phonetic realm, it does not clearly indicate how a speaker of
 150
the adopting language goes about modifying adopted foreign segments. However,
Silverman (1992) gives a very good description of how recipient language
constraints operate. He starts by identifying two levels at which constraints of the
native language operate on the incoming foreign words. The first level is what he
calls the Perceptual level. He describes this level as follows:

"The first level of loanword phonology consists primarily of a parsing of the

input signal into segment-sized chunks, for which native feature matrices
which most closely approximate their articulatory and/or acoustic properties
are provided" (1992:291 ).

Silverman hypothesises that there are certain constraints at this level which
influence the perception of the incoming forms by the native speaker. Thus he
regards the perception of phonetic material as language-set dependent, and as
such speakers will employ their language set when incorporating foreign words into
the native system. It is the native language system which determines how incoming
forms are perceived. The first constraint which he identifies at this level is the
constraint against foreign segments whose matrix does not exist in the recipient
language. In such a case the speaker of the native language will represent and
reproduce the native segment which most closely approximates the input in terms
of articulatory and /or acoustic properties. This constraint is referred to as the
native segment inventory constraint. Another constraint which operates at this level
is the tonal constraint.

Other adjustments such as the ones that occur on the syllable take place in the
Operational level. It is at this level .that perceived segments may undergo true
phonological operations triggered by native phonotactic constraints (Silverman
1992:293). These constraints account for the substitution of segments which are
perfectly acceptable to the phonemic inventory of the native language. The
changes in the syllable structure of adopted words may be caused by the native
syllable structure constraints and others. Some of these constraints may be
illustrated by the following example involving the Venda adoptive gireme <
 151
Eng.gram:

(i) The native segment inventory constraint which replaced the English vowel re
with the native vowel [e].
(ii) The native syllable structure constraint resulted in the insertion of the vowel
[ i ] between the cluster gr-.
This constraint also avoids the occurrence of a consonant in a word final
position, hence the vowel [e] is added to the final consonant m.
{iii) The tonal constraint requires the adopted word to conform to the language
tonal system, i.e. gireme.

More on these constraints will be discussed again in the sections that follow. The
levels and constraints discussed above can be represented in diagram form as
follows (extracted from Silverman 1992:293):

'
incoming · Perceptual level Operative Level Output
acoustic ~ Representation ' Representation H
input

Operative Level
Processes

, ...

Native segment Native phonotactic ·

and tonal constraints constraints and
preferences

For our study of adoptives it will be interesting to see how the constraints described
above operate in the different African languages of Southern Africa in modifying
adoptives from the same source language, for example, English and Afrikaans. As
 152
was indicated in Chapter 2, African languages have adopted a considerable
amount of words from these languages. It is therefore obvious that some of the
words from these source languages have been adopted in all of the African
languages spoken in the region.

Summary

In the aforegoing discussion different explanations of the phonological modification

of adopted words in a language· were scrutinized. As indicated, various
phonological theories have been employed in one way or another to explain this
complex process of phonological adaptation of foreign sounds to a language.
Firstly; we discussed the phonetic approximation theory. This theory seems to be
suitable to account for the substitution of foreign sounds which do not occur in the
native language sound inventory. Such foreign sounds are replaced by native
sounds which, phonetically are very close to them. However, some foreign sounds
which have equivalents in the recipient language may be replaced by different
sounds.

The phonemic approximation theory, as has already been indicated, seeks

explanations in allophonic variations. This explanation was found to be lacking in
that it cannot account for the phonological processes which are responsible for the
new resultant sounds in the native language. It was noted that this theory was too
simplistic. in trying· to explain the substitution of foreign sounds by allophonic
variation alone.

Another noteworthy contribution to the problem of the phonological adaptation of

foreign words comes from Generative grammar. The Generative approach
attempted to account for the phonological processes and niles which are
responsible for the adaptation of foreign sound segments to a language. According
to this approach foreign sounds are perceived in their underlying form. The
underlying forms are then transformed by means of phonological rules to
systematic phonetic forms in the receiving language. The problem with this
 153

approach lies in its abstractness in that some systematic surface forms seem to be
so different from the underlying forms from which they are said to have been
derived. In this approach a rule becomes an overpowerful tool which can be used
to transform any underlying form to the systematic phonetic form. The danger of
this approach to the study of the phonological adaptation of adopted words is that
segments which have been adopted from the source language may not be
accurately identified since a linguist can posit any segment as an underlying form
provided there are enough rules to justify its derivation. To constrain abstractness
in the generative approach, some phonologists like Stampe came up with a new
theory called Natural phonology. One good thing about this theory is its emphasis
on the natural processes which are a result of certain constraints on the part of the
speaker. However, the failure of this theory to provide a clear distinction between
rules and processes resulted in the exclusion of some morphophonological
processes which play a very important role in the phonological modification of
adoptives in a language.

Another explanation which was discussed is the autosegmental approach. The

advantage of this kind of approach lies in its representation. Certain phonological
changes which occur when foreign words are adopted into the native language can
be vividly described and illustrated. This approach can usefully account for the
adjustments of foreign words to the syllable structure of a language. This approach
may also be used in the study of tonal adaptation to the adopted words.

The final approach which was discussed is the recipient language constraints ·
approach. This approach has been advocated by linguists like Holden (1976),
Steinbergs (1984; 1985) and Silverman (1992). They regard the phonological
modifications to the adopted words as being due mainly to certain constraints in the
recipient language. Holden has indicated that these constraints have different
strengths. It is these constraints which prevent or allow the occurrence of certain
foreign sounds or sounds sequences in the adopting language. These various
constraints viz. the native segment inventory constraint, the native phonotactic
constraint, the native syllable structure constraint and the tonal constraint, which
 154

operate in the phonological adaptation of foreign words in Venda, will be

considered in our study of sound substitution, to be discussed in the following
section.

3.3.2 Sound substitution

When words are adopted from one language to another certain sounds may be
substituted by others. There are various factors which lead to the substitution of
sound segments in adopted words. Firstly, substitution occurs because of the lack
of the equivalant sound segments in the recipient language. The sounds and
forms which are incompatible with the sound segments of the adopting language
are replaced by native ones. This type of substitution seems to be triggered by the
native segment inventory constraint. In terms of this constraint foreign sounds
which do not occur in the recipient language sound inventory are barred from
occurring and are therefore substituted by native ones. However, the main
problem with this kind of substitution is "what principle does a borrowing language
follow in order to replace a source language sound which the borrowing language
lacks" (Steinbergs 1984: 118)?

The second type of substitution, and perhaps the most complicated, is the one
which affects sounds that are perfectly acceptable to the sound inventory of the
recipient language yet are substituted by others. This kind of substitution raises a
number of questions. In the first place it is difficult to· understand why foreign
sounds which are part of the recipient language's sound inventory are replaced by
different sounds. The second question relates to the problem of explaining the
factors which determine the choice of the substituting sounds. An attempt will be
made here to describe the factors which lead to substitution and those that
determine the choice of the substituting sounds. In discussing sound substitution
we shall first consider vowels and then consonants.
 155
3.3.2.1 Vowel substitution

Venda has only five basic vowels. Some of the languages from which Venda has
adopted words have seven or more basic vowels. In this section we shall only
consider the vowels from English and Afrikaans. Our choice of these languages is
motivated by the fact that most Venda adoptives have their origins in these two
languages. Morever, these languages have a wide variety of vowels. Some of the
vowels found in these languages are not found in the Venda vowel inventory. The
substitution of these vowels (i.e vowels which are not found in Venda) will be the
main concern of this study. Substitution occurs with both simple vowels and
diphthongs. We shall begin by looking at the substitution of simple vowels.

(a) Simple vowels

An attempt will be made in this section to discuss and illustrate the substitution of
the various vowel sounds in Venda adoptives. The focus here, as has already been
mentioned, will be on the substitution of English and Afrikaans vowels which have
no equivalents in Venda. Vowels from these languages which have replicas in
Venda and are therefore reproduced without any modification, will be ignored in
this study since their realisation is redundant Before discussing the substitution of
the various vowels we shall look at the English 1 and Afrikaans vowel inventori~s.

1. The English examples that will be used in this study will be based on the
Southern British Standard which is referred to as Received Pronunciation (RP). We
chose this British English dialect because (i) the English spoken in South Africa
originated from British English, (il) the RP dialect is widely 1:1sed and is well

documented (Lass 1987 :91 ).

 156
English has the following simple vowels:

i, 1, e (e), a, re, u, u, ::>, A, o, 3, a

The Afrikaans simple vowels are as follows:

2
Ia, ::>, re, e, a, i, o, u, e, ~~ y I

The substitution of simple vowels will be clearly illustrated in the examples that
follow. As mentioned above, we shall use adoptives from English and Afrikaans.
The vowels under consideration in the adopted words will appear in bold script.

1. The English vowel: re

The English· vowel re is realised as ( e ] or [ a ] in Venda.

stamp > tshitemmbe

,...

1. These vowels are based on the Received Pronunciation {RP). The symbols used
for these vowels, as well as the examples in which they occur, were taken from
various sources, namely the Everyman's English Pronunciation Dictionary {Jones
1956) , the Oxford Advanced Learners Dictionary, the Collins Concise Dictionary,
Giegerich 1992, Gimson 1970 and Roach 1983. The distinction between short and
long vowels is not relevant to this study, therefore this aspect of vowels has been
ignored in this study. In some cases the symbols as represented above are listed
in these works as phonetic symbols {i.e. between square brackets) and in others
as phonemic symbols {i.e. between slashes)

2. The Afrikaans vowels have been cited from De Vil!iers (1970) and 'Uitspraak-
woordeboek van Afrikaans' {by Le Raux en Pienaar 1970). In this list of vowels we
have ignored the length mark {:) used in their sources.
 157
gravel > giravhulo
cap > khephe
candle > khandela
"
dash > deshe

tap > thephe

cash > kheshe

telegram > telegiramu

Because the vowel re does not occur in Venda it is realised as [a] or [ e] in

Venda adoptives. This substitution is not accidental. The Venda vowels [a] and [e]
are the ones which are the closest to the English vowel ;:e.. The Venda vowel [ e ]
shares the feature [front] with the Engish vowel whereas the vowel ( re] shares the·
feature [low] with that of the English vowel. Thus the substitution in this case is
made on the basis of phonetic approximation. This can be seen from the position
1
of these vowels in the following vowel chart:

[a ]

1. The position of the English vowels in this chart and in the subsequent charts are
based on Roach (1983).
 158

2. The Afrikaans vowel: re

The vowel [re] is described by De Villiers (1970:58) as an abnormal midlow vowel

i.e. it shares the tongue position of a central vowel [a] and and a high [a]. It is also
characterised by a certain measure of roundness of the lips. This vowel is replaced
by the vowel [:>] in Venda. Let us look at the following examples:

Afr.sukkel > Ve.sogola 'suffer'

Afr.skuld > Ve.tshikolodo · 'debt'
Afr.brug > Ve.buroho 'bridge'

Considering the phonetic characteristics of the vowel [re] given above, it becomes
clear that the substitution of this vowel is mainly because of phonetic similarities
between the two vowels. Vowel [:>] is characterized by lip rounding in its
articulation. This roundness of lips agrees with the lip-rounding in the articulation of
the Afrikaans sound [re]. The chart below illustraes the position of these vowels:

[i J [u J

---
... .- . .
~ ..
......
, .......

[a ]

3. The English vowel: 3

This is a central vowel {i.e. a vowel produced when the centra! part of the tongue
is raised, in this case, to the mid position) which is produced with the lips in neutral
position. This vowel is found in the following examples:
 159
Eng.fern > Ve.feni
Eng .fence > Ve.fence
Eng.term > Ve.themo
Eng.germ > Ve.dzheme
Eng.jersey > Ve.dzhesi
Eng.jerk > Ve.dzhege
Eng.firm > Ve.feme
Eng .learner > Ve.lenasi
(licence)

The examples given above show that the English vowel 3 is replaced by e [e] in
Venda adoptives. This substitution is due to the fact that the two sounds are
phonetically close to each other. They. both are mid-vowels and are produced
without the lip-rounding. The closeness between these two vowels can be seen
from the following chart:

[i] [u]

[e ]

[a ]

4. The English and Afrikaans vowel: a

This vowel is often referred to as schwa. "Schwa is, in terms of its articulation,
neither high nor low, neither front nor back. It is a vowel produced with a neutral
setting of the articulators and is in this respect a 'minimal' vowel, involving, as it
does, no displacement of the articulators from the neutral position" (Giegerich
1992:68). This vowel does not occur in Venda and is realised in Venda as either
 160

[ e ], [ i ], [ u ], [ a ] or [;:, ]. This can be seen from the following examples:

fashion > fesheni

kitchen > khishi
station > tshititshi
/'

location > lokhishi

cupboard > khabodo

custard > khasitadi
"'"'
gravel > giravhulo
father > fada
A

customer > khasitama

A

culture > khalitsha

In the examples above it is interesting to note the various substitutions of the

English vowel ;:, . Its realisation is complex in the sense that it is realised by all
five Venda vowels. It is interesting to note that even in the following examples of
Afrikaans the vowel ;:, is realised by different Venda vowels viz. [a], [e], [i], [:)] and
[u].

Afr.politiek > Ve.polotiki 'politics'

Afr.ring > Ve.rinngi 'ring'
Afr.ink > Ve.ennge 'ink'
Afr.winkel > Ve.vhengele 'shop'
. Afr. tafel > Ve.tafula 'table'
A

Afr.kamer > Ve.kamara '

room'
Afr.betaal > Ve.badela pay'

The substitution of the vowel ;:, in both English and Afrikaans, as shown in the
examples above, is an interesting one. The question here is why is this vowel
realised by all the Venda vowels? To answer this question, the phonetic qualities of
this vowel and its position in the vowel chart need to be taken into consideration.
 161

As mentioned above, this vowel is a neutral vowel, therefore its substitution by all
the Venda vowels may be explained with reference to phonetic approximation. The
vowel a occupies the most central position in the vowel chart. Because of its
central position and its neutrality it can share phonetic characteristics with any
peripheral vowel in Venda. Nevertheless this account still does not explain what
determines the realisation of a specific vowel phoneme instead of another. One
way of explaining this problem is to look at the influence of other vowels within the
word in which this sound occurs. The following examples may throw some light in
this regard:

khishi < Eng.kitchen

tshititshi < Eng.station

kamara < Afr.kamer

fad a < Eng.father

The replacement of the vowel a by vowels [i] and [a] in the these examples
seems to be due to assimilation (i.e the schwa vowel has assimilated to the vowels
of the preceding syllables). However, the following examples seem to counter this
argument:

khabodo < Eng.cupboard

giravhulo < Eng.gravel

tafula < Afr.tafel

From these examples it appears that the preceding consonant is a factor rather
than the vowel. The sounds [b, 13, f] seem to have influenced the central vowel a
to be realised as a rounded back vowel in Venda. The common factor here is the
bilabial characteristic in
 162
both the consonant and the vowel. Both vowel [=>] and [ u ] are [+round] and like
their preceding consonants, involve the lips in their production. 1

The central position which the schwa vowel a occupies could be seen from the
following chart:

1
I

/
(
r

'I
\I
!
'
[c]

\ .__ .... .r
[a 1

5. The English vowel: A

This vowel is described by Giegerich (1992:73) as being unrounded and slightly

more back than front. It is fully low in English dialects like RP. This vowel is
replaced by [ a ] in Venda. Consider the following examples:

Eng.custard > Ve.khasitadi

, "'
Eng.pick-up > Ve.phikhaphu

Eng.bus > Ve.basi

1. The influence of bilabial consonants on the realisation of back vowels has also
been noted in other studies on other African languages (cf. Van ~Varmelo 1929
(Sotho), Kruger 1965 (N.Sotho), Hermanson 1991 (Zulu), Nkabinde 1968 (Zulu),
Khumalo 1987 (Zulu). According to them bilabial consonants or consonants with
bilabial characteristics may influence the realisation of a high back rounded vowel.
 163

Eng.cupboard > Ve.khabodo

Eng.customer > Ve.khasitama
.....
Eng.brush > Ve.bulatsho
Eng.republic > Ve.riphabuliki
Eng.judge > Ve.dzhadzhi
Eng.jumper cable > Ve.dzhampasi
Eng.jumble sale > Ve.dzhambulusele

The realisation of the English vowel A . as [a] in the examples above seems to be
due to phonetic approximation. Even though the English vowel is produced at a
slightly higher level, both are low vowels. In the articulation of both vowels there is
no rounding of lips. Therefore, the choice of the vowel [ a ] as a substitute for the
English vowel A seems to be determined by phonetic approximation. The
substitution of the vowel A is illustrated in the following chart:

[ i] [ul

[e ]
. ..
,-.,.,-..--
.::. • r < ~ I'. 1--Tti::C:--A~------4'~ [ 0 J
--- ---1- I
.~-
<· • •
...I,
• ,,
I ·•·· . ·. I
\ II

[a ]

6. The English vowel : o

The English vowel o becomes [;:,] or [a ]. These two Venda vowels are the
closest to this English vowel. The English vowel o shares the feature [back] with
the Venda vowel [;:,], and as a back vowel it is articulated with the rounding of lips,
 164
though to a lesser degree than vowel [:l]. The two vowels also share the feature
[back] i.e they are both back vowels. The fact that the English vowel o has much
in common with the Venda vowel [:l] shows that substitution is made here on the
basis of phonetic approximation. The same can be said of the the substitution of
the English vowel by the Venda vowel [a]. Although the two vowels differ with
regard to the part of the tongue that is used in their articulation, and are produced
at more or less the same level in the mouth, they are in fact both low vowels.
Therefore substitution is also made here on the basis of phonetic approximation.

Eng.polish > Ve.pholishi

Eng.policy > Ve.pholisi ·

Eng.pop > Ve.phophu (music).

Eng.posh > Ve.phoshi
Eng.porridge > Ve.phalishi

Eng.watch > Ve.watshi

In the examples above, the English vowel o is either replaced by the vowel [:l] or
[a] as shown in the last two examples. It is not clear from these examples as to
what determines the choice of either vowel. The word phalishi is more of a
colloquial word since there is a well-established word for porridge which is vhuswa.
Sometimes the word parenzha is used which is also an adopted word from the
English word "porridge". Consider the following chart for the various substitutions:

[ i] [u ]

[ f: ]

...,__..,
---
• l ........
...
' [a ]
 165
7. The English vowel: a

This vowel is realised as (a] in Venda. This realization is due to phonetic

approximation. Both are low vowels. Although the English vowel is produced with
the back of the tongue, it is very close to the central position where the Venda
..
vowel is produced. This reali,~ation can be seen from the following example:

Eng. father > Ve.fada "Roman Catholic priest"

I'-

Eng. party > Ve.phathi

Eng. park > Ve.phakha

Eng. parliament > Ve.phalamennde

Eng. passport > Ve.phasiphotho

In all the examples above the English vowel a is replaced by vowel (a] in Venda:
The phonetic relationship between the two vowels is illustrated in the chart below:

[u]

[e ] [c ]
,. ....... ~"'- \
"' , ",.
-
, 0 I'
/<..;,r . . 'tt.t· I

_-, ·.- -
t.... [a] '· ""'
'/

8. The English vowel : u

This high vowel is replaced by [u] in Venda:

Eng.sugar > Ve. tshugela

 166
Eng.goods (train) > Ve. gutsu
Eng.butcher > Ve. butshara
Eng.pudding > Ve.phudini
Eng.push > Ve.-phusha
Eng.pull > Ve.-phula

In all the examples above 'the English vowel u is replaced by [u]. This substitution
may be explained with reference to phonetic similarities between the two vowels.
Both these vowels are produced with the back part of the tongue and they also
share the features [high] and [round], though to different degrees. The back
position of these vowels and their height in the mouth is illustrated in the following
chart:

-l

[ E )

[al
9. The Afrikaans vowel: [ y :]

The vowel [y] is described by De Villiers (1970:58} as follows:

,,
"Die [y] van bv.minuut is 'n ABNORMALE HOE VOOR VOKAAL, d.w.s. dit het
die tongstand van (i] met gewoonlik effens geronde Iippe soos .bY [u]".

In the light of the above quote, it is not surprising that the Afrikaans vowel is
replaced by [i] or (u] in. Venda:

Afr.duur > Ve.dura

A .
"expensive"
 167
Afr.suur > Ve.swiri "orange"
Afr.huur > Ve.hira "hire"
· Afr.muur > Ve.miri "wall"
Afr.minuut > Ve.miniti "minute"

The Afrikaans vowel [y] is replaced by [i] in all the examples except in the first one.
In the first example, the vowel [yJ is replaced by the Venda vowel [u]. These
replacements may be explained by looking at the phonetic characteristics of these
two vowels. The two vowels are produced at a high level in the mouth. They are
both front vowels. even though their degree of frontness is not the same. This can
be seen from the following chart:

[ i]

II

.tal

From this chart we can see the proximity between the Afrikaans vowel [y] and the
the Venda [ i ]. The replacement of vowel [y] by [u] may also be explained. by the
fact that both are high vowels and that vowel [y] is produced with some form of lip
rounding. However, with regard to the choice of the vowel [u] in the first example
and not in others, there appear to be no clear explanation. The occurrence of this
vowel may therefore be regarded as an exception.

(b) Diphthongs.

Diphthongs were briefly discussed in section 3.2.1.2. Our main concern in this
section is to look at the phonological changes which foreign words containing
 168

diphthongs undergo when they are adopted into Venda. The main question here is
whether the diphthongs from foreign languages such as English will be accepted as
they are into Venda or whether they will be subjected to some changes. As was
mentioned, in the previous section, that diphthongs do not seem to be a. natural
feature of the African languages. Let's begin by looking at the sets of diphthongs
which are found in English and Afrikaans:

English Diphthongs: ai, au, ::li, ai, ea {e3), ua, au, ei

2
Afrikaans Diphthongs: a:i, oi, ui, eu, ai, rey, reu

Before we look at substitution, a few remarks will be made on the nature of the
diphthong. In defining the diphthong it was indicated that a diphthong is a two-part
vowel sound and that these sounds should be pronounced as a single syllable. The
phonetic symbols used for each diphthong represent the beginning and the end of
the vowel glide. According to Crystal (1992: 105)

"one element in the diphthong is always more sonorous than the other, if this
is the first element the diphthong is said to be 'falling' or 'descending', if the
second, it is 'rising' or 'ascending".

1. These vowels are based on the Received Pronunciation (RP). The symbols used
for these vowels and the examples of words in which they oocur were taken from
various sources, namely, the Oxford Advanced Learners Dictionary; the Collins
Concise Dictionary, Giegerich (1992), Gimson (1970) and Roach (1983).

2. The Afrikaans vowels have been cited from De Villiers (1970) and Uitspraak-
woordeboek van Afrikaans ( Le Raux en Pienaar 1976).
 169

He further describes sonority as

"the overall loudness of a sound, relative to others of the same pitch, stress
and duration. Sounds are said to have an inherent sonority, which accounts
for the impression of a sound carrying further" (1992:320 )

The difference in sonority, i.e. in pitch, stress and duration, we hope, will in one
way or another influence the substitution of diphthongs in adopted words in Venda.
Stress may determine the realisation of certain vowels in stressed or unstressed
syllables (Roach 1983:63). There are certain vowels, the 'schwa' vowel, for
example, which according to Roach (1983:63) does not occur in stressed syllables.
The vowels of a diphthong may also differ in terms of the duration they take in their
pronunciation. The first vowels of diphthongs usually have a longer duration than
the second ones. Their long duration may influence their perception, especially by
a foreign speaker of the language. These different characteristics will be taken into
consideration in our study of the substitution of the various diphthongs in Venda.

1. The English diphthong :[ai]

The following examples show how the English diphthong ai is adapted to Venda:

Eng.bicycle > Ve. baisigira

Eng.dice > Ve.daizi

Eng. ice-cream > Ve.aisikirimu

Eng.tie > Ve.thai I thayi

From the examples given above the English diphthong ai is realised with the
sequence of vowels [a+i] in Venda, where, the [a] vowel in Venda is a central
vowel and the English [ i ) is the same as that of Venda.

Consider now the following Venda examples that have been adopted from
 170
Afrikaans words that contain the diphthong [a:i] (the colon after [a] indicates
length):

Afr.slaai > Ve.tshilai 'slice'

/'

Afr.laai > Ve.laisa 'load'

"'
Afr.ba:ici > Ve.badzhi 'jacket'
Afr.fa:ici > Ve.fagi 'big drum'
Afr.onderba:ici > Ve.tshinorovhagi 'waist coat'

Note that in the first two examples, the Afrikaans diphthong is realised in Venda by
the same sequence of vowels as was the case with the English examples, but in
the last three examples the diphthong has been replaced by the single vowel [a]. In
these latter cases the vowel [ i ] is elided. The elision of the vowel [ i ] is not
unexpected here since it is preceded by a long vowel [a:] which appears to be
more stressed. This vowel becomes even more sonorous by the fact that it is the
first element of the diphthong. As already indicated, the sonority of the first vowel of
a diphthong causes the second vowel to be less audible.

2. The Afrikaans diphthong: [ rey ]

The Afrikaans diphthong [rey] presents a very interesting case. This diphthong is
replaced by a single vowel in Venda. Consider the following examples:

Afr.suiker > Ve.swigiri 'sugar'

Afr.kruiwa > Ve.girivhane 'wheelbarrow'
Afr.druiwe > Ve.ndirivhe 'grapes'
"

From these examples it appears the first vowel of the diphthong h·as been elided.
The Afrikaans vowel [y] is then realised as [ i ]. These two sounds are phonetically
close to each other (cf. vowel chart and discussion on p.166 of this study). The
elision of vowel [re] may be attributed to the high palatal tongue position of the
vowel [y]. This makes it to be more sonorous than its counterpart. Because of its
 171
high sonority it might have become more audible to the non-speakers of the
language than its counterpart.

3. The English diphthong: [ei]

Thr3 English diphthong [ei] is realised as [ei] in Venda.

En·g.tape > Ve.theiphi I theyiphi

Eng.bail > Ve.beili I beyili
Eng.cane > Ve.kheini lkheyini

In these examples the diphthong has been replaced by a sequence of the vowels
[e] and [i]. Sometimes a semi-vowel is inserted between the vowel sequences. The
question here is whether the vowel sequences in the above examples constitute
diphthongs. Some linguists feel that these sequences should be regarded as mere
sequences rather than diphthongs. This issue will be taken up later in this chapter.

In some instances the English diphthong [ei] may be replaced by a single vowel [ e]
in Venda:

Eng.gate > Ve.gethe

Eng.jail > Ve.dzhele

Eng.paper > Ve.phepha

Eng.cab!e > Ve.khevhele

Eng.cake > Ve.khekhe

The question that the two sets of examples raises is why in the first set of
examples the diphthong ei was replaced by a sequence of vowels [c:] + [i] and not
by a single vowel, as in the second set of examples. It is interesting to note that in
all the examples the first vowel of the diphthong is retained. The realisation of the
vowel [i] in the first set of examples may be explained by the influenced of the
insertion of the final vowel [i] which occurs in all the examples. It appears the final
 172

vowel [i] intensifies the sonority of the vowel [i] which is in the weaker diphthong
position. In the second set of examples, there is no insertion of a final vowel, and
so the vowel [i] of the diphthong is elided and the diphthong is replaced by a single
vowel [e]. The choice of this vowel seems to have been influenced by several
factors. In the first place, it is the same vowel [e] which occurs at the beginning of
the diphthong. As sueh it is more sonorous than its counterpart since the first
sound of the diphthong is usually produced with more intensity than that of the
. second one. Secondly, as the first segment of a diphthong, this vowel has a longer
duration than the second one. This again makes it more audible than the second
one.

4. The diphthong: [ai]

The Afrikaans diphthong [ai] is replaced by the single vowel [ i ] or [ e ] in Venda:

Afr.tapyt > Ve.tapita

A A
'mat'
Afr.spyker > Ve.tshibigiri 'nail'
Afr.pyp. > Ve.bibi 'smoking pipe'
Afr.kaptein > Ve.kapu~eni . 'captain'
Afr.trein > Ve.thireni 'train'
/"

The question here is what determines the choice of these two vowels in the
examples given above. The replacement of the diphthong by [i] (which represent
the second part of the diphthong) seems to contradict what we stated earlier on,
namely, that the first sound of a diphthong is more sonorous and is therefore likely
to dominate the second one. In this case it is interesting to note that the vowel
which occurs at the beginning of the diphthong is the schwa vowel [a ]. The
dominance of [i] in the examples above may be explained in the light of the nature
of the schwa vowel [a]. This vowel normally occurs in unstressed positions
(Giegerich 1992:73). According to him, this vowel may be entirely dropped in
certain words because it is not stressed when pronounced. Because of its low
sonority value, its lack of stress and short duration, this vowel does not seem to be
 173

easily perceived by non-speakers of the language, hence it is dropped in some

diphthongs. In view of the phonetic characteristics of the schwa vowel it is not
surprising that this vowel has been dropped in favour of the vowel [i]. Nevertheless,
the last two examples seem to counter the explanation given above. In these last
two examples the schwa vowel has been replaced by [e]. As has already been
indicated, the replacement of [ a ] by the vowel [e] is mainly due to phonetic
approximation. The vowel [i] (which represents the second part of the diphthong) is
elided in these last two examples. The deletion of this vowel may be explained by
the fact ·that this vowel occupies a weak position in the diphthong. It appears that
the substitution of the schwa vowel occurred before the deletion of the vowel [i].

5. The English diphthong: [oi]

The English diphthong is realised as [oi] in Venda.

Eng.join > Ve.dzhoini

Eng.toy > Ve.thoi

In these examples the English diphthong is replaced by the sequence [o+ij in

Venda. The vowels which constitute the English diphthong have been reproduced
by the same Venda vowel sequence. This substitution has been motivated by
similarity of phonetic features.

6. The English diphthong: [ea ]

The English diphthong ea is realised as ( ea I eya] in Venda.

Eng.square > Ve.tshikwea/ tshikweya

Eng. underwear > Ve.andawea/ andaweya

The first element of the English diphthong has been replaced by the vowel [e] and
 174
the schwa vowel [a] replaced by [a]. As was mentioned previously, the schwa
vowel can be replaced by any of the five vowels in Venda. The realisation of the
vowel (a] here also seems to be due to the proximity which this vowel shares with
the [ a ] . Consider the following vowel chart:

[e ] [~ ]

.... --
(a ]

7. The English diphthong: au

The English diphthong is replaced by both simple vowels and diphthongs. Consider
the following examples:

Eng.overall > Ve.ovarolo

Eng.overtime > Ve.ovathaimi

Eng.oversize > Ve.ovasaizi

Eng. protein > Ve.phurotheini

Eng.prose > Ve.phurosa

Eng.notes > Ve.notsi

Eng.notice > Ve.nothisi

Eng. telephone > Ve.thelefoni

Eng.phone > Ve.founi

Eng.phoneme > Ve.founimi

The English diphthong has been replaced by the vowel (:>] in all the examples with
the exception of the last two examples. The substitution of this diphthong is very
 175
interesting in the sense that (i) the diphthong is replaced by a single vowel, and (ii)
this vowel is not either of the two vowels in the diphthong. What seems to have
happened here is some form of coalescence between the two vowels. The [u]
vowel which perhaps is in a stronger position than the schwa vowel [<1], seems to
have a somewhat dominant role in the choice of the vowel [::>] in Venda and is
similar to the English one in terms of its roundness and backness.

8. The English diphthong: [au ]

The English diphthong is realised in Venda as follows:

Eng:blouse > Ve.bulausu

·Eng. pound > Ve.paundu

Eng.crown > Ve.khirauni

Eng.account > Ve.akhauthu

Eng.counter > Ve.khauthara

Eng.discount > Ve.disikhauthu

Eng.towel > Ve.thaula

Eng.half-a-crown > Ve.hafukoroni

In all the examples, except the last one, the diphthong au has been replaced by the
sequence [a+u] in Venda. The Venda vowels in this sequence are phonetically very
close to those of the English diphthong. The English vowel a, though not a central
vowel, shares phonetic features such as [low] with the Venda [a], while the vowel
[u] of the English diphthong shares the phonetic feature [round] with its Venda
counterpart. Therefore adaptation has been effected on the basis of phonetic
approximation. The replacement of the diphthong au by the vowel [:>] in same
Venda adoptives is difficult to explain. It appears that this word ·has been adopted
into Venda via Afrikaans. In Afrikaans the word for 'half-a-crown' is 'halfkroon'
(pronounced as halfkro:n). The pronunciation of the Afrikaans word seems to be
closer to Venda than the English one. The fact that the word 'half-a-crown' entered
the language via Afrikaans may be used as a possible explanation for the
 176
realisation of the vowel [ ;:, ] in the adoptive hafukoroni.

(c) General remarks on the substitution of vowels in Venda

Venda has adopted a considerable number of words from English and Afrikaans.
Some of the adoptives from these languages contain vowels which are not found
in the Venda vowel inventory. As was shown in the previous discussion, these
vowels are replaced by vowels tr1at are found in the language. The replacement of
these foreign vowels by native ones,· seems to be triggered by the native segment
inventory constraint. Studies of other languages (cf.Kaye and Nykiel 1972; Holden
1976; Yavas 1982; Steinbergs 1984; Silverman 1992) have identified different
constraints which operate in different languages. In Venda, the native segment
inventory constraint prevents the occurrence of any vowel which does not belong to
the language sound inventory. Various Venda vowels have therefore been used to
replace the foreign vowels. These replacements do not, however, occur in an ad
hoc manner. The substitution in these cases appears to be mainly due to phonetic
approximation. In other words, foreign vowels are replaced by vowels which are
closest to them in the adopting language. In some cases, the choice of the
substituting vowel is influenced by assimilation. The vowel which replaces a
foreign vowel assimilates to the preceding native vowel or to the succeeding one.
The substitution of the different simple vowels from English and Afrikaans can be
summarised as follows:

English vowels: re >[e] [a]

<) > [ a ], [ e ], [ i ], [ ;) ], [u]

A >[a],[e]
o > [ ;) ], [ a ]

o > [a]

3 > [e J
 177

Afrikaans vowels: a > [ a ]; [ e ], [ i }, [ : > }, [u]

y >[i ],[u}
ce >[::>].

In this section we also discussed the substitution of diphthongs from English and
Afrikaans. The English diphthongs can be summarised as follows:

Diphthong
I
J
centering
l
closing

l I
e/t\o
Ja ea ua
1\n
au au
u

(Loach 1983: 19)

The occurrence of diphthongs in African languages such as Venda, as has already

been mentioned, is still a matter of controversy. Diphthongs from English and
Afrikaans are adjusted in various ways in Venda. In some instances, as was shown
above, diphthongs are replaced by single vowels. The problem that arises concerns
the choice of the substituting vowel. From the analysis of the various diphthongs
different explanations were given. In some cases the first vowel of the diphthong is
chosen. If this vowel does not occur in Venda, it is replaced by another vowel
which then becomes a substitute for the diphthong. The choice of the substituting
vowel seems to be determined mainly by phonetic similarity, i.e. a foreign vowel is
replaced by one which is phonetically close to it. Let's take the English diphthong
cake ('keik) for example. This word is adopted as khekhe [khekhe] in Venda. The
diphthong in this word has been replaced by the single vowel [e]. The choice of this
vowel may be explained by factors like stress and duration. The stress in this
syllable falls mainly on this first vowel and the fact that it has a longer duration in
its pronunciation than the second one might have influenced its perception by the
 178

speakers of the adopting language. Sometimes the vowel which replaces the
diphthong is the result of coalescence between the two elements of the diphthong.
This can be seen in the adoptive ovarolo < Eng.overall /auvaro:ll. The vowels have
influenced each other, resulting in the vowel [:)] in Venda.

In certain instances diphthongs are realised as vowel sequences in Venda. In these

sequences vowels which do not occur in Venda are replaced by others closest to
them. The question here is whether or not such sequences constitute diphthongs.
The issue of the occurrence or non-occurrence of diphthongs in Venda has been
discussed, albeit briefly, at the beginnil}g of this chapter. There it was indicated that
diphthongs do not seem to be a common feature of Venda despite their alleged
occurrence in fast speech of some word forms. This question is raised here once
. more because of the many adopted words from English and Afrikaans which
incorporate diphthongs. Should this be regarded as the·'importationLof new foreign
sounds into the language? There are two views with regard to these sounds. The
first view rejects the notion that diphthongs occur in African languages. According
to this view, the English diphthong /ail would merely be regarded as a sequence of
separate vowels, i.e. /a+il rather than a diphthong (cf. Kruger 1965; Nkabinde
1968; Batibo 1993). The other view maintains that such sequences of vowels
constitute diphthongs since their articulation is exactly the same as that of the
diphthongs in the source language. Although this view may be applicable to some
languages, it does not seem to apply to Venda for several reasons. Vowels in
Venda are tone bearing units. Some vowel sequences have different tones, e.g thai
(HL), thai (HL). The vowel sequences in these examples are regarded as separate
syllables rather than diphthongs. However, in cases where the vowels of a
diphthong have the same tones, this principle doesn't apply.

There is also a tendency in Venda to insert a glide between the. vowel sequences
e.g. tshikweya, andaweya, thawula, phurotheyini, etc. The nature of the glide
depends on the nature of the vowels involved. The insertion of a glide between the
vowel sequence may, in a way, demonstrate the language constraint against the
 179
occurrence of diphthongs in Venda. Rising or falling tone may occur in some of the
vowel-glide sequences.

3.3.2.2 Consonant substitution

As already mentioned, substitution in Venda also occurs with consonants. Here we

shall focus mainly on the consonant sounds from English and Afrikaans 1 to
illustrate the type of consonant adjustment effected on Venda adoptives. However,
a few examples will also be taken from other African languages such as Nguni.
Some of the consonant sounds from these languages also occur in the Venda
sound inventory, others do not Substitution seems to occur in both the consonants
which have no replicas in Venda and also those which have replicas in the
language. We shall begin by looking at the kind of substitution which occurs when
equivalent sounds do not occur in the adopting language

1.1n our examples of the substitution of the different sounds we shall use various
symbols to represent English and Afrikaans consonants.· These symbols have been
cited from various sources, namely, Lass (1987), Giegerich (1992), the Oxford
Dictionary and the Collins dictionary). The Afrikaans symbols have been cited from
De Vilfiers (1970) and Uitspraak-woordeboek van Afrikaans by Le Raux & Pienaar
(1976). In some cases the examples given in English and Afrikaans have been
represented phonemicalfy in the above sources. In other cases they have been
represented phonetically. Since this study does not purport to represent a detailed
analysis of the phonetic and/or phonemic status of the different vowel qualities in
English and Afrikaans, the examples listed in this section are given as they were
represented in the works cited above. In some cases slashes are used, in other
cases round slashes are used and yet in other cases square brackets are used,
depending on the work from which they have been cited.
 180
(a) Substitution of consonants which do not occur in Venda

Substitution of this nature requires a comparative study of the sound inventories of

the source languages and the adopting language. If one compares English and
Afrikaans with Venda, one notes that there are a number of consonants from these
languages which are not found in Venda. The English language, for example, has
certain sounds, namely, lateral p], and the dentals [ n] and [a 1which do not occur
in Venda. Like English and Afrikaans, other African languages such as Zulu and
Shena, have certain consonants which do not occur in Venda. Zulu has, for
instance, clicks whereas Shena has implosives. These sounds do not occur in
Venda. The question of great interest, therefore, is what happens when words with
consonants which do not occur in Venda are incorporated into the language.

1. The English sound: th [e)

Consider the following adoptives from English:

Eng. theorem > Ve. theoremu -siyoremu

Eng. theology > Ve. thiolodzhi -siyolodzhi
Eng. theatre > Ve. thietha - siyetha
Eng. thermometer > Ve. themometha- semometha
Eng. athletics > Ve. athilethikhi
Eng. Sabath > Ve. Savhatha
Eng. thesis > Ve. thisisi
Eng.methodist > Ve. methodisi
Eng. mathematics > Ve. methemethiki
Eng. Lutheran > Ve. Lutere
A

From the examples given above the voiceless [ a ] may either change to the
dental ejective t [~, or to alveolar ejective [t] or it may be adapted to the aspirated
th [th] in Venda. The realisation of dental [t] in the last example above may be
1'..

attributed to the influence of the German language. The Germans were the first
 181

people to introduce the Lutheran church in Venda. With regard to the realisation of
the sound t [t]
,.. one may suspect the influence of the Afrikaans or German
languages. Therefore the word LuJ_ere might have been adopted to Venda from the
German language or via Afrikaans. In Afrikaans the sound [t] is unaspirated and
as such it is realised as the sound [t] which is also unaspirated to a certain degree
......
in Venda. In some instances the [s] sound is used as a variant of the English th
[ 6 ] as indicated in the above examples. This kind of substitution was triggered by
.,'\
the native segement
;:::../
inventory constraint This constraint replaces any foreign

sound which does not form part of the language inventory by the native one. In
both cases the choice of the substituting sound seems to be determined by
phonetic factors. The Venda consonants [th, s, t] which have substituted the
English consonant [ 6 ] are very close to it, hence we can speak of phonetic
approximation. A number of phonetic similarities between these sounds can be
noted. In some instances the original sound [ 6 ] is retained, especially by bilingual
speakers. For instance, the sound [ 6 ] in words !ike theory and theatre is very
often pronounced as in English.

2. The English sound: th [ IJ ]

The voiced English sound [ ~ ] is substituted by the sounds [d] and [d] in Venda:
"'
Eng. heathen > Ve. muhedeni
Eng. leather > Ve. leda
"'
Eng. father > Ve. fada
A

The motivation for this kind of substitution is the lack of equivalent sounds in
Venda. The sound [ ~ ] does not occur in the Venda sound inventory and is
therefore replaced by sounds which are very close to it. But the question here is
what determines the realisation of either of these substituting sounds. The
realisation of the dental sound d [d]
.....
in the last two examples may be explained with
reference to the influence of -er on the sound [ ~ ]. The -er sound makes the sound
[ ~ ] in the last two examples to be pronounced as a dental sound.
 182

3. The English lateral sound: I [ I]

The lateral [ I ] is another example of English sounds which do not occur in Venda.
This consonant may either change to the dental [ I ] or it may be replaced by the
tap sound [ r ]. However, this sound may be retained in some adoptives.

Eng.police > Ve. pholisa (tap) [r]

Eng.hostel > Ve. hositele (lateral/dental) [1]/ ~
Eng.telephone > Ve. thelefoni (lateral/dental) [I] I [I]
A

Eng.blouse > Ve. bulausu (dental/lateral) [I] I [ij

A
Eng.leave > Ve. livi (dental/lateral) [I] I [I]
Eng.lorry > Ve. Iori (dental/lateral) [1]/ [I]
A

Eng.towel > Ve. thaula (tap) [r]

Eng.wool > Ve. wulu (tap/lateral/ dental)[r]/[1]1 [ij
....
Eng. ball > Ve. bola (tap/lateral I dental) [r] I [I] 1[1]
A

Eng. pencil > Ve. penisela (tap) [r]

Here it is also interesting to see how the Afrikaans sound [ij is realised in Venda:

Afr.leer > Ve. Jeri (dental) 'ladder'

. ""'
Afr.lekker > Ve. ,..Iegere (dental) 'sweet'
Afr.tolk > Ve. dologi (tap) 'interpreter'
"' '
Afr.lepel- > Ve. lebula (tap) spoon'
Afr.plek > Ve. bulege (tap/lateral) 'place'
Afr.blik > Ve. lulege (tap/lateral) ' can'
Afr.geld > Ve. tshelede (tap/lateral) ' money'

From the examples given above we can note that the replacement of the lateral
sound IU in Venda is really quite complex. In some instances this consonant is
realised as a dental yet in other instances it is realised as a tap, lastly it may also
simply be retained as a lateral consonant. There seems to be no clear criterion
underlying the choice of the sound which replaces the English or Afrikaans sound
 ... 183

[ij in Venda. Traill (1993:7) says the following in this regard:

"In other respects, the evidence is that Venda employs the distinction
productively in adoptives from English and Afrikaans. The details of the
distribution of laminals and apicals in these adoptives suggests that there is
no fixed strategy for assigning the foreign sounds to one or other category".

The examples given above attest to the fact raised by Trail! here that the choice of
either the lateral or the dental sound to replace the lateral sound in Venda
adoptives is not clear.

4. The substitution of Nguni clicks.

Clicks do not occur in Venda. Therefore when Nguni words with clicks are adopted
into Venda, these sounds are replaced by sounds closest to them. The following
are examples of Nguni words which have click sounds.

Zu.umGqibelo > Mugivhela

Xh.amaXhosa > Mathosa
Zu.umqo!Tiboti > mutomboti

From these examples the Nguni clicks, namely qg[!g], xhPih], q[!], are replaced by
g[g], th[~h] and t[t] respectively. The replacement of these clicks appears to have
been made on the basis of phonetic approximation. In the first example the click [!]
is dropped and the [g] sound then remains in the place of the click. It seems the
voiced quality of the [g] makes it more audible than the click [!]. The choice of the
sound [th] to replace the click xhPih] appears to have been determined by the
alveolar and aspirated nature of this click. The click in the .last example is
articulated with some ejection. This may be the reason for its replacement by an
ejective sound t[t].
 184

(b) Substitution of consonants which occur in Venda

Substitution becomes more complicated in instances where the speaker of the

adopting language replaces the sounds which occur in his language by different
ones. There are many consonants from English and Afrikaans which are replaced
by other consonants in Venda even though they are found in the sound inventory
of the language. One would expect these consonants to be taken over without any
modification since they are already part of the adopting language's sound inventory.
The replacement of these consonants raises a number of questions. In the first
place, it is not clear as to why these sounds are replaced by different sounds. The
second question relates to what are the factors that determine the choice of the
substituting sounds. In attempting to answer these questions various examples will
be analysed.

1. The Afrikaans consonant k lkl

The voiceless Afrikaans sound k may be realised as voiced g[g] or voiceless (k] in
Venda. We shall begin by looking at the instances where k is replaced by g[g] in
Venda:

Initial position

Afr.kamp. > Ve.gammba 'camp'

Afr.kameel > Ve.gamela 'camel'
Afr.koffie > Ve.gofhi ' coffee'
'
Afr.koring > Ve.gurou wheat'
Afr.kettel > Ve.ge)tela 'kettle'
Afr.kerk > Ve.gereke 'church'
Afr.kam > Ve.gamu ·'comb'
Afr.kombers > Ve.gumbese 'blanket'
Afr.kapok > Ve.gambogo snow'
'
Afr. karretjie > Ve.gariki wagon'
 185

Afr.kook > Ve.guga 'cook'

Afr.kous > Ve.lugausu 'garment'

Medial position

Afr.speiker > Ve.tshibigiri 'nail'

Afr.sker > Ve.tshigero 'scissor'
Afr.bek.er > Ve.bigiri 'beaker
Afr.suiker > Ve.swigiri ' sugar'

Word final postion

Afr.blik > Ve.lulege 'tin'

Afr.plek > Ve.bulege 'place'
Afr.kapok > Ve.gambogo 'snow'
Afr.werk > Ve.vherega ' work'

Afr.week > Ve.vhege week'

Afr.boek > Ve.bugu 'book'
'
Afr.sak > Ve.saga sack'

In all the examples given above the Afrikaans k is adopted as g[g] in all the word
positions, i.e. initial, medial and word final position. Poulos (1990:517) attributes the
replacefTlent of the sound k by g[g] to the phonetic qualities of this Afrikaans sound.
According to him, the Afrikaans sound k is unaspirated. Because of this it is very
often adopted as g[g] in Venda which is also unaspirated.

Although in the examples given above the sound [k] is adopted as g [g], this is not
the case in the following examples:

Afr.kamer > Ve.kamara 'room'

Afr.kaptein > Ve.kaputeni 'captain'
"
Afr.kolonel ·> Ve.kolonele 'colonel'
 186
Afr.skip > Ve.tshikepe 'ship'
Afr.skuld > Ve.tshikolodo 'debt'
Afr.skelm > Ve.tshikeleme 'criminal'

The adoption of the Afrikaans k as ejective sound [k] in the first three examples is
not easy to explain. One possible explanation could be that these words were
adopted by bilingual speakers who wanted to reproduce the Afrikaans sound as it
is pronounced in the source language. After all, these are recent . adoptives.
Therefore this substitution may be . explained by phonetic approximation. The
substitution in the last three examples may be explained by the consonant cluster
sk. In this cluster the Afrikaans sound k is unaspirated and is therefore replaced by
the ej~ctive sound [k] which is closest to it.

The English consonant lkl

The English phoneme /kl can be realised differently in Venda. It can be realised as
an ejective sound [k] or as an aspirated [kh] depending on the environment it
occupies in a word. These realisations can be clearly seen from the following
examples. We shall first look at instances where the phoneme /kl is realised as an
aspirated sound in Venda:

Eng.cabbage > Ve.khavhishi

Eng.cabl.e > Ve.khevhele
Eng.chemist > Ve.khimisi
Eng.cake > Ve.khekhe

Eng.computer > Ve.khomphiutha

Eng.catalogue > Ve.khathalogo
Eng.curry > Ve.khere
Eng.cup > Ve.khaphu

From the examples given above it is apparent that the English phoneme /kl is
 187

realised as an aspirated sound [kh] in Venda where it occupies the initial position in
a word. This does not come as a surprise since this sound is realised in a similar
manner in English. However, in the following examples the English non-aspirated
sound [k] is realised as an aspirated sound in Venda:

Eng.cream > Ve.khirimu

Eng. cricket > Ve.khirikhethe
Eng.Christmas > Ve.Khirisimasi

However, in the following examples English non-aspirated sound is realised as an

ejective sound despite the fact that it occupies the initial position of a word.

Eng.class > Ve.kilasi

Eng.club > ve:kilabu
Eng.clinic > Ve.kiliniki

It appears that the realisation of the phoneme /kl as an ejective sound in Venda is
due to the manner in which the consonant cluster cl is articulated. In the
articulation of this cluster the sound I immediately follows I kl. As a result the sound
lkl is pronounced as the ejective sound k[k] of Venda. The same realisation of the
phoneme /kl occurs where it is preceded by an [s] sound in the word initial position:

Eng.sch~me > Ve.tshikimu

Eng.school > Ve.tshikulu
Eng.Scottish > Ve.tshikotshi
Eng.skirt > Ve. tshikete
Eng.skipper > Ve.tshikipha
Eng.scooter > Ve.tshikuta
Eng.scorns > Ve. tshikontsi

In these examples the English sound /kl occurs in sound clusters, and as such it
loses its aspirated quality.
 188

The English sound k may also be realised as an ejective sound [k] in medial and
final position of the word. We shall first look at the examples in which this sound
occurs in the medial position:

Eng.market > Ve.makete

Eng.basket > Ve.basikete
Eng.casket > Ve.khasikete

The following examples show the realisation of lkl as in word final consonant
clusters:

Eng.mask > Ve.masiki

Eng.clerk > Ve.kilaka

Although the examples given above seem to create the impression that the English
phoneme is always realised as an ejective sound in medial and word final position,
this is not the case with the following examples:

Eng.packet > Ve.phakhethe

Eng.park > Ve.phakha
Eng.cake > Ve.khekhe

Eng.khakhi > Ve.khakhe

Eng.truck > Ve. ,...thirakha

Eng.track > Ve. thirekhe
1\

The realisation of the aspirated sounds in both medial and final positions may be
explained with reference to the phonological process of consonant harmony. In the
adoptives given above there is always an aspirated sound in the preceding or
following syllable. In a word like phakhethe there are three aspirated sounds
viz.[ph], [kh] and [th]. One may therefore argue that the realisation of the sound
 189

[kh] in Venda is due to the influence of the preceding aspirated sound [ph]. This
influence is called consonant hannony and can be illustrated with the following
diagram:

(a) phakhethe (b) khekhe

1
Feature Tier +ASP +ASP

2
Laryngeal Tier)
l\ I
I \ I
I \ I
I \ I
I \
3
Segmental Tier ph-a-kh-e-th-e kh-e-kh-e
(cf.Khumalo 1987:11)

3. The Afrikaans and English consonant phoneme : I b I

The Afrikaans and English sound [b] may be realised as [ p ] or [b] in Venda. Let's
first consider instances where [b] is retained in Venda:

Afr.band > Ve.bannda 'belt'

Afr.boek > Ve.bugu 'book'
Afr.bout > Ve.baudu 'bolt'

1. A phonological/eve! comprised of distinctive features only.

2. A level of distinctive features that are related to the larynx.
3. A level comprising of phonological units such as phonemes.
 190
Afr.botel > Ve.bodeio
A
'bottle'
Afr.botter > Ve.bodoro 'butter'
1'\

Afr.baadjie > Ve.badzhi 'jackert'

Eng.bar > Ve.bara

Eng.ball > Ve.bola
Eng.bus > Ve.basilbisi

In these examples the sound b occurs in the initial and the stressed syllable. In this
position the sound b becomes audible enough to be correctly approximated with·
the sound [b] on Venda. However, in the following examples, the sound b is
replaced by the Venda sound vh[{3].

Afr.broek > Ve.vhurukhu 'trouser'.

Afr.brood > Ve.vhurotho 'bread'
Afr.boor > Ve.vhoro 'drill machine'
Afr.boer > Ve.livhuru 'boer/farmer'
A

Afr.bybel > Ve.bivhili 'bible'

Eng.cabbage > Ve.khavhishi
Eng.cabble > Ve.khevhele
Eng.rubber > Ve.ravhara

The realisation of the sound vh[{3] in the first two examples may be explained by
morphological factors. In these examples the Afrikaans consonant cluster br is
interpreted as the vhu- prefix of noun class 14. It is not clear as to what causes
the b in vhoro and livhuru to be replaced by vh[{3]. The adoption of b as vh[{3] in
the remaining examples above may be explained by the fact that the sound occurs
in medial position and in unstressed syllables. Because of this the sound b
becomes 'softened' in such a way that it is perceived in Venda as the sound
vh[{3].
 191

4. Afrikaans consonant: t

The Afrikaans consonant may be realised as dental sounds [d]

,. and alveolar [d]:

Afr.tent . > Ve.dennde 'tent'

Afr.tou > Ve.deu 'rope'
Afr.tolk > Ve.dologi
A
'interpreter'
Afr.tafel > Ve.dafula 'table'
A

· Afr.trekker > Ve.deretere

A /'
'tractor'

The voiceless consonant [t] is realised as a voiced alveolar [d] in the first two
examples and as voiced dental sound [~] in the last three examples. The
realisation of the sounds [d] and [d] in these examples may be explained by the
/\

nature of the Afrikaans sound [t]. This sound is unaspirated and can therefore be
approximated to the Venda sound [d] or [d]
1\
which though voiced, is also
unaspirated. These substituting sounds also occur in medial:

Afr.stoel > Ve.tshidulo 'chair'

Afr.steen > Ve.tshidina 'brick'
Afr.kettel > Ve.gedela
,.... ·kettle'
Afr.skottel > Ve. tshigodelo
A
'household'
Afr.hospital > Ve.sibadela 'hospital'

In the first two examples, the sound [t] originally occured in a cluster. As has
already been mentioned, adopted nouns which begin with s+C cluster are assigned
to class 7. In these examples, the sound [t] is replaced by the alveolar sound [d]. ln
medial position, except the last example, the sound [t] is replaced by the alveolar
sound [d]. In word final position, the sound [t] is replaced either by the dental [~] or
alveolar [d]:

Afr.pot > Ve.bodo 'pot'

Afr.koerant > Ve.gurannda ·newspaper'
.1\
 192

Afr.skiet > Ve.tshigidi 'gun'

Afr.tent > Ve.dennde 'tent'

From the examples given above, it appears that the Afrikaans sound t is replaced
by the sound [d] or [d] in all the positions, i.e. word initial, medial and final position.
A

However, Afrikaans It/ can also become the voiceless dental [t] in Venda, as shown
in thefollowing examples: "'

Afr.tamaties > Ve.tamatisi

.......
'tomatoes'
A
.
Afr.tapeit > Ve.tapita
/'- r-
'tile'
Afr.trein > Ve.tireni ~train'
"""
Afr.tronk > Ve.tironngo
,.... 'jail'
Afr.tie > ·ve.tie 'tea'

The choice of the dental sound [t] seems to be mainly due to phonetic
I'

approximation. The Afrikaans sound [t], as already indicated, is unaspirated.

Therefore,. it is replaced by the sound [t] which, even though is produced with some
/"

from of ejection, is an unaspirated sound. In the last example the substituting

sound is the alveolar [t]. However, it is not clear why the [t] has been replaced by
the ejective sound [t] in the last example. The influence of the high vowel [i] which
immediately follows the sound [t] may be suspected to have influenced the
realisation of an ejective sound.

The dental sound [t] is also realised where the Afrikaans [t] occur in a cluster:
I'

Afr.straat > Ve. tshitarata 'street'

" A

'
Afr.stokies > Ve. tsh itoki
,...., si court'
Afr.strop > Ve. tshitiropo
,...... 'strop'
Afr.stand > Ve.tshitannde
/'"
'stand'
Afr.streep > Ve.tshitirepe 'stripe'
/'
 193

Afr.stasie > Ve.tshitasie

A
'station'
Afr.straaf > Ve.tshitarafu
/"'
'punishment'

In a nutshell, the Afrikaans sound [t] is replaced by the sounds [d], [,5!], [t] and !!)·
All these sounds are phonetically closer to the Afrikaans sound [t], though to
different degrees.

5. The English consonant : It I

In English the phoneme Jtl has two allophones, viz. unaspirated [t] aspirated [th].
These allophones may constitute phonemes in Venda since they can function
distinctively. The phoneme /tl is realised as t[t] or th[th] in Venda. Let's begin by
looking at the instances where the sound Jtl occurs in word initial position:

Eng.tank > Ve.thannge

Eng.tennis > Ve.thenisi
Eng.towel > Ve.thaula
Eng. tie > · Ve.thai
Eng.toy > Ve.thoi

The English /tl in the examples above is an aspirated sound since it occupies the
word initial position. This sound is adopted without any change into Venda.
However; in the following examples the English sound ltl is adopted as th[th] even
though it is not at the initial position of a word:

Eng.toilet > Ve.thoilethe

Eng.ticket > Ve.thikhithi
Eng.deposit > Ve.diphosithi
Eng.plate > Ve.phuleithi

The realisation of the aspirated sound [th] in word final position in the above
examples seems to be influenced by the aspirated sounds in the preceding
 194
syllables. In this case the aspirated sound which is found in the initial or medial
position influences the ttl in final position in such a way that it too becomes an
aspirated sound. This is illustrated by the fact that if one uses an ejective sound in
the initial or medial position the [t] in the final position is also realised as an
ejective sound t[t] in Venda. As already mentioned, this process is called consonant
harmony. Consonant harmony was discussed in the previous sections. What is
interesting here is the fact that the influence may occur across a syllable:

Eng.deposit > Ve.dipositi cf. diphosithi

Eng.plate > Ve.puleiti cf. phuleithi
Eng.report > Ve.ripoto cf . riphotho

The English sound t may be realised as non-aspirated in initial position especially

where it occurs in a cluster:

Eng.store > Ve. tshitolo

Eng.steam > Ve.tshitimu
Eng.stamp > Ve.tshitemmbe
,;...

Eng.station > Ve. tshititshi

,....

In these examples the English sound t is unaspirated and is therefore realised as

an ejective t[t] in Venda.

4. The Afrikaans sound !pi

The Afrikaans sound [p] may be realised as an ejective [p] or as a voiced sound [b]
in Venda. We shall begin by looking at instances where [p] is realised as [b]:

Afr.pad > Ve.bada 'path'

Afr.plaas > Ve.bulasi 'farm'
Afr.pyp > Ve.bibi 'pipe'
 195

Afr.pot > Ve.bodo 'pot'

Afr.krap > Ve.garaba '
scrape'
Afr.pas > Ve.basa
.pass'
Afr.papa > Ve.baba 'daddy'
Afr.seep > Ve.tshisibe 'soap'

From these examples, the Afrikaans p is adopted as b[b] when it occurs in word
initial and final position. It would appear the consonant p is approximated with the
voiced bilabial [b] in Venda when it occurs in the .initial position of a word. In the
examples such as bibi, bodo, garaba, baba the sound [p] is replaced by the voiced
sound b[b] even in the word final position. The occurrence of this sound in this
position of a word seems to be due to the influence of other voiced stop sounds
which occur word-initially. However, there are examples which are exceptions to
this explanation, e.g. pennde < Afr. pent In this example the sounct p occurs in
the initial position yet it remains voiceless. The consonant p is usually adopted as
an ejective p[p] when preceded by voiceless stops. Let us consider the following
examples:

Afr.skrop > Ve.tshikoropo 'piece job'

Afr.strop > Ve.tshitiropo 'strop'
A

Afri.kamp > Ve.kompo 'hostel'

The word. tshisibe < Afr.seep seems to be an exception to the explanation given
above. In this example the final stop sound p is replaced by a voiced consonant
[b]. However, it is interesting to note that in this example there is no preceding stop
in the source word.

The consonant p may alsobe realised as an ejective sound p[p] when it occurs in a
cluster word-initially. This can be seen from the following examples:

Afr.span > Ve;tshipano 'span'

Afr.spyt > Ve.tshipeti
""
 196
Afr.spook > Ve.tshipuku 'ghost'
Afr.plan > Ve.pulane 'plan'

7. The English consonant: !pi

This phoneme can be realised in Venda as p[p] or ph[ph]. Let us start by

considering instances where p is realised as ph[ph]:

Eng. pipe > Ve.phaiphi

Eng. polish > Ve.pholishi
Eng. porridge > Ve.phalishi
Eng. pind > Ve.phaindi
Eng. parliament > Ve.phalamennde
Eng. pavement > Ve.phevumennde
Eng. pepper > Ve.phiriphiri
Eng. puppet > Ve.phaphethe
Eng. passport > Ve.phasiphotho

In these examples the English sound p occurs word-initially and it is aspirated. In

Venda it is adopted as the aspirated sound ph[ph]. However, in some examples
given above, namely phaiphi, phiriphiri and phaphethe the ·aspirated sound ph[ph]
is also realised in medial position. This may be explained by the influence of the
aspirated sound in the initial position. This sound may be adopted as a non-
aspirated sound in the initial position when it occurs in a cluster, e.g. :

Eng.sports > Ve.zwipotso

Eng.spelling > Ve.tshipelini
Eng.speed > . Ve.tshipidi
Eng.spoon > Ve. tshipuni
Eng.spade > Ve.tshipeidi
 197

The English sound p may be adopted as an ejective sound in the medial and final
positions of a word:

Eng. report > Ve.ripoto

Eng. recipe > Ve.risipi
Eng. deep > Ve.dipe
Eng. stop > Ve.tshitopo

In a nutshell, the English unaspirated sound p is adopted as an ejective sound p[p],

and an aspirated ph as the aspirated sound ph[ph] in Venda adoptives. Substitution
·in this case has been motivated by phonetic approximation.

B. The Afrikaans sound : lx/

The Afrikaans sound x may be realised as a [ij'h], [x] or [fi] in the adopted word:

'
Afr.geld > Ve.tshelede money'
Afr.agent > Ve.axennde 'lawyer'
Afr.gemmer > Ve.xemere 'ginger'
'
Afr.saag > Ve.saha saw'
Afr.brug > Ve.buroho 'bridge'

In the first example the Afrikaans [x] is adopted as [ij'h]. The sound /xl may be
pronounced as xelt, 'xeldg, (celt) (cf. Le Roux en Pienaar 1976 :65; Kruger 1 965:62).
The sound [ij'h] in Venda might be a result of the influence of the variant sound
[ij'h] in the last example ( personal communication with Professor Louwrens). In the
adoptives axennde and xemere the sound /xl is reproduced without change. But
this is not the case with the last two examples. In these examptes lxl occurs in
word final position and is adopted as voiced [fi]. The voicing in the Venda sound
h[fi] may be attributed to the final vowels [a] and [J] which are voiced by nature.
 198

9. The Afrikaans sound /wl

This sound is usually adopted as vhl13] in Venda.·

Afr.werk > Ve.vherega 'work'

Afr.week > Ve.vhege ' week'
Afr.winkel > Ve.vhengele 'shop'
Afr.kruiwaen > Ve.girivhane ' wheelbarrow'

The explanation to the replacement of the sound w by 1131 in these examples may
be explained by looking at the phonetic characteristics of t.hese sounds. The
Afrikaans sound w is described by Kruger ( 1965:48) as "'n stemhebbende bilabiaal-
velere wrywinglose kontinuant met sentrale vrylating". This sound is softer (less
tense) than the Venda v[v] yet it is stronger (more tense) than the- Venda semi-
vowel w[w]. Therefore 1131 seems to be the closest sound to the Afrikaans sound w.
These sounds share bilabial qualities, i.e. they both involve the lips in their
production.

(c) General remarks on the adaptation of consonants in Venda

In the aforegoing discussion consonant substitutions in Venda adoptives has been

discussed. We began by discussing the substitution of the consonants which do not
have equivalents in the adopting language. These consonants are th[9], th[t'5] and
1[1], These sounds are restrained from occurring in the Venda language by the
native segment inventory constraint Therefore they are replaced by sounds which
are phonetically closest to them.

The substitution of consonants from other languages which have equivalents in the
Venda language was also discussed and included the following consonant
substitutions:
 199

Afrikaans sound k > Ve .g[g], k[k]

English sound k > Ve. k[k], kh[kh]
Afrikaans sound b > Ve. vh[{3], b[b]

English sound b > Ve. vh[{3], b[b]

Afrikaans sound t > Ve. d[d], d[d], t[t], t[t]

A r..
English sound t > Ve. t[t], th[th]
Afrikaans sound p > Ve. b[b], p[p]
English sound p > Ve. p[p], ph[ph]
Afrikaans sound 9 > Ve. tsh[t.fh], x[x], h[fi]

Afrikaans sound w > Ve. vhfll]

The substitutions in these consonants have been motivated by phonetic factors.

There are close similarities between the foreign sound and the one that replaces it
in the source language.· The replacement of consonants which are part of the
sound inventory of the source language with different ones may be explained with
reference to the enviroment in which that sound occurs. For instance, certain
sounds were replaced because of the influence of other preceding sounds. Most of
the stop sounds which occur in clusters were retained unchanged in Venda
adoptives.

3.3.3 Syllable adjustments

The syllable structure of Venda was discussed in section 3.2.4. The focus here will
be on the syllable structure adjustment to adopted words which do not conform to
the Venda syllable structure. We shall begin by looking at adjustment on consonant
clusters and then at adjustments to closed syllables, i.e. word final syllables which
end with consonants.

(a) Adjustments to consonant clusters

It was already indicated at the beginning of this chapter that consonant clusters,
except for the nasal compounds, are not characteristic of the Venda language.
 200
Therefore, when consonant clusters occur in the source language, they are
modified in one way or another when adopted into Venda. Here we shall be looking
at the different strategies which are employed by the Venda language to modify
consonant clusters from foreign languages. Only clusters in words from English and
Afrikaans will be considered in this section. We shall begin by looking at clusters
which occur in the word initial position ( i.e. clusters which serve as onsets).

Clusters which occur in word initial position

The following English and Afrikaans consonant clusters can occur in the initial
position of a word where they serve as onset syllables:

/prJ, /br/, /pll, ltrl, ldrl, Jgrl, /kll, kr/, /gll, /bll, /fU, /fr I, ls+CI etc.

1. English consonant cluster: /prJ

The consonant cluster /pr/ is simplified (i.e. broken up) by inserting a vowel which
is either [::>] or [u]. This can be clearly seen from the following examples:

Eng.profit > Ve. purofitil porofiti

Eng.prophet > Ve.mupurofita/ muporfita
Eng.prosecute > Ve. phurosikhutha
Eng.project > Ve.purodzhekiti
Eng.promote > Ve.-puromota/ poromota
Eng.program > Ve.purogireme /porogireme
Eng.prose > Ve.purosa I porosa
Eng.protein > Ve. puroteini/poroteini

•
Eng.president > Ve.puresidennde
Eng,press >· Ve.purese

Eng. principal > Ve. phirisipala

 201

Eng.printer > Ve.phirintha

Eng.priest > Ve.mupirisita
Eng. prefect > Ve.phirifekhithi

From the examples given above the following observation can be made:
(i) All the pr clusters are broken up by an epenthetic vowel.
(ii) The choice of the epenthetic vowel is determined by the process of assimilation.
Assimilation in this case can be seen as occurring in two ways. First, the
epenthetic vowel assimilates to the preceding consonant. In most of the examples
given above, the realisation of the epenthetic vowels u and o is due to the
influence of the preceding bilabial consonants. The influence of bilabial consonants
on the realisation of back epenthetic vowels has been noticed in other languages
such as Northern Sotho (Kruger 1965), Sotho (Warmelo 1929), ShoAa {Chimhundu
1982), Zulu (Nkabinde 1968). Back vowels, like bilabial consonants, involve lips in
their production. This may be the reason for a back vowel, particularly u, to be
inserted after the bilabial consonants.

Assimilation can occur in vowels across syllable boundaries. Some of the

epenthetic vowels have variants, e.g puromota- poromota, purosa -porosa. In these
examples u alternates with o. The realisation of o in poromota and porosa is due
to assimilation to the succeeding vowel. Vowel assimilation can also be seen in
example_s such as phirisipala, phiritsha, mupirisita. The epenthetic vowei i in these
examples is due to assimilation to the succeeding vowel. The insertion of
epenthetic vowels within the consonant cluster pr can be represented in the word
profrt as follows {Carlisle 1991 :79):

II\
c c
/~ v c v c Underlying representations

p
I I I I
r o
I
f
 202

c v
/~ v c v Epenthesis

p
I
u r
I 0
f I\ \
f i t a

1\
\I
Resyllabification

p u r 0 f t a

2. English consonant cluster: /br/

Eng.break > Ve.bureiki

Eng.brush > Ve.bulatsho
Eng. brake > Ve.biriki

The epenthetic vowels p] and [u] in the above examples have simplified the
consonant cluster /br/. In the first two examples the epenthetic vowel is [u]. The
influencing factor here is the bilabial consonant [b]. However, this is not the case in
the third example. Here the bilabial consonant [b] is followed by a front vowel [i]. It
appears that this vowel is a result of vowel assimilation. The vowel [i] of the
diphthong ei which succeeds the cluster br has retrogressively influenced the
realisation of the epenthetic vowel p].

3. English consonat cluster: lpU

The consonant cluster /pU is separated by inserting the vowel [u]. As has already
been mentioned, this epenthetic vowel is determined by the preceding bilabial
consonant [p] and also by vowel assimilation.
 203

Eng.plan > Ve.pulene

Eng.plank > Ve.bulannga
Eng.plastic > Ve.pulasitiki
Eng.pleat > Ve.piliti
Eng. plot > Ve.puloto
Eng. plug > Ve.pulaga
Eng.plumber > Ve.mupulambara
Eng.planet > Ve.pulanete
Eng.plate > Ve.pulethi
Eng.plantation > Ve. pulanthesheni

The epenthetic vowel [i] in the adoptive piliti requires a brief comment. The choice
of this vowel. is determined by the vowel [i] of the succeeding syllable.

4. English consonant cluster : /tr/

The consonant cluster /tr/ is simplified by inserting the vowel [i]. The choice of this
vowel is determined by the alveolar consonant t which precedes .

Eng.trunk > Ve.thirannga

A

Eng.truck > Ve.thirakha

A

Eng.tray > Ve.thirei

.....
Eng.traffic > Ve.thirafiki
"'
Eng.triangle > Ve.thiraengele
"'
Eng.trophy > Ve.thirofi
A

Eng.tropic > Ve.thiropika

"'

The influence of the alveolar consonants on the realisation of epenthetic vowel [i]
can be explained by considering the similarities between these types of sounds.
Alveolar sounds are produced when the tongue is in a high position and with the
 204
lips spread. The vowel [i] also shares these articulatory features. It seems that this
epenthetic vowel is preferred in this cluster because of these phonetic similarities
(Hermanson 1991 ).

5.English phoneme cluster: ldrl

The consonant cluster /dr/ like the /tr/ cluster is simplified by inserting the high
vowel [i]. The choice of this vowel is also determined by the alveolar consonants or
dental [d].
/\

Eng.driver > Ve.diraiva

A

Eng.drum > Ve.diramu

A

Eng.drill > Ve.dirili

"
Eng. drug > . Ve.diraga
A

Eng.dribble > Ve.dirivhuala

A

Eng.drawer > Ve.dirowara

"
Eng.drama > Ve.dirama
,..

The fact that some of the English alveolars are realised as dentals in Venda does
not affect our hypotheis of the influence of the alveolar sounds on epenthetic vowel
[i]. The alveolar of the source language is the one that seems to influence the
realisation of [i].

6.English consonant cluster: /grl

The vowel p] is interposed between the consonant cluster /gr/ in the following
examples:

Eng.gravel > Ve.giravhulo

Eng.grease > Ve.girisi
Eng. grammar > · Ve.girama
Eng.gram > Ve.gireme
 205

Eng.grey > Ve.gireyi

In some cases the consonant cluster is broken up by the epenthetic vowel [u]:

Eng grocer > Ve.gurozara

Eng.group > Ve.gurupu
Eng.groovy > Ve.guruvi
Eng. grudge > Ve.giradzhi

The choice of the epenthetic vowel in the examples given above seems to be due
to the influence of the vowel which immediately follows the cluster. It appears the
epenthetic vowel [u] occurs when the gr cluster is followed by a back vowel, i.e. u
or a diphthong such as [~u]. In all the positions where the cluster is not followed by
a back vowel, the epenthetic vowel is i[i].

?.English consonant cluster : /kU

The consonant cluster /kU is separated by the epenthetic vowel i[i] or u[u]:

Eng.club > Ve.kilabu

Eng.class > Ve.kilasi
Eng.climate > Ve.kilima
Eng.clinic > Ve.kiliniki
Eng.clutch > Ve,kilatshi
Eng.club > Ve.kilaba
Eng.clerk > Ve.kilaka
Eng.closet > Ve.kulosethe
Eng.clock > Ve.kuloko

The choice of the epenthetic vowel also seems to be determined by the nature of
the vowels that immediately follow the cluster. From the examples given above the
epenthetic vowel [u] occurs in cluster which are immediately followed by a back
 206
vowel while [i] occurs when the cluster is followed by any other vowel which is not
a back vowel.

B.English consonant cluster : /gV

The cluster JgV is broken up by inserting the vowels [i] or [u]. The choice of the
epenthetic vowel [u] seems to be mainly determined by the influence of the vowel
following the consonant cluster.

Eng. globe > Ve.gulupu

Eng.glue > Ve.guluu
Eng.glottal > Ve.gulotala
Eng.glucose > Ve.gulukhousi

The insertion of the epenthetic vowel i[i] in the first example seems to be
determined by the non-back vowel u[ v ] which follows the cluster gl. In all the other
examples, as already mentioned the realisation of the epenthetic vowel u[u] seems
to be a result of the influence the back vowels following the cluster.

9. English consonant cluster : /bV

The consonant cluster /bV is broken up by inserting the back vowel [u], as can be
seen from the following examples:

Eng.blouse > Ve.bulausu

Eng.block > Ve.buloko
Eng.blue > Ve.buluu
Eng.blazer > Ve.buleizara

In all the examples given above, . the epenthetic vowel [u] is inserted. The
occurrence of this vowel in the first three examples may be explained by the
influence of the back vowels which follow the cluster bl. However, But this
 207
explanation is countered by the epenthetic [u] in the last example. The cluster in
this example if followed by the diphthong ei. This diphthong has nothing to do with
backness. ln this instance it appears the influence of the bilabial consonant [b]
offers a better explanation. As already mentioned, the bilabial consonants play an
important role in the realisation of the epenthetic vowel [u] in Venda adoptives,
even overriding the potential influence of the vowel following the consonant cluster.

10. English consonant cluster : ltv

The cluster ltv is broken up by the epenthetic vowel [u] and [i] in Venda:

Eng.flour > Ve.fulauru

Eng.flat > Ve.fulethe
Eng.flute .> Ve.fuluthu
Eng.flag > Ve.fulaga
Eng.flask > Ve.fulasiki
Eng.flannel > Ve.fulanele
Eng.flu > Ve.fuluu
Eng. floor > Ve.fuloro
Eng.flea market > Ve.filiimakete
Eng.flicker > Ve.-filikhara

The epenthetic vowels [u] and Pl have been used in these examples to separate
the consonant cluster fl. The realisation of the epenthetic vowel [u] appears to be
due to the influence of the bilabial characteristics of the denti-labial sound f. This is
not the case in the last two examples. Here the vowel [i] has been used to break
down the cluster. The choice of this vowel is due to assimilation with the
succeeding vowel. Although vowel assimilation may also be use.d to explain the
realisation of the epenthetic vowel [u] in words such as fuluthu and fuluu, this does
not rule out the influence of the bilabial characteristics of the consonant f in these
examples.
 208
11.English consonant cluster: fr

The adjustment of the consonant cluster /fr/ in the following examples is

interesting. The cluster /fr/ is broken up by the epenthetic vowels [i] and [u]:

Eng .fridge > Ve.firidzhi

Eng .frill > Ve.firili
Eng. France > Ve.Furannsi I Fura
Eng.Frank > Ve.Furennge
Eng .frame > Ve.fureme
Eng .fresh > Ve.fureshe

While the epenthetic vowel [u] can be explained with reference to bilabial
characteristics of the consonant [f], the epenthetic vowel [i] cannot be explained by
this factor. The epenthetic vowel [i] is realised in the first two examples. It is
interesting to note that all the other syllables (i.e medial and final ) in these first two
words are characterised · by the vowel [i]. Therefore the occurrence of the
epenthetic vowel [i] in the cluster fr may be regarded as being due to assimilation
with the succeeding vowel i.e. vowel [i].

12. The onset consonant clusters discussed above a/so occur in Afrikaans.

The way in which the Afrikaans consonant clusters are treated is no different to
that of English. The vowels interposed between the Afrikaans clusters· are the
same as those in English adoptives. Therefore the Afrikaans consonant clusters
will not be treated separately here since this will not bring any new evidence to this
study. Instead, we shall only give a few examples:

Afr.profeet > Ve.mupurofita/ muporofita 'rophet'

Afr.brood > Ve. vhurotho 'read'
Afr.plaas > · Ve.bulasi 'arm'
Afr.trein > Ve.tireni 'rain'
A
 209

Afr.drom > Ve.cJJromu 'rum'

Afr.draad > Ve.dirata
/". 1\
'wire'
Afr.blaar > Ve.bulara 'tobacco leaf

13. English and Afrikaans consonant cluster: /s+C/

The consonant cluster /s+C/, is very common in English and Afrikaans adoptives in
Venda. The following illustrate some of these clusters:

/sp-/, /st-1, /sk-1, /sl-/, /sn-/ etc.

Eng.station > Ve.tshititshi

A
"station"
Eng.stamp > Ve.tshitemmbe "stamp"
"
Eng.scheme > Ve.tshikimu
Eng.Scottish > Ve. Tshikotshi
Eng.scarf > Ve.tshikhafu

The examples given above show the adjustment of consonant cluster /s+C/ in the
word initial position (i.e. onset syllables). These clusters are adjusted by replacing
the consonant s with the noun prefix tshi- of class 7. The replacement of s in this
cluster by the prefix tshi- seems to have been motivated by morphological factors
rather than phonological ones. Thus assimilation to the following vowels does not
play a rol.e in these examples. One would expect the sound s to be retained and
the epenthetic vowel to be inserted between the cluster, or the s to be deleted, as
is the case with verb roots. Instead the s sound is interpreted as the prefix tshi-.
The same changes occur to Afrikaans clusters with /s+C/ as onsets. Consider the
following examples:

Afr.speiker > Ve. tshibigiri nail'

Afr.stoel > Ve.tshidulo 'chair'
Afr.steen > Ve.tshidina 'brick'
Afr.skuld > Ve.tshikolodo 'debt'
 210

Afr.skeer > Ve.tshigero 'scissor'

A few remarks on the adjustment of consonant clusters to adopted verb roots

Although the adjustment of clusters in some adopted verb roots is the same as that
of nouns, this is not the case with all roots. The consonant clusters in certain verb
roots are reduced to single consonants by means of deletion. Consider the
following examples:

Eng.discharge > Ve.-tshatsha

Afr.skuld > Ve.-koloda

Eng.stamp > Ve.-temmba

/'

Afr.skeer > Ve.-gera

Afr.spook > Ve.-puka

Afr.sterk > Ve.-tereka

A

Afr.straf > Ve.-tarafa

Afr.span >
"'
Ve.-pana
Eng.scheme > Ve.-kima

In the first example the initial syllable has been deleted. In the other examples it is
only the first element of the cluster, namely, s which has been deleted. If these·
clusters had been in nouns, the prefix tshi- of class 7 would have been used to
break up the cluster. From these examples we can see that clusters are adjusted
in a unique way in some verb roots. Consonant deletion also occurs in verb roots
which have no initial clusters, e.g.:

Eng. report Ve.-pota

The first syllable has been deleted in the example given above.
 211
The adjustment of clusters that occur in medial and word-finally

The changes that occur to clusters which occur in the initial position sometimes
differ from that of clusters that occur in medial and word final position. Let's look at
the following English examples: .

Eng.fullscap > Ve.fulusikepe /fu!uskepe

Eng.custard > Ve.khasitadi /khastadi
"""
Eng.passport > Ve.phasipoto /phaspoto
Eng.basket > Ve.basikete /baskete
Eng.biscuit > Ve.basikitsi /baskitsi
Eng.bastard > Ve.basitere
A.
/bastere

Although the /s+C/ consonant clusters in the initial position of the adopted nouns
generally become /tshU, this is not the case in these example. The lsi sound is
retained and an epenthetic vowel may be inserted or it can be left out, as can be
seen from the adoptives in the second column. In the adoptives in the second
column the cluster is retained. The insertion of the epenthetic vowel, as already
indicated, is due to a constraint against the sequence of consonants in one
syllable, yet the occurrence of consonant clusters in the examples given above
shows a relaxation on this language constraint. The consonant constraint against
the occurrence of clusters may be violated by people who have a thorough
knowledge of the source language sound system. Such people often prefer to
reproduce the sound clusters without modifications.

The consonant clusters which occur in the final position of words may also be
modified. Consider the following Afrikaans examples:

Afr.dorp > Ve.dorobo

A
'town'
Afr.kart > Ve.garata
A
card'
Afr.verf > Ve.firifa 'clean (floor)'
Afr.tolk > Ve.dologi ·interpreter'
 212
Afr.melk > Ve.melege 'milk'
Afr.kerk > Ve.kereke 'church'
Afr.hark > Ve.haraga 'rake'
Afr.geld > Ve.tshelede 'money'

In these examples the following clusters occur:

1-rpl, 1-rtl, 1-rfl, 1-lkl, 1-rkl, /-ttl, etc.

In all the examples the cluster is broken up by a vowel which in each example is
determined by the vowel of the preceding syllable. In other words, the choice of the
epenthetic vowel is due to the process of assimilation, whereby the epenthetic
vowel assimilates to the vowel of .the preceding syllable.

{b) Adjustment to words which end with closed syllables

Certain words adopted from foreign languages such as English and Afrikaans end
with consonants. This type of syllable is referred to as a closed syllable. Closed
syllables are not permitted in Venda, with a few exceptions of some ideophones.
Therefore, whenever a foreign word which ends with a consonant is adopted to
Venda, some modifications to such a syllable are inevitable. The modifications
occur in different ways. In some instances a vowel is inserted whereas in others
the final.consonant is deleted. These processes will be discussed below. We shall
begin by looking at the instances where the vowel is added onto words that end
with consonants.

(i) Vowel epenthesis

Words which end with a consonant, with the exception of ideophones, are not
acceptable in Venda. Such words or syllables are adjusted by adding a vowel onto
them. Consider the following examples:
 213

Afr.broek > Ve.vhurukhu 'trouser'

Afr.brood > Ve.vhurotho 'bread'
Afr.boor > Ve.vhoro 'drill machine'
Afr.boer > Ve.livhuru
,.... ·boer/farmer'
Afr.bybel > Ve.bivhili 'bible'
Afr.band > Ve.bannda 'belt'
Afr.boek > Ve.bugu 'book'
Afr.bout > Ve.baudu 'bolt'
Afr.brug > Ve.buroho 'bridge'
Afr.botter > Ve.bodoro
A
'butter' .
'
Afr.lekker > Ve.legere sweet'
A.
Afr.plek > Ve.bulege 'place'
Afr.blik > Ve.lulege 'tin'
Afr.geld > Ve.tshelede 'money'
Afr.kart > Ve.garata
,... 'card'

The choice of the epenthetic vowel in these examples is mainly due to vowel
assimilation. The epenthetic vowel is determined by the vowel of the preceding
syllable(s). However, the epenthetic vowel may also be determined by the nature
of the preceding consonant, as can be seen from the following examples:

Eng.glove > Ve.gilavu

Eng.steam > Ve.tshitimu
Eng.drum > Ve.diramu
,....
Eng.scheme > Ve.tshikimu

In all these examples the epenthetic vowel is [u] and is preceded by a bilabial
consonant or a consonant with bilabial characteristics. As already mentioned,
bilabial consonants or consonants with bilabial characterics seem to have some
influence in the realisation of the back vowel [u] (cf. Van Warmelo 1929, Whiteley
1967; Nkabinde1968; Hermanson 1991 ). The vowel [u], like the bilabial
consonants, involve the lips in its production.
 214
The bilabial conditioning can also be realised with Afrikaans adoptives:

Afr.kam > Ve.gamu 'comb'

Afr.dam > Ve.damu 'dam'
Afr.riem > Ve.rimu 'ream'
Afr.drom > Ve.diromu
,... 'drum'
Afr.blom > Ve.bulomu 'flower'

In the following examples the bilabial conditioning did not apply. Instead the
epenthetic vowel assimilated to the preceding vowel.

Eng. gram > Ve.gireme

Afr skelm > Ve.tshikeleme 'crook'

It is not clear in these two examples as to why the epenthetic vowel did not
assimilate to the bilibial consonant /m/ to be realised as a back vowel. In other
words, it is not clear what prevented the application of a bilabial conditionig rule in
these examples since in all of them there is the bilabial consonant /m/ which
provides the necessary condition for it to apply. In some words, 'blom' for
example, the epenthetic vowel may be determined by vowel assimilation as well as
assimilation to the preceding consonant. Thus the Afrikaans word "blom" may be
pronounced as bulomu (bilabial conditioning) or bulomo (vowel assimilation). One
speaker .may use these forms interchangeably. It seems there is no clear principle
which determines the application of any of these types of assimilation.

The choice of the epenthetic vowel in the word final position may also be
influenced by alveolar consonants:

Afr.trein > Ve.tireni 'train'

Afr.dres > Ve.diresi 'address'
Afr.klas > · Ve.kilasi ·'class'
 215
The realisation of the epenthetic vowel [i] in these examples may be explained by
the preceding alveolar nasal [n] and [s]. The influence of alveolars on the
realisation of front vowels has already been discussed in the previous sections.

A few remarks on the adjustments of word final consonants in verb stems

The manner in which final consonants are adjusted in adopted verbs is different·
from that of the other word groups.

Afr.duur > Ve.-dura

.expensive'
,....
Afr.kerk > Ve.-kereka ·attend church service'
Afr.tolk > Ve.-do!oga
I'
'interpret'
Afr.leer > Ve.-lera 'learn'
Afr.terk > Ve.-tereka
I'
'strike'
Afr.werk > Ve.-vherega ' work'

Afr.skuld > Ve.-koloda 'borrow'

Afr.hark > Ve.-haraga ' rake'

Eng.report > Ve.-pota

Eng.deposit > Ve.-diposita

Eng.brake > Ve.-birika
Eng.brush > Ve.-burasha

Eng.drill > Ve.-diri!a

A

Eng.short > Ve.-shotha

In all these examples the final vowel is [a]. In Venda verbal stems end with vowel
[a] in positive constructions. Therefore the realisation of the vowel [a] in these
examples cannot be explained by phonological factors, but by morphological
factors. All the phonological rules discussed above do not seem to apply here. If
the verbal stems given above are used in negative constructions, the final vowel
will change to [i]. This indeed shows that the final vowel in adopted verbal stems is
determined by morphological factors.
 216

(ii) Consonant deletion

Although the nasal consonant can form a syllable in Venda, it cannot do so as a

coda (the consonant which occupies a final position in a syllable). The following
examples show how English coda, a nasal in this regard, is adjusted in Venda:

Eng.station > Ve.tshititshi

A

Eng.location > Ve.lokheshi

Eng.kitchen > Ve.khishi
Eng.garden > Ve.ngade

The segment [n] in the examples given above is deleted. The reason for such
deletion appears to be due to the fact that it resembles the locative suffix -ni of
Venda. It seems that the speakers of the language delete this consonant to avoid
confusion between nouns and locatives. The word tshftitshi
/""".
is a noun whereas
tshftitshini ·at the station' is a locative form derived from the noun tshftitshi. From
A A

the nouns lokhishi, khishi, ngade the following locatives can be derived
respectively, viz. lokhishini 'in the location', khishini 'in the kitchen', ngadeni 'in
the garden'.

It is clear that the deletion of the final consonant [n] is determined by semantic and
morphological factors rather than purely phonological factors.

3.3.4 Tonal adaptation of adoptives in Venda

As was mentioned in the previous section, Venda is a tone language. Words

'
adopted from foreign languages are adapted to the Venda tonal system. Here we
shall only consider words adopted from non-tonal languages such ·as English and
Afrikaans. Studies done so far on the adaptation of words from non-tonal
languages have established that the stressed syllables in such languages are
realised as high tone in African languages (cf. Kruger 1965; Chimhundu 1982; Britz
1983). In this section we shall see whether this is also applicable to Venda
 217
adoptives. The words adopted from other languages appear to be subjected to
Venda tonal patterns. Nouns, for example, have their own specific tonal pattern,
particularly with regard to prefixes. Poulos has this to say with regard to nouns:

" ... one can make· the blanket statement that Venda noun prefixes are low-
toned in nature, with the exception of class 2b, which is high. In other words,
the basic tone on the noun prefix is generally low {L)" {1990:569).

Noun prefixes are assigned to adopted nouns. Some of these noun prefixes are
overtly marked whereas others are not. We shall first consider the assignment of
tone to nouns with overtly marked prefixes:

Eng.typist > Ve.muthaiphi

Eng.planner > Ve.rriupulani
Eng. machine > Ve.mutshini
Eng.apple tree > Ve.muapula
Eng.stamp > Ve.tshitemmbe
A

Eng.station > Ve.tshltitshi

A

Afr.skool > Ve.tshikolo 'school'

Afr.boer > Ve.livhuru 'farmer'
A

Afr.boom > Ve.mubomo 'tree'

All the prefixes in these examples are marked with a low tone. This therefore
indicates that the Venda prefix tonal pattern has not been affected by the
incorporation of adopted nouns. What is interesting in these examples is the
correlation between the stressed syllables in the source languages and the .
assignment of a high tone in Venda. In cases where the onset in the source
language consists of consonant clusters, the first segment is realised as a prefix.
This prefix always has always a low tone. The second segment of the cluster
becomes part of the stem and, if it is stressed in the source language it is assigned
a high tone in Venda. In all the examples given above, the prefixes are overtly
marked and have low tone. All the stressed syllables which come after the prefixes
 218

are marked with high tone. While the examples given above seem to give the
impression· that stressed syllables are assigned. a high tone in Venda, there are
other examples which do not comply to this generalisation:

Eng.deacon > Ve.mudik6ni

Eng. president > Ve.mupurisldennde
Eng.prophet > ve.muporofita
Eng.lorry > Ve.lori

The discrepancy in the realisation of tone in these examples may be explained by

considering Afrikaans as the source of these words instead of English. Let us
consider their Afrikaans counterparts:

Afr.diaken > Ve.mud1k6ni

Afr. presi'dent > Ve.purisiderinde
Afr. pro'feet > Ve.muporoffta·
Afr.lor'rie > ve.lori

The stressed syllables in these examples are realised as high tones in Venda.
Therefore, tone can also help in distinguishing between the adoptives from
Afrikaans and English, especially with regard to words which are similar in these
two languages.

Some of the adopted nouns which have no overtly marked prefixes but begin with
a stressed syllable are realised as high tones and the non-stressed ones as low
tone. Let us look at the following examples:

Afr.kamp > Ve.gammba 'camp'

Afr.pampier > Ve.bammbiri 'paper'
Afr.winkel > Ve. vhengele 'shop'
Afr.kamer > ve.kamara 'room'
Afr.kam > ve.gamu 'comb'
 219

Afr.draad > ve.dirata

A
'wire'
>
"'
·Afr.kettel ve.gedela 'kettle'
"'
Eng.cup > ve.khaphu
Eng.cake > ve.khekhe ·
Eng.team > Ve.thimu
Eng.driver > Ve.dfraiva
A

Eng. truck > ve.thfrakha

A

Eng.drama > Ve.dlrama

A

Eng.drill > Ve.dirili

A

Eng.club > ve.kilabu

In the above examples stressed syllables of the source languages tally well with
the high tone syllables in Venda. However, the following examples seem to deviate
from the generalisation made above. There does not appear to be a clear
explanation for these exceptional forms.

Afr.kerk > Ve.gereke 'church'

Eng.referee > Ve.refiri

Eng. caravan > Ve.kharavani

Eng. committee > Ve.komiti

Eng.company > ve.khamphanf

Eng. recipe > Ve.rlsfp[

The assignment of tone on verbal stems also seems to be based on the stress-
tone relationship:

Afr.duur > Ve.-dura ' expensive'

""
Afr.ry > Ve.-refla 'drive'
Afr.laai > ve.-laisa 'load'
Afr.kerk >
"
ve.-kereka 'attend church service'
Afr.tolk · > ve.-d616ga ·interpret'
/\.

Afr.leer > ve.-lera 'learn'

 220
Afr.werk > ve.-vherega ' work'

Afr.skuld > ve.-k616da 'borrow'

'
Afr.hark > ve.-haraga rake'
Eng.report > ve.-p6ta
Eng.deposit > Ve.-dlp6sita
Eng.brake > ve.-blrlka
Eng.brush > ve.-burasha
Eng.drill > Ve.-dirila
"
Eng .freeze > Ve.-firiza
Eng.short > Ve.-sh6tha

As can be ascertained from the above examples, the stress on the initial syllables
of English and Afrikaans words is usually realised as high tone in Venda. All the .
final syllables are realised as low tones.· This is typical of the tonal behaviour of
verbal stems in Venda when used in isolation. The tonal pattern of a verb stem
may change when certain morphemes such as verbal extensions are added to it.

What is interesting about the above examples is the assignment of tone to syllables
which are not stressed in English and Afrikaans. The unstressed syllables would
be expected to be realised as low tones in Venda, but this is not always the case
in the examples given above. Unstressed syllables are realised as both high and
low tones. This can be seen from adoptives such as dip6sita (LHHL) <
Eng.dep()sit. Here, the high tone occurs in the unstressed syllable -si-. The
question, therefore, is what determines the assignment is such cases. In some
instances the high tone, which corresponds to the stressed syllable of the source
language, influences the tones of the succeeding syllables. In this case the high
tone spreads to the adjacent syllables. Tone spread in the adoptive deposita can
be illustrated as follows:
 ..

221
L H L L

\
di po si
I ta

L H L L

I
di po
l',j
~i ta Low Tone deletion

H L L

di
.
po
\',\si ta High Tone Spread

In this example a high tone is spread to the adjacent low tone syllable to the right.
The question may arise as to how one determines the tone of a syllable in a word.
The guiding principle here is stress. As has been mentioned, the stressed syllables
are more often than not perceived as high tones in Venda.

The following are more examples of words in which the high tone is spread to the
adjacent low tone syllables:

dzha.mu (cf.Eng.jam)
khisf (cf.Eng.kiss)
bugu (cf.Afr. boek)
bege (cf.Eng.bag)
m6d6r6 (cf. Eng .motor)
"
tshlk616 (cf.Afr. skoal)
phathf etc. (cf.Eng.party)
 222
3.4 Concluding remarks on the effects of adoptives on Venda phonology

The adoption of foreign words in Venda seems to have influenced the Venda
sound system in many ways. Some of the language constraints against foreign
forms appear to have been weakened. Certain new sounds like [8], [0], [I] seem to
be infiltrating the sound inventory of the language.

The effect of adoptives on the Venda sound system is evident from the way
consonant clusters are adjusted to the language. Whereas consonant clusters are
broken up by an epenthetic vowel in some cases, in others they are retained,
especially in recent adoptives eg. -grivhuala, khrimu, klabu, blasi etc. This is a
violation of the native phonotactic and syllable structure constraints.

Despite these influences, the Venda language appears to have coped well .in
modifying other phonological aspects of adopted words. The tonal constraint
seems to have applied in most of the cases.

3.5 Conclusion

An attempt was made in this chapter to discuss the phonological adjustments of

adopted words in Venda. Before the discussion on the phonological adaptation of
foreign words to .Venda was undertaken, the Venda sound system was outlined
and different language constraints were discussed and exemplified. In discussing
the phonological changes that adopted words undergo, different explanations were
offered, namely, the phonetic approximation theory, the phonemic approximation
theory, the generative approach, natural phonology, autosegmental phonology and
the language constraints hypothesis. The autosegmental approach was used to
explain some of the changes that have occurred to adopted words. · With regard to
the phonological modification of the adopted words processes such as substitution,
assimilation, consonant harmony, consonant cluster simplification, adjustment to
the syllable structure,· etc. were discussed. An attempt was also made to show the
adaptation of foreign words to the Venda tonal system. In conclusion, it appears
 223

that although the language is coping well in adapting foreign word forms to its
sound system it has also been affected by these foreign influences to some
degrees. New sounds and sound sequences have been introduced to the
language. This could well indicate that there is a weakening of some constraints in
the Venda language.
 224

CHAPTER 4

4.0 GRAMMATICAL ADJUSTMENT IN ADOPTIVE FORMS

4.1 Introduction

Words adopted from other languages are adapted to the Venda grammatical
structure. The word 'grammar' is used here to refer to both morphology and syntax.
With regard to morphology we shall be looking at the morphological adaptation of
foreign words to Venda. Here we shall consider the morphological adaptation of
adopted words in different parts of speech such as nouns, adjectives, verbs,
adverbs and conjunctives. We shall also look at some syntactic aspects of the
adopted words. As our point of departure , we shall look at nouns since most of the
words which have been adopted into Venda are nouns. But before discussing the
adaptation to adopted nouns, the morphology of the Venda noun will be briefly
outlined. This will help to provide a better understanding of the changes that occur
to adopted nouns.

4.2 Morphology and some syntactic aspects of the adoptive nouns in Venda.

4.2.1 The morphology of the Venda noun

A noun can be described as ,.a word or group of words that refers to a person,
place, or thing or any syntactically similar word" (Collins English Dictionary). In this
regard a noun is defined according to its meaning and its syntactic behaviour. With
regard to meaning, our focus will be mainly on the noun prefixes. The prefix
indicates the general characteristics of a thing, person, animal or general category
to which the thing named belongs in the mind of the speaker (Westphal 1945:90).
This can be illustrated by the following examples:
 225
Muvenda
;.
(a Venda person)

,..
Tshivenda (Venda language or culture)
Vhuvenqa (some Venda characterictics)
Kuven)ta (small Venda person)

Different noun prefixes have been usecf in these examples with the noun stem -
venda. The use of the different prefixes with this noun stem results in a shift of the
"
meaning of the stem.

Secondly, the noun prefix has a classificatory function. In this case noun prefixes
may be seen as occurring in different classes. This has long been established by
Bantu linguists such as Sleek, Guthrie and Meinhof like, just to mention a few.
According to Meinhofs ( 1906 ) system Venda noun classes can be listed as
follows:

Class Prefix
1 mu-
1(a) (/)

2 vha-
2(a) vho-
3 mu-
4 mi-

....li-
5
6 ma-
7 tshi-
8 zwi-
9 N-
10 dziN-
11 lu-
14 vhu-
15 u
16 fha-
 226

17 ku-
18 mu-
19 ku-
21 di-
I'

Nouns in Venda are therefore said to belong to anyone of these noun classes.
However, according to Westphal (1946:47) "the principle of this classification is not
known yet and will probably never be determined completely". What has been
established so far is that noun stems may be assigned to the different classes on
the basis of meaning and number. We have already indicated that prefixes affect
the meaning significance of the noun stems. Therefore noun class prefixes have
traditionally been taken to represent different semantic significances. Each noun
stem is assigned to a class prefix according to the semantic significance which the
. -
speaker intends to express. Although Hendrikse and Poulos (1992) have come up
with a new approach, they still acknowledge the role of semantic significance in
the classification of nouns. Their approach will be discussed in relation to adopted
nouns later in this chapter. They summarise the semantic significances of the
different noun class prefixes of the Bantu languages as follows (here, only the
classes that apply to Venda have been considered) :

Class 1 Human beings

Class 1(a) Proper names
Kinship
Personifications
Class 2. Regular plural of pass 1
Class 2(b) Regular plural of Class 1(a)
Class 3 Natural phenomena
Body parts
Plants
Animals
Class 4 Regular plural of Class 3
Class 5 Natural phenomena
 227
Animals
Body parts
Collective nouns
Undesirable people
Augmentatives
Derogatives

Class 6 Regular plural of Classes 5 and 14

Mass terms and liquids
Time references
Mannerisms
Modes of action

Class 7 Body parts

Tools, instruments and utensils
Animals and insects
Languages
Diseases
Outstanding people ·
Amelioratives
Derogatives
Diminutives
Augmentatives
Curtatives (Shortness and stoutness)
Mannerisms
Class 8 Regular plural of Class 7
Class 9 Animals
People
Body parts
Tools, instruments and household effects
Natural phenomena
Class 10 Regular plural of Class 9
 228

Class 11 Long, thin entities

Languages
Body parts
Natural phenomena
Implements, utensils and other artefacts
Augmentatives
Derogatives
Class 14 Abstracts
Collectives
Location terms
Infinitives
Class 15 Infinitives
Class 16 Location
Class ·17 Location
Class 18 Location
Class 20 Derogatives
Augumentatives
Diminutives
Mannerisms
Class 21 Augumentatives
Derogatives

.(Hendrikse and Poulos 1992:199-201)

From the list given above we can see that "the traditionalists presented a typical
interpretation of the system by assigning specific diverse significances to each and
every class" (Hendrikse and Poulos 1992:202). Besides showing the semantic
significances associated with different classes of noun prefixes, the 'list also shows,
as already mentioned above, that noun class prefixes which have the same
meaning may differ in terms of number. In this regard certain noun prefixes indicate
singularity whereas others indicate plurality. Therefore noun stems may be
appropriately assigned to different noun classes according to number. However,
 229
the problem with regard to classifying nouns according to number is the fixed order
of the singular-plural relationship. In this system nouns which belong to a singular
class, let's say Class A are expected to belong automatically to its corresponding
plural class, class B. Although most nouns seem to follow this trend there are those
which do not seem to adhere to this principle, as will be seen later in our
discussion of adoptive nouns. Thus some linguists like Zawawi (1979) feel that the
allocation of nouns into different noun classes should not necessarily mean that
nouns are put into rigid and fixed classes. In other words noun stems can be
changed from one class to another. The traditionalist approach also seems to
create the impression that nouns can be clearly classified according to their
semantic significances which are expressed by the noun prefixes; this does not
seem to be the case. We have seen from the aforegoing discussion that the
principle of semantic significance, though it may be useful in the classification of
certain nouns in Ve.nda, is not without problems. There is still the problem of
certain significances that are found in more than one class. If we take, for instance,
a semantic significance such as Body parts, we note that they occur in nouns of
Classes 3 & 4, 5 & 6, 7 & 8, 9 & 10 and 11. The difficulty here concerns the
allocation of a noun stem referring to body parts to a particular class. This difficulty
will be clearly illustrated in the section on the allocation of adoptives to different
noun classes. It is because of problems such as these that some linguists have
sought other ways of classifying the nouns of a language. Amongst these scholars
are Denny and Greider, as cited in the article on "A continuum interpretation of the
Bantu noun class system" by Hendrikse and Poulos (1992), who came up with
other principles which may be used in the classification of nouns. According to
them the bulk of the noun prefixes may be associated with configurational or shape
meanings. These criteria yield to the following classes of noun prefixes:

(a) Prefixes which denote kinds of entities

These prefixes are used in nouns which denote things which have "real or distinct
existence" (Collins Concise Dictionary). Nouns in these classes will, according to
them refer to kinds of people, tools and animals. Therefore most of these nouns
 230

are those which are found in classes 1/2, 7/8 and 9/10 (Hendrikse and Poulos
1992:201 ).

(b) Prefixes which denote spatial and shape configurations

Denny and Creider (1986) describe configuration classes as those in which class
prefixes denote characteristics such as the shape of the entity designated. Here
properties such as roundness, length and size play an important role. With regard
to shape certain entities may have a solid shape (e.g 3/4, 5/6 or an outline shape
such as the nouns of class 9/10, 11/10. "The contrast between 'outline' and 'solid'
shape is a contrast between objects which have clear profiles, edges or boundaries
such that there is a difference between an outside and an inside, and objects which
do not have this characteristic" (Denny and Creider 1986:217). We will not go into
detail with this approach since our objective is to merely lay a background for our
discussion of adoptive nouns. Hendrikse and Poulos (1992:202) summarise the
significance of Denny and Greider's approach as follows:

"Denny and Creider, in their analysis, make a significant contribution by

abstracting away from the details of each class and attempt to regroup some
of the classes in terms of general semantic notions such as kind and shape"
(Hendrikse and Poulos 1992:202).

Even though, Hendrikse and Poulos commend the contribution made by Denny and
Creider to the classification of noun prefixes, they express their dissatisfaction with
the fact that the analysis does not show the ·deeper underlying principle of the
whole system. They therefore propose an approach which is based on four
fundamental semantic parameters, namely:

(a) Concreteness (Classes 1,2,3,4,5,6,7,8,9,1 0)

(b) Attribution (Classes 11, 12, 13,19,20,21 ,22)
(c) Spatial orietation (Classes 16,17,18,23)
(d) Abstractness (Classes 14,15)
 231
They summarise their hypothesis in the following manner

"We would, in the meantime, like to hypothesize in terms of the views

expressed above that the speakers of these languages systematize the
referents of their noun universe in terms of at least these four semantic
parameters. Thus although cultures may differ as to the detail of the
subcategories distinguished within each one of these parameters, they all
seem to organize the noun universe in terms of major categories based on
the four parameters referred to above" (Hendrikse and Poulos 1992:203).

The advantage with this approach ·is that it attempts to provide some basic
principles which underlie a speaker's perception and classification of the entities
designated by the different nouns. According to this approach

"... homo sapiens sorting and categorization of these objects primarily alludes
to the perceptual (i.e. concrete) properties of these objects. Furthermore,
these properties are qualitatively evaluated by homo sapiens, hence the
sorting of objects in terms of selective qualitative attributes. He also perceives
objects relative to some or other fixed point both in space and time, hence
spatial categorization of objects. Finally, abstracts such as qualities and
activities are brought into the noun universe by means of nominalizations,
hence the derivational mechanisms in the system" (Hendrikse and Poulos
1992:204).

The four parameters are considered to have some cognitive basis rather than just
occurring arbitrarily. The "notion of concreteness may form the basis of the whole
system ... a concrete entity is one that can be perceived by the various senses."
(Hendrikse and Poulos 1992:204). According to them the concrete ·entities can be
seen, heard, touched, smelt and tasted. Using the concreteness criterion, they
were able to establish a continuum which starts from one extreme pole comprising
entities which can be perceived by all five senses to the other end where none of
the senses is involved. They regard the Bantu nouns as organized along this
 232
continuum. At the one end of the continuum will be the nouns with the most
concrete perceptual saliency and at the other extreme of the continuum will be
nouns which express abstract notions. They represent the continuum for the Bantu
noun universe as follows:

Concreteness Attribution Spatial Orientation Abstract

------>------------------------------------------------------------------------------------>
Five senses Two senses One sense No senses

(Hendrikse and Poulos 1992:204)

From· this continuum it is clear that at least five senses are involved in the
perception of concrete things, though not. simultaneously. Only two senses are
involved in the notion of attribu.tion. Attribution has to do with the perception of size
and shape. In this regard visual and tactile senses are involved. According to
Hendrikse and Poulos "attributive nouns can metaphorically take on an abstract
significance of emotive perceptions (e.g. where size and shape elicit certain
attributes, hence the common use of augmentatives and diminutives to express
derogation and amelioration). Finally, there are the spacial and abstract nouns.
Since they cannot be perceived by any sense, abstract nouns are mainly due to
derivational mechanisms in the language system.

Hendrikse and Poulos's proposal appears to have come close to providing a

solution to Westphal's problem of finding the underlying principle to the
classification of nouns in Bantu. In the following discussion, we shall assess the
extent to which this approach can explain the assignment of adoptive nouns to the
different noun classes in Venda.

4.2.1.1 Morphological adjustment in the adoption of nouns

The adoptive nouns in Venda are subject to various morphological adjustments in

Venda. Zawawi (1979:61) raises very interesting and thought provoking questions
 233

pertaining to the grammatical or morphological adaptation of foreign nominals to a

language. Some of the questions he raises will be paraphrased below since they
are pertinent to our study:

(1) Which 'foreign' words have been incorporated into the language lexicon?
(2) To which of the traditional "Noun Classes" have the nouns of non-Bantu
origin been allocated?
(3) What criteria have been used in assigning these nouns to these classes?
(4) Do these criteria lead us to understand better the morphological and syntactic
structures of the language?
(5) Has the introduction of nouns of non-Bantu origin to a Bantu inventory of
nouns affected the traditional classification of the language "Noun classes,"
either by changing the morphological patterns of their cocncordial agreements
or by introducing new morphological forms?
(6) What other changes in the grammatical patterns have occurred as a result of
the introduction of these nouns of non-Bantu origin?
(7) Is it possible to formulate new rules on the basis of these new patterns which
will incorporate the whole lexicon of the language nouns, Bantu and non-
Bantu?
(9) Is it possible to predict the grammatical patterns of future nouns of non-Bantu
origin?

Our discussion in this chapter will address the issues raised by these questions.
First we shall consider the question of the allocation of prefixes to adoptive nouns.
It is an established fact that when words of non-Bantu origin are incorporated into a
Bantu language they are modified to fit in with the prefix system of the language.
The prefix system of Venda and other African languages was briefly discussed in
the previous section and will serve as a frame of reference for the subsquent
discussion on the prefixes assigned to adoptives.
 234

4.2.1.1.1 Affixing class prefixes to adopted nouns

This section will attempt to address the questions raised in (1), (2) and (3) above,
all of which revolve around the issue of assigning class prefixes to adoptives. Only
adopted nouns from English and Afrikaans will be taken into consideration. As
already mentioned, nouns are assigned to classes when they are adopted. The
main question here is what determines the choice of one class prefix for a
particular adopted noun stem over another. There are various factors which serve
as criteria for the assignment of prefixes to adopted nouns. These factors will be
discussed herebelow:

(a) Conversion of initial element to independent prefix

Studies done in other African languages on the morphological adaptation of foreign

words (Van Warmelo 1929; Whiteley 1963; 1967; Hansford & Hansford 1989;
Chimhundu 1982; Poulos 1990) have shown that in some adoptives the first
syllable contains sound elements which are similar or identical to the sound/sounds
of one or other of the noun prefixes, and as such the original syllable undergoes
slight modifications in order to be accommodated into the class system of Venda.
This can be clearly seen from the following examples: ,

( 1) Afr.meneer > Ve.mu_Qeri (1) 'a male missionary'

vhaneri(2) ' male missionaries'
"'
(2) Eng. machine > Ve.mutshini (3)
mitshini (4)
Eng.mat > Ve.methe (3)
mimethe (4)
Afr.mandjie > Ve.manngi (3) 'basket'
mimanngi (4) 'baskets'
Afr.mark > Ve.maraga (3) 'market'
mimaraga (4) 'markets'
Afr.moer > Ve.muru (3) 'nut (on a bolt)'
 235

mimuru (4) 'nuts'

Afr.motor > Ve.modoro (3) 'car'
"'
mimo<toro (4) 'cars•
'
Afr.museum > Ve.mesiamu(3) museum'
mimesiamu(4) 'museums•
Eng.match > Ve.metshe (3)
mimetshe (4)
Eng.matches > Ve.matshesi (3)
mimatshesi( 4)
Eng.map > Ve.mepe (3)
mimepe· (4)
Eng. mineral > Ve.munirale (3)
minirale (4)

The prefixes in the above examples have been allocated on the basis of the close
similarity between the initial consonant of the adopted nouns and the first
consonant of the Venda class prefix mu-. It seems that the class prefix is allocated
to the singular form only, and then the plural prefix is allocated on the basis of the
class to which the singular noun belongs. This criterion of allocating prefixes to
adopted noun stems on the basis of initial syllable resemblance is not without
problems. Firstly, even though the adopted nouns all have the segment /m/ as the
initial syllable, they are allocated to class 1/2 or class 3/4, e.g. muQeri (Afr.meneer),
plural vl1aQeri and mutshini (Eng.machine), plural mitshini respectively. This shows
that even though the allocation of the class prefix to adoptive nouns is made on the
basis of the initial sound resemblance, the actual class to which the noun is
-
assigned may be determined by other factors. In order to determine to which class
the adopted noun is assigned, other factors such as semantic content also need to
be considered. The second problem that one encounters is the fact that there are
many other adoptive nouns which do not seem to adhere to this principle of initial
syllable resemblance. This can be illustrated with the following examples:
 236

Eng.mile > Ve.maila (9 )

dzimaila (1 0)
Eng.million > Ve.milioni (9)
dzimilioni (1 0)
Eng.metre > Ve.mithara (9)
dzimithara (1 0)
Eng.mini > Ve.mini (9)
dzimini (1 0)

It is not clear why the adoptive nouns in the above examples have not been
allocated to class 3/4 since their initial syllables resemble these class prefixes.

The approximation hypothesis may be used to account for the allocation of prefixes
in the following examples. In these examples, the consonant clusters begin with s.
The s sound is realised as the prefix tshi· in Venda. It appears that when s occurs
before a stop consonant, it is approximated to the the prefix tshi·.

. Eng.stamp > Ve.tshitemmbe

""
zwitemmbe
"' 'school'
Afr.skool > Ve.tshikolo
'
zwikolo schools'
Afr.stasie > Ve.tshitasi
A
'mission station'
zwitasi
A
'mission stations'
Afr.straf > ve.tshixarafu 'punishment'
zwitarafu 'punishments'
A

Eng.station > Ve.tshititshi

A

zwititshi
,...
Eng.scheme > Ve.tshikimu
zwikimu
Eng.square > Ve.tshikwea
zwikwea
 237
Eng.Scottish > Ve. Tshikotshi
Zwikotshi
Eng. skipper > Ve.tshikhipha
zwikhipha ·
Eng.scooter > Ve.tshikuta
zwikuta

As can be noted from above examples, the sound s in English and Afrikaans
clustersis approximated to the nominal prefix tshi-. The prefix zwi- is a result of
plural formation. Another example is that of Afrikaans words which begin with the
consonant b. Such nouns are placed in class 14 which has the prefix vhu-.

· Afr.brood > Ve. vhurotho 'bread'

Afr.broek > Ve. vhurukhu 'trouser'
Afr.brief > Ve. vhurifhi 'letter'

Although this principle seems to apply well to the examples we have given above,
it is still not clear why this is not the case with other adoptive nouns such as the
ones given below. The consonants of the initial syllables in the examples below are
very close in production to the Noun Prefix tshi-of class 7. Therefore, according to
the approximation hypothesis, these adoptive nouns would be expected to be .
assigned to this class. As is shown below, this is not the case; instead these
nouns are placed in Class 9.

Eng.change Ve.tshintshi (9)

Eng.chain Ve.tshaini (9)
Eng.cheese Ve.tshisi (9)
Eng.chisel Ve.tshesela (9)
Eng.champion Ve. tshampioni (9)
Eng. chewing-gum Ve.tshingamu (9}
Eng.charge-office · Ve. tshatshaofisi (9)
Eng. chamber Ve.tshimba (9)
 238
Eng.cheque Ve.tsheke (9)
Eng.checkers Ve.tshekasi (9)
Eng.chalk Ve.tshoko (9)
Eng. chocolate Ve.tshokoleithi (9)
Eng.chimney Ve.tshimini (9)

The failure of the approximation hypothesis can also be seen from the following
examples:

Afr.leer > Ve. leri ( 9)'!earn'

"'
Afr.lepel > Ve. lebula ( 9)'spoon'
Eng .lorry > Ve. Iori (9)
Afr.boor > Ve. vhoro (9) 'drill'
Afr.boer > Ve. livhuru (3) 'farmer'
A

In the first two examples one would expect the noun to be placed in class 5,
because of the occurrence of the I consonant in the initial position (cf.Noun prefix .-..li~
of class 5). However, it should be noted that some adopted words which begin with
the sound I are assigned to Li- class e.g.:

Afr.lekker > Ve.legere 'sweet'

I'

Eng.litre > Ve.litha

/"

In the examples vhoro and Jivhuru one would expect these nouns to be assigned to
noun class 14 (where vhu-is the prefix). But this is not the case. These nouns are
instead assigned to class 9 and 3 respectively. The former can be explained by the
zero prefix class principle and the latter by taking into consideration semantic
criteria. These two principles will be discussed in detail in the following sections.
From the examples given above it appears that the initial consonant principle does
not apply automatically in all cases. In some instances it is overridden by other
factors such as the semantic and zero prefix principles.
 239

(b) Allocation of class prefix on the basis of the semantic content of the adoptive
noun

Adopted nouns may also be assigned prefixes on the basis of their meaning. In
fact this is the traditional way of assigning prefixes to nouns. This approach was
discussed under section 4.2.1, where it was indicated that noun prefixes carry
semantic significances. As a result, certain adopted nouns are assigned to certain
prefixes on the basis of their meaning. For instance, adoptive nouns which refer to
human beings would, according to this criterion, normally affix the noun prefix mu-
of Class 1 in singular and vha- of Class 2 in pluraL This can be illustrated by the
following examples:

Afr.boer > Ve.muvhuru 'farmer'

Afr.menner > Ve.muneri
.missionary'
"
Eng.prophet > Ve.muporofita
Eng.evangelist > Ve. muevangeli
Eng. deacon > Ve.mudikoni
Eng.president > Ve.muphurusidennde
Eng.zionist > Ve.muzioni
Eng.bishop > Ve.mubishopho
Eng.student > Ve. mutshudeni (Tshitshudeni)
Eng.heathen > Ve.muhedeni
Eng. christian > Ve.mukiresite

There are, however, some adoptives which denote human beings but which are not
assigned to this class (cf.Chimhundu 1982). Some of these adoptive nouns, as will
be seen below, are assigned to class 5 and others to class 7 and 9. There are
various factors that determine the allocation of the adopted nouns to these noun
classes, namely, the zero prefix principle and the approximation hypothesis
principle .

Eng.judge > Ve.dzhadzhi (5)

 240
Eng.driver > Ve.diraiva
... (5)
Eng.teacher > Ve.thitshere (5)
Eng. inspector > Ve. tshipikitere (7)
A

Eng.doctor > Ve.dokotela (5/9)

Eng.nurse > Ve.nese (5)
Eng. conductor > Ve.khondai (9)
Eng.police > Ve.pholisa (5)
Eng. principal > Ve.phirisipala (5/9)
Eng.clerk > Ve.kileke (5/9)
Eng.lawyer > Ve.loyara (5/9)
Eng. professor > Ve.phurofesa (1/9)
Eng .lecturer > Ve.lekhitshara (1/9)
Eng.cook > Ve.guga (5/9)
Eng.minister > Ve.minisita (1/9)
Eng.soldier > Ve.liswole (5)
""

From the examples given above, it may be noted that although the semantic
principle may explain the allocation of some of these adopted nouns to their
respective classes, in others it cannot. The adopted nouns given above fall into
different classes, namely, classes 1, 2, 5, 6, 7, 8, 9,and 10. Some of the adoptive
nouns which are found in classes 5 & 6, 7 & 8 do not express the semantic
significances that are normally expressed by native words in such classes. For
example, the noun thitshere and ~iraiva which are allocated to classes 5 and 6, do
not have the pejorative connotations usually expressed by nouns of these classes.
The same applies to the adoptive noun tshipikitere; this noun does not have any
"
derogatory meaning or diminution. These adoptives, therefore, seem to defy the
principle of grouping nouns into rigid nominal semantic categories. This fact was
also illustrated in our discussion of the allocation of class prefixes· on the basis of
the resemblance to the initial element of the class prefix. There we noted that the
principle of semantic significance is very often not applicable. Because of the
influence of adoptive nouns, certain classes which used to consist of nouns with
common semantic characterictics now include many nouns which do not fit the
 241

class definitions at all (cf. Zawawi 1979:68). For instance, the allocation of nouns
like vhurotho, vhurifhi, vhulege, vhurukhu to class 14 has nothing to do with
abstractness, which seems to be the dominant feature of this class.

(c) Allocation of adoptive nouns to a zero prefix class

Most studies on the morphological adaptation of non-Bantu words to Bantu

languages have shown that most of the adopted nouns are allocated to Class 1 (a),
5 & 9 on the basis of the zero prefix. By 'zero' prefix we are referring to adoptive
nouns which cannot be identified with any class prefix. Therefore according to the
zero prefix principle such nouns are assigned to those noun classes with nouns
that do not normally have overtly marked prefixes. Generally speaking, adoptive
nouns are allocated to 1 (a) if the referent is human and 5 & 9 if the referent is non-
human (Chimhundu 1982:66). These adoptive nouns could not fit into any of the
other classes by virtue of the nature of their initial consonant. The following are
examples of nouns that belong to Class 1 (a):

Eng .father > Ve.fada

......
Eng.referee > Ve.refiri
Eng.captain > Ve.kaputeni
"'
In Class 5 we find the following nouns which have been allocated to this class on
the basis of the zero prefix principle:

Eng.paper > Ve.phepha

Eng.gate > Ve.gethe
Eng.basket > Ve.basikete
Eng.glove > Ve.gilavu
Afr.tent > Ve.dennde 'tent'
Afr.tou > Ve.deu 'rope'
Afr.dam > ·ve.damu 'dam'
Afr.drom > Ve.diromu 'drum'
A
 242
Afr.blom > Ve.bolomu 'flower'
Afr.papier > Ve.bammbiri 'paper'

It appears that most of the adoptive nouns allocated to a class on the basis of the
zero prefix principle are found in Class 9. This may be so because most of the
native nouns whch have unmarked prefixes in Venda are also found in this class.
The following nouns have been allocated to class 9 on the basis of having prefixes
which are not overtly marked:

Eng. gravel > Ve.giravhulo

Eng.fashion > Ve.fesheni
Eng.kitchen > Ve.khishi
Eng.cupboard > Ve.khabodo
Eng.custard > Ve.khasitadi
A ,-..

Eng.pick-up > Ve.phikhaphu

Eng.jug > Ve.dzhege
Eng.polish > Ve.pholishi
Eng.porridge > Ve.phalishi
Eng.watch > Ve.watshi
Eng.diamond > Ve.daimane
Eng.bicycle > Ve.baisigira
Eng.rice > Ve.raisi
Eng.mule > Ve.meila
Afr.tapeit > Ve.tapita 'plastic tiles'
"' .....
Afr.suiker > Ve.swigiri '
sugar'
Afr.trein > Ve.thireni 'train'
"
Afr.donkie > Ve.donngi 'donkey'
Afr.perd > Ve.bere 'horse'

The adoptive nouns allocated to this class seem to cover different semantic fields
such as tools, instruments and household utensils.
 243

Hendrikse and Poulos's approach seems to throw some light on this complex
problem of the allocation of class prefixes to adoptive nouns. According to this
approach speakers of a language sort and categorize whatever they perceive. The
"sorting and categorizing of these objects primarily alludes to perceptual (i.e.
concrete) properties of these objects" (1992:204). This criterion of concreteness
appears to play an important role in the allocation of class prefixes to adopted
nouns. All the adopted nouns which are allocated to classes 1, 2, 3, 4, 5, 6, 7, 8, 9
and 10, primarily denote concrete entities, as the examples below show:

E.ng.heathen > Ve. muhedeni (1)

pl. vhahedeni (2)

Afr.motor > Ve. modoro (3) ' a car'
A

pl.mimodoro
,... (4) 'cars'
Eng.gate > Ve. gethe {5)
pl.magethe (6)
Eng.inspector > Ve.tshipikitere (7)
"'
pl.zwipiki~ere (8)
Afr.kamp > Ve.gammba (9) 'camp'
pl.dzigammba (1 0) 'camps'

Hendrikse and Poulos hypothesis seems to gives us some clue as to what actually
happens when a speaker perceives foreign word· forms. It is interesting to note that
most of the adoptive nouns in Venda, particularly those from English and Afrikaans,
represent foreign objects which have been introduced to the Venda people. The
distribution of adoptive nouns in the classes given above depends on various
factors. For instance, the physical properties of some of these objects may in one
way or another influence the speaker in assigning an adoptive noun to a particular
class. As already indicated, allocation may be made on the perception of things
that are human as against non-human. In other instances, adoptives are allocated
to certain classes according to the speaker's perception of their initial sounds,
which are equated with certain class prefixes. Although the concreteness criterion
is useful in the classification of some adopted nouns, in others it does not seem to
 • •
244

.ough adopted nouns such as vhurukhu, vhurotho, vhurifhi,

tS, they are allocated to class 14 which is associated with
case, the allocation of the adopted nouns to this class is
initial sound resemblance rather than their parameter of

of foreign objects or concepts may be influenced by other

namely, size and shape. And according to Hendrikse and
ze and shape may elicit certain attributes such as
inutives, which may express derogation and amelioration.
oy the way certain nouns referring to certain tribes or ethnic

sa person) (5) pi.Mathosa (6)

te person) (5) pi.Makhuwa (6)
Jian person) (5) pi.Magula (6)

the following in this regard:

~elings towards other ethnic groups are sometimes reflected

which certain nouns are assigned. Thus, not all names of
J to other tribes or ethnic groups occur in classes 1 and 2
where they would be expected to be found. In fact, where
other prejudice or adverse feeling shown towards a person
then class 5 (with plural li-), rather than class 1 would be
"

It seems that most of the adoptive nouns in class 5 & 6 are allocated to thes~
classes to denote attributes such as size and shape which invoke certain attitudes
in the speaker:
 245

Afr.boek > Ve.libugu

A
{5) 'big book'
mabugu {6) 'big books'
Eng.stamp > Ve.litemmbe
AA
{5) 'big stamp'
matemmbe
/'
(6) 'big stamps'
Eng.skirt > Ve.likete (5) 'big skirt'
"'
makete (6) 'big skirts'

The adopted nouns given above were allocated to this class to denote the size and
shape of the concrete things which they designate. The adoptive bugu is allocated
to class 9 on the basis of the zero prefix principle. But here it is allocated to class 5
and 6 in the singular and plural respectively on the basis of the size and shape.
The last two examples should be allocated to class 7 in the singular and 8 in the
plural on the basis of the initial syllable resemblance principle. Instead in this case
these nouns are allocated to classes 5 and 6 on the basis of the size and shape of
the things they denote. With regard to spatial orientation, there does not seem to
be an adoptive noun which is allocated to such class prefixes according to these
properties. With regard to the last principle of abstractness, some adoptives like
vhurotho, vhurukhu, vhurifhi, which denote concrete things, are allocated to class
14 which is associated with abstract nouns. As already indicated, these nouns are
assigned to class 14 on the basis of their initial sound which resembles the vhu- of
class 14. However, some abstract nouns which belong to this noun class may be
derived from other concrete nouns:

vhukhuruku ·crookedness' < khurukhu (Eng. crook)

vhukeleme
.criminality ' < tshikeleme (Afr.skelm)
vhurobane
.robbery' < u roba (Eng.to rob)
vhukiresite ·christianity' < Khiresite (Eng. Christ)

In this section, an overview was given of Hendrikse and Poulos's approach to noun
class classification. This was then applied to the analysis of adoptives to
demonstrate a possible approach to the classification of adopted nouns in a
language. The semantic criterion seems to play a very important role in the
 246

allocation of adopted nouns to different classes in Venda. As already indicated,

under the semantic parameter of concreteness, there are at least ten noun class
prefixes. Adopted nouns are allocated to these classes according to factors such
as semantic content, the initial syllable resemblance principle, the zero prefix
principle, etc. With the second parameter of attribution, size and shape may
provide some clues as to the class to which a particular adoptive noun can be
allocated to. However, as was indicated in the aforegoing discussion, some nouns
are allocated to certain classes on the basis of phonological factors. Usually, these
are adoptive nouns which have been allocated to the class prefixes on the base of
their initial sound elements which resemble those of the respective class prefixes
e.g.:

Eng. inspector > Ve.tshipikitere

A
(7)
Eng. teacher > Ve.thitshere (5)
Eng.driver > Ve.diraiva
/'
(.5)

The first example does not express a diminutive connotation as is the case with
other nouns found in this class while the last two examples, although they belong
to class 5, do not have derogatory connotations.

4.2.2 Some syntactic aspects of adoptive nouns in Venda

In our definition of the noun we indicated that a noun may be defined according to
its syntactic behaviour. With regard to syntactic behaviour, we shall be looking at
the way in which and the extent to which adoptive nouns generate agreement
concords. In Venda every noun class governs its own concords. Although the focus
in this section is on the adoptive nouns, the influence of other African languages on
Venda will also be considered. We shall begin by looking at the syntactic behaviour
of adopted nouns from English and Afrikaans. Let's look at the following passages:

Mapholisa u bva ~uvha la mulovha vha vhukati na u to9._a muthu we a

vhulaha munna-wa-vhane
A
Vho-Ramalida, rabulasi a divheaho
A
Seshego hothe.
A
 247
Mapholisa ~ Soviet Union madekwe vho fara muriphabuliki mu1ivhalea a kha
di bvaho u bviswa tshiduloni (Maumela and Madiba 1992:11 0).
"'
(The police have since yesterday been busy looking for a man who has killed
the late Vho-Ramalida, a well-known farmer throughout Seshego.

The police in Russia last night arrested· a well-known republican who has
been ousted from his position just recently.)

From the ·paragraphs given above, we find the following adoptive nouns:
mapholisa, rabulasi and muriphabuliki. In the first paragraph, the subject concord
vha is used to bring about agreement between the antecedent mapholisa and the
predicate. The use of the concord vha is very surprising here since it is a concord
which belongs to class 2. The concord a of class 6 would have been expected as
is the case in the second paragraph. In the second paragraph the concord a is
used as well as vh-. One would have expected the zero concord o to have been
used instead of vh-. The other adoptive nouns, viz, rabulasi, muriphabuliki have
been used with concords which correspond to their classes. The different concords
used with the adoptive noun mapholisa illustrate the gender conflict in the
concordia! system of the language as a result of adoption. It seems that the use of
concords is sometimes determined by what the speaker wants to convey. The
concord vha is used to show that although the adoptive noun mapholisa has been
allocate<;! to class 6 it still denotes human beings. The speaker may use the
concords of class 1 and 2 with the noun mapholisa to show his or her respect to
them as human beings, or else use the concords of class 5 and 6 to express the
derogatory connotation often attached to police. The gender conflict also occurs in
some native nouns in Venda:

Khosi ya fhano vha vhusa vhathu vhavho zwavhu~. A vha fani na dzinwe
khosi dzine dza fara vhathu sa dziphuli. Khosi dzi itaho nga u rali dzi sala dzi
dzothe vhathu vho pfuluwa. Avha khosi yashu vha funa vhathu vholhe sa
~ .
 248
vhana vhavho. Mahosi manwe vha a tama nyan~ano ine ya vha hone
shangoni lashu.
"'
(The chief of this place rules the people well. He is not like other chiefs who
treat people like slaves. Such chiefs remain alone when the people have
migrated away. This, our chief loves all the people like his children. Other
chiefs admire the cooperation in our place).

In this passage the nouns khosi (cl.9 ), dzikhosi (cl.10) and mahosi (ci.S) have
been used with different concords. The noun khosi is used with the concord vha of
the nouns of class 2. The concord i would have been expected here. The use of
the concord vha indicates that the noun khosi denotes a human being and it also
shows respect. The noun mahosi has been used also with the concord vha of class
2 instead of the concord a. However, in some instances a speaker may use the
concord a to express certain semantic· significances, namely, derogation or
augmentation. The noun khosi has also been used with the concord dzi. This
concord, in the first place, indicates plurality, and secondly shows the class to
which this noun belong.

Whilst the adopted nouns from English and Afrikaans have had little impact on the
Venda syntax, other languages, namely Shona, Northern Sotho and Tsonga, have
had considerable influence on the syntax of the language, particularly in those
areas where there is strong contact with these languages. The contact between
Venda and these language was discussed in Chapter 1 and so, in order to avoid
repetition, only a few examples are given here.

The influence of Shena is found in the standard language and in some instances,
is only confined to certain sections of the Venda speech community, in this case,
the Northern and North-eastern areas of Venda. The influence of Shona on
Venda, as was shown in Chapter 1, also occured at the syntactic level. In proverbs,
riddles, folktales and songs one finds constructions such as the following:
 249
Tsha nkunda ndilipo (Riddle).
The construction ndilipo is typically Shona. In Venda it should be Ndi henefho.

Vhomme vha ka enda pi (from the song Funguvhu)

(Where the mother have gone to?)
What is interesting in this example is the use of the Shona past tense maker ka.

Pembela ndi ku pe nombe (from the folktale Mukukulume na dongololo).

"Dance excitedly so that I should give you cattle"
Vha ,...
do sala vha no thenga zwavho (from the folktale Mukegulu we a vha a sa funi
riwana wawe a tshi mala) (They will remain enjoying with me).

Again in these songs we find the use of the concord ku instead of the Venda
concord u, as well .as the use of the present tense marker no. The use of the
concord ku and formative no also characterise the Venda spoken in areas like
Tshikundamalema and Manenzhe. Consider the following examples:

Munna a no ya fhi? "Where is the man going to?"

Nwana a no gala pano. "The child stays here."
Ndi no Ia vhuswa. "I eat food"
"'
Ka Mutele a ku endi nwana. "Hamutele is not good for a child visit"

The use. of the formative no and ku is the result of intensive contact with the
Lembethu, who inhabited the area before the Venda.

The influence of Tsonga was also discussed in Chapter 1. Only a few examples
will be given here to illustrate its influence on Venda syntax.

A ndi tshi khou da

A
musi ndi tshi mu vhona.
(cf.Tso. A ndzi(ni) karhi nita loko ni n'wi vona)
"I was coming when I saw him".
 250
A ndo rna vheya hafha.
(cf.Tso. A ndzi mavekile !aha.
"l had put them here".

Matombo ama mo kwashea.

(cf.Tso.Maribye lawai lama rna pandzekile.
"These stones have broken"

The use of the concord rna was discussed in chapter 1. The use of constructions
such as a ndo rna, a ndzi etc. shows the impact of Tsonga on the Venda spoken
in these areas. When it comes to the past or perfect tense construction however,
Venda still uses the perfect tense marker -o instead of changing the verb stem, as
is the case in Tsonga.

Northern Sotho has also influenced Venda syntax in areas such as Mulima and
Mashamba. In these areas one finds constructions such as the following:

Gi tswenya vhanna khoroni.

(cf.NS.Ke tswenya banna kgorong)
Wena u nwile mafhi (cf. Mulaudzi 1992)

In these examples, the concord Gi (cf.NS. ke) is used instead of the Venda
concord .ndi. In the second example, the pronoun wena is used instead of ene.
What is interesting here is the formation of the perfect tense construction. The
suffix -ile is used with the verb stem -nwa to mark tense on the verb. The addition
of new morphemes such as -ile shows the degree of Northern Sotho influence on
the Venda spoken in these areas. According to Thomason and Kaufman (1 988)
the influence on the syntax of a language only occurs when there is strong or
intensive contact between languages.
 251

4.3 Adopted words and other word categories

Adoptives are also found in other parts of speech such as adjectives, verbs,
conjunctives and adverbs.

4.3.1 Adjectives

Venda has adopted a number of words pertaining to colour and number from other
languages like Afrikaans, English and African languages. We si1all begin by
looking at the adoptives which have to do with colour.

Colour tenns

The contact between Venda and other languages has led to the adoption of colour
terms which previously did not exist in Venda. Consider the following examples:

Eng. blue > Ve.buluu

Eng. green > Ve.girini
Eng.brown > Ve.buraweni
Eng.maroon > Ve.meruni
Eng.pink > Ve.pinki
Eng.grey > Ve.gireyi

Although some of these colour terms have native counterparts such as lutombo for
blue, muroxwe for girini, there is a strong tendency to use adopted words rather
than native ones. The adopted terms for colours are preferred because of their
specialised meaning. Venda has no terms which distinguish between the different
kinds of blue as in English where there is dark blue, sky blue, etc. Colours such as
pinki and meruni are in fact new to the Venda language. Otherwise the colours
which are common in Venda are mutswuku (red or pink), mudala (blue, green),
mutshena (white) and mutswu (black). From these examples we can also note the
difference which exists between the morphology of the adopted and the native
 252
adjectives. The native adjectives are formed by the adjectival concord and
adjectival stem, e.g

Adjective adjectival concord adjectival stem

mutswuku mu -tswuku
mudala mu -dala
mutshena mu -tshena

This is not the case with the adopted forms. Adoptives such as pinki, buluu,
meruni, girini do not show the adjectival concord even when they occur with nouns
which have overtly marked prefixes, e.g.:

Vhathu vha vhanna a vha koni u fhambanya muvhala mutswuku na wa pinki.

U do pfa muthu a tshi ndo ambara vhurukhu vhutswuku ngeno o ambara
"'
meruni. Vhaiiwe vho ambara tshikhipha tsha girini u 9_o pfa vha tshi ri vho
ambara buluu. Khavho buluu na girini zwi a fana ndi muvhala mudala.

(Men do not distinguish between red and pink colours. You will hear a man
saying he has put on a red trouser, yet the trouser is pink. Some claim that
they have put on green skippers whereas they have put on blue ones. To
them blue and green is one and the same colour).

From these passage we find adopted words such as pinki, meruni, girini and buluu.
All these adjectives do not show the adjectival concord of the nouns they qualify.
Native words such as mutswuku and mudala have the adjectival concords which
are similar to the prefix of the nouns which they qualify.

Number tenns

Although the Venda have their own number system, there is a significant number of
counting numbers adopted from other languages, English in particular. The most
 253
interesting adoptives in this regard pertain to money:

Musalauno tshelede yo wa tshileme. Kale muthu o vha a tshi kona u renga

vhurukhu nga hafukoroni. Ano maduvha
A
a hu tshe na vhurukhu ha fhasi ha
flflthi rannda. Saga ya mugayo ine zwino ya vho vhidza eithi rannda yo vha i
disheleni fhedzi. Muthu o vha a tshi ri o fara sigisi paundu a pfa o fara
tshelede vhukuma. Lofo ya vhurotho muthu o vha a tshi kona u i renga nga
thubobo.Muthu o vha a tshi kona u renga kholomo nga thupaunzu-theni. Ano
maduvha a huna kholomo ya fhasi ha thu-thauseni rannda.
A .

(Nowadays the money has lost value. In the olden days one could buy a pair
of trousers with half-a-crown. These days there are no trousers which are less
than fifty rands. A bag of mielie meal which now costs eighty rands used to
be ten shillings. A person with six pounds used to feel that he had a lot of
money. One could buy a loaf of bread with two bob. One could buy cattle with
two pounds ten. These days there are no cattle for less than two thousand
rand.)

From this passage, the following adoptives can be identified:

Eng.ten shillings >Ve.disheleni cf.sheleni dza fumi

Eng.two bob >Ve.thubobo cf. masheleni mavhili
Eng.six pounds >Ve.sigisipaundu cf.bonndo thanu na nthihi.
Eng. two pounds ten >Ve.thupaunzu-theni cf.bonndo mbili na sheleni dza
fumi.
Eng.flfty rands >Ve.flfrthi rannda cf.rannda dza mahumi matanu.
Eng.two thousand rannda >Ve.thuthauseni rannda cf.rannda dza zwigidi
zwivhili.

Some of these adoptives seem to have been introduced into the Venda speech
community during the time the British monitory system was still used. This can be
seen from the use of words such as pounds, bobs and shillings.
 254

Sometimes Venda counting numbers may appear to be very long, especially when
it comes to big numbers. Therefore adopted forms are preferred to native ones.
Poulos (1990:81) expresses this point as follows:

"Another consideration worth noting, where exposure to various languages is

involved, is the importance of choosing the simplest and perhaps shortest
way of expressing a particular concept".

If we compare the following forms we will realise that the adopted forms are more
concise compared to the native ones:

Mafobvu a thwendi-faifi o dzhena banngani a tswa rannda dzi linganaho frfrthi-

faifi milioni. Vhanna avha vho vha vha tshi khou tshimbila nga Cressida ya
thirii-litha. Mapholisa vho kona u fara muthihi we a vha a tshi khou shavha
nga milenzhe. Khae vho kona u wana rannda dzi linganaho naini-hundirede-
frfrthi-thauseni. Vhathu vha humbelwa u thusa mapholisa kha u fara mafobvu
aya. Muthu ane a do
A
neya vhutanzi
r. A
nga hune mafobvu aya a vha hone, u do
A

livhuwiwa nga thu-milioni rannda.

(Twenty-five thieves entered the bank and stole fifty-five million. These men
were travelling in a three-litre Cressida. The police managed to arrest one
man who was attempting to escape on foot. He was found with nine hundred
and fifty thousands rands in his possession. People are requested to help the
police to arrest these thieves. Any person who can give information about
these thieves will receive a reward of two million rands.)

From this passage the following adoptives can be identified:

thwendi-faifi (Eng.twenty-five) cf.mahumi mavhili na Athanu

flfrthi-faifi milioni · (Eng.fifty-five million) cf.zwigidi zwa zwigidi zwa mahumi
matanu na zwitanu.
,...
"
 255
thirii litha (Eng.three- litre) cf.litha tharu
I' I

naini-handirede-flfdhi-thauseni (Eng.nine-hundred and fifty-thousand) cf.zwigidi

zwa ma~ana matanu na mana na mahumi
A A..

matanu.
""
thu-milioni (Eilg.two million) . cf.zwigidi zwa tshigidi zwivhili

The examples given above demonstrate that the Venda counting numbers may
sometimes become very cumbersome for speakers of the language. The use of
adopted forms may thus be regarded as a more concise way of expressing the
number of things referred to.

4.3.2 Adopted verbal roots

There is a considerable number of verbal roots in Venda that have been adopted
from other foreign languages. These verbal roots came into the language in various
ways. Firstly, adopted roots may be derived from verb forms in the source
language. Secondly, adopted roots may be derived from other word categories of a
source language. Last but not least, verbal roots may be derived from adopted
nouns in the recipient language. We shall begin by looking at instances where
adopted roots are derived directly from the source language verb forms:

Afr.werk > Ve.-vhereg- 'work'

Afr.laai > Ve.-lais-
A
'load'
Afr.ry > Ve.reil- 'drive'
Afr.pos > Ve.-pos- 'post'
Afr.saag > Ve.-sah- saw'
Afr.sukkel > Ve.-sogol- 'suffer'
Afr.leer > Ve.-ler- 'learn'
Afr.tolk > Ve.-dolog- 'nterpret'
"
Afr.hark > Ve.-harag- 'rake'
Eng.win > Ve.-win-
Eng.loose > Ve.-luz-
 256
Eng.record > Ve.-rikhod-
Eng .faint > Ve.-fenth-
Eng.iron > Ve.-ain-
Eng.polish > Ve.-pholish-
Eng.change > Ve. -tshintsh-
Eng.join > Ve.-dzhoin-

All the adopted roots given in the examples above have been derived directly from
the source language verb forms. The following verb roots have, unlike the ones
above, been derived from nouns:

Eng.ruler (n) > Ve.-rul- ·underline with a ruler'

Eng.torch (n) > Ve.-thotsh- 'shed torch light on something'
Eng.private (n) > Ve.-phuraiveth- ·work as unqualified teacher'
Eng.mechanic (n) > Ve.-khanikh- 'fix vehicles'
Eng.Coolie (pn) > Ve.-gul- 'rob'
Afr.Sondag (pn) > Ve.-sondah- ·attend Sunday church service'
Afr.kerk (n) > Ve.-kerk- ·attend church service'
Afr. sleutel (n) > Ve.-lodel-
A
'lock'

The examples given above show instances where verb roots are derived from
nouns. In examples such as -khanikh- and -lode!- the initial syllables have been
"'
deleted .. The problem with these verbal roots is ascertaining whether such
derivations were made directly from the source language nouns or whether the
nouns were adopted first and the verbal roots were then derived from them. The
latter view seems to be the case with the examples given above. Verbal roots such
as -gul-, -phuraiveth-, -khanikh-, -sondah-, and -loge!- do not occur in the source
languages. These verb roots have been derived from the adopted nouns:

Eng.private > Ve.phuraivethe :-phuraveth- 'temporary teaching'

Eng.mechanic > Ve.makhanikhe :-khanikh- 'fix car engines'
Afr.Sondag > Ve.Sondaha: :sondah- ·attend Sunday church
 257
service'
Afr.sleutel > Ve.tshilodelo
,.... -lode I- 'lock'
"
The examples given above show how verbal roots are derived from adopted nouns.
The process of deriving verb roots from nouns is also common to native words in
Venda.

Adopted roots can be derived from foreign language adjectives. Consider the
following examples:

Eng.right > Ve.-raith- "make a right mark"

Eng.cheap > Ve.-tship- "become cheap"
Eng.wrong > Ve.-ronng- "get a wrong mark"
Eng.bad > Ve.-bed- "perform badly in high-jump" -
Afr.duur > Ve.-dur-
A
"expensive" .

Further derivations can be made from the adopted roots by means of verbal
extensions. For example, from the adopted root -pholish- we can derive various
verbal roots by using different verbal extensions, e.g. :

-pholish-el- ·polish for or on behalf of

-pholish-is- ·cause or help to polish'
-pholish-an- 'polish each other'
-pholish-ulu- ·polish over again'
-pholish-iw- 'be polished'
-pholish-es- 'polish intensively'
-pholish-e- · polishable'
-phoilsh-ekay- ·polish without care'

The examples above illustrate the fact that adopted verbal roots may be used with
different verbal extensions, just as is the case with native verbal roots. Adopted
verbs may, like the native verbs, be used in sentence constructions with or without
 258

object elements. Verbs which are followed by direct objects are referred to as
transitive verbs whereas those which are not followed by direct objects are called
intransitive verbs. The transitivity of the verb roots with different extensions can be
illustrated by the following examples:

Vhatukana vha ~o ri pholishela zwienda. A vha 1o9_i muthu ane a tou vha
pholishisa vha sa funi. Vha tou u kombetshedzwa vha a tambisa pholishi vha
sokou pholishekanya ho!_he-ho~e. Huriwe vha pholishana ngayo mavhudzi.
. Vhatukana vhone vha 1o9._a zwienda zwi no pholish~a. Zwienda zwo
pholishiwaho ndi zwavhu5!i. Hone wa zwi pholishg_§a zwi a tshinyala.

(The boys will polish the shoes for us. They do not like a person who forces
them to polish against their will. If they are forced (to polish) they waste polish
by just polishing everywhere. Sometimes they polish each other's hairs. The
boys like shoes which are polishable. The polished shoes are beautiful.
However, if they are overpolished they get spoiled.)

Given that there is no difference in transitivity between adopted verbs and native
ones, we will not discuss this aspect any further.

4.3.3 Adverbs and conjunctives

4.3.3.1 Adverbs

Adverbs are, according to Poulos (1990:397), words which

" ... provide more information about the actual perfomance or carrying out of an
action or state .... they may indicate, inter alia, how, when, where and why an
action or state is performed".

Further, he describes the function of the adverb as follows:

 259
"There are also other secondary roles that an adverb may perform, such as,
for example, when it can add emphasis to a qualificative form in the sense of
"very" e.g. very bad/good/strong and so on".

Since some adverbs in Venda are derived from other parts of speech or word
categories like nouns, pronouns, qualificatives, etc., it is common to find adopted
words in their formations. Adoptive nouns may be used with the adverbial prefixes
to form adverbs in Venda. Consider the following examples:

Vele o kwasha tombo nga hamula (Afr.hammer)

(Vele broke the stone with a hammer) .

..
Nwana o kandwa nga modoro
X
(Afr.motor)
(The child has been overrun by a car).

Vhakoma vho tumulwa tshan~a nga mutshini (Eng.machine)

(Vhakoma 's hand has been cut by the machine).

Vhatukana vha do tuwa nga Swondaha (Afr.Sondag)

"' "'
(The boys will leave on Sunday).

Adoptives are also found in the formation of locatives. According to Crystal

(1992:233) locatives are words which express "the idea of a place of a state or
action". In the locatives below either prefixes or suffixes are added to adoptive
nouns, as illustrated by the following examples:

Mulalo o dzula kha bola (Eng.ball).

(Mulalo has sat on a ball).

Vhana vho dzula kha bugu (Afr.boek)

(The children have sat on a book).
 260
Vhatukana vha khou ya doroboni (Afr.dorp)
(The boys are going to town).

Tshikoloni hu a takadza {Afr.skool)

(It is nice to be at school).

In the first two examples the locative prefix kha is used with adoptive nouns
whereas the locatives in the last two examples are derived by means of the
locative suffix -ni. The use of either the prefix kha or the suffix -ni depends on what
the speaker wants to convey. The prefix kha is used to mean 'at, on, to, in, from'
etc. The suffix -ni may also be used to mainly to mean 'in, to'. The use of the
locative prefixes and suffixes in adoptives is not different from native locatives.

The most noteworthy influence of foreign languages on this categor-y is that of the
adoptive adverb badi (Eng.bad). This adverb is often used to mean -very, much or
many. Its meaning is equivalent to the Venda form -nga maanda.
)\
Its use can be
seen from the following examples:

--
Mulovha ro Ia maswiri badi
.....
(Yesterday we ate oranges very much).

Musidzana uyu o naka badi

(This lady is very beautiful).

Ba nzi Jo rwiwa ba di mulovha

A --

(Banzi was bitten very much yesterday).

The adoptive badi has been used in the examples given above to express the
notion of -very. In this case there is a shift of meaning. It is also interesting to note
that the adopted adverb badi is used in the examples given above without
adverbial prefixes, as is the case with other native words. The equivalent Venda
meaning, namely, nga maanda,
~
has a prefix nga followed by the noun maanda.
A
 261

Another adopted adverb which is used very often in Venda is the word futhi. This
word has been adopted from Zulu. In Venda it can be used as an adverb and as a
conjunction. Here we shall only give the examples where this word is used as an
adverb. (Its use as a conjunction is discussed in the next section.)

Banzi ~o farwa futhi nga mapholisa

(Banzi was arrested again by the polices).

Ndi do da futhi matshelo

/'.. A --

(I will come again tomorrow).

Vha ni rwa futhi ni mmbudze

(If they beat you again, you should tell me).

In these examples the adopted word futhi conveys the repetition of the meaning
expressed by the verb. Although there is some relunctance by the Venda language
planning bodies to recognize this word as part of. the language, it is used so
frequently that it cannot be ignored.

4.3.3.2 Conjunctions

Conjunctions are described by Poulos (1990:445) as "words which introduce

clauses", Here we shall be mainly concerned with conjunctions adopted from other
languages. The following conjunctions have been adopted from other languages:

Afr.maar > Ve.mara 'but'

Eng.and > Ve.ende
Zu.futhi > Ve.futhi 'moreover'

These conjunctives can be used in sentences as in the following examples:

 262
Ema ndo mu mala, mara a thi mu funi. Nda funa ndi nga mu ),_ala, ende ha
ngo funzea. Vhabebi vhanga futhi a vha mu funi.

(I married Ema, but I do not love her. I may divorce her if I want, and she is
not educated. My parents, moreover, do not like her.)

The conjunctions given above have their native counterparts. The conjunctive mara
is used for fhedzi, ende for nahone and futhi for hafhu. As to the question why
speakers use the adopted forms of these conjunctions instead of the native ones,
the answer can be explained by extra-linguistic factors. Some speakers use these
forms to display their knowledge of languages like English, Afrikaans and Zulu. It
has been indicated in Chapter 2 that certain words have been adopted into Venda
for prestige.

4.4 The effects of adoptives on Venda grammar

From the aforegoing discussion it has become apparent that the adoption of foreign
word forms into Venda has affected the language's grammatical structure in one
way or another. Nevertheless, the language appears to· have successfully adapted
most of the foreign word forms to its own grammatical system, except in certain
areas which have been considerably impacted by other African languages such as
Northern Sotho dialects, Tsonga and Shona. However, in most cases the influence
is only restricted to particular sections of the speech community and never occurs
in the standard language. This influence is not like that of English and Afrikaans,
which impact on the standard language. Here we shall only consider the influences
of foreign languages on the standard language. The effects of the adopted words
on the grammatical structure of the standard form are not very great. Nevertheless,
they will be discussed under separate heading for the sake of clarity. We shall
begin by looking at the influence of adoptives on the morphology of the noun, and
then the influence on other parts of speech
 263
4.4.1 The effects on the morphology of the noun

It was indicated in the aforegoing discussion on the morphology of the noun that
adopted nouns seem to have adapted to the noun structure of Venda. In this case
nouns from languages such as English and Afrikaans affix prefixes when they are
taken over to Venda. Such prefixes may or may not be overtly marked. After
receiving prefixes, these nouns are then assigned to a particular class prefix.

Although Venda has successfully incorporated the adopted nouns to its prefix
system, it was affected in some ways. First, the impact can be clearly seen in the
semantic significances of certain noun prefix classes. Traditionally, certain class
prefixes are known to express certain specific semantic significances. In the
preceding discussion it was shown how certain adopted nouns are allocated to
noun class prefixes. Instead of being allocated to a class prefix on the basis of their
meaning, they are sometimes allocated to a class prefix on the basis of the
resemblance of their initial sound segment to a class prefix. This in a way weakens
the semantic approach to the classification of nouns. There are many adopted
nouns which have been allocated to class 7 because they start with a s+Consonant
cluster, yet they do not carry the meaning significance normally expressed by
nouns of that class. Other adopted nouns have been allocated to class 9 on the
basis of their zero prefix. Zawawi (1 979: 134-135) describes this kind of change in
the following manner:

"As a result of the assimilation of these non-Bantu elements the semantic

function of some of the indicators has changed. These changes range from a
blurring of a common feature in those nominals which were considered to
form a group to the extension of semantic function in those units which
originally had only one function .... lt may be that the original ·meaning of the
indicator (traditional "Prefix") is still maintained even though a new reference
or a new extension of meaning has been added".
 264
The effect of adoptive nouns can also be noted in deviations in the class system
with regard to singular-plural pairings in Venda. Traditionally class prefixes are
grouped according to singular and plural forms. In other words each singular class
is said to have its corresponding plural form. However, it appears that adoptives
have introduced singular-plural pairings which are new to the language. This can
be observed froni the following examples:

Singular Plural
Mutaliana Mataliana
A
. "Italians"
A

MuphothOgisi Maphothogisi "Portuguese"

Mudzheremane Madzheremane "Germans"
Muindia Maindia "Indians"
Mugu Ia Magula "Indians"
Mundevhele Mandevhele "Ndebeles"

These examples have to do with names which are indicative of nationality or tribes.
It is very strange that class 1 should take class 6 as its plural. Normally nouns that
belong to class 1 take class 2 as their plural and nouns which take the plural form
in class 6 are mostly those which belong to class 5 in the singular. Some of the
nouns given above may take class 5 in the singular as a variant form, but this only
occurs in instances where the derogatory significance is intended. In the plural
there is no choice, the prefix remains the same, namely, ma-. These irregular
singular-plural pairings may also be observed with names used to refer to different
denominations:

Singular Plural
Muroma Maroma 'members of the Roman Catholic'
Mudatshi Madatshi 'members of the Dutch church'
Musavhatha Masavhatha 'members the Sabbath church'
Muzioni Mazioni 'Zionists'

It is not possible to use the plural prefix Vha- of class 2 in the examples given
 265
above. The use of the rna- prefix does not necessarily express the derogative
meaning normally carried by such a class. Where the class prefix-5 is used in the
singular it expresses some derogative connotations.

One other interesting example is that of the adoptive noun thitshere (Eng.teacher).
This noun belongs to class 1(a) in the singular or to class 5 if the derogatory sense
is intended. But in the plural it becomes mathitshere (teachers) which belongs to
class 6. It is difficult to understand why this noun is not assigned to class 2 in the
plural since this is the normal plural class for nouns that belongs to class 1 in the
singular.

What we can observe from the examples discussed above is a change in singular-
plural pairings of some noun classes, and this has resulted in the emergence of
new genders. Chimhundu (1982:75) regards this change "as evidence of an
expansion of the native class-number gender system".

4.4.2 The effects on the concordia! agreement system

The passages given in 4.1.2 clearly demonstrate how the incorporation of adopted
nouns has further complicated the Venda concordia! agreement system. Here it
was illustrated that there does not seem to be a fixed correlation between the
adopted noun and the concordia! agreement it generates. It appears that the
agreement concords are mainly determined by semantic factors rather than by
syntactic factors. In the examples given in 4.1.2 the adopted noun mapholisa
(Eng.police) was used with the agreement concord a in one instance and with the
concord vha in another instance. This vacillation in the use of agreement concord,
as indicated in the aforegoing discussion, is not a novelty to the Venda language.
Native nouns such as khosi may be used with the concord i or vha in the singular
and dzi, vha and a in the plural. Therefore it is not necessary for this study to
elaborate further on this issue here.
 266

4.4.3 Enrichment of the grammatical system of the language

The adoption of foreign words into Venda has enriched the language in various
ways. Many words which belong to different parts of speech have been added to
the language. There are many nouns, adjectives, verbs, adverbs and perhaps a
few conjunctions which have been added to the language. All these words enrich
the grammatical system of the language.

4.5 CONCLUSION

In this chapter we have been looking at the grammatical adaptation of adoptives in

Venda. We looked at the manner in which adopted words, particularly from English
and Afrikaans, are adapted to the Venda grammatical system. We started by
looking at the morphological adjustment that takes place when the adopted nouns
from non-Bantu languages are incorporated into the Venda language. An attempt
was made in this chapter to review the approaches used by Bantu linguists for the
classification of nouns in African languages. The traditional approach which is
based mainly on the semantic content of the nouns was first discussed. This
approach attempts to provide semantic details for every class prefix. Secondly, the
approach proposed by Denny and Greider (1986) was briefly outlined. In this
approach nouns may be allocated to different classes according to configurational
and shape meanings. Lastly, the hypothesis advocated by Hendrikse and Poulos
(1992) was set out. According to these two linguists, nouns may be classified
according to four fundamental semantic parameters, namely concreteness,
attribution, spacial orientation and abstractness.

In the discussion of the morphological adaptation of the adopted nouns, it was

pointed out that various factors are involved in the allocation of adopted nouns to
different class prefixes. Here we discussed three different kinds of allocation of
class prefixes, namely, allocation on the basis of the initial syllable resemblance to
a class prefix, allocation on the basis of zero prefix and allocation on the basis of
semantic content. Problems that are encountered when using these criteria were
 267
also discussed. We also discussed the syntactic behaviour of adopted nouns, i.e.
the manner in which adopted nouns, like the native ones, generate agreement
concords. An attempt was also made to describe the grammatical adaptation of
words in other parts of speech such as adjectives, verbs, adverbs and
conjunctives. Lastly, the effects of adoptives on the grammatical system of Venda
were discussed. Here it was establishe(j that Venda grammar has, in one way or
another, been affected by the adoption of foreign words, though to a lesser degree
than the Venda phonological and lexical systems.
 268

CHAPTER 5

5.0 GONCLUSION

This study has attempted to assess the changes that have taken place in the
Venda language due to the influence of other languages. The influence of these
languages on Venda came about as a result of contact. Some of these contacts
occurred in the distant past whereas others are more recent or are still occurring.
This study commenced with an overview of the historical background of the Venda
language with the purpose of establishing the earliest contacts. To this end an
attempt was made to establish the origins of the Venda. Different theories about
the origins of Venda were discussed. Here it was established that the Venda, as
they are known today, may have originated from Central Africa as well as locally.
Certain Venda groups, the Singos for example, probably originated from
somewhere in Central Africa. However, recent studies (Loubser 1988, 1989, 1991)
have established that the first Venda people in the area north of the Limpopo
originated locally as a result of the interaction or amalgamation between the Sotho-
Tswana and the early Shona groups. This interpretation is supported by historical,
archaeological as well as linguistic evidence.

Different contacts between the Venda language and other languages during
migration were also discussed. The main focus here was on the Singo group.
Although this group probably originated from Central Africa, it has not been
possible to trace any contacts with languages spoken in that area. The picture only
becomes clear when the Singo arrived in Zimbabwe. Here they came into contact
with Western Shona groups, namely the Kalanga and Rozwi, and other Shona
groups such as the Karanga. There is enough historical, archaeological and
linguistic evidence to support argument that contact between Venda and Shona
speaking people took place. The contact between Shona and Venda, as was
indicated in this study, also occurred on the southern side of the Limpopo. Some
Shona groups such as the Lembethu and the Twanamba inhabited this area before
the Singo and other Venda groups arrived. When the Singes and other Venda
 269

groups arrived, they then come into con~act with these Shona groups.

Venda also came into contact with other languages spoken in this region, namely,
Nguni, Sotho, Tsonga, English and Afrikaans. The Venda contact with these
languages came about in various ways. In some instances the contact came about
as a result of geographical proximity, and in others through urbanisation or
westernisation. These languages have influenced Venda to various degrees. The
influence of these languages is mainly on the lexicon. The term 'adoptives' was
used in this study to refer to the linguistic forms transfered from other languages to
Venda. The process of transfer is referred to as 'adoption'. The main concern of
this study has been to look at the· adaptation of foreign forms to Venda and their
impact on the language. The adaptation of foreign words occur on the semantic,
phonological and morphological levels.

With regard to the semantic level, the adopted words seem to fall into· various
semantic domains and adopted words have adapted semantically to Venda in
various ways. Certain words have been taken over without any change of meaning.
However, in some words there is a change of meaning. The meaning of adopted
words in Venda has either been extended or narrowed. In some adoptives meaning
extension occurred by adding figurative meaning to such words. Other adopted
words have had their original meaning radically shifted, i.e. the original meaning of
such words has been completely lost and a new meaning has been added to those
words. Other semantic changes occur in the emotive value of the adopted word.
Some adopted words acquire a pejorative connotation whereas others acquire an
ameliorative connotation. The native words may also undergo semantic changes to
adapt to new situations. The meaning of some native words may, just like the
adopted words, be extended to accommodate new concepts which have been
introduced into the speech community.

The adoption of foreign words into Venda has affected the language's semantic
system in various ways. First, the adoption of foreign words into Venda has
enlarged the language's lexicon. Hundreds of new words have been added to the
 270
language. The addition of new words has also increased the number of synonyms
in Venda. Some adopted words are used to replace taboo words in the language.
In other words, adoptives are sometimes used to replace words which are
forbidden or regarded as offensive in the languge, consequently, the adoptive
forms are regarded as being more acceptable or less offensive than their taboo
counterparts. Adoptives may also have negative effects on the semantic system of
the language. The adoption of foreign words has resulted in many homonyms in
the Venda language. Homonyms usually cause confusion in the meaning or usage
of some words. Adoption has also resulted in the loss of native words in Venda.

The adaptation of foreign words also occurs on the phonological level. This study
has shown that most of the adopted words in Venda have been adapted to the
sound system of the language. The process of substitution was first considered.
Here we discussed the replacement of sounds which do not occur in the sound
inventory of the Venda language as well as the substitution of sounds which
already exist in the language. The replacement of sounds which do not occur in the
Venda sound inventory seems to be triggered by the native segment constraint
This constraint was shown to be active in the substitution of both vowels and
consonants. However, as has already been mentioned, even though some sound
segments occur in the sound inventory of Venda, they are for one reason or
another replaced with different sounds. In most cases, the choice of the substituting
sound seems to be determined by phonetic approximation, i.e. a foreign sound is
replaced by one which is phonetically closest to it.

Phonological changes also occur to the syllable stucture of adopted words. Words
with a syllable structure which is not acceptable to Venda undergo some form of
modification. Words with consonant clusters are adjusted by means of the insertion
of an epenthetic vowel between the cluster. Adjustment also occurs·to words which
end with a consonant. Venda does not permit words to end with a consonant, with
the exception of ideophones. A vowel can thus be added to the final consonant or
the consonant can be deleted. The choice of the epenthetic vowel is determined by
both phonological and morphological factors. With regard to phonological factors,
 271

assimilation seems to play an important role in the choice of the epenthetic vowel.
The epenthetic vowel assimilates to the vowel of either the preceding or
succeeding syllable. However, assimilation may occur across a syllable boundary.
The epenthetic vowel may also be determined by the nature of the preceding
consonant. Bilabial consonants are usually followed by rounded epenthetic vowels
whereas alveolar consonants are usually· followed by unrounded ones.
Morphological factors also play a role with regard the epenthetic vowels in adopted
verbal stems. All the adopted stems, like the native ones, end with vowel -a in
positive constructions and -i in negative constructions.

Adopted words from non-tonal languages such as English and Afrikaans are
assigned tone when they are incorporated into Venda. It appears that the tone
pattern follows the stress pattern of the word in Jts source language, i.e. stressed
syllables are realised as high tones in Venda. Where adoptives do- not seem to
follow this pattern, some tonal changes are suspected. In this case the high or low
tone is spread to the adjacent syllable.

The impact of adoptives on the Venda phonological system was also assessed.
Here it was established that although the Venda language is coping well in
adapting foreign word forms to its sound system, it has also been influenced to
some extent in that new sounds and sound sequences have been introduced to the
language.

Adopted words in Venda have also undergone some adjustment at the grammatical
level (i.e. morphological and syntactic level). The focus here was mainly on
adopted words from English and Afrikaans. These languages have a grammatical
system which is different from that of Venda or other African languages. First, the
morphological adaptation of adopted nouns was examined. Adopted nouns from
non-Bantu languages acquire prefixes when they are taken over to Venda. The
choice of these prefixes is determined by various factors. The class prefix has been
assigned to adopted nouns in Venda on the basis of the initial syllable resemblance
to a class prefix, the semantic content of the adopted word and on the basis of the
 272

zero prefix in some classes. The morphological modification to adopted words in

other parts of speech, namely adjectives, verbs, adverbs and conjunctives was also
discussed. An attempt was made to describe the syntactic behaviour of adopted
words such as nouns and verbs. With regard to the former, it was established that
adopted nouns, like native nouns, generate agreement concords. Verbs on the
othElr hand may affix different verbal extensions to express different meanings.
Adopted verbs may occur in transitive as well as intransitive constructions. The
effects of adoptives on the grammatical system of Venda were also assessed. It
was established that the adoption of foreign words in Venda has affected the
grammatical system of the language in a number of ways. The semantic
significances of certain classes have been affected by the incorporation of foreign
words into the language. The singular-plural pairings of some classes have also
been affected. Some nouns which belong to class 1 in the singular take the prefix
rna- of class 6 in the plural form. Furthermore, when this prefix is used with such
nouns it does not express derogatory connotation, as would be the case with the
native nouns which belong to this class. The addition of new nouns, adjectives,
verbs, adverbs and conjunctives has enriched the grammatical system of the
Venda language.

This study has attempted to provide as complete a picture as possible of the

changes that occur to adopted words in Venda and the impact of such words on
the Venda language. We hope that the evidence on adoptives in Venda presented
here will help in the interpretation of the ongoing linguistic changes, not only in
Venda, but in other languages too. This study also provides a good background for
corpus planning. Corpus planning refers to a planned intervention to change the
corpus of a language, i.e. its vocabulary, morphology, spelling and otherwise.
Therefore, this study may throw some light on the effectiveness of adoption as a
method of language development. The fact that there are many ·variations in the
modification of adopted words also shows the need for further research on
standardisation principles.
 273
M A P N 0.

( This map was taken from Phillipson 1977:1 09)

~1
'···~
l

··..., [ .............
i.... ~.l·

HYPOTHETICAL PROGRESSION of Bantu-speakers over some 2,000

years sees the language arising among a Neolitbh: people before 1000 B.C.
(1). A dual movement then seems to have brought Bantu-speakers, some
using iron, beyond the forest (J, 5). An east-to-west movement (4) rein-
forced southward expansion of languages ancestral to the Western Group
(6). Early in the first millennium A.o. the eastern stream expanded to the
south (7, 8). The western stream (9) gave rise to a center (10) that sent forth
languages ancestral to the East~rn Group (11) in the 11th century .-~.o.
 274

rA A P tJ 0. II

{ This map was taken from Doke 1931)

THERN RHODESIA.
ANO PORTIOM Of'

fGUESE EAST AFRICA

5W(WING
I
t-
----,.---I.L-~'<----t
'lll: DISTRIBUTION' Of!'
~-SP~G PEOPLES
~··"'·:''"'... ·==r. ~

R C A
 275
MAP N 0. III
(A map showing the contact between Venda, Sotho and Tsonga in the South)
(Warmelo 1935)

NB. Green dots represent Tsonga speaking people.

Round brown dots represent Sotho speaking people.
Square brown dots represent Venda speaking people)
 276

MAP N 0. I V

(A map taken from Warmelo (1935) showing the contact between Venda and
Tsonga in the North East.
NB.Brown dots represent Venda speakers.
Green dots represent Tsonga speakers.

t. S I,.
QI400

~
''
'
t
...

...
..

. .
-
 277

BIBLIOGRAPHY

Alpers, E.A.
1970 "Dynasties of the Mutapa-Rozwi complex". Journal of African
History XI (2):221-240.
Anderson, S.R.
1985 Phonology in the Twenieth Century. Theories of Rules and
Theories of representation. Chicago: The University of Chicago
Press.
Arlotta, A.
1972 Introduction to historical linguistics. Boston: Boston Houghton
Miffin Company.
Bach, E and Harms, R.
1972 "How do langues Get Gracy Rules" In: R.Stockwell & R. Macaulley
(eds.), Linguistic change and Generative Theory, 1-21.
Bloomington: Indiana University Press.
Batibo, H.M .
1993 "Does Kiswahili have diphthongs: A case of interpreting foreign
sounds in an African languages." (A paper read at Alasa
Conference.)
Baumbach, E.J.
1970 'n Klasifikasie van die Tsonga dialekte van die Republiek van
Suid-Afrika . Unpublished Ph.D thesis. University of South Africa:
Pretoria.
Beach, D.N.
1984 The Shona and Zimbabwe 1900-1950. Gwelo:Mambo Press.
Beuster, C. "Das Volk der Vawenda." Zeitschritt der Gese!!schaft fur
1879 Erdkunde zu Berlin, vol. 14:236-240.
Bjarkman ,P.C.
1976 Natural Phonology and Loanword phonology (with selected
examples from Miami Cuban Spanish). Unpublished Ph.D thesis.
University of Florida, Florida.
 278

Blacking, J.A. R.
1967 Venda children's songs: A study in ethnomusicological analysis.
Johannesburg: Witwatersrand University Press.
Breal, M.
1964 Semantics, Studies in the Science of meaning. New York: Dover
Publication.
Britz, R.M. H .
1983 Vreemdetaalbeinvloeding op die leksikon van lsiXhosa.
Unpublished Ph.D thesis, University of Port Elisabeth: Port
Elizabeth.
Bryant, AT.
'1963 Bantu origins. The people and their language. Cape Town:
Africana Specialist.
Burling, R .
1970 Man's many voices. Language in its cultural context. New York:
Holt, Rinehart and Winston.
Carlisle, R.S.
1991 "The influence of Environment on Vowel Epenthesisin Spanish I
English lnterphonology." Applied Linguistics 12(1 ):76-95.
Cassimjee, F.
1992 "Fundamental principles of Venda tonology." In: Gowlett (ed.)
(1992:12- 51).
eaton-Thompson, G.
1970 The Zimbabwe Culture: Ruins and Reactions. London: Frank Cass
and Company.
Cheng, R.L.
1985 "Group interest in treating words borrowed into Mandarin and
Taiwanese" Anthropological Linguistics 27(1):177-1188.
Chimhundu, H.
1982 Adoption and adaption in Shona. Unpublished Ph.D thesis, .
University of Zimbabwe: Zimbabwe.
 279
Chomsky, N. & Halle, M.
1968 The sound pattern of English. New York: Harper and Row.
Clark, C.
1981 An introduction to the History of the English language. London:
Stanley Thornes.
Clements, G.N and Keyser, S.J.
1983 CV Phonology. A Generative Theory of the syllable. Cambridge:
The MIT Press.
Cluver, A.D. de V.
1989 A manual of terminography. Pretoria: Human Sciences Research ·
Council.
Coetsem, F.
1988 Loan Phonology and the Two Transfer Types in language contact.
The Netherlands: Floris.
Cole, D.T.
1953. "Fanagalo and the Bantu languages in South Africa." African
Studies, 12: 1 - 9.
1990 "Old Tswana and new Latin." South African Journal of African
Languages, 10(4): 345-353.
Collins Concise Dictionary. 1989. Glasgow: William Collins & Sons.Ltd.
Craig, C.(ed.)
1986 Noun classes and categorization. Amsterdam: John Benjamins.
Crystal, D.
1992 An Encyclopaedic Dictionary of language and languages. Oxford:
Blackwell.
Dalby, D. (ed.)
1965 Language and History in Africa. London : Frame Cass & Co.
Davison, P.
1984 "Lobedu material culture". Annals of the South African meseum,
94-201.
 280

Denny, P.J & Creider, C.A

1986: "The semantics of noun classes in Proto-Bantu". In: Craig
(1986:217 -239)
Desmond, C.
1969 The discarded people: An account of African resettlement.
Braamfontein: The Hague.
De Villiers, M.
1970 Afrikaanse klankleer: Fonetiek, Fonologie en Woordbou. Cape
Town: A.A .Balkema.
Doke, C.M.
1931 Report on the Unification of the Shona dialects. Hertford. Stephen
Austin.
1954 The Southern Bantu languages. Oxford: Oxford University Press.
Ehret, C.
1967 "Cattle keeping and ·milking in Eastern and Southern
Africanhistory: the linguistic evidence."Journa/ of Afr.ican History,
8:1-17. 1968:"Sheep and Central Sudanic people in southern
Africa". Journal of African History, 9:213-221.
1972 "Outlining Southern African History: a re-evaluation, A.D.100-
1500". Ufahamu, 3 (1 ):9-27.
1973 "Patterns of Bantu and Central Sudanic settlement in Central and
Southern Africa." Trans-African Journal of History, 3:1-71.
1976 "Linguistic evidence and its correlation with archaeology". World
Archaeology, 8(1):5-18.
Emslie, C. F.
1977 Phonological aspects of loanwords. Unpublished M.A.dissertation.
University of the Orange Free State: Bloemfontein.
Fortune, G.
1954 "A Rozi Text with Translation and Notes". NAOA 31:67-91
Giegerich, H.J.
1992 English Phonology. An introduction. Cambridge: Cambridge
University Press.
 281

Gimson, A.C
1970 An introduction to the Pronunciation of English. London: Edward
Arnold.
Gottschling, E.
1905 "The Bavenda: A sketch of their history and customs". Journal of
the Royal Anthropological Institute, 35:365-386.
Gowlett, D.F.
1992 African linguistic contributions. Pretoria: Via Afrika.
Greenberg, J.H.
1963 "The Language of Africa" . International Journal of American
Linguistics, 29(1 ), Part II.
1972 "Linguistic Evidence regarding Bantu Origins" Journal of African
History, 8(2): 189-216.
Grundler, W.
(n.d) Geschichte der Bawenda- Mission in Nord- Transvaal. Berlin:
. Buchhandlung der Evangelischen Missiongesellschaft.
Guthrie, M
1971 Comparative Bantu.(Vo1.1). London: Cregg International.
Halle, M and Vergnaud, J.R
1980 "Three dimensional phonology" Journal of Linguistic research 1: 83-
105.
Hannan, M.
1959 . Standard Shona Dictionary. Harare: The College Press.
Hanan, S.
1973 "The study of English Loa-words in Mordern French". Computers
and the Humanities, Vll:389-398.
Hansford, K & Hansford, G.
1989: "Borrowed words in Chumburung". African · Languages and
Cultures 2(1 ):39-50.
Harris, J.W.
1983 Syllable structure and stress rules. Bloomington: Indiana University
Club.
 282

Haugen, E.
1950 "The analysis of limguistic borrowing". Language 26: 210-231.
1955 Bilingualism in the Americas: A Bibliography and research guide.
Alabama: University of Alabama Press.
Heine, B.
1991 "Towards Language Planning in a Post-Apartheid South Africa.
On planning Indigenous Languages: A challenge". (A paper read
at the First International LICCA conference)
Hendrikse, A. P. and Poulos, G.
1992 "A continuum interpretation of Bantu noun class system." In:
Gowlet, (1992:195-205).
Hermanson, E.A.
1991 The Transliteration of New Testament Proper nouns into Zulu.
Unpublished M.A dissertation. University of- South Africa:
Pretoria.·
Hoffer, B.
1980 "English Loanwords in Japanese: Some cultural implications."
Language Sciences 12(1):1-21.
Holden, K.
1976 "Assimilation rates of borrowing and phonological productivity".
Language 52 (1): 131-146.
Hornby, A.S.
1974 . Oxford Advanced Learner's Dictionary of Current English. Oxford:
Oxford University Press.
Hungerford, H., Robinson, J. and Sledd, J.
1970 English Linguistics. An Introductory Reader. lllinois:Scott,
Foresman.
Hyman, L.M.
1970 "The role of borrowing in the Justification of phonological
Grammars" Studies in African Linguistics 1(1 ): 1-48.
 283
Inskeep, R.R.
1965 Preliminary investigation of a proto-historic cemetery at Nkudzi
Bay, Malawi. Zomba: National Museum of Zambia.
Jones, D.
1956 Everyman English Pronunciation Dictionary. London: J.M. Dent.
Janes-Phillipson, R.
1972 Affinities between Venda and other Southern Bantu languages
Unpublished Ph.D thesis, University of London: London.
Katamba, F..
1989 An introduction to phonology. London: Longman.
Kaye, J. and Nykiel, B.
1979 "Loanwords and abstract phonotactic constraints" The Canadian
Journal of Linguistics 24:71-93.
Khumalo, J.S.M.
1984 "A preliminary Survey of Zulu adoptives". African Studies 43(2):
205-216.
1987 An Autosegmental account of Zulu phonology. Unpublished Ph.D
thesis. University of Witwatersrand: Johannesburg ..
Kiparsky, P.
1982 "Metrical structure assignment is cyclic". Linguistic Inquiry
10:421-442.
Knappert, W.J.
1970 "Contribution from the study of Loanwords to the Cultural History
of Africa". In Dalby (1965:78-88 ).
1972 An anthology of Swahili. Love Poetry. Los Angels: University of
California Press.
Krige, J.D.
1937 "Traditional origins and tribal relationshipOs of the Sotho of the
Northern Transvaal". Bantu studies, Xl:321-357.
 284

Kruger, W.J
1965 Leenwoorde van Afrikaanse oorsprong in Noord-Sotho. 'n Studie in
Transsitematisasie. Unpublished M.A dissertation. University of
Pretoria: Pretoria.
Kunene, O.P
1963 "Southern Sotho words of English and Afrikaans Origin." Word 19:
347-375.
Kunene, C.L. & Sukumane, J.G.B.
1987 "Borrowing and neologism as a means of bridging the gap
between foreign concepts I experiences and Siswati terminology in
applied branches of knowledge." Logos 7(2): 115-125.
Ladefoged, P.
1982 A Course in Phonetics. New York: Harcourt Brace Jovanovich.
Langacker, R. W.
1968 Fundamentals of Linguistic Analysis. New York: Harcourt Brace
Jovanovich.
Lanham, L.W. & Prinsloo, K.P.
1978 Language and Communication Studies in South Africa. Cape
Town. Oxford University Press.
Lass, R.
1984 Phonology: An Introduction to Basic concepts. Cambridge:
Cambridge University Press.
1987 The Shape of English: Structure and History. London: Dent.
1992 "Harmony and assimilation: Suprasegmental aspects". (Handouts
given at Unisa Departmental seminar). Pretoria.
Leech, G.
1974 Semantics. The Study of Meaning. Harmondworth: Penguin.
LeRoux, T.H. and Pienaar, P.O.
1970 Uitspraakwoordeboek van Afrikaans. Pretoria. J. L. van Schalk.
 285
Lestrade, G.P.
1927 "Some notes on the ethnic history of the Vhavenda and their
Rhodesian affinities." South African Journal of Science 24:486-
495.( Also in Van Warmelo (1960:XX-XXVII)
1932 "Some notes on the Political organization of the Venda-speaking
people." (Also in Van Warmelo (1 960 :I-XI).
Loubser, J.
1988 Archaeological contribution to Venda ethnohistory. Unpublished
Ph.D thesis. University of Witwatersrand: Johannesburg.
1989 "Archaeology and Early Venda History". South African
Archaeology Society. Goodwin Series 6: 54-61.
1990 "Orala traditions, archaeology and the history of Venda mitupo".
African Studies , 49:13-37.
1991 The ethnoarchaeology of Venda-speakers in South Africa.
Bloemfontein: Bloemfontein National Museum.
Louwrens, L.J.
1993 "Semantic change in loan words". South African Journal of African
Languages 13(1 ):8-16.
Lovins, J. B.
1975 Loanwords and the phonological structure of Japanese. Tokyo:
Indiana University Linguistics Club.
Lyons, J.
1977 Semantics (Vol 1). Cambridge: Cambridge University Press.
Madiba, M.R.
1989 A Preliminary Survey of adoptives in Venda. Unpublished Honours .
article. University of South Africa: Pretoria.
Marodzi, F.P
1924 "The Barozwi" NADA 2:88-91.
Mathivha, M.E.R.
1966 The tone me patterns of the Venda noun . Unpublished M.A
dissertation. University of South Africa. Pretoria.
 286
1972 A Survey of literarry achievement in Venda. Unpublished
O.Litt.thesis, University of the North: Turfloop.
Mathivha, R.N.
1985 The Berlin Missionary venture in Education at Tshakhuma: Venda
(1872-1954). An historical-critical survey. Unpublished M.Ed.
dissertation. University of the North . Torfloop.
Maumela, T.N. & Madiba, M.R.
1992 Gonda ....Ia Tshivenda,
"' murole 8. Pretoria: Via Afrika.
Meinhof, C.
1906 Grundzuge einer verglichennden Grammatik der Bantusprachen.
Berlin: Dietrich Reimer.
Mokgokong, P. C .
1966 A dialect-geographical survey of the phonology of the Northern
Sotho area. Unpublished M.A.dissertation. University of South ·
Africa: Pretoria.
1975 Context as a determinant of meaning with special reference to
Northern Sotho. Unpublished Ph.D thesis, University of South
Africa: Pretoria.
Mojela, V.M.
1991 Semantic changes accompanying Loan-words in the Northern
Sotho Lexicon. Vista University : Pretoria.
Motenda, M.M.
1940 "Ramabulana". In: Van Warmelo (1940: 154-180.)
Mudau, E.F.N
1940 "Ngoma-lungundu and the early invaders of Venda". In: Van
Warmelo (1940: 154-180).
Muhlhausler, P.
1979 "Growth and structure of the lexicon of New Guinea Pidgin.
Research School of Pacific Studies". Pacific Linguistics Series C.
No. 52.
 287
Mulaudzi, P.A.
1992 "Is Venda language homogeneous or heterogeneous?" (Paper
read at the Departmental meeting of African Languages).
University of South Africa: Pretoria.
Mullan, J.
1969 The Arab builders of Zimbabwe. !969. Rhodesia: Salisbury:
Rhodesia Mission Press.
Ngcongwane, S.D.
1985 "Fanagalo and the Zulu language. The languages we Speak"
Publication of the University of Zululand, Series B 54:1-13 (Also in
South African Journal of African languages 1983, 3(1 ):62-71 ).
1987 "Some instances of language mixing in South Africa". (Paper read
at International Symposium on Professional terminology in African
Language.)
Nida, E. A.
1969 Componential Analysis of Meaning. The Hague: Mouton.
Nkabinde, A.C
1968 "Some aspects of foreign words in Zulu". Communication of the
University of South Africa 59:5-21 ..
1968 The Adaptability of Zulu to new situations. Unpublished M.A
dissertation. University of South Africa: Pretoria ..
Okonkwo, C.J.E.
1977 Towards an intergrated theory of language Planning (with special
emphasis on Africa). Unpublished Ph.D. thesis. State University of
New York at Buffalo. New York.
Palmer, F.R.
1976 Semantics: A new outline. Cambridge: Cambridge Un_iversity
Press.
Partridge, E.
1973 Usage & Abusage. A guide to good English. Harmondsworth:
Penguin.
 288

Phillipson, D.W.
1977 "The Spread of Bantu languages". Scientific American 236(1 ): 106-
114.
Pop!ack, S. & Sankoff, D.
1984. "Borrowing: the synchrony of intergration" Linguistic 22: 99-135.
Poulos, G.
1986 "Venda: Only study guide for VEN 302-K". Pretoria: University of
South Africa.
1990 A Linguistic analysis of Venda. Pretoria: Via Afrika.
Ralushai, V.N.M.N.
1977 Conflicting accounts of the Venda History with particular reference
to the role of mutupo in social organization. Unpublished
Ph. Dthesis, Belfast: Queen's University.
Rassman, E.
1977 "A Preliminary Survey of adoption in South Sotho" Afrcan Studies
36:3-30.
Rauch, I. & Scott, C.T.
1967 Approaches in Linguistics methodology. Madison, Wisconsin:
University of Wisconsin Press.
Roach, P.
1983 English Phonetics and Phonology. A Practical Course. Cambridge:
Cambridge University Press.
Robins, R.H.
1975 General Linguistics. London: Longman ..
Robinson, K.R.
1971 "The Iron Age in Malawi: A brief account of recent work". In :
Pachai (ed.): The Early history of Malawi 49-69. London:
Longman.
1979 The Nkhotakota Lake Shore and marginal areas, Malawi: An
Archaeological reconnaissance. Zomba : Malawi Government
Department of Antiquities 19.
 289
Santiago, A.O.
1978 "Phonemicity of Pilipino orthographic system: A Blessing or A
Curse?" Pacifies Linguistics Series 47:319-338.
Schfield, J.F.
1937 "A preliminary study of the pottery of the Bantu tribes of South
Africa". South African Journal of Science, 39:256-281.
Serjeantson, M.S.
1968 A history of foreign words in English. London: Routledge & Kegan
Paul.
Silverman , D.
1992 "Multiple scansions in loanword phonology: Evidence from
Cantonese*." Phonology 9(2):289-328.
Sinton-Schoettler, G.
1971 "The Sotho, Shena and Venda: a study in cultural continuity".
International Journal of African historical studies, 1:1-17.
Stampe, D. L.
1972 A dissertation on natural phonology. Doctoral dissertation.
Chicago. University of Chicago.
Stayt, H.A.
1931 The Bavenda. London: Oxford UNiversity Press.
Steinbergs, A.
1984 "Loanword Incorporation processes: Examples from Tshiluba"
Studies in the Linguistic Sciences 14(2): 115-124.
1985 "The Role of MSC's in Oshikwanyama Loan Phonology." Studies
in African Linguistics 16(1 ):89-1 00.
Stern, G.
1931 Meaning and change of meaning. Bloomington: Indiana University
Press.
Thea!, G.M.C.
1916 History of South Africa before 1795. London:Williams
Claws & Sons.
 290

Thomason, S.G & Kaufman, T.

1988 Language Contact, Creolization and genetic Linguistics. Berkeley:
University of California Press.
Trail!, A
1993 "Perceptual salience and the case of Venda apicals and laminals".
(Paper delivered to the Alasa Conference held at the University of
the Witwatersrand).
Ullman, S.
1957 The Principles of Semantics. Oxford: Basil Blackwell & Mott Ltd.
1970 "Change of meaning" In: Hungerford et al. (1970: 457-491)
Van Jaarsveld, F.A
1975 From Van Riebeeck to Vorster: An introduction to the history of
the Republic of South Africa. Johannesburg: Voortrekkerpers.
Van Warmelo, N.J.V.
1929 "European and other influences in Sotho." Bantu Studies Ill: 405-
421.
1935 A preliminary survey of the Bantu tribes of Souht Africa.
Pretoria:Government Printer.
1937 "The classification of cultural groups" In : W.O. Hammond- Tooke
(1937:56-84), The Bantu speaking peoples of Southern Africa.
London: Routledge and Kegan Paul.
1940 The Copper miners of Musina and the early history of the
Zoutpansberg. Pretoria. Government Printers.
1960 Contributions towards Venda History, Religion and tribal Ritual.
Pretoria: Government Printer.
1989 Venda Dictionary. Tshivenda-English. Pretoria. J.L van Schaick.
Van Wyk, E.B.
1978 "Language Contact and Bilingualism". In: Lanham (1978:29-152)
Venda Terminology and OrthograpiJY No.3. 1980., Department of Education and
Training 1980
 291

Von Sicard, H.
1952 Ngoma Lungundu. Eine Afrikanische Bundeslade. Uppsala: Studia
Ethnographica Upsalensia.
Waldron, R.A.
1967 Sense and sense development. London : Andre Deutsch .
. Ward, I. C.
1972 The phonetics of English . Cambridge: Heffer & Sons.
Webber, E.A. W.
1988 "Introduction to Kalanga/ Ulima" (Notes of the Introductory Course
on Kalanga I Ulima held at the Enhlanhleni Lutheran Theological
Seminary).
Webster's New Collegiate Dictionary. 1980. Springfield, Mass. G.& C. Merriman
Company.
Weinreich, U.
1953 Languages in Contact : Findings and problems. New York.
Publication of the Linguistic Circle of New York No. 1.
Wentzel, P.J.
1983 The relationship between Venda and Western Shona (Vol./11).
Pretoria: University of South Africa.
Wesseman, R.
1908 The Bavenda of Spelenken. London : African World.
Westphal .E.
1946 A Scientific analysis of the phonetics, morphology and syntax of
the Venda language. Unpublished M.A dissertation, University of
Witwatersrand :Johannesburg ..
Whiteley ,W.H.
1967 "Loanword in Linguistic description: A case study from Tanzania,
East Africa". In: Rauch et.al (eds.) (1967:125- ·
143).
Wilson, M.N
1969 "The Sotho, Venda and Tsonga". In: Wilson et.al (1969: 131-132)
 292

Wilson, M.N. & Thompson, L. (eds)

1969 · The Oxford History of South Africa (Vol. 1). South Africa to 1870.
Oxford: Clarendon Press.
Yavas, M.
1982 "Natural Phonology and borrowing assimilations". Linguistics
20:123-132.
Yule, G.
1985 The study of language. Cambridge: Cambridge University Press.
Zawawi, S.H
1979 Loanwords and their effect on the Classification of Swahili
nominals. Leiden: J.Brill.