Pages

African Religions: Fred Welbourn's


Ancestors index pages or Nurelweb

Copyright F.B. Welbourn 1968
CHAPTER TWO
The natural and the supernatural

Few Britons have seen bacteria. We rarely think about them unless they cause
disease. And even today many people are ignorant that bacteria are essential to
healthy organic life. If we have to deal with them, whether to kill the malign
bacteria which cause dysentery or to increase their benign activity in a compost
heap, we consult a specialist. Normally, if we think about them at all, we
simply take them on trust. But most of us, if challenged, would say that they
are a natural and inescapable part of our environment.

No traditional Baganda have seen a ghost (Baganda ghosts are not expected to be
seen. They work in other ways). Although custom directed that shrines to the
dead should be tended regularly, they usually were neglected unless the ghost of
the deceased caused trouble. The benign activity of ghosts was so much taken for
granted as to be hardly mentioned except on very special occasions. If, however,
a ghost caused trouble a specialist was consulted. But, if challenged,
traditional Baganda would say that ghosts were a natural and inescapable part of
their environment. There was nothing "supernatural" about them. Indeed, the
contrast between "natural" and "supernatural" cannot be expressed in the Luganda
language.


The idea that religion has to do with the supernatural, that is with miracles
and the breaking of "natural law", is deeply ingrained in Western thinking. We
think of the "natural" and "supernatural" as two entirely different realms to
such an extent that some of us find it extremely difficult to believe that a
supernatural world exists at all.

Contrast our way of thinking with two folk-tales which show how differently the
Baganda thought about what we would call "the supernatural":

When Kintu, first king of the Baganda, came to the country, he was alone except
for his cow. He ate its dung and drank its urine and enjoyed its company. One
day, sliding down the rainbow from the sky, which was ruled by a king called
Ggulu, came Ggulu's sons and daughters to have a look at the earth. One
daughter, Nnambi, fell in love with Kintu and determined to marry him. Her
brothers told their father, who advised them to steal Kintu's cow so that he
would die. But Nnambi saw what happened and took Kintu to the sky to recover it.

Ggulu set Kintu impossible tasks: to eat 10,000 helpings of food and beer; to
use a copper axe to cut a rock into firewood; to collect a pot of drinking water
from the dew; and finally to find his own cow among 40,000 others. All these
Kintu accomplished and drew wondering praise from Ggulu. He married Nnambi; and
they went back to earth to breed the Baganda.

Whatever the moral of this story, it is clear that the wonder-worker was Kintu
the man, not Ggulu the "ruler of the sky" whom we would call a god. Kintu and
Ggulu dealt with each other as any Baganda suitor would deal his potential
father-in-law. Earth and sky, man and heavenly being, do not represent
different, incomparable, levels of existence in this story but are part of a
single, intercommunicating, whole.

Nnambi's brothers, Walumbe, followed the happy pair to earth and, out of
jealousy, became the spirit of disease and death. He was eventually chased away
and lived in a hole in the earth. But he was finally responsible for all death;
and all ghosts had to visit him before returning to their clan graveyards. On
one occasion a hunter, called Mpobe, followed his dog, which was chasing an
edible rat; and the rat ran down Walumbe's hole. At the bottom was a village and
many people and, a little further on, an old man who was Walumbe. Mpobe was, not
unnaturally, afraid; and he knelt down, as any Muganda would kneel to a chief.
He explained how he had come and was allowed to go home on condition that he
told nobody where he had been. Else, Walumbe would kill him. In the end, of
course, he blabbed to his mother and Walumbe came to claim him. Mpobe asked if
he might first sell his things, buy a cow and eat it. Permission was given; and
Mpobe made the deal last over several years before he finally had to pay his
debt.

There is no suggestion here that death not something to be avoided. But Death
does not appear as a terrible figure but as a patient creditor-having certainly
the last word but, dealing with man on almost equal terms.

The recent experience of a Marakwet woman in Kenya, interviewed in the 1960's,
tells the same story. In a dream she was visited by the ghost of her dead son,
who asked her to get him his favourite fruit. This she did, leaving it at the
spot he indicated. The next night he again asked for fruit. She replied, 'If you
can't come home without making a nuisance of yourself, why don't you stay where
you are?' So the third night he just came for a chat.

‘There is nothing more natural than the supernatural;’ and any attempt to
understand the religious experience of traditional Africans must begin from a
reversal of our normal secular Western, but not particularly Christian, idea
that the "natural" and "supernatural" are different. Certainly, there is a
recognition, in many African societies, of a hierarchy of power. The Creator
God, if there is a belief in him or it, is the source of this power, as he or it
is the source of all things. But power is found, in descending order, in lesser
spirits, in ancestral ghosts, in chiefs, who are often the focus of
communication with the spirits, in witches and sorcerers. Finally in ordinary
men and women, animals, plants and inanimate things.

Sometimes this power is personalised. The Baganda, for instance, have many
stories of pregnant girls, deserted by their lovers, who gave birth to water
instead of a child. The water became a river; and the girl's spirit might catch
unwary travelers and drown them. Near one village in Uganda there lived a
leopard which kept the people in its care and warned them if other leopards were
coming to steal their stock. An animal's horn, which had been filled with
suitable ingredients and empowered by a spirit, could be used by its owner to do
jobs for him at a distance and could speak to him.


Sometimes the power has to be put into things, as into the horn and articles of
sorcery which will be discussed later. But sometimes it is inherent. A Buganda
man who killed the animal after which his clan is named was believed to have
killed his clan totem and automatically died. If a pregnant woman laughed at a
lame person, her child would be born lame. If a sheep, a goat or a dog got onto
the roof of a house, the inhabitants would leave it at once, saying it was
unlucky to live there. All these things were 'taboo.'


When this power was first described by Europeans, it was called by the
Melanesian word mana. The Melanesians used mana to explain any exceptional
excellence or skill in men - the power of a chief, the success of a warrior,
skill in rearing pigs or raising large crops of food. The attempt to obtain mana
was thought to be the basis of Melanesian religion. A similar idea was later
found in Polynesia, where mana was described as an all-pervasive psychic force
behaving very much like electricity. People and things which were positively
'charged' could pass it by contact to one that was 'negative'. Unless this
process was properly controlled, damage might result. A positively charged
chief, for instance, might come into contact with a commoner. The chief would
lose some of his power; and the commoner might be injured. An example of this
type of reaction in 2 Samuel 6.6f. Therefore 'taboos' were imposed to prevent
the fatal contact.


But the electrical metaphor, which was so effectively used to describe the
action of mana, can be used to illuminate the relation between 'natural' and
'supernatural'. Sir Arthur Eddington once pointed out that a physicist's
description of matter in terms of electrons and protons might easily give the
impression that a chair consists largely of wide-open space - hardly a suitable
support for sitting. The common-sense account is entirely different. Common
sense can normally disregard the electrical basis of matter. Electrical forces
become of concern to common sense when they are not properly controlled - when
lightning strikes, for instance - or when they are harnessed to human welfare to
produce effects which are impossible without them. To harness them requires
specialist skill; lightning may be extremely frightening. But we do not
therefore regard electricity as supernatural. It is an inescapable part of our
natural environment with special powers for good or ill. It is very much in this
way that traditional peoples understand the power which operates in taboos and
magic, through the ghosts and the spirits. It underlies all life; but 'common
sense' can normally disregard it. When out-of-the-way things happen, or when a
man needs special power for a particular purpose - to deal with misfortune or to
seek unusual success - he becomes aware, as we become aware of electricity, of
something which he believes to be around him and available all the time.
But a word of warning is necessary. Electricity can be dangerous as well as
useful. The Doctor Who stories suggest how the popular imagination fears
electricity when it gets into the hands of men who use it for their own evil
ends. An electric iron may give a fatal shock. So a man who can control mana may
be feared as well as admired. If he can use it to improve his own crops, he can
equally well use it to do harm to others. Among the Lugbara of Uganda, a man who
consistently has better crops than his neighbours is liable to be accused of
witchcraft and punished.

There are three ways in which men have looked at the universe (other ways may be
possible). One is that of materialism, which sees all phenomena as matter
organised in less or more complicated forms but, because it is matter,
ultimately subject to control by man. The second, which has been described in
this chapter, sees all phenomena as the expression of a mysterious force, which
may or may not be personalised, which men may try to control for their own
advantage but to which they are ultimately subject. The third divides phenomena
into ‘natural’ and ‘supernatural’, the one subject to man's control the other
wholly mysterious. In order to understand the ‘religions’ described in this
book, it is necessary to try to suspend judgement as to which of theses views of
the universe is true and to enter as fully as possible into the point of view of
the second.



Return to:Atoms and Ancestors index pages, the Africa pages, or to Nurelweb

Cushitic pastoralists Explorations in the Prehistory Roger Blench

Cushitic pastoralists


Introduction

The origin and homeland of the major linguistic groupings in Africa,
has been a subject of controversy since the first tentative attempts to
classify the more than 2000 languages of the continent. Although
most scholars are now agreed on the assignment of most languages
to one or other of the phyla present in Africa, the intemal arrangement
of the subgroups within each phylum is very much open to discussion.
Models of interna1 structure influence the historical interpretation of
ethnolinguistic diffusion; if one branch of a phylum is considered
especially close to another then historical models must account for
its speakers' contiguity at some time in pi-ehistory.

Of particular interest in this respect is the Afroasiatic phylum, both
because it is so widespread in Africa and the Near East and because
its intemal structure is as yet very unclear. Afroasiatic has a somewhat
ambiguous status among the major language phyla of the world. As
the grouping that includes not only several languages sanctified by
major world religions, but also the earliest written language, it has
benefited from a massive research and publication effort in certain
rather specific areas. It also has old-established traditions of scholarship
that have not always had a positive effect on innovative research.

40 V L'homme et l'animal dans le bassin du lac Tchad

One branch of Afroasiatic that presents a specific problem is Chadic,
the family of 150 + languages centred on Lake Chad but spreading
from the borders of Sudan to northwestein Nigeria. Chadic is clearly
the most intemally diversified subgroup of Afroasiatic and perhaps
for that reason might be considered as the most ancient branching.

However. linguistic geography suggests rather strongly that it is indeed
an intrusive group reaching the region after the establishment of the
Nilo-Saharan and Niger-Congo phyla (see maps in Perrot 1988;
Crozier & Blench 1992; Blench 1993a, 1997a). Since its nearest
relatives are geographically remote (Berber or Cushitic) it has often
been suggested that speakers of the Proto-Chadic were mobile
pastoralists of some type. This has never been substantiated either
linguistically or archaeologically and no date has been proposed for
such a movement.

The purpose of this paperl is to put forward a mode1 to account for
the position of Chadic within Afroasiatic and to suggest a time and
a route whereby Chadic pastoralists could have ai-rived at the Lake
Chad area. Linguistic support foi- this hypothesis is presently limited
to livestock tei-minology; full confirmation of this idea could only
come from much more detailed comparative work within Afroasiatic.
The paper explores the history of ideas concerning the interna1
classification and membership of Afroasiatic and to a lesser extent of
Nilo-Saharan, since the two phyla intei-penetrate in the crucial region
of Central Africa. It then considers the domestic stock in this region
of Central Africa and sets out the linguistic evidence for connections
between Afroasiatic subgroups. Finally, archaeological evidence that
can be linked to the pioposed migrations is reviewed.


1 I would like to thank the organisers of Méga-Tchad for allowing me to
present this long and somewhat complex paper, a preliminary version of
which was given at SOAS in 1995.1 would like to thank David Appleyard,
Jean-Charles Clanet, Richard Hayward, Hermann Jungraithmayr and Kay
Williamson who have commented on various versions of it. Lionel Bender
has been the source of stimulating debates on the subject of the classifi-
cation of Nilo-Saharan and Afroasiatic, while not commenting directly on
the text of the paper.

R. BLENCH-The westward wanderinqs of Cushitic pastoralists 41 V
Afroasiatic and Nilo-Saharan

Classification

Historical Views of Afroasia tic

Ruhlen (1 987: 87 ff.) gives a useful concise history of the classification
of the languages that constitute the phylum. The kinship of Hebr-ew,
Arabic and Aramaic was recognised as early as the 1530s, and Ludolf
pointed out the affinity of Ethiosemitic with the near Eastern languages
in 1702. The name "Semitic" was proposed in 178 1 by von Schlozer.
Berber and some of the Chadic languages, notably Hausa were added
during the course of the nineteenth century. The earliest version of
Afroasiatic as presently understood probably appears in Müller (1 876-
87) who linked Egyptian, Semitic, Berber, Cushitic and Hausa, the
only known Chadic language at the period.

A phylum under the name Afroasiatic goes back to Joseph Greenberg
(1963). Previously, the preferred name was "Hamito-Semitic", an
unfortunate conjunction both clumsy and redolent of suspect racial
theories. Hamito-Semitic is by no means expunged from the lexicon:
hence the confus in^ titles of various collections of conference
proceedings (cf. Bynon 1984). Even disregarding the "Hamitic
hypothesis" Hamito-Semitic gives a primacy to Semitic that is entirely
without linguistic justificationz. Other proposed names include
Afrasian, Lisramic (Hodge 1976) and more strangely, Lislakh. These
have not been widely adopted and Afroasiatic will be used here.


Afroasiatic has been the subject of a number of overviews, beginning
with Müller (op. cit.). Historically, the most important of these have
been Cohen (1947) and Diakonoff (1988). Hodge (1 971, 1976)
represents a summary of the situation in the early 1970s. In 1995,
two very different perspectives on Afroasiatic were published, both
accompanied by substantial data tables (Ehret 1995: Orel & Stolbova
2 Much the same has been the case with Sino-Tibetan, where the written
record of Chinese came to be regarded as evidence for its primary split
with the largely unwritten Tibeto-Burrnan languages. As Van Driern (1995)
has recently shown, this is not supported by the linguistic evidence, which
suggests that Chinese should be classified with Bodic.

42 V L'homme et l'animal dans le bassin du lac Tchad
1995). The implications of these works have yet to be fully absorbed,
but the contrast between them is that Ehret is concerned to justify a
particular view of Afroasiatic phylogeny (Fig. 2, p. 43), while Orel
and Stolbova are oriented towards etymologies. Ehret argues for a
particular version of Afroasiatic phonology while Orel and Stolbova
take for granted that proposed by Diakonoff.


An aspect of Afroasiatic that is worth noting is the important role
played by scholars whose focus has been text. Inteipretations have
been, even more than usual, a miiror to the intellectual preoccupations
of each scholarly generation. Ancient Egyptian has always been
inteipreted by Mediterraneanist scholars and this is reflected in the
interpretations of the sound-system. The undoubted African
contribution has been largely ignored or implicitly denied.


Despite this, it is fair to Say that there has been a revolution in the
treatment of Afroasiatic largely brought about by the massive growth
in studies of African languages. Greenberg (1 963) was responsible
for the establishment of this phylum in its pi-esent foi-m. His particular
contribution was the dethronement of Semitic fi-om its foimerly central
position. and the emphasis he placed on its i-elations with the languages
of Africa.


The Interna1 Phylogeny of Afroasiatic


Gi-eenberg's hypotheses marked an important development, but in
one way they remained resolutely old-fashioned; they left the inteinal
stiucture of Afroasiatic unexplored. Greenberg's classification allowed
five CO-ordinate branches, with Cushitic subdivided into five further
CO-ordinate branches. This is represented in Figure 1.

I I I I l
Semitic Berber Ancient Egyptian Cushitic Chadic
l I l I I
Northern Central Eastern Western Southern
1 Figure 1
The principal subdivisions of Afroasiatic in Greenberg (1963).


R. BLENCH-The westward wanderings of Cushitic pastoralisls 43 V
This scheme broadly follows Cohen (1 947) although Cohen included
only Hausa as an example of a Chadic language and offered no specific
hypothesis about Cushitic. Greenberg was undoubtedly the first
researcher to outline Chadic as a distinct language family, eliminating
the typological elements that had confused Lukas' classification.
The most significant development since this period has been the
recognition that Greenberg's "Western Cushitic" is quite distinct from
other branches of Afro-Asiatic. To mark this, it has been renamed
Omotic (Bender 1975, 1988). Most scholars have accepted the
coherence of Omotic as a group and agree on its assignment to
Afroasiatic. Some researchers would prefer to retain Omotic within
Cushitic, but these are now in a minority. In the case of the other
branches of Cushitic, there has also been considerable discussion
about whether it really constitutes a family and Beja, Ethiopian
Cushitic and Southein Cushitic are often treated as distinct branches.
Ehret (1987) has proposed a "proto-Cushitic" making explicit the
hypothesis that these branches form a unity. New data on Dahalo have
made its usual classification with South Cushitic less evident and
some writers now wish to make it an independent branch of Cushitic
(Tosco 1991 ).
Despite a wealth of documentation, attempts to put a structure to the
groupings within Afro-Asiatic have been relatively few. Within Chadic,
the interna1 classification of the most ramified of the subgroups of
Afroasiatic has proved particularly complex. Greenberg (1963) left
Chadic with nine rather ill-defined subgroups, but Newman and Ma
(1966) made a major breakthrough in proposing three divisions.
Newman (1977) later expanded this to four with the separation of the
Masa group, although Tourneux (1 990) has argued that Masa should
be re-incorporated in Central Chadic. Barreteau and Jungraithmayr
(1993) in a study combining lexicostatistics with proposed lexical
innovations, have split West Chadic into two CO-ordinate groups,
opposing Hausa and the Plateau Chadic languages, such as Ron, with
the Miya-Warji and other northeastern languages such as Ngizim.
It is interesting to note from the point of view of intellectual history
that the first proposa1 to specifically link Cushitic and Chadic seems
to have been made in 1909 by Leo Reinisch, the great Austiian scholar
of the languages of the Hom of Africa. Reinisch noted that these
languages were linked with Semitic and Egyptian and concluded on

44 V L'homme et l'animal dans le bassin (lu lac Tchad
ground of linguistic geography that Afroasiatic ("Chamitische" in his
terminology) must have originated in Africa. Reinisch's conclusions
are rarely cited and were probably far more unwelcome in 1909 than
Greenberg in the 1950s and even Greenberg had to face considerable
opposition. Arelated and challenging view of the Afroasiatic homeland
was put fonvard by Behrens (1985) who used linguistic evidence,
especially livestock terminology, to suggest that the homeland of
Berber was far from its present centre of gravity. Behrens argued for
a region of Western Sudan with subsequent diffusion both West and
north some 6000 years BC.
Most recently, there have been a number of developments that have
yet to be fully evaluated. The most important of these are:
a) The proposal that Ongota, a moribund language 3 of southwestem
Ethiopia constitutes a valid seventh branch of Afroasiatic (Fleming
et al. 1992).
b) Blaiek (in press) has proposed that Elamite. an extinct language
of the Ancient Near East, either constitutes a seventh branch of
Afroasiatic or is CO-ordinate with it. Elamite is usually classified
with Dravidian, spoken in South India, but does show clear
resemblances with Afroasiatic. Blaiek proposes a structure where
Afi-oasiatic is related to Dravidian at a higher level and Elamite
foims a bridge between the two. Whether the links between Elamite
and Afroasiatic reflect a genetic relationship or are simply a case
of extensive loanwords, remains to be explored.
Ehret's (1995) schema of the intemal structure for Afroasiatic is fairly
similar to the models proposed formally or informally by other
researchers and 1have adapted sonle of his proposed names for the
nodes (e.g. North Afroasiatic and Erythraic). Figure 2 shows a
composite view of Afroasiatic incorporating my own views and some
of the recent proposals made conceining Elamitic, On, Dota etc.
Bender (1 997) has also proposed a radically new structure for
Afroasiatic ("upside-down Afi-asian" in his teiminology). His revised
tree is as follows (Figure 3).
3 Ongota has only 6 speakers as of 1997, down from the 15 reported in
1992.

R. BLENCH-The weslward wanderings of Cushitic pastoralists 45 V
Proto-Afro-Asiatic
Elamite?
Erythraic
1

I
North-Afroasiatic
I
~esl-1 East Beja Aga* E. Cushitic S. Cushitic erter Egyptian Semitic OmoÉ Ongola?
1 Figure 2
Proposed Revised Afroasiatic Classification
Afroasiatic
I
1
Central
Macro-Cushitic
r-1
Chadic Egyptian-Coptic Berber Sernitic Cushitic Ornotic
1 Figure 3
The lnternal Structure of Afroasiatic according to Bender (1997)


46 V L'homme et l'animal dans le bassin du lac Tchad
Bender proposes a homeland for Afroasiatic (the region where Chad,
Sudan and Libya meet today) and a date (10,000 BP). Perhaps even
more startlingly, he canvases the possibility that Indo-European is
somehow an offshoot of his "Macro-Cushitic". Whether these
suggestions will be taken on board by the scholarly commiinity will
depend on the presentation of fuller evidence than is given in his short
article.
Nilo-Saharan
The Nilo-Saharan language phylum remains the least-known and most
controversial of African language groupings. Since its initial delineation
by Greenberg (1963) there have been a series of studies, piincipally
by Bender (1991 b, 1996a, b) and Blench (1995a). An unpublished
classification by Ehret is reviewed in Bender (1 996a). Figure 4 shows
one mode1 of the relationship between the various branches of Nilo-
Saharan; quite different structures are given in some of the literature
cited. This is a much more controversial topic: but for the present
argument, this is marginally relevant: al1 that needs to be taken on
trust is that there are large numbers of fragmented Ni1o.-Saharan
languages presently in the region between Chadic and Cushitic.
Proto-Nilo-Saharan
Kuliak
Shabo
Berta Kunarna Kornuz
Fur Maba
Saharan Songhay 1
I
Sudanic
Kado
(= Kadugli-Krongo)
Central Sudanic East Sudanic
1 Figure 4
Interna1 Phylogeny of Nilo-Saharan: Minimal Hypothesis.


R. BLENCH-The westward wanderings of Cushitic pastoralists 47 V
Trees, Peoples and Origins
This paper treats non-phylogenetic views with limited sympathy.
Languages are spoken by people and communities split and diversify
in real historical time for a multiplicity of reasons. Although the
interplay of factors that underlie these processes remains only partly
understood, it is only by seeking to apply sociological models of
known processes that we can hope to mode1 the past. Historically
speaking, divisions in communities are a common process. Apastoral
society divides as one group goes to seek pasture and water elsewhere.
An agricultural community divides as one body of villagers go to
seek new agricultural land.
Languages usually spread by two complementary processes, language-
shifting and physical expansion. The Hausa and Fulfulde languages
of West Africa are good examples of these processes at work. Hausa
has largely spread in historical time through the Hausaization of
agricultural populations, a process still at work today. Fulfulde,
however, has spread across West-Central Africa through the physical
movement of pastoralists with their herds. There is no reason to suppose
these processes were not as common in the past as they are today.
The Inter-Saharan Hypothesis
Much of the conventional literature on the diffusion and spread of
Afroasiatic assumes, implicitly or explicitly, a trans-Saharan route
for the development of Chadic. Links with Berber and Egyptian
abound in the literature and the analogy with the medieval Islamic
trade-routes is extended into an unknown past. This paper argues,
that while trans-Saharan routes were of importance. the present-day
distribution of Chadic languages and their immediate affinities in
Afroasiatic can best be understood by assuming that speakers of the
proto-language migrated from east to West. from the Nile to the Niger.
to exaggerate slightly. To distinguish this from the conventional view
1 propose to cal1 this the "inter-Saharan" coi-ridor.
If Chadic and Cushitic languages do have a privileged relationship.
then this is best explained by the assumption that Chadic speakers

48 V L'homme et i'animal dans le bassin du lac Tchad
broke away from a branch of Cushitic and moved westwards. In view
of the intemal diversification of Chadic this must have been several
thousand years ago. Given the long distances involved, it seems likely
that this migration took place in the context of a pastoral subsistence
system. In other words, a group of Cushitic speakers, herding cattle,
sheep and goats, began to drift westward (Blench 1995b).
The inter-Saharan coi-ridor is today largely in the Republics of Sudan
and Chad. Today and presumably in the past it was inhabited by Nilo-
Saharan speakers. If such a migration took place, then one
confirmatory piece of evidence should be the scattered presence of
livestock terms in Nilo-Saharan languages al1 the way between the
Nile and Lake Chad. The data tables given below provide some
evidence that this is indeed the case.
Leo Reinisch pointed out in the early part of the century that there
are striking lexical correspondences between Nile Nubian and
Cushitic. Work on the prehistory of Nubian and the languages of the
Nile Valley by Bechhaus-Gerst (1 98415, 1989, 1999) has made this
more historically probable. She shows that when Nobiin speakers
reached the Nile Valley (by Ca. 1500 BC) they encountered resident
speakers of Cushitic languages from whom they borrowed a large
number of words, most strikingly those connected with livestock
production (goat, sheep, hen, pig, dung, stock enclosure, milk etc.).
The languages that are apparently the source of these loanwords are
Highland East Cushitic (Haddiya etc.) rather than Beja or the Agaw
languages which are today geographically closer.
Q

P Pastoralism and Domestic Aninials
Why Domestic Animals?
One approach to exploring the history of a language phylum is to
examine in detail a semantic field that illuminates some aspect of the
subsistence strategies of its assumed speakers. In the case of Khoisan,
for example, it would be sensible to look in detail at animal names
and hunting technology. In the case of Afroasiatic, livestock
terminology provides a useful window, since it has long been obseived

R. BLENCH-The westward wanderings of Cushitic pastoralists
49 V
that many lexical items are widely distributed through the phylum.
Livestock is also useful because:
a.
it is probably older than cultivation in Africa: cattle, donkeys, cats
and guinea-fowl are indigenous domesticates:
b.
it is represented in rock-art and it is bettes attested archaeologically
than cultivated plants;
c.
maRy Afroasiatic speakers are still pastoralists.
Terms for domestic animals occupy a curious halfway house between
cultural and basic lexicon. Domestic animals are ancient, but their
exact antiquity is often in question. Therefore their presence at the
period when a hypothetical proto-language is spoken remains doubtful.
The Principal Pastoral Species:
Cattle, Sheep and Goats
The ancestry of domestic cattle remains one of the most disputed
topics in the broader debate over domestication. The most
comprehensive overviews of the origin of the traditional cattle breeds
of Africa are Epstein (1971) and Epstein and Mason (1984). Wild
cattle seem to have been present in the Ancient Near East and
Northeast Afi-ica as late as 5000 B.C. and the earliest African cattle
presumably derive from these. Muzzolini (1983b) has reviewed the
evidence for cattle in Ancient Egypt and Gautier (1987) has
synthesised the archaeological evidence for Northern and Middle
Africa. Blench (1993b) represents an oveiview of the existing evidence
from cattle breeds and races. MacDonald and MacDonald (1 999)
represents a comprehensive recent summary of the archaeozoological
evidence for West-Central Africa.
Very early dates, before 9000 BP, are postulated for cattle in the
Eastern Sahara (Gautier: 198 1: 336, 1984: 69). Wendorf & Schild
(1984: 420) note comparable domesticated cattle from Syria by the
tenth millenniurn BP. Breunig et al. (1993) and Breunig and Neumann
(1996) give dates of > 3000 BP (uncalibrated) for the bones of
domesticated cattle in Boino.
Many early representations in rock-art of cattle in the Ancient Middle
East, Egypt and the Sahara show cattle with some sort of hump.

50 V L'homme ef l'animal dans le bassin tlu lac Tchad
Muzzolini (1983a. 1991) concludes that there are sorne apparently
early images of humped cattle in Saharan rock-art which do not fit
with the late introduction of zebu and therefore advmces the hypothesis
of an independent evolution of humpedness in the Sahara. The present-
day humped breeds of West Africa almost certainly cornbille genetic
rnaterial from the indigenous breeds and the incorning zebu. Recent
work on the cattle DNA does appear to suggest a dual domestication
in the Indian and NE Afi-ical Near Eastern regions (Loftus et al. 1994).
Goat
The goat. Capra I~ircus aegagrus, evolved 7 million yeass ago. but it
was probably not dornesticated until 10,000 years ago in the Mesolithic
period of the Ancient Near East (Gautier 198 1 : 336: Mason 1984b).
Goats were certainly kept in Egypt after 5000 BC and presumably
spread to sub-Saharan Africa shoi-tly after that. The site at Haua Fteah,
Cyrenaica in North Afi-ica, has srnall ruminant bones datiny from the
6800 BP with no associated cattle and Kadero. near Khartum. has
both cattle and small iurninants at 6000 BP (Gautier 1981 :336).
As with goats. sheep are descended from an ancestral Near Eastern
wild sheep and domestic foms xe recorded in Iraq as early as 11,000 BP.
In Africa, they first occur as domesticates in the eastern Sahara at
7000 BP and at Haua Fteah in North Africa at 6800 BP (Gautier 198 1:
336). Muzzolini (1990) has reviewed the evidence for sheep in Sahasan
rock ait and his revision of the chronology placing the first appearance
of sheep rather later. at 6000 BP, seems generally accepted.
Associated Species:
Donkeys, Dogs and Guinea- fo wl

The wild ass, Equus asinus africanus, is indigenous to the African
continent and is usually divided into a chain of races of subspecies
spi-eading from the Atlas rnountains eastwards to Nubia, down the
Red Sea and probably as far as the border of present-day Northein
Kenya (Groves 1966, 1986: Haltenorth & Diller 1980: 109; Kingdon

R. BLENCH-The westward wanderings of Cushitic pastoralists 51 V
1997). Four notional races, atlanticus. africanus, taeniopirs and
somaliensis are located approximately as shown in earlier studies
(e.g. Haltenorth & Diller 1980). However, two of these, atlanticus
and tueniopus have been rejected more recently and indeed the
proposed arlanticus race tuins out to have been based on misidentified
zebra bones (Kingdon 1997: 311). The extent to which the wild ass
penetrated the interior of Africa is controversial, but it is generally
considered unlikely that it evei- occurred in sub-Saharan regions.
Groves (1986) argues that the wild ass extended into the Near East
in ancient times and CO-existed with the onager, Equus hemionus.
Blench (1999 a) summarises the recent evidence for the history of
the donkey in Africa.
The main features differentiating races of wild ass are the amount
and type of stripes and the shoulder crosses. However, their charac-
terisation may be somewhat bluired, since populations that survived
into historical times have almost cei-tainly crossed with feral donkeys,
leading to a merger of characteiistics. Civil war in both Somalia and
Eritrea may mean that the fragile populations marked have
disappeared or are severely threatened. There are two doubtful
populations of wild ass near Siwa oasis in Egypt and further south
towards the Sahara proper.
Records of domestic donkeys begin in Egypt in the fourth millennium
B.C. with cleai. representations of working donkeys by the middle of
the next millennium (Epstein 1971: 392). At about the same period
there are textual records of extremely large herds of donkeys, many
of which were apparently used for portage. The expeditions to Punt
(Ethiopia) consisting of large trade caravans usually included
numerous donkeys (Kitchen 1993). Donkeys from the second
millennium BC occur at Shaqadud in the Butana grasslands of Sudan
(Peters 1991). Donkeys were found in the faunal assemblages at
Carthage in the Roman period (1-4th centuries AD) (Levine 1994).
The earliest record of a donkey in West Africa is at Siouré in
Senegambia (MacDonald and MacDonald 1999). The stratigraphy
of this site appears to be reliable and the donkey bone is dated to
between 0-250 A.D. After this? the next finds of donkey bones are at
Akumbu in Mali with a date of 600-100 A.D. However, such finds
are extremely rare even in sites, such as Tegdaoust, where there have
been extensive finds of other domestic species.

52 7 L'homme et l'animal dans le bassin du lac Tchad
The ancestry of the domestic dog remains uncertain and a number
of canids may be implicated in present-daytypes (Clutton-Brock,
1984). The dog is not native to Africa and was introduced at an
unknown period in the past. Epstein (1971, 1) who reviewed this
question at length, shows that dogs were known in Egypt in the pre-
Dynastic period and so could have been brought across the desert in
prehistoric times. It is likely that there have been multiple
introductions from different sources, although the only race found
in Central Africa is what Epstein calls the "pariah dog". Dogs are
kept everywhere in Africa for hunting and security purposes. Frank
(1 965) has exhaustively reviewed the literature on domestic dogs in
Africa, and Epstein (1 97 1) has examined the evidence for the
evolution of the African dog.
The crested or helmet guinea-fowl, Numida meleagris galeata, Pallas,
is part of the native fauna of West Africa. It is distributed from
Senegambia to Cameroon and is also found in a part of Western Zaire.
It was presumably domesticated long ago, although the larger domestic
races closely iesemble their wild countei-parts. There are several wild
species and geneia of guinea-fowl in West and East Africa, notably
N. nzeleagris nzeleagris in Sudan and Ethiopia, but apparently only
N. nzekagris galeata has been domesticated (see Donkin 199 1, Map 1).
Wild guinea-fowl are still regularly trapped as a source of food and
their eggs are raided in the bush. Mongin and Plouzeau (1 984) present
an overview of recent scholarship on the guinea-fowl worldwide while
Ayeni (1 983) summarises existing infoimation for West Africa. Donkin
(1991) is an "ethnogeographical" study of the guinea-fowl that
synthesises a great deal of scattered material, especially on the
iconography of the guinea-fowl in the Mediteiranean. Blench (1999 b)
summarises the recent evidence for the history of the guinea-fowl in
Africa.
The history of the domestic pig in Africa remains highly controversial.
Although the wild pig, Sus scrofa, is native to north Africa, and its


R. BLENCH-The westward wanderings of Cushitic pastoralists 53 V
range extends along the Atlantic Coast to the Senegal River, there is
no evidence that it was ever domesticated in Africa (Epstein 1971, II).
Pigs are usually thought to have been domesticated in Anatolia and
the earliest archaeological finds of pigs date back to 7000 BC.
Domesticated pigs were kept in the Ancient Near East and Egypt from
the end of the fifth millennium BC (Epstein 197 1, II: 340). Pigs were
known along the North African littoral, and seem to have spread down
the Nile at least as Sennar, where they are still kept (Spaulding &
Spaulding 1988). Pigs cannot be herded and are generally not kept
by pastoralists unless they settle. Since pigs cannot survive by grazing
for more than part of the year and depend on grown food they are
usually kept by settled farmers. There is evidence that semi-feral pigs
spread into the Omotic-speaking regions of the Ethiopian borderland
and westward at least as far as Kordofan (see map of sites where pigs
were recorded in Spaulding & Spaulding (1988)) and may have spread
to West-Central Africa along a corridor from Darfur to Lake Chad.
Blench (1999 c) summarises the recent evidence for the history of
the domestic pig in Africa.
!Linguistic Evidence
This section sets out the principal base forms proposed to illustrate
the inter-Saharan connection. 1have given apparent or actual cognates
in Berber, Egyptian and Semitic where these have been proposed
rather than omit evidence that may run contrary to the argument
proposed here. I have not given the source of the data for each
attestation to keep the references to manageable length. In most cases
these are standard published sources and are listed in the references.
#+a,"cow, cattle"
West and Central Chadic attest a form something like $a-with likely
cognates in East Chadic (Jungraithmayr and Ibriszimow 1994,I: 43).
Southein Cushitic also has a voiceless lateral: #J-, in the same Ci slot
(Ehret 1987: 80).
Related terms seem to be found in Semitic but not in Berber or
Egyptian, if the 411 coi~espondence holds. Cohen (1 947: 182) presents
an #1-series for Semitic, including Akkadian lu and Soqotri le3e:

54 V
L'homme et l'animal dans le bassin du lac Tchad
Acronyms/Toponyms etc.
'Central Africa' here refers to the area presently
encompassed by Chad, Cameroon and Central
African Republic.
Orthography
Spellings can be phoneniic (where the language
has been analysed in depth), phonetic (where the
foim given is the surface form recorded in field-
work) or orthographic (talen froni earlier sources
with inexplicit rules of transcription). The follo-
wing table gives the fornis used here and their IPA
equivalents;
This Other IPA
Work Orthographic (1993)
Y j
c ch ts
j dj d3
5 dl. zl, 2 B
4 il, hl, SI. S 4
Words extractcd froni French sources have been
normalised to niake coniparison easier.
Tone and stress marks
The exact sigiiificance of tone-niarks varies froni
one language to another and I have used the
conventions of the authors in the case of published
languages. The usual conventions are :
High
Mid unniarked
-.
Low
Rising
v
,.
Falling
In non-tonal languages, such as South Arabian,
stress on vowels is niarhed with an acute accent,
a convention 1 have retained.
In Afroasiatic languages with vowel length distinc-
tions, only the first vowel of a long vowel is tone-
marked. Some nineteenth century sources, such as
Heinrich Barth, use diacritics to mark stress or
length. These have been 'translated' into modein
notation to avoid theconfusing implication of tone-
marking.
Vowel Length
Long vowels are usually niarked by doubling in
African languages but are often transcribed with a
macron in Semitic etc. All long vowels have been
transcribed by doubling Io make comparison
simpler.
Reconstructions
A word prefaced by # represents a pseudo-recons-
Lniction. in other words a form denved from inspec-
tion of roots that looks probable, but has notbeen
rigorously established through sound-correspon-
dences. This contrasts with *, used to indicate
reconstructions from systeniatic sound-corres-
pondences.
Acronyms
*
Reconstmction established froni coniplete
analysis of sound-change
# 'Quasi-reconstruction' established from
quick inspection of CO,Onates
BC Benue-Congo
BES Berber-Egyptian-Semitic
C Consonant
Eth Ethiopic (unlocated Ethiopian root)

HEC Highland East Cushitic
N Nasal
NC Niger-Congo
NS Nilo-Saharan
PAA Proto-Afroasiatic
PC Proto-Cushitic Ehret, 1987
PEC Proto-Eastein Cushitic Ehret, 1987
PO Pinto-Oniotic
PS Prolo-Seniitic
PWS Proto-West Sudanic Westerniann, 1927
s/r sniall ruminant (in tables)
V Vowel

R. BLENCH -The westward wanderings of Cushitic pastoralists 55 V
Phylum Family Branch Language Attestation Gloss
A A Cushitic Agaw Bilin lawi cow
East Gedeo lali cattle
Oromo loon cattle
West Rift Iraqw 3 ee cow
Chadic West Ngizim 4à cow
Semitic
Central
Central
Ga'anda
Akkadian

lu'um
cow
wild bull
l1
bull
Iuu
Arabic la'an bull
Jibbali
(=Shahri)
Ié'llh6ti cow 1
Kuliak II, 13 cow l
1Table 1
Attestations of #+a, "cow, cattle".

which may form a cognate set. These may: however, refer to the wild
bull, still present in the Middle East and Arabia in the fifth millenium
BC. Leslau (1938: 61) points out that the Hebrew persona1 name
"Leah" is almost certainly cognate with these forms. The common
Ethio-Semitic #lam for cow is something of a puzzle (Appleyard
1977: 26). Semitic scholars seem generally unwilling to connect this
with the lateral fricative roots in Cushitic. As a result it has been
suggested that the 1-!z-nz roots meaning "food" in Arabic and "shark"
in Soqotri (lehenz) are cognate. A semantic coi~espondence between
"shark" and "cow" has a certain Greenbergian charm, but Akkadian
Lu'ui11 "wild bull" is surely more likely. Leslau (1979, II: 379) also
notes a comparison with Arabic lihm, "aged ox".
#saa, "cattle"
This root is a suppletive plural for "cow", i.e. "cattle" throughout
Eastern Cushitic and Beja. Hudson reconstmcts *sa?a for Highland
East Cushitic and Ehret (1987: 61) has reconstiucted *Sua!- for Proto-

56 V L'homme et l'animal dans le bassin du lac Tchad
Cushitic. An interesting further possible link suggested by Beja is
with words for rhinoceros. Beja Se pl. Sa for rhinoceros is certainly
close to the common plural for cattle Sa'. Since these two animals
would have inhabited the same ecological niche in the pre-
domestication period, such a semantic shift is at least plausible.
This root is analysed by Pilszczikowa (1960) who links it with words
for "sheep" in Semitic and Egyptian. Behrens (1985: 179) and
Jungraithmayr and Ibriszimow (1994,I: 43) assume that the roots with
lateral fricatives in Ciare also cognate. This analysis is not adopted
here.
Although attested in Hausa, sâa is an isolated citation and it seems
likely that this is a loanword, possibly from Berber or directly from
Arabic. The same may apply to the isolated Kotoko witness the source
of which may be Shuwa Arabic, which has sâ 'a, meaning "wealth in
livestock"4. Other attestations related to shaanui~occur in Old Semitic
langages, for example, Akkadian sa'ni*m,and in Berber. Tamachek
1 Phylum Farnily Branch Language Attestation Gloss 1
AA Cusliitic Beja Beja Sa. pl. 3a'a CON
Eastern Sidanio saa cow
A Far saga
Proto-Cushitic *J~~,. (Ehret)
Chadic West Hausa siiniyii
sâa pl. sliaanuu
Central Kotoko h'sââ
Semitic Akkadian ga'num
Shuwa Arabic sâ'a
Brrbrr Tamachek eesu. pl. eeswaan
1 Tamazight esu cow 1
NS Kuliak Tepeth saa cattle kraal
C. Sudnnic Sara Ndoka sa+ cow
Modo si cow
1Table 2
Attestations of #saa, "cattle".

R. BLENCH -The westward wanderings of Cushitic pastoralists 57 V
eeswaan "cattle". The nasals seem to be added in Berber-Egyptian-
Semitic forms. Althoiigh these roots clearly CO-exist in Afroasiatic,
the absence of widespread attestations for s-in Chadic do suggest
boirowing from Berber.
#k-1-rn, "bull"
The Chadic #k-m-(bull) resembles closely the common Agaw term
for "cattle", something like #karn-(Appleyard 1984: 39). Jungraith-
mayr and Ibriszimow (1994,I: 43) consider this connected with more
widespread Chadic roots for "meat". Cushitic forms usually have
#k-Y-IIIso Agaw may have shortened this. Cohen (1 947: 112) noted
a common Afroasiatic k-for "bull" though he speculated that it was
possibly a widespread loan. Although this word occurs throughout
West Rift it so closely resembles the Ethiopian foims that it is probably
Phylum Family Branch Language Attestation Gloss
AA Omotic N. Ometo Maale k'6lnio cattle
S. Onieto Koy ra Ltymo cattle
Cushitic Agaw Bilin kam cattle
East Gedeo korma bull
Arbore koll catt le
Southrrn Iraqw karamao Steer
Chadic West Kulere kyààl cattle
Karrkare kwàni bull
Central Hwana k wÈl bull
East Lele kdl-b& cattle
Mubi kiyi cattle
Ancient Egyptian km3 bull
l
Saharan Kanuri k5nni calf
1
(O) This root also occurs in a number 1 Table 3
of Bantu languages in Tanzania and I assume
Attestations of #k-1-m, "bull"
these are loans from West Rift languages.
4 Also a cornmon association in Indo-European; see "cattle" and "capital"

58 V L'homme el l'animal dans le bassin du lac Tchad
a recent loan rather than an old retention, since the practice of using
pack-oxen is probably not as old as domestication.
Phylum Farnily Branch Language Attestation Gloss
AA Ornotic S. Ometo Zayse-Zergula gal6
Cushitic Brja tagar
Agaw Bilin gar
East Burji giree
Arbore goran heifer
Somali agor bull calf
Chadic West Mburku $'wi cow
Znnr gààl cow
Central Guduf dayalr bull
Vulum gàrii bull
Semitic West Ugaritic '-3.1 calf
Hebrew Teegel calf
Egyptian Coptic ~gol calf
NS E. Sudanic Nubian Nobiin ;or calf
1Table 4
Attestations of #gor, "calf"

#gor, "calf"
This root was suggested by Bechhaus-Gerst (1 999) as a loanword
into Nobiin. However, it clearly is more widespread as Table 4 shows,
assuming the Chadic forms are indeed cognate.
The Zayse-Zergula citation may be a single loanword, since this is
not a common form for calf in Omotic. The Chadic forms are almost
certainly cognate with each other, but less certainly cognate with the
Cushitic forms.

R. BLENCH-The westward wanderings of Cushitic pastoralists 59 V
#b-g-r, "male ruminant"
Phylum Family Branch Language Attestation Gloss
A A Ornotic Mocha big0 sheep
Shinasha baggoo sheep
Cushitic A;aw (= Central) Bilin bagga sheep
Chadic West Karekare bùgùrk sheep
Kwaarni rnb6Mri sherp
Central Racarna b6;Gré sheep
Musgu béggere he-;oat
East Mokilko bû-ggàR cattle
Kera ku-purki he-;oat
Sernitic Centriil Shuwa Arabic bagar cattle
South Harsüsi bakarat cattle
Bzrber Tarnachek Ni;er A-beg;ug sheep
NS Saharan Saharan Zaghawa b66guri young
rnamied man
E. Sudanic Nubian Nobiin fag goa
1Table 5
Attestations of #b-g-r, "male ruminant".

Jungraithmayr and Ibriszimow (1 994,I: 8 1) cite this as #b-k-r, a pan-
Chadic root and describe this as a Wanderwort. Given its widespread
distribution in Afi-oasiatic and the antiquity of goat domestication,
there is no reasons why this should be so. However, they also (op.
cit. 148) give #baga for "sheep" in Central Chadic and these roots
must almost cei~inly be combined. The Berber citation (fi-om Behrens,
1985: 167 ex Heinrich Barth) is unusual and not otherwise attested
in Berber; it may therefore be a loanword. Bechhaus-Gerst (1989)
also argues that the West Rift forms such as Lraqw be'i are also related,
although this involves vowel changes and the assumption that the
deleted consonant is /g/.The -ri consonant common to Semitic and
Chadic is curious, suggesting a source in Ethiopic not synchronically
attested.

60 7
i'homme et l'animal dans le bassin du lac Tchad
#kol, "goat"
Phylum Family
Branch Language Attestation Gloss
A A Omotic
N. Ometo Maalc k6le goat
South Karo k'oli goat

Cushitic East
Burji k'al-60 goat
Yaaku k311-&h castratr goat
Rendille kelex castrate goat
Chadic West Koîjar Loor largl
castrate goat
Bade akun goat
Central Dera kwiarào goat
Uroovin xwun goat
Ysdina kàanio goat
Kuliak 1k ka1 goat
Kadu Central Katcha k3r3mak goat
E. Sudanic Suimic Bodi koloy goat
Temein Dese kwjrarnàl he -g09t
E. Nilotic Turkana
a-korai 209t
S. Nilotic Proio-Kalenjin "~WEr he -goai
Snharari Kanuri kaliwo virgin she-goal
(O) These may be loans from Kanuri kanyîand thus indirectly or unconnected
1Table 6
Attestations of #kol. "goat"

#t-m-k, "sheep"
The base form #t-ln-koccurs in Afroasiatic, Saharan, and Niger-Congo
(not cited in the table) and is definitely a Wanderwort. Newman
(1 977: 3 1) proposes *tanzki for sheep in proto-Chadic and
Jiingraithmayr and Ibriszimow (1 994,I: 148) note its presence in al1
branches of Chadic, but they do not consider the exteinal Afroasiatic
lookalikes to be cognate. The word for a two-yeai- old sheep in Teda-
Daza, durna, is cognate with the Kanuri term di'nzi and lamb turna as

Fi. BLENCH-The westward wanderinqs of Cushitic pastoralists
with the Berti tanzi. The boirowings into Niger-Congo would have
come from multiple introductions in the West African Sahel at the
termini of trans-Saharan routes.
Phylum Farnily Branch Language Attestation
AA Cushitic East Oromo turnzamaa
Chadic West Hausa tunkiyii.
pl. tumaikii
Central Bade taarnan.
taniahun
Hiei of Kiria tlmbaka
Tpala tàrnâk
Masa Masa dirniina
East Mubi turnik
Kera taamagi
Berbsr Wargla adanimani
NS C. Sudanic Moru-Madi Moru temilé
Kadu Eastern Krongo diirnà
1 Saharan Kanuri dirni
Kanuri tania
Beni tami
1Table 7
Attestations of #t-m-(k),"sheep".

The third literal, -k-,only occurs in Chadic and is presumably an early
affix or compound. As the Oromo citation seems to be isolated, without
further evidence the provenance of this root as Erythraic must remain
doubtful. However, the Nilo-Saharan citations look convincing,
providing sonle evidence for the base form further East in Central
Africa. The Berber citation is interesting, since this word explicitly
applies to hair sheep that have been bi-ought from Mali and Niger
(Delheure 1987: 53). It is likely that al1 such forms in Berber are
loanwords fi-om Chadic or even Saharan.
61 V
Gloss
castrate
sheep
sherp
sheep
sheep
sheep
sheep
sheep
hair sheep
sheep
female zoat
Sernale larnh
lamb

62 V L'homme et l'animal dans le bassin du lac Tchad
#'aare, "srnall ruminants"
Phylum Family Branch Language Attestation Gloss
AA Omotic Koyra ?&are Rock of shwp
Cushitic Beja Beja ano. annee e we
East Burji aray
Rendille 'a@ s'r
South Gorwaa aaraa ;o;its
Chadic West Hausa irà-ir8 lune-lc;;ed sir
East Lele ore ;o;its
Semitic Guraze Muher Br2z
Berber Guanche ara ;oat
Kuliak 1k ri goat
E. Sudanic E. Jebel Gaam àar shcep
Nubian Meidob arar ram
Nyiman: Dinik Ér sheep
Maba Masalit Br i i-am
Fur Fur w-ùri ewr
Saharan Za;hii~,a aro ewe
66rù flock
arro he-zoat
1Table 8
Attestations of #'aare, "small ruminants"

The Guanche citation is curious and may well be just coincidence.
Perhaps related is a root that floats between sheep and goat, onu,
which appears as a word for ewe in Beja but surfaces in Gurage as
"goat". These could be a subset of Proto-Semitic #n-z.
#xorge, "he-goat"
Ehret (1987: 22) reconstructs Proto-Cushitic *?erg-for "small
ruminant" but this is most likely "he-goat" to judge by the predominant

R. BLENCH -The westward wanderings of Cushitic pastoralists 63 V
1 Phylum Farnily Branch Language Attestation Gloss 1
AA Ongota Ongota orgai-ko goat
Omotic Omet« Wolaitta lorggé he-goat
Cushitic East Harso orkakk6 he-goal
Saho xarge he-goat
Borana orge hzifer came1 (!)
South Asax 'a'aku sheep
T'roto-Cushitico *?org-small ruminant
Chadic West Hausa àwiakii she-goat
Ngizim iakù soat
Central Ndrenie àwik goat
l1 "Proto-Chadicoo
7 -
l NS Saharan
East Dangla
Teda
i~skà
*a(w)ku
orko
goat
goal
zoat
l1
1
(0) Ehret (1987: 22)
("0) Newman (1977).
1Table 9
Attestations of #xorge, "he-goat"

gloss. This particular root has both Omotic and even Ongota cognates>
although these could well be loans (Table 9).
Since this form does not otherwise occur in Nilo-Saharan, the Teda
attestation may be a recent loanword.
Newman (1977) gives #a(w)kuas a proto-Chadic reconstruction, but
there seem to be sufficient attestations of a Iateral in C2 position in
Cushitic to add this to the reconstruction. Jungraithmayr and
Ibriszimow (1 994,I: 43) give the root as "wk-and similarly attest its
presence in al1 branches of Chadic. These are almost certai~ily cognate
with the Cushitic kor-roots. Indeed it is possible to speculate that
kor-and org-are in fact the same root with metathesis.

64 V L'homme et l'animal dans le bassin du lac Tchad
#m-r-k, "castrated small ruminant"
1 Phylum Family Branch Language Attestation Gloss 1
A A Omotic N. Ometo Wolaitta mara larnb
Koy ra mari ram
Chadic West Polci niaar gout
Tangale rnara castrate goat
Masa Masa marak castrate goat
East Birgid mar bu Il
NS Maba Masalit mar-Sa bull
Fur Fur Fur m aà lanib
Tama Tama mi bull
C. Sudanic Modo mùrikù castrated sheep
E. Sudanic W. Nilotic Dinh -Bol- aniiil shssp
E. Nilotic Tes0 e-rnerek2k ran~
Nubian Birgid niar ra rn
1 Suimic Murle rnerkee ram
1Table 10
Attestations of #m-r-k. "castrated ruminant"

This root is so widely attested that it is suiprising to find no evidence
for Cushitic.
The historical and archaeological evidence for the wild ass or donkey
does not appear to suggest either early domestication or transmission
to West Africa. The linguistic evidence, however, is clear. The #k-r
root is spread from Omotic to West Chadic, with intervening Nilo-
Saharan attestations and is also largely apparently absent in BES
which Ilas a series of quite distinct roots.
The most likely history of this root is that it originally developed as
a word applied to "wild ass". probably in Ethiopia. Bender (1 988:
152)reconstructs proto-Omotic "krrrfor ass. Skinner cites *dAn&i~Ar
for proto-Cushitic based on forms such as Bilin daxllara. The dV-
prefix strikingly links Southern Cushitic and Agaw and is apparently

R. BLENCH-The westward wanderings of Cushitic pastoralists
65 V
Phylum Family
Branch Language Attestation
AA Omotic
Gimira Benr Non kur
Mao Hozo
kuuri
Southem Karo
uk'uli
Agaw Bilin
daxwara
Cushitic Eastern
Saho okaalo
West Rift Iraqw
daqwaay
Chadic West
Karekare k6or6o
Central Vulum
kùré
Masa Pevc
koro
East Nanrere
kuri
NS E. Sudanic
W. Nilotic Mabaan tmrm
Temein Keiga-Jiriu kul-kjq

C. Sudanic Sara Mbay k6ro
Saharan Kanuri k6ro
1Table 11
Attestations of #k-r, "donkey".

not attested in Eastein Cushitic at all. The Mabaan form is only cognate
if initial t- coi~esponds to k-. Although the West Chadic forms closely
resemble those of Masa and East Chadic, they may be loanwords
from Kanuri.
This is an extremely widespread root through the Horn of Africa, and
appears virtually unchanged in niimerous East Cushitic and Omotic
langiiages. This suggests that it is probably a widespread loanword
and should not be reconstructed to Proto-Ciishitic. The Ethio-Semitic
languages have a different word, cognate with the Near Eastern Semitic
root h-171-r:argiling that the ancestral speakers of these languages
already had a domestic donkey when they crossed the Bab el Mandeb.
The most probable source for harre are the Oromoid words for "zebra".
Zebras are not part of the fauna of the highlands but they are
widespread in the lowlands south of the Ethiopian Plateau and are

66 V L'homme et i'animal dans le bassin du lac Tchad
very familiar to pastoral groups such as the Borana. Borana has harre
dida for zebra, with dida meaning "outdoors" or "open air" The term
harre was probably originally a word for zebra in lowland Oromoid
and was transferred to donkey once it was fully domesticated. The
zebra would then become the "donkey of the plains". Formations
such as Konso harr-etita for "zebra" would be calques of the Borana
expression, already using the borrowed word for donkey. The
development of the donkey as pack animal is probably reflected in
the Beja harri "anything ridden, from a came1 to a train".
In the Hoin of Africa, an old root for the wild ass #kuur-was largely
displaced by #harre when the domesticated donkey developed
economic significance. The term #harre was probably borrowed from
terms in lowland Oromoid originally applied to "zebra".
Arguing historically from terms for "dog" presents a special problem;
these words have an astonishing similarity across much of Eurasia.
For example, proto-Omotic for dog is *kan(Bender 1988: 145) closely
resembling Proto-Indo-European *bon-(Rabin 1982: 27). Similar
forms are also found in proto-Austronesian and Chinese. Newman
(1 977) proposes #kar-for the original proto-Chadic, forms of which
also show up in Nilo-Saharan and is identical to English "cur".
Jungraithmayr and Ibriszimow (1 994, 1: 49) note the widespread
1-eflexes of this root across Chadic and consider it may reflect a "Central
Saharan areal lexeme". Linguistically. therefore, probably the only
useful evidence comes from compounded or affixed forms.
If, as Bender (1975: 159) and Skinner (1977: 187) suggest, this root
is common Afroasiatic, then Akkadian k-l-b,Arabic kalb and Kabylé
akelbun al1 form part of a cognate set. The k-l-b root is also applied
to wolves in Eurasia (e.g. South Arabian languages) but this is probably
a secondaiy meaning as wolves are absent in Africa. The South Semitic
languages, such as Mehn and Soqotri, explicitly apply the same word
kalb to both "dog" and "wolf'. The third radical, -b,is now generally
considered to be an affix marking wild animals and would not
necessarily travel with the remainder of the word. Rabin (1982: 27)
notes that forms such as Latin canis may be direct loans from
Afroasiatic. Historically speaking, given the Middle Eastein origins

R. BLENCH-The westward wanderings of Cushitic pastoralists 67 V
Phylum Family Branch Language Attestation Gloss
AA Ornotic Common *kana
Cushitic Beja Beja keluus PUPPY
Agaw Bilin zadar]
East Saho kare
Gawwada xar-o
Konso kuta
South Asax kite
Chadic West Hausa kàree
Central Bata kade
Kada kara
East Mokilko zédè
Sokoro kuyo
Sernitic Central Ugaritic k-1-b
South Soqotri kalb
Berber Kabyle akelbun PLlPPS
NS Kuliak Trpeth kudo'
Il
Saharan Kanuri
Teda
kari
kidii
ll
1 able 12
Attestations of #k-r, "dog"
of the dog, this is not improbable. Agaw terms for dog, such as Bilin
gadag, seem to resemble Central Chadic foims very closely, although
this may be accidental similarity.
Another root with some promise is #k-t-r for "puppy". This lexical
item is much more rarely recorded, and therefore less certain. However,
this root has the advantage that it does not appear to be con-espondingly
widespread across the world in the same way as the basic terms
for "dog".
The #t-t-1forms are only found in Gurage and probably loans from
Cushitic.

68 V L'homme et l'animal dans le bassin du lac Tchad
Phylum Farnily Branch Language Attestation
AA Cushitic East Haddiya tuu!iilla
l
l
South Gorwaa kut~ikuti
l
l
Chadic West Karekare tiiutùu
Central Bura kutiru
East Bidiya kurkido
Semitic Gurage Zway ~ulalla
Saharan Kanuri kut~~rii
1Table 13
Attestations of #k-t-r, "puppy"

The most comrnon root in Chadic is #sVb-vVn which Skinner (1 977:
192-3) shows is spread throughout the family. Skinner argues from
this that the word has spread recently, but this seems unlikely as the
guinea-fowl is indigenous to the region. Newman (1977) also notes
this root and proposes #zaban for proto-Chadic while Jungraithmayr
and Ibriszimow (1994,I: 84) propose #z-b-1. Strikingly, the foi-ms in
Cushitic are very similar. The common Ethiopic i-oot appears to be
#z-g-r,widespread in Cushitic and Ethio-Semitic; whether its witnesses
in Omotic are more than sporadic loans remains to be seen. This root
also means "spotted" in many languages. Ehret (1987: 54) suggests
*zagr-for proto-Cushitic, but Agaw foims have -n-in the C, slot and
centralised vowels in VI and V2.
Linguistic evidence suggests thai some of the pigs in West Africa
were introduced at an early period by the Portuguese, "unimproved
Iberian swine", as Epstein has it. Loanwords from Portuguese porco
are widely found in the coastal region of Nigeria (Williamson, P.C.).
But there is also evidence for a chain of teims stretching from Eastein
Burkina Faso to the Sudan-Ethiopian borderlands that appear to be
unrelated to European introductions. Spaulding & Spaulding (1988).

R. BLENCH-The westward wanderinas of Cushitic oastoralists 69 V
Bechhaus-Gerst (1999) and Blench (1999 c) have made preliminary
compilations of the evidence (Table 14).
1 Phylum Farnily Branch Language Attestation 1
Kornan Anej kutum
ES Nyirnang kudur
Old Nobiin kutun
Maba Aiki ginvà wart-hog (?C)
Saharan Kanuri godu warthog
Kadu Karndanz b-oburuk. pl. k-aburuk
Kordofanian Ori; kàdifi
Benue-Congo Nupe kutsü
Kwa Fon agurusa
Gur Dagbane kurutJu
Bantu #CB #-$du wild pis
Ornotic Kefa gudinoo
A A Sernitic Sudan Arabic kadruuk
Chadic Hausa ;hduu
1 Table 14
Attestations of #-kutu, "pig".

This root appears in Nilo-Saharan. Niger-Congo and Afroasiatic and
can also be applied both to the warthog and the bush-pig
(Potamochoerus porcus). Manessy (1972: 314) points out that the
chain of lexemes connecting to the Gur languages can be traced
through dialect and obsolete terms for domestic pig given in Koelle.
It was also cited by Gregersen (1972: 86) who used this as evidence
for a proposed "Kongo-Saharan" grouping (wrongly, given that it is
clearly a widespread cultural loan). Gregersen (op. cit.) also mentions
Greenberg's suggestion that the Saharan form was loaned into *PB.
Schadeberg and Elias (1 979: 84) observe that this root has been loaned
into Sudariese Arabic to give kudruuk.
The linguistic evidence is rather compelling; it suggests strongly that
the small black pigs of the interior of Africa were indeed part of an

70 V Chomme et l'animal dans le bassin du lac Tchad
ancient pig-keeping culture that spread across Central Africa from
the Nile. Pigs were kept in a semi-fera1 manner either roaming
throughout the year or only being confined during the growing season.
The rise of Islam drove pig production into pockets, and the intro-
duction of larger European breeds which crossed freely with the local
pigs has virtually obscured their genetic heritage. The pig, the
"democratic philosopher of the Medieval Sudan" needs to be
highlighted as a significant element in African subsistence strategies.
It may be, however, the transmission of the domestic pig was
chronologically and culturally distinct from the pastoral movement
proposed in this paper; pastoralists usually eschew pigs because they
cannot move long distances.
-$
iThe "Inter-Saharan" Hypothesis
Tentative Historical lrnplications
A rather unexpected consequence of the study of domestïc animal
names is the numbei- of common lexical items shared between Cushitic
and Chadic. This tends to confirm the studies of Mukarovsky (1990,
in press) on numerals and body parts. If this is correct, then Cushitic
and Chadic may shai-e a special relationship and be opposed to Berber-
Egyptian-Semitic or "North Afroasiatic" (Ehret 1995). The links
between Cushitic and Chadic would then be the result of a migration
of Cushitic speakers westward. This is a considerable distance and
might be explained by the gradua1 migration of pastoralist peoples.
The example of the Ful6e pastoralists who have expanded from
Senegambia to the borders of Sudan in the last millennium show that
such a migration can occur (Blench, 1995b, 1999d). The animals
accompanying this migration would have been three species of
ruminant: cattle, goats and sheep. More controversially, donkeys,
dogs, pigs and guinea-fowl may also have been associated with this
movement, although perhaps not kept as pastoral species.
Speakers migrated from the Nile Valley to Lake Chad, as would the
Shuwa Arabs, millennia later. Languages related to present-day Chadic
were presumably once spoken in a strip across present-day Sudan but

R. BLENCH-The westward wanderings of Cushitic pastoralists 71 V
were later eliminated by movements of Nilo-Saharan speakers
(Map 1). Little-known Chadic languages such as Kujarke, spoken in
Western Sudan, may well be the last suiviving remnants of this process
(Blench, in press, b).
Archaeological Correlations
Such a significant long-distance movement of actual population as
implied by this model should have archaeological correlates. One of
the distinct problems in relating linguistic to archaeological evidence
is the patchy nature of excavation. Sudan is relatively well covered,
but data for Chad and the relevant regions of Ethiopia and Nigeria
remain spasse indeed. To seek sites or traditions that might provide
material evidence for such a movement, certain parameters must be
established. The linguistic data provides no inteinal evidence for
dating although the model has to allow sufficient time for the intemal
diversity of Chadic languages to develop. Such a movement of pastoral
peoples must also lie within the known parameters of ruminant
domestication.
A likely candidate for the wandering Cushites is the Leiterbatzd pottery
tradition that has been identified in the Eastern Sahara, most
specifically in the Wadi Howar, which is a now dry river system that
stretches oves 1000 km between Eastern Chad and the Nile Valley 5.
The Howar ends just beyond the Sudanese border and the proposed
migrating pastoralists would then have faced a substantial obstacle
in the shape of the Ennedi and Biltine mountainous regions which
run Noith-South. However, there is a gap between these two outcrops
which would permit pastoral migration, and the herds would then
pick up the Wadi Hawach and thence a seiies of smaller wadis, iunning
towards Lake Chad.
Leiterband traditions were first identified by Kuper (1 98 1) as distinct
from Nubian C-group pottery. They have been subsequently studied
in more detail by Keding (1 993) who argues that this tradition shows
its strongest links with the Khartum Neolithic, out of which it may
5 1 am grateful to Jean-Charles Clanet, who encouraged me to examine
the geography of this region more closely.

72 V L'homme et l'animal dans le bassin du lac Tchad
develop. Keding shows that the pottery traditions are strongly
associated with cattle-keeping and indeed complete cattle skeletons
have been found in pits on Leiterband sites. From this she argues that
the makers of the pottery were pastoralists who also supplemented
their diet with fish, at that period widely available in the rivers. This
pastorallfishing economy is extremely familiar today from the Nilotic-
speakers in the region. such as the Dinka. Map 1 shows the projected
route of the speakers of proto-Chadic as well as the approximate
locations of the wadis referred to above.
O 1 000 km /
-1
-.,7
Cushltrc (taday) Cushific (presumed former exlensian)
1Map 1
Proposed migrations of Chadic-speakers.

Leiterband traditions have yet to be convincingly dated directly, but
if the chronological sequence linking it with the Khartum Neolithic
is correct, then it would begin to develop appioximately 4000 BP.
This would suit the present hypothesis extremely well: if the Cushites

R. BLENCH-The westward wanderings of Cushitic pastoralists 73 V
began their westward movement from Ethiopia some 6-5 000 years
BP they may have been responsible for the Khartum Neolithic
(beginning 5700 BP) and then gradually spread westwards along the
Wadi Howar some 4000 years ago. The increasing aridity after this
period severed the links with the Nile Valley allowing an independent
evolution of decorative styles. In the meantime, the continuing
westward drift reaches Lake Chad Ca. 3000 BP. This would then link
with the earliest dates for cattle in this region at about this period
(Breunig et al. 1994).
v Conclusion
Various models of the internal structure of Afroasiatic have been
pi-oposed, most notably those of Fleming (1983), Ehret (1 975, 1999,
Stolbova and Orel (1 995), Bender (1997) and Blakek (in press). The
terminology of domestic livestock suggests strongly that Cushitic
and Chadic share a special relationship and that this is reflected in
the terminology for species of domestic animals. As names for
domestic animals are notoi-iously susceptible to loaning, the
demonstration of such links is far fi-om constituting proof of the
specific Cushitic-Chadic relationship. However, other evidence also
supports this notion, pointing to an avenue for further investigation.
Bibliography
APPLEYARD Linguistics: Papers of the Third
D.L., 1977 -
A comparative approach International Hamito-Semitic
to the Amharic lexicon. Malibu, Congress. Amsterdam,
Afroasiatic Linguistics, 5,2, Undena John Benjamins :33-68.
Publications,
AYENI J.S.O., 1983 -
APPLEYARD Studies of the Grey-Breasted
D.L., 1984 -
The internal classification of Helmeted Guineafowl (Numida
the Agaw languages: a comparative meleagris galeata, Pallas) in Nigeria.
and historical phonology. In Bynon J. World's Poultry Science Journal
(ed.) Current progress in A fro-Asiatic 39 (2) : 143-151.

74 V L'homme et l'animal dans le bassin du lac Tchad
BARRETEAU
D.,
JUNGRAITHMAYR, -
H., 1993
Calculs lexicostatistiques et
glottochronologiques sur les langues
tchadiques. In Barreteau D. & von
Graffenried C. (eds.) Datation et
chronologie dans le bassin du lac
Tchad.Paris, Orstorn, Colloques
et séminaires : 103-140.
BARTHH., 1862 -
Collection of vocabularies of Central
African languages. Gotha, Justus
Perthes.
BECHHAUS-GERST -
M., 198415
Sprachliche und historische
Rekonstruktion irn Bereich des
Nubischen unter besonderer
Berücksichtigung des Nilnubischen.
Sprache und Geschichte in Afrika,
6: 7-134.
BECHHAUS-GERST -
M., 1989
Nubien und Kuschiten im Niltal.
Sprach- und Kulturkontakte im 'no
man's land'. Afrikanische
Arbeitspapiere, Sondernurnrner 1989.
BECHHAUS-GERST -
M., 1996
Sprachwandel durch Sprachkontak?
am Beispiel des Nubischen im Niltal.
Koln, Rudiger Koppe.
BECHHAUS-GERST -
M., 1999
Linguistic evidence for the prehistory
of livestock in Sudan. In Blench R.M.,
MacDonald K.C., The origin and
development of African livestock.
London, University College Press :
449-461.
BEHRENSP., 1985 -
Wanderungsbewegungen und
Sprache der frühen saharanischen
Viehzüchter. Sprache und Geschichte
in Afrika, 6: 1335-21 6.
BENDERM.L., 1975 -
Omotic: A New Afroasiatic Language
Family. Carbondale, University
Museum, Southern Illinois University.
BENDERM.L., 1982 -
Livestock and Linguistics in North and
East African Ethnohistory. Current
Anthropology, 23: 31 6-318.
BENDERM.L., 1988 -
Proto-Ornotic: Phonology and
Lexicon. In Bechhaus-Gerst M.,
Serzisko F. (eds.), Cushitic-Omotic:
Papers from the International
Symposium on Cushitic and Omotic
Languages. Harnburg, Buske Verlag :
121-162.
BENDERM.L. (ed.), 1991a.-
Proceedings of the Fourth Nilo-
Saharan Conference. Bayreuth, 1989.
Harnburg, Buske Verlag.
BENDER M.L., 1991b -
Subclassification of Nilo-Saharan
In Bender M. L, 1991 a: 1-36.

BENDERM.L., 1994 -
Aroid (South Omotic) Lexicon.

Afrikanische Arbeitspapiere,
38: 133-1 62.
BENDERM.L., 1996a -
The Nilo-Saharan languages:
a comparative essay.
Munich, Lincorn Europa.
BENDERM.L., 1996b -
Nilo-Saharan '95. Proceedings
of the 6th Nilo-Saharan Linguistics
Conference, 1995. Afrikanische
Arbeitspapiere, 45: 1-26.
BENDERM.L., 1997 -
Upside-down Afrasian. Afrikanische
Arbeitspapiere, 50: 19-34.
BLAZEKV., 1999 -
Elarn: a bridge between ancient near
east and Dravidian India? In Blench
R.M., Spriggs M. (eds.), Archaeology
and Language, IV. London,
Routledge : 48-78.
BLENCHR.M., 1991 -
The dessication of Lake Chad in
1990. Méga-Tchad Bulletin, 9 112.

R. BLENCH-The westward wanderings of Cushitic partoralists 75 V
BLENCH R.M., 1993a -
Recent developments in African
language classification and their
implications for prehistory. In Shaw T.,
Sinclair P., Andah B., Okpoko (eds.),
The Archaeology of Africa. Food,
Metals and Towns. London,
A. Routledge : 126-138.
BLENCH R.M., 1993b -
Ethnographic and linguistic evidence
for the prehistory of African ruminant
livestock, horses and ponies.
InShaw T., Sinclair P., Andah B.,
Okpoko (eds.), The Archaeology
of Africa. Food, Metals and Towns.
London, A. Routledge : 71-103.
BLENCHR.M., 1995a -
Is Niger-Congo simply a branch
of Nilo-Saharan? In Nicolai R.,
Rottland F. (eds), Proceedings
of the Fifth Nilo-Saharan Linguistics
Colloquium, Nice, 1992.Koln,
Rudiger Koppe: 83-130.

BLENCHR.M., 1995b -
A History of Domestic Animals in
Northeastern Nigeria. Cahiers de
Sciences Humaines, 31 (1): 181 -238.

BLENCH R.M., 1997a -
Language studies in Africa.
In Vogel J.O. (ed.) Encyclopaedia
of precolonial Africa. Walnut
CreekILondonlNew Delhi, Altamira:
90-1 00.

BLENCH R.M., 1997b -
Neglected species, livelihoods and
biodiversity in difficult areas: how
should the public sector respond?
Natural Resource Perspective
Paper 23. London, Overseas
Development Institute.
BLENCH R.M., 1998 -
Le West African Shorthorn cattle
au Nigeria: histoire, répartition
et productivité. In Seignobos C.,
Thys E. (éd.), Des taurins et des
hommes, Cameroun, Nigeria. Paris,
Orstom, Latitudes 23 :249-292.
BLENCHR.M., 1999a -
A history of donkeys and mules
in Africa. In Blench R.M.,
MacDonald K.C. (eds), The origin and
development of African livestock.

London, University College Press :
339-354.

BLENCHR.M., 1999b -
Minor livestock species in Africa.
In Blench R.M., MacDonald K.C.
(eds), The origin and development
of A frican livestock. London,
University College Press: 314-337.

BLENCHR.M., 1999c -
A history of pigs in Africa.
In Blench R.M., MacDonald K.C.
jeds), The origin and development
of African livestock. London,
University College Press: 355-367.

BLENCHR.M., 1999d -
Why are there so many pastoral
peoples in East Africa? In Azarya V.,
Breedveld A., de Brujn M., van Dijk H.
(eds.), Pastoralists under pressure:
Fullie societies confronting change
in West Africa. Leiden, Brill: 29-49.

BLENCHR.M., 1999e -
The languages of Africa: macrophyla
proposals and implications for
archaeological interpretation. ln
Blench R.M. and Spriggs M. (eds.),

Archaeology and Language, IV.
London, Routledge: 29-47.
BRADLEYD.G., MACHUGH D.E.,
LOFTUS R.T., Dow R.S., HOSTE C.H.,
CUNNINGHAM -
E.P,. 1994
Zebu-taurine variation in Y
chromosomal DNA: a sensitive assay
for genetic introgression in West
African trypanotolerant cattle
populations. Animal Genetics,
25: 7-12.
BREUNIGP., GARBA U.,
WAZIRI, l., 1991 -
Recent archaeological surveys
in Borno, NE Nigeria. Nyame Akuma,

37: 10-17.

76 V L'homme et l'animal dans le bassin du lac Tchad
BREUNIGP., BALLOUCHE
A.,
NEUMANNK., ROSING F.W.,
THIEMEYERH., WENDT P.,
VAN NEER W. 1993 -
Gajiganna. New data on early
settlement and environment in the
Chad Basin. Berichte des Frankfurter
Sonderforschungsbereichs 268.
Frankfurt, 1992, 11 : 51-74..
BREUNIGP., K. NEUMANNW., 1996 -
Archaeological and archaeobotanical
research of the Frankfurt University
in a West African context. Berichte
des Frankfurter
Sonderforschungsbereichs 268.
Frankfurt, 1995, Vlll : 181-191.
BYNONJ., 1984 -
Berber and Chadic: the lexical
evidence. In J. Bynon (ed.) Current
progress in A fro-Asiatic Linguistics:
Papers of the Third International
Hamito-Semitic Congress,
Amsterdam, John Benjamins :
241-290.
CLUT~ON-BROCK -
J., 1984
Dog. In Mason, I.L. (ed) Evolution of
Domesticated Animals. London,
Longman: 198-21 1.
COHEN D., 1970 -
Dictionnaire des racines sémitiques
ou attestées dans les langues
sémitiques. Fasc 1, II. La Haye,
Mouton.
COHEN M., 1947 -
Essai comparatif sur le vocabulaire et
la phonétique du Chamito-Sémitique.
Paris, Honoré Champion.
CONNAHG., 1981 -
Three Thousand Years in Africa.
Cambridge, Cambridge University
Press.
CROZIER D.. BLENCH R.M., 1992 -
Index of Nigerian Languages.
Second Edition. Dallas, SIL.
CYFFER N., HUTCHINSON J. (eds),
1990 -
Dictionary of the Kanuri Language.
Dordrecht, Foris.
DELHEUREJ. 1987 -
Dictionnaire Ouargli-Français. Paris,
SELAF.

DENHAMD., CLAPPERTON
H.,
OUDNEYW., 1828 -
Narrative of Travels and Discoveries
in Northern and Central Africa.
London, John Murray
(3rd ed. in 2 vols).

DIAKONOFFI.M. 1988 -
Afrasian Languages. Moscow, Nauka.

DONKIN R.A. 1991 -
Meleagrides: an historical and
ethnogeographical study of the
Guinea fowl. London, Ethnographica.
DOORNBOSP., BENDER M.L.., 1983 -
Languages of Wadai-Darfur.
In Bender M.L. (ed.), Nilo-Saharan
Language Studies. Michigan State
University Press : 43-79.
DRIEM G. van, 1995 -
Black Mountain Conjugational
rnorphology, Proto-Tibeto-Burrnan
rnorphosyntax, and the linguistic
position of Chinese. In Nishi Y.,
Matisoff J. A., Nagaro J. (eds), New
Horizons in Tibeto-Burman Morpho-
syntax, Osaka, National Museum
of Ethnology, Senri Ethnological
Studies 41 :229-259.
EHRETC., 1987 -
Proto-Cushitic reconstruction.
Sprache und Geschichte in Afrika,
8: 7-1 80.
EHRETC., 1995 -
A froasiatic: a historical reconstruction.
University of California Press.

EPSTEI~IH., 1984 -
Ass, mule and onager. In Mason I.L.
(ed) Evolution of Domesticated
Animals. London, Longrnan: 174-184.

EPSTEINH., 1971 -
The origin of the domestic animais

77 7
R. BLENCH-The westward wanderings of Cushitic partoralists
Corporation (2 vols).
EPSTEINH., MASON I.L., 1984 -
Cattle. In Mason I.L. (ed) Evolution
of Domesticated Animals. London,
Longman :6-27.
FLEMINGHAROLDC., 1965 -
The age-grading cultures of East
Africa: an historical enquiry. Ph.D.
University of Pittsburgh.
FLEMINGH.C., 1983 -
Chadic external relations. InWolff E.,
Meyer-Bahlburg H. (eds.), Studies in
Chadic and A froasiatic Linguistics.

Hamburg, Helmut Buske : 17-31.
FLEMING H.C., YILMA A., MITIKU A.,
HAYWARDR., MIYAWAKI
Y. MIKESH P.,
SEELIG J.M., 1992 -
Ongota or Birale: a moribund
language of Gemu-Gofa (Ethiopia).
Journal of Afroasiatic Languages, 33:
181-225.
FRANK B., 1965 -
Die rolle des hundes in afrikanischen
kulturen. Wiesbaden, Franz Steiner.
FRATKINE., K.A. GALVIN,
E.A. ROTH, 1994 -
African pastoralist systems.
An integrated approach. Boulder
and London: Lynne Rienner
Publishers.
GALATYJ.G., P. BONTE, 1991 -
Herders, warriors and traders:
Pastoralism in Africa. Boulder,
Westview Press.
GAUTIERA., 1981 -
Contributions to the Archaeozoology
of Egypt. In Wendori F., Schild R.
(eds.), Prehistory of the Eastern
Sahara. New York, Academic Press,
317-344.

GAUTIER A. 1984 -
Archaeozoology of the Bir Kiseiba
Region. In Wendorf F. and Schild R.
(eds). Cattle-keepers of the Eastern
Sahara. Dallas, Department of
University :49-72
GAUTIERA., 1987 -
Prehistoric Men and Cattle in North
Africa: a dearth of data and a surfeit
of models. In Close A.E. (ed.),
Prehistory of Arid Nofth Africa.
Essays in Honor of Ffed Wendorf.

Dallas, Southern Methodist University
Press : 163-187.

GREENBERG -

J. H., 1963
The Languages of Africa. The Hague,
Mouton for Indiana University.
GREGERSEN
Edgar A., 1972 -
Kongo-Saharan. Journal of African
Languages, 1 1, 1 :69-89.
GROVESC.P., 1966 -
Taxonomy. In Groves C.P., Ziccardi F.,

A. Toschi. (eds), Sull'Asino Selvatico
Africano. Supplement to Richerche
di Zoologia Applicata alla Caccia,
5, 1:2-1 1.
GROVESC.P., 1986 -
The Taxonomy, Distribution
and Adaptations of Recent Equids.
In Meadow R.H., Uerpmann H.P.
(eds), Equids in the Ancient World.
Wiesbaden, Ludwig Reichert: 11-51

HALTENORTH -

T., H. DILLER, 1980
A field guide to the mammals of
Africa, including Madagascar.
London, Collins.
HODGE C.T., 1971 -
Afroasiatic: A survey. The Hague,
Mouton.

HOOGE C.T., 1976 -
Lisramic (Afroasiatic): an overview.
ln Bender M. L., Fleming H.C. (eds.),
The non-Semitic languages of
Ethiopia. Michigan University Press :
43-65.

HUDSON R. ined. -
English-Beja dictionary. ms

HUTCHISONJ.P., 1981 -
The Kanuri Language: a reference

78 V L'homme et l'animal dans le bassin du lac Tchad
grammar. African Studies Program,
University of Wisconsin, Madison.
JUNGRAITHMAYR
H., NAGEL G., 1991 -
West African savannah. Culture,
Language and Environment
in an historical perspective.
Frankfurt-am-Main.
JUNGRAITHMAYR
H.,
IBRIS~IMOW D., 1994 -
Chadic lexical roots.
Berlin, Reimer, 2 vols.

KEDINGB., 1993 -
Leiterband sites in the Wadi Howar,
North Sudan. In Krzyzaniak L.,
Kobusiewicz L., Alexander J. (eds).

Environmental change and human
culture in the Nile Basin and Northern
Africa until the Second millennium
B.C. Poznan Archaeological Museum:
371 -380.
KINGDONJ. A. 1997 -
The Kingdon field guide to African
mammals. San Diego, Academic
press.
KITCHEN K.A., 1993 -
The land of Punt. In Shaw T.,
Sinclair P., Andah B. and Okpoko A.
(eds), The Archaeology of Africa.
Food, Metals and Towns, London,
Routledge: 587-608.

KOELLES.W., 1854 -
African Native Literature. London,
Church Missionary Society.

KUPERR. 1981 -
Untersuchungen zur
Besiedlunsgeschiche der ostlichen
Sahara. Vorbericht über die
Expedition 1980. Beitrage zur
Allgemeinen und Vergleichenden
Archaologie, 3: 2 15-275.
LESLAU W., 1938 -
Lexique Soqotri (sudarabique
moderne). Paris, Klincksieck.
LESLAUW., 1963 -
Etymological Dictionary of Harari.
Berkeley & Los Angeles, University
of California Press.
LESLAUW., 1979 -
Etymological Dictionary of Gurage.
Wiesbaden, Harrassowitz, 3 vols.
LETHEMG.L., 1920 -
Colloquial Arabic, Shuwa dialect of
Bornu, Nigeria and of the region of
Lake Tchad. London, Crown Agents.
LEVINEMarsha A., 1994 -
The analysis of Marnmal arid bird
remains. In Hurst. H.R. (ed.),
Excavations at Carthage, the British
Mission. Vol II, 1. Oxford, Oxford
University Press forthe British
Academy : 314-317.
LEWICKIT., 1974 -
West African food in the Middle Ages.
Cambridge University Press.
LOFTUS R.T., MACHUGH D.E.,
BRADLEYD.G., SHARP P.M.,
CUNNINGHAM
E.P., 1994 -
Evidence for two independent
domestications of cattle. Proceedings
of the National Academy of Science,
9 1 :2757-2761.
MACDONALD
K.C., R.H. MACDONALD,
1999 -
The Origins and Development of
Dornesticated Anirnals in Arid West
Africa. In Blench R.M.,
MacDonald K.C (eds), The origin
and development of African livestock.
London, University College Press:
127-162.
MANESSYG., 1972 -
Les Noms d'Animaux domestiques
dans les langues Voltaïques. in
Langues et Techniques, Nature
et Societé. Approche Linguistique.
Paris, Klincksieck, 1 :301-320.
MASON I.L. (ed.), 1984a -
Evolution of Domesticated Animals.
London, Longman.
MASON I.L. 1984b -
Goat. In Mason I.L. (ed.), Evolution

R. BLENCH-The westward wanderings of Cushitic partoralists 79 '1
of Domesticated Animals. London,
Longman :85-99.
MAUNY R., 1954 -
La savane nilo-tchadienne, voie
de pénétration des influences
égyptiennes et orientales?
In Conferencia internacional
de Africanistas occidentales,
II. Madrid :83-1 15.
MEEK C.K., 1931 -
Tribal Studies in Northern Nigeria.
London, Kegan Paul, Trench
and Trubner.
MONGIN P., PLOUZEAU M., 1984 -
Guinea-fowl. In Mason I.L. (ed.)
Evolution of Domestica ted Animals.
London, Longman: 322-325.
MUKAROVSKY
H., 1976-1 977 -
A study of Western Nigritic. 2 vols.
Wien, Institut für ~~~~tologie
und Afrikanistik, Universitat Wien.
MULLERF. 1876-87 -
Grundriss der Sprachwissenschaft.
Wien, 4 vols.
MUZZOLINI A., 1983a -
L'Art Rupestre du Sahara Central:
Classification et Chronologie.
Le Boeuf dans la Préhistoire africaine.

Thèse de troisième cycle, université
de Provence, 2 vols.
MU~ZOLINI
A., 1983b -
Les types de boeufs domestiques
dans l'Égypte ancienne. Bulletin
de la Société méridionale
de spéléologie et de préhistoire,
XXIII: 55-74.
MUZZOLINI A., 1990 -
The Sheep in Saharan rock art.
Rock Art Research, 7, 2: 93-1 09.

MUZZOLINI A., 1991 -
Les débuts de la domestication
au Sahara et les gravures rupestres
les plus anciennes ("école bubaline").
Préhistoire Ariégeoise, XLV 1:
211-233.

NACHTIGAL,
G. (ed. and trans. A.G.B.
and H.J. Fisher), 1980 -
Sahara and Sudan II. London, Hurst.
NEWMANP., 1977 -
Chadic Classification and
Reconstructions. Afroasiatic
Linguistics, 511.
NOYE D., 1989 -
Dictionnaire Foulfouldé-Français.
Paris, Geuthner.
OREL V. & STOLBOVA -
O., 1995
Hamito-Semitic Etymological
Dictionary. Leiden, Brill.
PERROTJ. (ed.), 1988 -
Les langues chamito-sémitiques.

Les langues dans le monde ancien
et moderne. Paris, CNRS, vol. 3.
PETERSJ. 1991 -
The faunal remains from Shaqadud.
In Marks A.E. and Mohammed-Ali A.
(eds.), The Late Prehistory of the
Eastern Sahel, Dallas, Southern
Methodist University Press :197-235.

PHILLIPSON -

D.W., 1993
The antiquity of cultivation
and herding in Ethiopia. In Shaw T.
Sinclair P., Andah B. and Okpoko A.
(eds), The Archaeology of Africa.
Food, Metals and Towns, London,
Routledge :344-357.
PILCZ~ZIKOWA -
N., 1960
Le Haoussa et le chamito-sémitique
à la lumière de l'Essai comparatif
de Marcel Cohen. Rocznik
Orientalistyczny, XXIV, 1 :97-1 30.
RABIN C., 1982 -
Ron-Semitic etymologies.
In Jungraithmayr H. (ed.), The Chad
languages in the Hamitosemitic-
Nigritic Border area.. Berlin, Reimer
Verlag : 24-31.
REINISCH L., 1909 -
Das personliche Fürwort
und die Verbalklexion
in den Chamito-Semitischen
Sprachen. Wien, Alfred Holder.


80 V L'homme et l'animal dans le bassin du lac Tchad
RUHLENM., 1991 (revised edition) -
A Guide to the World's Languages:
Volume 1: Classification. London,
Edward Arnold.
RYDERM.L., 1984 -
Sheep. In Mason I.L. (ed.), Evolution
of Domesticated Animals. London,
Longman : 63-84.

SCHADEBERG

T. C., ELIAS P., 1979 -
A Description of the Orig Language.
Tervuren: MRAC.

SKINNERN.A., 1977 -
Domestic Animals in Chadic.
In P. & R.M. Newman (eds.), Papers
in Chadic Linguistics, Leiden,
Afrika-Studiecentrum :175-198.

SKINNERN.A. 1981 -
Loans in Hausa and pre-Hausa: some
etymologies. In H. Jungraithmayr
(ed.), Berliner Afrikanistiche Vortrage.
Berlin, Reimer: 169-202.

SMITH A.B., 1980 -
Domesticated cattle in the Sahara
and their introduction into West
Africa. In Williams M.A.J., Faure H.
(eds) The Sahara and the Nile.
Quaternary environments
andprehistoric occupation in northern
Africa, Rotterdam, Balkema, chap. 20.

SMITH A. B., 1992 -
Pastoralism in Africa. Hurst, London
with Ohio University Press, Athens
and Witwatersrand University Press.

SPAULDINGJ.L., J. SPAULDING, 1988 -
The democratic philosophers of
the Medieval Sudan: the pig. Sprache
und Geschichte in Afrika, Band 9:
247-268.

STENNINGD., 1959 -
Savannah Nomads. London, Oxford
University Press for the International
African Institute.
TOSCO M., 1991 -
A Grammatical Sketch of Dahalo.
Hamburg, Buske.
TOURNEUXH., 1987 -
Les noms des équidés en Afrique
centrale. In Seignobos C. (ed.),
Le Poney du Logone. Paris, Institut
d'élevage et de médecine vétérinaire
des pays tropicaux: 169-205.
TOURNEUXH., 1990 -
La place du Masa dans la famille
tchadique. In Murakowsky H. (ed.),
Proceedings of the Fifth international
Hamito-Semitic Congress. Volume 1 :
Hamito-Semitic., Berber, Chadic.

Vienna, Afro-Pub :249-260.
WENDORFF., SCHILD R., 1981 -
Prehistory of the Eastern Sahara.
New York, Academic Press.
WENDORFF., SCHILD R. 1984 -
Cattle-keepers of the Eastern Sahara.
Dallas, Department of Anthropology,
Southern Methodist University.
WILLIAMSON -
K., 1970
Some food plant names in the Niger
Delta. International Journal of
American Linguistics, 36: 156-1 67.
WINKLERH.A., 1938, 1939 --
Rock-drawings of Southern Upper
Egypt. 2 vols. London, Egypt
Exploration Society.
ZEUNER F.E., 1963 -
A History of Domesticated Animals
London, Hutchinson.