. 2
( 12)


Egypt in the late ¬fth millennium bc, followed by smelting of copper ore in
the ¬rst half of the fourth millennium and the use of bronze (a harder cop-
per alloy) after its invention in western Asia during the third millennium, the
period when Egyptians also carried copper-smelting into Nubia. Iron-smelting
was introduced to Egypt by Assyrians and Greeks during the sixth century bc.
The earliest evidence of the technique further south at Meroe dates to the same
period or slightly later, but the region is unlikely to have transmitted it across
the Sahara because Meroe™s industry became substantial only after the birth of
Christ, its techniques differed from those in sub-Saharan Africa, and research
has found no archaeological evidence of a transmission route southwards. A
more likely source was Carthage. The Phoenicians were the great metalworkers
of the ancient Mediterranean, both in bronze and iron. They worked iron at
Carthage during the eighth or seventh century bc, but there is no direct evi-
dence that they transmitted their skills southwards to West Africa, where iron-
smelting technology was to differ greatly from that in Mediterranean lands. In
West Africa, claims that copper and iron were smelted in modern Niger during
the second millennium bc have not gained general acceptance, but copper was
certainly smelted at Akjoujt in Mauritania during the middle of the ¬rst mil-
lennium bc, while iron-smelting of the same date or slightly later took place in
Niger and northern Cameroun and at Taruga in the centre of modern Nigeria,
a site of the widespread Nok culture whose makers used ground-stone axes,
exploited oil-palms, and produced sub-Saharan Africa™s oldest known sculp-
ture of human ¬gures and other objects in terracotta (pottery). While this
sculpture was being made, iron-working was spreading through West Africa,
where dates from the fourth or third century bc have been found at Nsukka
in southern Nigeria and at sites in Gabon and Congo-Brazzaville. There is also
evidence of early iron-smelting in the Great Lakes region of East Africa. Smelt-
ing furnaces in Rwanda and Burundi appear to date back to a period before
400 bc that cannot be more accurately determined by radiocarbon. There are
similar or perhaps slightly earlier dates from Katuruka in northwestern Tan-
zania. Again the technology differed from that in North Africa or the Middle
Linguistic evidence suggests that the ¬rst ironworkers in the Great Lakes
region spoke Nilo-Saharan languages, but their skills and much of their tech-
nical vocabulary were adopted by the Bantu-speakers who reached this region
from the west at the beginning of the ¬rst millennium bc. Their smelting sites
are associated with a pottery style, Urewe ware, whose derivatives later spread
widely through eastern and southern Africa where Bantu languages are now
spoken. In Rwanda there are indirect indications that both ironworking and
pottery were also associated with the cultivation of sorghum and millet and
the keeping of goats and (from at least the third century ad) cattle, showing
Impact of metals 35

that the Bantu had added to their forest agriculture a range of food-producing
activities suited to savanna life. Such a combination could have permitted pop-
ulation growth and might explain why Bantu-speakers came to prevail over
the Nilo-Saharan-speakers from whom they probably gained their new agri-
cultural skills and livestock. Pollen analysis suggests extensive deforestation of
the Lake Victoria region from the late ¬rst millennium bc, possibly in part
for agriculture and iron-smelting. Linguistic evidence suggests that between
about 500 bc and ad 500 agriculturalists colonised almost the entire region
surrounding the lake.
Yet this was only a small part of an expansion by which agriculture, iron-
working, livestock, and Bantu languages spread from the Great Lakes region
to nearly every corner of eastern and southern Africa. The earliest movement
may have been southwards into the Upper Zambezi Valley, where cattle remains
and pottery derived from Urewe ware have been found from about the second
century bc. Thence the culture expanded westwards into the savannas of
modern Angola and eastwards into modern Malawi, Zambia, and Zimbabwe
during the ¬rst centuries of the Christian era. In the latter region, it met other
Bantu-speakers whose ancestors had dispersed in a more easterly direction
from the Great Lakes region. Expanding from that area about 2,000 years
ago, they had spread eastwards at accelerating speed, marking their arrival
by a type of pottery, derived from the Urewe tradition, which extended across
modern Tanzania to the Indian Ocean coast, arriving there soon after the birth
of Christ. The makers of this pottery practised an iron-working technology with
similarities to that of the Great Lakes region and favoured especially the fertile
soils beneath East Africa™s mountain outcrops, avoiding the plains occupied by
stone-using pastoralists. While inland groups spread further southwards into
Central Africa, those who reached the Indian Ocean moved rapidly down the
coast of modern Mozambique, exploiting shell¬sh and other marine resources,
reaching modern Maputo by the second century ad and soon penetrating as far
south as Durban. By the late ¬rst millennium ad, Bantu-speakers had reached
the Great Kei River in South Africa, but that was their limit, for their sorghum
staple was a summer rainfall crop unsuited to the winter rains of the western
Cape and Namibia. They left that region to Khoisan peoples, some of whom
acquired cattle (perhaps from Bantu neighbours) and came to call themselves
This brief account can do little justice to the complexity of the process by
which the new culture spread through the southern half of Africa. This was
no simple mass migration by conquering, culturally superior Bantu. Often
different features of the new culture reached a region at different dates. Some
Khoisan forager-hunters seem to have found pottery the most useful of the
innovations and adopted it even before food production reached the region.
Others, like the Khoikhoi, adopted food production itself. Yet neither was
36 africans: the history of a continent

this merely a transmission of new cultural practices and languages from one
already established population to the next. The speed of diffusion down the
eastern coast to South Africa suggests a true population movement, probably
in small and uncoordinated bands, as does the predominance achieved by
Bantu languages and non-Khoisan genetic markers.12 At this early stage, Bantu-
speaking colonists were not farmers slowly expanding cultivation by nibbling at
the fringes of the bush; they were mobile pioneers, probably still heavily reliant
on foraging and hunting, who selected only the land best suited to their farming
technology, avoided arid plains in favour of better-watered environments, and
abandoned ¬elds ruthlessly once their virgin fertility was lost. The process
has been studied in detail near the Victoria Falls in modern Zambia. Pottery-
makers entered the area by the third century ad, bringing agriculture, cattle,
iron, and copper. There were perhaps no more than a thousand of them. They
selected microenvironments where they could utilise their skills and build their
thatched wattle-and-daub huts in compact villages, averaging perhaps ¬fty
metres across, a pattern generally adopted by Bantu frontiersmen in eastern
and southern Africa. If they possessed cattle, they penned them at the centre of
the village. When the surrounding ¬elds were exhausted, the pioneers simply
moved on to the next suitable microenvironment, with no suggestion at this
stage of returning to a village site after a period of fallow, much less of adapting
their modes of exploitation to changed circumstances. Not until the late ¬rst
millennium ad did they begin to return to former village sites after long periods
of fallow, indicating that the agricultural colonisation of eastern and southern
Africa was giving way to more settled communities.

Christianity and Islam

while bantu-speaking peoples were colonising southern
africa, the north was entering one of its greatest historical periods. Per-
haps only in pharaonic times had it been more central to human progress than
in the third and fourth centuries ad, when it was the intellectual spearhead
of Christianity, and again 800 years later, when it was the pivot of Islam and
a commercial network encompassing most of the Old World. This leadership,
already threatened, was destroyed during the fourteenth century by the demo-
graphic catastrophe of the Black Death, from which the region took 500 years
to recover. But in their time of greatness, North Africans adapted Christianity
and Islam to their own cultures and transmitted both religions to Black Africa,
where centuries of internal development had prepared social environments for
their reception and further adaptation.

christianity in north africa
Legend said that St Mark himself brought Christianity to Alexandria in ad 61.
In reality the church in Jerusalem probably sent missionaries to Alexandria™s
large Jewish community. The ¬rst ¬rm evidence of Christianity there comes
from an early second-century controversy between Jews who had and those
who had not accepted the new faith. Shortly afterwards Christianity expanded
beyond this Jewish nucleus. By ad 200 there was a Greek-speaking church
under a Bishop of Alexandria, with many Christians in Upper as well as Lower
Egypt. They saw Christ as a great teacher in the Greek manner; their ¬rst major
theologian, Origen (c. 185“253/4), believed that man should elevate himself
towards God through wisdom and asceticism. Once the ¬rst bishops outside
Alexandria took of¬ce early in the third century, Christianity spread among
Egyptians as well as Greeks. By 325 Egypt had ¬fty-one known bishoprics
and the Bible was widely available in the vernacular Coptic language (ancient
Egyptian written in Greek script). The chief leaders of popular Christianity
were monks: ¬rst individual hermits like St. Antony, who lived in the desert
from about 285 to 305, then disciplined communities pioneered in c. 321 by
Pachomius. Monasticism may have had models in ancient Egyptian priestly
asceticism, just as the Coptic Church™s elaborate charity inherited an ancient
38 africans: the history of a continent

tradition of famine relief. Both exempli¬ed the indigenisation of Christianity
at a time when Egypt™s old religion and culture were disintegrating. In 312
Constantine made Christianity the Roman Empire™s of¬cial religion. Later in
that century, the authorities persecuted traditional priests and either closed
their temples or converted them into churches or monasteries. By ad 400
perhaps 90 percent of Egyptians were Christians.
Further west, Christianity may have reached the Maghrib through Greek or
Roman rather than Jewish networks. The ¬rst ¬rm evidence of its existence is
the execution of twelve Christians at Carthage in ad 180 for refusing to sacri¬ce
in honour of the emperor. Such early Christians appear to have come from
every rank, age, and sex in urban society. Christianity offered fellowship across
social divisions in increasingly strati¬ed towns, just as it offered literal bodily
resurrection in a purposeless world and spiritual protection in a dangerous
world. In place of the multitudinous spiritual forces (daemones) and human
sorcerers whom pagans feared, Christianity pictured a dualistic con¬‚ict
between God, who protected the faithful, and the Devil, whose forces included
all aspects of paganism. Christianity did not threaten social rank and its teach-
ing generally passed from older to younger people, but it fed upon con¬‚icts of
generation and gender in complex, patriarchal households, as it would later
in tropical Africa. Among the ¬rst North Africans to be martyred, in the arena
at Carthage in 203, was a well-born, twenty-year-old wife and mother named

We walked up to the prisoner™s dock. All the others when questioned admitted
their guilt. Then, when it came my turn, my father appeared with my son,
dragged me from the step, and said: ˜Perform the sacri¬ce “ have pity on your
Hilarianus the governor . . . said to me: ˜Have pity on your father™s grey
head; have pity on your infant son. Offer the sacri¬ce for the welfare of the
˜I will not,™ I retorted.
˜Are you a Christian?™ said Hilarianus.
And I said: ˜Yes, I am.™ . . .
Then Hilarianus passed sentence on all of us: we were condemned to the
beasts, and we returned to prison in high spirits.1

Persecution was sporadic until 249“51, when the Emperor Decius, a soldier
who thought Christianity was corrupting the state, launched more thorough
repression. Martyrs were especially numerous in prosperous North Africa
because the Church was growing most quickly there, with at least 150 bishoprics
concentrated especially in the ancient colonial zone around Carthage but also
scattered generously further south in Byzacena and west in Numidia. During
Christianity and Islam 39

5. Christianity and Islam.
40 africans: the history of a continent

the next half-century, Christianity spread rapidly in the countryside, espe-
cially in Numidia, the inland plains of modern Algeria, which were then being
planted with olives. In this settler country of estates and Berber villages, Chris-
tianity was a religion of protest, infused with Berber traditions of statelessness
and honour, which forbade man or woman to betray loyalties from fear of
pain or death. When Diocletian launched his Great Persecution in 303 in a
desperate attempt to restore the old Roman order, church leaders were required
to surrender the scriptures for destruction. Those who complied, the traditores
(surrenderers), were subsequently denied recognition by zealots who created
a schismatic church under the leadership of Donatus, their candidate for the
bishopric of Carthage. Whereas the Catholics found followers especially among
urban notables and in the Romanised farming region near the coast, Donatist
leaders, although themselves mostly Latin-speaking urban intellectuals, won
support chie¬‚y among the non-Roman lower classes of the towns and, espe-
cially, the Berber cultivators and labourers of Numidia. Many Donatist churches
there had a local martyr™s body beneath the altar. The coincidence of religious
and agrarian con¬‚ict bred violent zealots, the Circumcellions (those ˜around
the shrines™), often perhaps seasonal labourers, who defended Donatist insti-
tutions and terrorised exploitative landlords and Catholic clergy. Donatism
predominated in the Maghrib throughout the fourth century. Its repression
was eventually organised by St. Augustine of Hippo (in eastern Algeria), who
condemned it as narrow, provincial, schismatic, and socially subversive. In
ad 411 Donatism became a criminal offence and the Catholic Church, now
increasingly integrated with the Roman state, intensi¬ed repression. The Van-
dal invasion of North Africa in 429 interrupted this, but persecution resumed
when Byzantine rule was established in 533. Donatism was gradually con¬ned to
its Numidian strongholds, but there it survived until the seventh-century Arab
To this day the Coptic Church of Egypt dates events not from the birth of
Christ but from ˜the era of the martyrs™. Yet it forgave its traditores and suffered
only brief schism. Its crisis came later, following the Council of Chalcedon of
451, which tried to shore up the disintegrating Roman Empire by declaring
the primacy of the bishops of Rome and Constantinople (the new imperial
capital) at the expense of Alexandria and by adopting a characterisation of
Christ “ that He had two distinct but inseparably united natures, divine and
human “ acceptable to Rome but anathema to Alexandria. Overt schism came
in 536, when the Emperor Justinian tried to impose a pro-Chalcedonian hierar-
chy upon Egyptians who now proclaimed the Monophysite (one-nature) faith.
Byzantine persecution of Monophysites prevented united Christian resistance
to the Muslim invasion of 639, which destroyed the pro-Chalcedonian hier-
archy but left Coptic Christians as protected tributaries concerned ever more
exclusively with survival.
Christianity and Islam 41

christianity in ethiopia and sudan
The Coptic Church had wide regional in¬‚uence. Its ¬rst engagement was in
Ethiopia. Following the collapse of Daamat between the ¬fth and third centuries
bc, several small successor states occupied the northern Ethiopian plateau. The
growth of Red Sea trade in Ptolemaic times enriched the region and linked it
to Mediterranean developments through its chief port at Adulis, famed for its
ivory. During the ¬rst century ad, at a time of unusually generous rainfall, a
kingdom emerged at Aksum. This went on to reunite the region, inheriting
much South Arabian culture and embellishing its capital with palatial stone
buildings, tall stone stelae marking royal graves, and a surrounding belt of
rural villas. Two centuries later the kingdom struck coins on Roman models.
Its seamen voyaged as far as Sri Lanka.
The introduction of Christianity to Aksum is traditionally attributed to
Frumentius, a young Christian trader kidnapped en route from Tyre to India.
He became tutor to the future King Ezana, who of¬cially adopted Christianity in
about 333, after Frumentius had been consecrated in Alexandria as Aksum™s ¬rst
bishop. This tradition oversimpli¬es a complex process, for Christianity was
only one of several religions (including Judaism) at Ezana™s court; more than a
century after his supposed conversion, a successor recorded the sacri¬ce of ¬fty
captives to Mahrem, local god of war. Ezana had probably sought to patronise all
religions, including Christianity, whose prominence on his coins suggests that
he displayed it especially, but not exclusively, to foreigners. Because Christianity
reached Aksum from Alexandria, the Ethiopian Church became Monophysite
and was headed by Coptic monks from Alexandria until the mid-twentieth
century. Moreover, because Christianity ¬rst in¬‚uenced the court, it became
a state religion, gradually extended among the people by priests and monks
with royal backing. Between the ¬fth and seventh centuries, the scriptures were
translated into Ge™ez (the Semitic lingua franca of Aksum, written in a script
derived from the South Arabian culture), Christianity and Aksumite authority
spread further southwards on the Ethiopian plateau, and pagan temples in
Aksum and Adulis became churches. But from the late sixth century, Aksum™s
prosperity declined, ¬rst perhaps because warfare between Byzantium and
Persia dislocated trade, then owing to Muslim expansion that destroyed Adulis,
and ¬nally because increasing reliance on agriculture coincided with declining
rainfall. Aksum struck its last coins in the early seventh century. The king who
died in 630 was buried not in the capital but further to the southwest, where
the merging of Aksumite and indigenous Cushitic cultures was to create the
historic church and kingdom of Ethiopia.
Christian origins in Nubia differed from those in Aksum, partly because
Nubia immediately adjoined Christian Egypt. After the collapse of Meroe dur-
ing the fourth century ad, Nubian-speaking rulers created three kingdoms in
42 africans: the history of a continent

the Nile Valley: Nobatia in the north with its capital at Faras, Makuria in the
centre with its headquarters at Old Dongola, and Alwa in the south based on
Soba (close to modern Khartoum). Symbols on pottery and other objects sug-
gest Christian in¬‚uence from Egypt by at least the ¬fth century, but the initiative
for systematic conversion came from the Byzantine court, where rival parties
sent both Catholic and Monophysite missionaries to Nubia. The Monophysite
reached Nobatia ¬rst, in 543, ˜and immediately with joy they yielded themselves
up,™ as the chronicler John of Ephesus recorded, ˜and utterly abjured the error
of their forefathers, and confessed the God of the Christians.™2 Evidence of vil-
lage church-building and adoption of Christian burial con¬rms this account,
although pagan temples survived in Nobatia for another two centuries. Alwa™s
rulers were also keen to link themselves to the larger world. When the mission-
ary Longinus arrived there from Constantinople in 580, ˜he spake unto the king
and to all his nobles the word of God, and they opened their understandings,
and listened with joy to what he said; and after a few days™ instruction, both
the king himself was baptized and all his nobles; and subsequently, in process
of time, his people also.™3
The Nubian kingdoms remained Christian for nearly a thousand years.
Nobatia and Alwa were Monophysite from the ¬rst and Makuria soon became
so. Nubian bishops drew their authority from Alexandria and the Church dated
events by the Coptic era of the martyrs. Yet Byzantium also exercised a pow-
erful in¬‚uence on elite culture. The beautiful murals in the cathedral at Faras,
excavated from the sand during the 1960s, began in Coptic style but gradually
changed to Byzantine, although they also displayed distinctive local features.
The liturgical language was Greek; only slowly were parts of the liturgy and
Bible translated into Nubian, written in the Coptic form of the Greek alphabet.
Church architecture suggests that the liturgical role of the laity diminished with
time. Kings were in priestly orders and bishops held state of¬ces in the Byzan-
tine manner. Some historians attribute the ultimate disappearance of Nubian
Christianity to a failure to adapt as fully to the local culture as did Ethiopian
Christianity, which was more isolated from external in¬‚uence. Nubian paint-
ings, for example, always depicted Christ and the saints with white faces in
contrast to Nubians, a distinction not drawn in Ethiopian art. Yet the different
fates of the two Churches owed more to different relationships with Islam.

islam in north africa
The expansion of Arab power and the Islamic religion following the Prophet
Muhammad™s death in ad 632 was the central process in world history for
the next 400 years. During that time Islam became the predominant faith
throughout North Africa and established footholds in both West and East
Africa. In doing so it not only tied the north permanently to the wider history
Christianity and Islam 43

of the Old World, but it began to reintegrate sub-Saharan Africa into that
history for the ¬rst time since the desiccation of the Sahara.
Some four thousand Muslims commanded by Amr ibn al-As invaded Egypt
in December 639. Within less than three years, they had conquered the Byzantine
Empire™s richest province. They were helped by deep antagonism between
Byzantine rulers and Monophysite subjects, who con¬ned their resistance to
defending their villages. But the Muslims™ chief strength was the disciplined
conviction that characterises the zealots of a new faith. ˜We have seen a people
who prefer death to life and humility to pride™, a later historian imagined the
Byzantines saying. ˜They sit in the dust, and they take their meals on horseback.
Their commander is one of themselves: there is no distinction of rank among
them. They have ¬xed hours of prayer at which all pray, ¬rst washing their hands
and feet, and they pray with reverence.™4 In 643 their momentum carried Amr
ibn al-As and his horsemen into modern Libya. Four years later they defeated
the main Byzantine army near Sufetula (Sbeitla) in modern Tunisia and gained
access to the fertile heartland of successive imperialisms in North Africa, more
rural then than in Roman times and somewhat depopulated by a great plague
in 542, but still rich in grain and olives. At this point, however, the conquest
faltered, owing to con¬‚ict over succession to the Caliphate. When expansion
resumed in 665, the main leader, Ukba ibn Na¬, bypassed North Africa™s coastal
cities and in c. 670 founded Kairwan in the Tunisian hinterland as the capital of
a new Muslim province of Ifriqiya (Africa). Then he drove westwards through
the inland plains until he rode his horse into the Atlantic, declaring that he
had fought his way to the end of the world in God™s name. On his way back,
however, his army was annihilated by a Berber coalition led by Kusayla, a chief
of the Tlemcen region, who went on to capture Kairwan. This opened a new
period in the conquest. For four centuries the Berber peoples of the inland
plains and mountains had been regaining strength from Romans, Vandals, and
Byzantines. Now they mounted the stiffest resistance the Arabs met during
their conquests, restricting Arab power to the colonial heartland of Ifriqiya.
When a Muslim army ¬nally conquered western Algeria and Morocco early
in the eighth century, it was a largely Berber army, as was the expedition that
conquered Spain in 711 “12. Islamic predominance in Berber territory meant
Berber predominance in Islam.
In North Africa, alone in the continent, Islamisation drew its initial impulse
from conquest, but the victors seldom compelled the conquered to accept
their faith. Their concern was to establish an Islamic social order, in the con-
¬dence that individuals would gradually conform to it. In Egypt, therefore,
they offered Christians either client status as Muslim converts or toleration as
protected tributaries (dhimmi) in return for land and poll taxes, as was ini-
tially preferred by most Copts, on whom the Arabs at ¬rst relied to administer
Egypt™s complex society. By the eighth century, however, Arab immigrants had
44 africans: the history of a continent

increased and Christians were gradually excluded from of¬ce, as one of several
social and economic pressures to adopt Islam. By 717“20 so many Copts were
becoming Muslims to escape the heavy taxes needed to ¬nance Arab wars that
converts were declared still liable to the land tax. At the same period, of¬cial
business ¬nally came to be conducted in Arabic. The Coptic language sur-
vived temporarily in the countryside but eventually became purely a liturgical
language, while the Coptic Church itself lived on its past as a religion of sur-
vival, periodically harried by the authorities and unable to rival the conviction,
authority, and modernity of Islam. By the fourteenth century probably fewer
than one-tenth of Egyptians were Christians.
The Umayyad Caliphate, which lasted until 750, was effectively an Arab king-
dom led by the Meccan aristocracy. Egypt in particular was dominated by an
Arab garrison. When the Abbasids gained power in 750, however, they relied on
non-Arab nationalities and moved their capital eastwards to Baghdad, thereby
encouraging North African autonomy. By the late ninth century, power in
Egypt lay with Turkish military governors and their multiethnic mercenaries,
who had supplanted the Arab horsemen of the heroic age. Further west, in
the Maghrib, separatist tendencies were even stronger. The Berbers retained
their language and, according to the great Tunisian historian Ibn Khaldun,
apostasised a dozen times during their ¬rst seventy years of Islam. Certainly
they displayed the same egalitarianism, puritanism, and particularism as had
inspired the Donatist schism. At least one Christian community survived for
a thousand years. A group in the Atlantic Plains of Morocco claimed to pos-
sess a Koran in the Berber language and maintained its heterodoxy until the
eleventh century. But the chief vehicle of Berber aspirations was Kharijism, an
extreme wing of Islam born in 657 during the civil war that created the Umayyad
Caliphate. It taught the absolute equality of Muslims, the right of any worthy
Muslim to be elected Imam of the whole community, and consequently the
duty to reject the existing, illegitimate Caliphate. Kharijites escaping persecu-
tion in the east left for the Maghrib in c. 714, winning more support among
Berbers than anywhere else, especially, it appears, among former Christians. In
740 they launched a revolt in Tangier, led by a former water-carrier, sparking
turmoil that eventually overthrew the Umayyads. When the Abbasids proved
equally repressive, Kharijites formed several zealous communities in the North
African hinterland, especially at Tahert in western Algeria, which from 761 “2
became the core of a Kharijite state. In 789“90 a refugee descendant of the
Prophet, Idris, created a kingdom based in Fes that became the chief vehicle
of Islamisation in northern Morocco. Throughout these disturbances, the cen-
tre of Abbasid power in the Maghrib and almost the only area of extensive
Arab settlement remained Ifriqiya, but there, in 800, an Arab governor estab-
lished the hereditary Aghlabid dynasty. Thereafter the Maghrib was effectively
Christianity and Islam 45

During the following ¬ve centuries, North Africa bred several of Islam™s most
creative dynasties. The ¬rst, the Fatimids, were a Shia family claiming descent
from the Prophet through his daughter Fatima. They came to power in Kair-
wan in 910 on the back of a Berber revolt, incorporated the Aghlabid kingdom,
temporarily overran much of Morocco in 958“9, and went on in 969 to take
Egypt peacefully from its Turkish military rulers, completing the Berber recon-
quest of North Africa and building Cairo as a capital ¬t for a Fatimid Caliph.
Despite their heterodox origins, the Fatimids had no radical programme. They
had gained power in Ifriqiya at a time of unprecedented prosperity once the
Arab conquest was stabilised. The traveller al-Yakubi (d. 891) wondered at Kair-
wan™s wealth, with its ¬‚ourishing textile industry, growing gold imports from
West Africa, surrounding market gardens, and supplies of fruit from the coast,
grain from the northern plains, olives from the Sahel, and dates from Saharan
oases. Townsmen owned great estates carved out by victorious ancestors and
worked by the slaves for which the region was famed. Initially the slaves were
Berbers captured during the conquest; thereafter, they were white and black
slaves imported from Europe and tropical Africa. Cultivation of sorghum and
hard wheat expanded southwards, famine was virtually unknown during the
tenth century, and population almost certainly increased. Mediterranean trade
was largely in Muslim hands, thanks to the Fatimid ¬‚eet, which sacked Genoa
in 934“5. When this wealth enabled the Fatimid army of Slav mercenaries and
Berber auxiliaries to capture Egypt, prosperity shifted to the new capital. The
records recovered from the Cairo Geniza “ where Jews deposited unwanted
papers to avoid destroying any bearing the name of God “ show that immigrant
Fatimids were followed by merchants from the Maghrib seeking their fortunes
in what now became the centre of the Islamic world. ˜It was the heyday of
the bourgeoisie™, their historian has written,5 a commercial world dominated
by family ¬rms of many faiths, operating through partnerships and agencies
spread throughout the Mediterranean, pro¬ting from a freedom of movement
and religious toleration that caused Jewish merchants to call Fatimid Egypt ˜the
land of life™. This bourgeoisie dominated a strati¬ed but mobile society with an
exceptional level of craft specialisation, many female slaves in domestic service,
and numerous paupers. Cairo despised and exploited the countryside, where
Arab rule introduced sugar, cotton, and rice, encouraged multicropping, and “
after an initial hiatus during the conquest “ probably stimulated population
growth, which by the fourteenth century was regaining Ptolemaic levels. In
the meantime, however, exploitation of the countryside may have contributed
to severe famine in 1062“73, which was the ¬rst symptom of Fatimid decline.
Twenty years later their dominions were con¬ned to Egypt. In 1171 they were
overthrown by their Kurdish Vizier, the great Saladin.
In Ifriqiya the shift of power and prosperity to Egypt led the Fatimids™ Berber
lieutenants, the Zirids, to renounce their allegiance in 1048. Tradition alleged
46 africans: the history of a continent

(without foundation) that the Fatimids replied by encouraging the Banu Hilal
and other nomadic Arab tribes who had entered Egypt to move on westwards
into Ifriqiya. The Hilal, wrote Ibn Khaldun, ˜gained power over the country
and ruined it.™ In 1057 they sacked Kairwan. The Zirids shifted their capital and
their attention to the seaboard, losing control of the interior. Transport was
disrupted and gold caravans dispersed to reach the coast at several points, espe-
cially further to the west in Morocco. Berber pastoralists retreated westwards.
Cultivators withdrew into mountain strongholds. A huge swathe of former
Berber plains was permanently Arabised, the nomads™ dialect becoming its
vernacular Arabic. The effects of this ˜Hilalian invasion™ have no doubt been
exaggerated. It was more an in¬ltration than an invasion. North Africa™s rain-
fall and cultivated area had probably been contracting since the ¬fth century
ad and would reach their nadir in the fourteenth century.6 Loss of naval con-
trol of the Mediterranean to the Byzantines during the tenth century deprived
Ifriqiya of its northern slave supply, which further damaged the rural economy
and contributed to repeated famines after 1004. These and the Zirids™ political
weakness brought commercial decay to Kairwan even before the Banu Hilal
sacked it. Their depredations were consequences as well as causes of a collapse
from which Ifriqiya never recovered. By the 1090s the former granary of Rome
was becoming dependent on imported Sicilian wheat.
Initially the chief bene¬ciary was the previously fragmented western
Maghrib, where nomad ambitions coincided with economic diversi¬cation
and the full internalisation of Islam among its Berber converts to produce
a period of great splendour. It began with the Almoravid movement, which
originated among the nomadic Sanhaja Berbers of southern Morocco and the
western Sahara, long overshadowed by their more settled Zanata rivals to the
north and gradually losing their long-standing control of trade in the west-
ern desert. The Sanhaja were largely oral Muslims until the eleventh century,
when their leaders sought further instruction from rigorous teachers anxious
to root out the Shiite and Kharijite legacies so powerful in the Maghrib. Abdal-
lah ibn Yasin began to teach among the Sanhaja in c. 1039, gathered a following
of zealots and tribesmen, and launched them against Zanata supremacy. In
1070 they created a new capital at Marrakesh. By 1083 they had conquered the
whole Maghrib west of Algiers. Three years later they entered Muslim Spain
to organise its resistance to Christian expansion. This military supremacy was
backed by capturing much of the West African gold trade and by developing the
grainlands of Morocco™s Atlantic Plains. Prosperity enabled the Almoravids to
introduce into Morocco the elegant Islamic culture of southern Spain, which
is still resplendent in the architecture of Marrakesh. This attracted puritan
criticism, while others resented the regime™s ruthlessness in enforcing ortho-
doxy and its reliance on the tribes that had initially supported Abdallah ibn
Christianity and Islam 47

These criticisms animated the Almohad (Unitarian) movement, which was
to supplant the Almoravids. It arose not among nomads but among their
long-standing enemies, the Berber agriculturalists of the Atlas Mountains. Its
leader, Muhammad ibn Tumart, was born there in about 1080 but educated
in Baghdad, where he learned to criticise the Almoravids™ legalistic rigour
and to admire instead the personal spirituality then entering Islam through
the mystics known as su¬s. Returning to his mountain home, he was declared
Mahdi by his fellow Masmuda tribesmen and in 1128 led them in a jihad against
the dominant Sanhaja nomads and all corruptions of the faith. They took
Marrakesh in 1147 and Ifriqiya in 1160, checking the expansion of the Banu
Hilal and uniting the Maghrib for the ¬rst time under a single Berber regime.
Almohad rule was rigorously Islamic, Christianity was virtually eradicated
from the Maghrib, and Jews found Almohads exceptionally intolerant. But they
were less legalistic than the Almoravids, enabling su¬ brotherhoods to establish
themselves during the late twelfth century throughout the region, where they
were to become the core of popular Islam. The decline of the overextended
Almohad empire began with its defeat by Christian forces in Spain in 1212
and was compounded by its inability to control nomadic tribes, notably the
Arab pastoralists whom the regime had deported from Ifriqiya to the Atlantic
Plains, thereby initiating further Arabisation of former Berber territory. In
1269 a Zanata tribe already dominant in northern Morocco, the Banu Marin,
captured Marrakesh, transferred the capital to Fes, and ruled Morocco for two
centuries as the Marinid dynasty.
Marinid rule witnessed general decline in the Maghrib. An Almohad succes-
sor dynasty, the Hafsids, ruled Ifriqiya until the Ottomans conquered it in the
sixteenth century, while Zayyanids, another Zanata dynasty based in Tlemcen,
exercised such central authority as existed in western Algeria. Reliant on mer-
cenary troops rather than their subjects and based in northern cities dependent
on maritime commerce controlled by Europeans, these regimes grew away from
a countryside increasingly dominated by Arab pastoral tribes and su¬ broth-
erhoods. Most important of all, the demographic growth that had underlain
the Fatimid and Almoravid regimes was checked during the thirteenth century
and dramatically reversed in 1348 when the great plague known in Europe as
the Black Death reached the Maghrib from Sicily.
In Egypt, likewise, the Black Death ended nearly four centuries of prosperity
and power unequalled since the New Kingdom. The Fatimids had initiated this
prosperity. Saladin revitalised the state after he seized power in 1171, making
Egypt the champion of Islam against Crusaders and Mongols. In 1250 his Ayyu-
bid dynasty was overthrown by its Mamluk troops. These were slaves purchased
as children from the horsemen of the Eurasian steppe, rigorously trained in
Islam and the military skills of mounted archers, and then freed to become
professional soldiers loyal to one another and to their former masters, forming
48 africans: the history of a continent

a caste so exclusive that even their sons were barred from it. This system was
designed to combine the virtues of nomadic valour and civilised organisation.
Mamluk generals ruled Egypt until 1517. They reorganised its land into ¬efs
from which of¬cers drew tribute to support themselves and their men. They
extended irrigation and cultivation, raised medical skill to new levels, and
were Egypt™s greatest builders since the Ptolemies. The expenditure of their
great households made early fourteenth-century Cairo the ˜metropolis of the
universe, garden of the world, swarming core of the human species™,7 as Ibn
Khaldun later described it.
Yet this prosperity was already threatened. By the early fourteenth century,
the international trading system stretching from Flanders to China, with Cairo
at its core, was breaking down as the Mongol Empire disintegrated in Central
Asia, leaving Egypt as a channel through which oriental goods passed to the
increasingly dominant economies of Europe. Christians had enjoyed naval
control of the Mediterranean since the tenth century. Italian trading cities such
as Genoa and Pisa made commercial treaties with North African rulers from the
1130s. Portuguese and Aragonese (Catalan) mercenaries served the same rulers
from the 1220s. Christian friars of the Dominican Order established a house at
Tunis in 1250. In 1284“6 Aragon made two islands off the Tunisian coast Europe™s
¬rst African colonies since Vandal times. By then European traders regularly
frequented North African cities, siphoning away the gold trade and damaging
Cairo™s textile industry by their competition. Europe was outpacing the Islamic
world in technology, business organisation, and agricultural production on
virgin land no longer available in North Africa.
Relative decline became crisis when the Black Death reached Egypt along
the trade routes from the Asiatic steppe. Egypt had suffered sporadic plague
since the last great epidemic in the sixth century, but that had been bubonic
plague, transmitted by ¬‚eas from rats to humans, whereas the Black Death was
also the more infectious pneumonic plague, passed aerially from one person
to another, making death even more common, rapid, horrible, and certain.
Nobody understood the means of transmission and no effective countermea-
sures were taken; all that religious leaders could counsel was prayer, charity,
and digni¬ed resignation. In eighteenth months, the epidemic killed perhaps
one-quarter or one-third of Egypt™s population.8 For urban working people,
the consequence was higher wages, but the immobility of irrigated agriculture
enabled Mamluks to respond to rural depopulation by trying to squeeze an
undiminished revenue from fewer cultivators, although unsuccessfully in the
long term, for shortly after 1517 Egypt™s rulers collected less than one-¬fth of
the land tax paid in 1315. Resilient agriculture and control of trade between Asia
and Europe enabled Egypt to survive the Black Death better than the rest of
North Africa and the Middle East, but economic decay was nevertheless grave
and coincided with recurrent warfare between Mamluk groups and the decline
Christianity and Islam 49

of the whole military class as ¬rearms rendered their skills obsolete. Most dev-
astating of all was that pneumonic plague remained recurrent after the Black
Death, unlike in Europe. During the next 160 years Egypt suffered twenty-eight
plague outbreaks, which were probably more destructive cumulatively than
the Black Death itself. They continued until the early nineteenth century. The
Maghrib suffered equally, Tunisia enduring ¬ve plague epidemics during the
seventeenth century alone. This demographic catastrophe ended North Africa™s
time of greatness and moved Ibn Khaldun “ who lost both parents during the
Black Death “ to preserve the memory of a vanished world:

in the middle of the eighth [fourteenth] century, civilization both in the East
and the West was visited by a destructive plague which devastated nations and
caused populations to vanish. . . . Civilization decreased with the decrease of
mankind. Cities and buildings were laid waste, roads and way signs were
obliterated, settlements and mansions became empty, dynasties and tribes
grew weak. The entire inhabited world changed. . . . Therefore, there is need
at this time that someone should systematically set down the situation of the
world among all regions and races, as well as the customs and sectarian beliefs
that have changed for their adherents.9

trade and islam in west africa
The Arab conquest of North Africa led to the transmission of Islam across the
Sahara to the West African savanna. Agriculture and iron-using existed here
before the birth of Christ, but the ¬rst Muslims who knelt their camels on the
northern fringe of the savanna also found towns and a regional trading system
that appear to have been predominantly local inventions. Indeed, the chief
reason why trans-Saharan trade grew so swiftly in the early Islamic period was
probably that it linked two ¬‚ourishing regional economies.
The best evidence for this comes from archaeological excavations at Old
Jenne, a site in modern Mali on the southern edge of the internal Niger delta,
where ¬‚oodplain agriculture met transport routes northwards to the savanna
and southwards to the forest. From about 2000 bc, declining rainfall drew
¬shermen from the southern Sahara and farmers from settlements like Dhar
Tichitt towards the increasingly habitable Niger Valley. A settlement with iron-
working existed at Old Jenne by the third century bc and a substantial town
with crowded cemeteries from ad 400. The town was built in sun-dried mud
in a style whose simplicity may partly explain such early urbanisation, as in
the ancient Near East. By the ninth century, Old Jenne occupied forty-one
hectares, was surrounded by a two-kilometre wall, and had some sixty-¬ve
other settlements within a four-kilometre radius. Similar clusters existed else-
where in the valley, especially around Dia and Timbuktu.10 These settlements
50 africans: the history of a continent

show occupational specialisation, but their graves reveal little social differen-
tiation and there is no evidence of powerful rulers or major public buildings,
suggesting that the region achieved social complexity without state formation
in a manner later repeated by some other African stateless societies. Townsmen
owned copper objects made no closer than the southern Sahara, although they
apparently did not weave cloth, a skill probably introduced by Muslim traders.
The earliest archaeological ¬nds of gold in sub-Saharan Africa also come from
seventh- or eighth-century Jenne. In the mid ¬rst millennium ad, however,
its inhabitants possessed almost nothing of Mediterranean origin except glass
beads. In other words, Old Jenne was part of an extensive West African trading
system with some limited external contacts. Glass beads in warrior graves from
the ¬rst to the seventh centuries ad in northern Burkina Faso similarly suggest
a low level of exchange across the desert.
This pre-Islamic commerce may help to explain one of the mysteries of
African history: the discovery at Igbo-Ukwu, in southeastern Nigeria, of the
grave-goods buried with a ninth-century ruler or ritual leader, including bronze
artefacts made from local metals, in African style and showing a superb tech-
nical skill that was both distinctive and arguably unequalled elsewhere in the
world at the time. Their symbolism, especially the use of animal motifs, shows
remarkable continuity with that employed by the Igbo people of the area a
thousand years later. But Igbo-Ukwu also shows that by the ninth century West
Africa was less isolated from the outside world, for its grave-goods included
over 100,000 glass beads, some probably from Egypt or even India.
By contrast, the two Hellenistic glass beads found in the late pre-Christian
deposits at Old Jenne suggest that trade across the Sahara can then have been on
only the smallest scale. The Garamantes of the Fezzan had exported ivory and a
few slaves northwards in Roman times. That was an especially arid period in the
Sahara and West African savanna, but by about ad 300 rainfall was increasing
and Berbers were exchanging their horses for the camels that enabled them to
open the desert to trade. Its growth is suggested by Ukba ibn Naªs actions as
he led his men westwards into the Maghrib. In 666“7 he broke off southwards
to reconnoitre the road to Fezzan, probably the main source of slaves. Sixteen
years later, he made a similar excursion into southern Morocco. A subsequent
expedition there in the 730s returned with gold, probably from the Bambuk
gold¬eld on the upper Senegal. This was the period when West African gold
was ¬rst used in Mediterranean coins.
On the slave route, a new trading base was established in the eighth century
in the eastern Fezzan at Zawila, a Kharijite Berber settlement that became the
main supplier of black slaves to Ifriqiya, Egypt, and the Middle East. It enjoyed
a relatively easy desert crossing to the northern environs of Lake Chad. Here the
main suppliers of slaves were the Zaghawa, a largely pastoral people, mentioned
by an Arab author before 728, who controlled a loose confederation known as
Christianity and Islam 51

Kanem, possibly created as early as the late sixth century. Lacking gold, Kanem
and its successor, Borno, were to be the main suppliers of slaves from the West
African savanna to the Islamic world for a thousand years, buying in return
horses to facilitate further slave-raiding. Many slaves probably went to the
Aghlabids who ruled Ifriqiya in the ninth century and relied on black slave
soldiers, as did their Fatimid and Zirid successors. Kanem was ¬rst mentioned
in 872 by al-Yakubi as one of West Africa™s three main savanna kingdoms, along
with Ghana and Gao to the west.
Ghana, centred in the east of modern Mauritania, was a kingdom of the
Soninke people, black speakers of a Niger-Congo language. It was ¬rst men-
tioned in an Arabic source of 788“93 emanating from the Berber Kharijite
community at Tahert, who pioneered trans-Saharan trade with the western
savanna just as their counterparts in Zawila developed the trade with Kanem.
A trade route ran westwards from Tahert to Sijilmasa in southern Morocco
(founded in 757“8) and then southwards to Awdaghust and Ghana, following
the easiest desert crossing parallel to the Atlantic coast. This and its strategic
position to the northeast of the Bambuk gold¬eld gave Ghana its importance,
although it sought to control gold trade rather than gold production. Ghana™s
royal town, not yet discovered, was said to be ten kilometres from a traders™
town thought to have been identi¬ed at Koumbi Saleh, where excavation has
shown urban occupation and northern trade from the ninth to the ¬fteenth
century, although on a site that had already been occupied in the mid ¬rst
millennium ad. Writing in Spain in 1067“8 from travellers™ testimonies, the
geographer al-Bakri described Ghana™s court at its apogee:

The king has a palace and a number of domed dwellings all surrounded with
an enclosure like a city wall. . . . The king adorns himself like a woman round
his neck and on his forearms, and he puts on a high cap decorated with
gold and wrapped in a turban of ¬ne cotton. He sits in audience or to hear
grievances against of¬cials in a domed pavilion around which stand ten horses
covered with gold-embroidered materials. Behind the king stand ten pages
holding shields and swords decorated with gold, and on his right are the sons
of the [vassal] kings of his country wearing splendid garments and their hair
plaited with gold.11

The king was not a Muslim, although many of his ministers were.
At the time al-Bakri was writing, Ghana was challenged from the west by the
Takrur kingdom on the Senegal, which siphoned away Bambuk™s gold to feed the
newly created Almoravid empire. But Ghana™s chief rival lay to the east at Gao, a
town ¬rst mentioned in the early ninth century and situated on the River Niger
where modern Mali and Niger meet. Gao was probably a city of the Songhay
people. Its main settlement, containing the royal capital, was supplemented by a
commercial quarter supplying the caravans that travelled northwards through
52 africans: the history of a continent

Tadmekka and Wargla to Tahert and the North African coast. Caravans seldom
undertook this two-thousand-kilometre crossing as a single journey. Rather,
coastal merchants traded their cloth and copper southwards to an entrepˆ t like
Tahert on the northern desert edge. There the trade was taken up by men, chie¬‚y
Berbers, who lived in the desert and transported goods across it along a line of
oases, gathering their produce of dates, copper, and especially salt “ so highly
valued at Gao that it was used as currency “ until they reached either another
entrepˆ t on the southern desert fringe, such as Tadmekka, or pushed further
south to an African town like Gao. A merchant sought partners or agents at
each stage of the journey, perhaps men of the same community “ Kharijites
were ideally organised for this “ or even kinsmen. In the thirteenth century, two
brothers of the Maqqari family lived at Tlemcen (close to the coast of western
Algeria), another at Sijilmasa, and two at Walata (on the southern desert edge),
cooperating in the family business and investing in wells along the route.
Exports of gold from the western savanna appear to have increased steadily.
In the eighth century, the only gold mints in North Africa were at Kairwan and
Fustat (in Egypt). But gold coins were the mark of a Caliph. When the Fatimids
in Ifriqiya, the Umayyads in Spain, and then the Almoravids and Almohads in
Morocco aspired to that status, all began to coin gold. In the eleventh century,
the Almoravids alone had twenty-one gold mints in Spain and the Maghrib
and the trade attracted southern European merchants to establish themselves
in North African coastal cities. Initially they exported Islamic gold coins to
supplement Europe™s silver currency. Then Genoa and Florence in 1252, Venice
in 1284, and northern European states in the earlier fourteenth century began
to coin gold, sparking a late medieval gold rush. Black slaves also appeared in
southern European markets during the fourteenth century.
The expansion of the gold trade contributed to a shift of power within West
Africa away from Ghana, whose location on the desert edge probably also
suffered when a period of desiccation began after about 1100. By then a new
gold¬eld ¬‚ourished at Bure on the headwaters of the Niger among Mande-
speaking people, who possessed a number of small chiefdoms. When Soninke
groups from a successor state to Ghana sought to dominate them, a hunter and
warrior named Sunjata Keita led Mande resistance and created the kingdom
of Mali during the ¬rst half of the thirteenth century. Its capital was close
to the northeastern edge of the Bure gold¬eld. Its suzerainty came to extend
nearly 2,000 kilometres from the Atlantic coast to the middle Niger, taking four
months to cross, according to a long-term resident. Yet Mali was not only a
larger and more important state than Ghana. It was centred not at the desert
edge but in the agricultural lands of the Upper Niger Valley. It marked a further
stage in West Africa™s reintegration with the Old World.
Early fourteenth-century Mali was of¬cially an Islamic state and was so
recognised in the Islamic world, where its rulers participated conspicuously in
Christianity and Islam 53

the Pilgrimage. In 1352“3 the great traveller Ibn Battuta admired its people™s
˜assiduity in prayer and their persistence in performing it in congregation and
beating their children to make them perform it™, but he was less impressed by the
coexistence of such non-Islamic practices as masked dancing, public recitation
of pagan traditions, self-abasement before the king, eating of unclean foods,
and scanty female clothing.12 Because Islam was not only a religion but a social
order, Africans necessarily adopted it only gradually. Whereas conquest had
created the conditions for this in North Africa, in the west the initial agency was
trade, especially by Berber Kharijites. There was little lasting Kharijite in¬‚uence
in West Africa, except in mosque architecture, but most desert-trading peoples
were probably Muslims by the tenth century and traders were probably also
among the ¬rst to accept the new religion further south, for they had the most
contact with foreign Muslims, pro¬ted by joining an international community,
and were little involved in the agricultural rituals central to indigenous religions.
Cultivators, whose circumstances were exactly the opposite, probably resisted
Islam most strongly. Rulers, concerned to preserve political unity, generally
patronised all their subjects™ religious activities in an eclectic manner. That
was clearly the case in Mali and also in eleventh-century Gao, whose ruler
had been the ¬rst in tropical Africa to accept Islam, sometime before ad 1000,
followed by Takrur (before 1040) and Kanem (in c. 1067), while Ghana appears
to have adopted orthodox Sunni Islam under Almoravid pressure during the
1070s. The eleventh century was a breakthrough period for Islam, as it was
on the East African coast, although the extent of conversion varied greatly. In
Ghana and Gao, Islam seems long to have been con¬ned to traders and the
court, but in Takrur and Kanem it spread more quickly to the common people
and aroused con¬‚ict between Islamic teachers and the practitioners of magic
(closely associated with ironworking) who had previously served the throne.

trade and islam in east africa
Whereas Islam reached West Africa across the world™s harshest desert, it trav-
elled to East Africa along the easily navigated trade routes of the Indian Ocean.
A mariner™s guide shows that during the ¬rst century after Christ traders from
southern Arabia and the Red Sea penetrated down the East African coast to
˜Rhapta™, somewhere in modern Kenya or Tanzania, where the main export was
ivory. Iranian pottery of the ¬fth to seventh centuries not only appears at sites
on the coast from the Horn of Africa to Chibuene in southern Mozambique,
but it has been found some sixty kilometres inland of Bagamoyo in Tanzania,
suggesting that the Indian Ocean trade was already linked to a regional com-
merce comparable on a smaller scale to that of the middle Niger. This doubtless
underlay the ¬rst appearance of Islam in East Africa at Shanga, a settlement
in the Lamu Archipelago off the northern coast of Kenya. Excavation here has
54 africans: the history of a continent

revealed what was probably a wooden mosque, accommodating only about
ten worshippers, roughly aligned towards Mecca and associated with eighth-
century pottery and radiocarbon dates. This was the ¬rst of nine mosques of
gradually increasing size (the last three in stone) erected on the site over three
centuries. Whether the builders were local people or immigrants is unknown,
but their ¬rst mosque was at the centre of an agricultural settlement where
most pottery was identical to that used by nearby African communities and
everywhere on the coast during the following centuries. Shanga also imported
small quantities of Iranian pottery and Chinese stoneware, the latter probably
coming via the Persian Gulf. From the ninth century, Shanga used and probably
produced silver coins.13
Shanga suggests the establishment within an eighth-century African com-
munity of a small nucleus of Muslims “ indigenous or alien “ who gradually
converted their neighbours. On neighbouring Manda Island, by contrast, the
small town built in the ninth century appears to have been the work of alien
settlers, possibly from Siraf on the Persian Gulf, who employed Middle Eastern
building styles, supplemented the local coral stone with burnt bricks imported
from Arabia, and were exceptionally well supplied with foreign pottery. From
the beginning, therefore, the East African coastal culture showed its endur-
ing tension between indigenous and imported elements. Other settlements of
the ninth and tenth centuries extended from Gezira (south of Mogadishu) to
Chibuene. Evidence of trade with the interior is strongest in the south, where
a fragment of imported glass has been found at a seventh-century site near
Victoria Falls and imported beads of the next two centuries have appeared
in southern Zimbabwe, northern Botswana, and eastern Transvaal. The chief
export, as al-Masudi found on the coast in 916, was ivory, which went via the
Persian Gulf to India and China. Indications of ivory exports of that period
have been found in the Limpopo Valley. Mangrove poles were another impor-
tant commodity for the treeless Persian Gulf. Slave exports from the coast are
¬rst mentioned in the tenth century. Al-Masudi reported that ˜The Zanj have
an elegant language and men who preach in it™,14 probably referring to Swahili,
the one among a cluster of Bantu languages spoken on the Kenya coast that
was probably used in overseas trade and carried southwards with the trade to
become the coastal lingua franca.
From about ad 1000, the Islamisation and commercial development of the
coast accelerated. At least eight coastal settlements built stone mosques dur-
ing the eleventh and early twelfth centuries. This expansion may have resulted
chie¬‚y from the Islamic world™s growing prosperity and lust for gold, which al-
Masudi ¬rst mentioned in 916 as an export from ˜Sofala™, meaning the Mozam-
bique coast, which acted as the outlet for gold produced by ancestors of the
Shona peoples of modern Zimbabwe, where archaeological evidence con¬rms
the beginnings of mining at that period. Yet the trading system extended beyond
Christianity and Islam 55

even the Islamic world. Between 1050 and 1150, China™s imports of African pro-
duce increased tenfold.
The most striking evidence of eleventh-century commercial expansion was
the foundation of a Muslim dynasty at Kilwa on the southern Tanzanian
coast, hitherto a ¬shing village. Coins bearing the inscription of ˜The majestic
Sultan Ali bin al-Hasan™, remembered in local tradition as the founder of Kilwa,
have been discovered in contexts suggesting a date around 1070. They were in
the tradition of Shanga and the new dynasty may have come from the Lamu
Archipelago. It was overthrown two centuries later by the Mahdali, possibly
Yemeni settlers claiming descent from the Prophet. Their rule brought Kilwa
to its greatest prosperity in the early fourteenth century, when their gover-
nor on the Sofala coast controlled the gold trade and their coins penetrated
to Great Zimbabwe. The Mahdali doubled the size of Kilwa™s Great Mosque,
built a magni¬cent palace, caravanserai, and slave barracoon known as Husuni
Kubwa, and won a reputation for conspicuous largess in a culture whose mate-
rialism coexisted with a piety noted by Ibn Battuta when he visited the city
in 1331. Yet he remembered Kilwa as a town of wood and thatch, for around a
core of stone houses were the simpler huts of the ˜Zanj of very black complex-
ion™ who comprised most of its estimated ten thousand to twenty thousand
people. Some were slaves, for Ibn Battuta noted that the Sultan of Kilwa ˜fre-
quently makes raids into the Zanj country™.15 The ruler was probably of mixed
race and, like the contemporary Sultan of Mogadishu, knew Arabic but spoke
Swahili, which was then still largely free of Arabic loan-words. Kilwa™s ruler
made pilgrimage to Mecca in 1410“11 and Muslims were probably the great
majority of Kilwa™s foreign visitors, for the ¬rst ¬‚eet of Chinese ˜treasure ships™
did not reach East Africa until 1417“19, only shortly before the Ming govern-
ment abandoned overseas adventures, while trade with India remained only
a background factor until the ¬fteenth century, when Gujarat™s growing pros-
perity encouraged its merchants to bring their cloth and copper to East African
ports. By then Kilwa was in decline. Its rulers abandoned Husuni Kubwa in the
later fourteenth century. The reasons are unknown. One could have been the
Black Death, but local traditions do not mention it and other coastal towns
prospered until Portuguese seamen reached East Africa in 1498.

islam in sudan
The Arabs had barely conquered Egypt when their forces entered Christian
Nubia in 641 and met ¬erce resistance from its famous bowmen. A further
costly invasion ten years later deterred the Arabs from again attacking ˜those
people whose booty is meagre, and whose spite is great™. Instead they made
a truce, the baqt of 652, with the kingdom of Makuria, which undertook to
deliver 360 slaves a year in return for Egyptian products and agreement to
56 africans: the history of a continent

respect each other™s traders. For the next ¬ve hundred years slaves, doubtless
acquired from the pagan south, were Nubia™s chief export. Arabs settled in the
Christian kingdoms as traders, miners of gold and precious stones, and, from
the ninth century, pastoralists. Egypt™s Fatimid rulers from 969 to 1170 relied
on black slave soldiers and their rule coincided with the apogee of Christian
Nubia. It was a time of high Nile ¬‚oods when the southern kingdom of Alwa
had ˜an uninterrupted chain of villages and a continuous strip of cultivated
This prosperity began to crumble when Saladin ousted the Fatimids in 1171
and slaughtered their African slave army, undermining the mutual advantage
linking Egypt and Nubia. Nubian forces raided southern Egypt. Saladin replied
by attacking northern Nubia. When Arab pastoral tribes in Egypt rebelled in
1253, his successors drove the dissidents into Nubia. Islamic communities in
the Christian kingdoms had been expanding slowly for several centuries; now
they were swollen by turbulent nomads at a time when river levels were falling,
the Christian ruling families were divided, and Nubian society as a whole
was increasingly militarised. The crisis began in Makuria in 1268 when an
usurper appealed for Mamluk recognition. Repeated dynastic war and Egyptian
intervention followed. In about 1316 a Muslim gained Makuria™s throne and
the cathedral at Old Dongola was converted into a mosque. Fifty years later a
King of Makuria is mentioned for the last time. ˜No vestige of royal authority
has remained in their country,™ wrote Ibn Khaldun, ˜since the system of Arab
nomadism turned them from their own system through utter disorder and
unceasing warfare.™17 Meanwhile the Arabs had gained access to the higher
rainfall and better pastures of Alwa. They took control there during the ¬fteenth
century, only to fall themselves under the suzerainty of the Funj, Africans of
obscure origin who conquered the area in 1504 and rapidly adopted Islam. The
next three centuries were a period of poverty and disorder when the Funj ruled
as far as the Third Cataract, while the valley further north was dominated by
meks, robber barons controlling stretches of river from mud-brick castles. But
it was also a period of Arabisation and Islamisation, when nomads and Muslim
teachers created the basic pattern of the modern northern Sudan. The last
report of Christians in Nubia was in 1742, although village women appealed to
the Virgin in time of need even in the late twentieth century.

Ethiopian Christianity survived Islamic expansion, chie¬‚y because it was
more remote from Islamic power. Underlying Ethiopian history between the
ninth and sixteenth centuries was continuing colonisation of former Cushitic-
speaking territory in the highlands by Semitic-speaking cultivators. By the
ninth century, the kingdom™s core was no longer in Aksum but further south
Christianity and Islam 57

in modern Wollo, whose indigenous Cushitic people spoke Agaw languages.
In 1137 an Agaw prince seized the throne and created the Zagwe dynasty, which
ruled until 1270, claiming legitimacy from such conspicuously Christian cre-
ations as the rock-hewn churches at Lalibela, laid out as a new City of Zion
around a stream named Yordanos and a hill named Calvary. Christian settle-
ment was drawn southwards by higher rainfall and the lure of trade through
the eastern lowlands to the coast at Zeila, exchanging slaves, gold, and ivory for
salt from the lowlands and imported Islamic luxuries. Muslims controlled this
trade and the peoples along the route gradually adopted Islam, ¬rst the Cushitic-
speaking Somali peoples of the eastern lowlands, then Semitic-speakers on the
southeastern highland fringes, creating two of sub-Saharan Africa™s earliest
Islamic states in eastern Shoa and Ifat.
As Semitic-speakers colonised still further southwards beyond Agaw coun-
try towards Amhara and Shoa, southern forces overthrew the Zagwe dynasty
in 1270 and installed Yikunno Amlak, who claimed descent from Solomon
and the Queen of Sheba. His grandson, Amda Siyon (1314“44), was Ethiopia™s
greatest warrior king. He conquered Ifat, forcing its Muslim leaders to create
a new emirate further east in Harar. He also extended the Christian kingdom™s
southern and western borders at the expense of non-Christian Cushitic regions
and of peoples preserving Aksum™s ancient Jewish traditions whose long resis-
tance to royal control consolidated them into the Beta Israel (Falasha) com-
munity. The Solomonic kingdom in its classic form was chie¬‚y Amda Siyon™s
Thanks to royal chronicles and ecclesiastical documents, Solomonic Ethiopia
is the earliest black African society that can be analysed in detail. It was organ-
ised chie¬‚y for the control of nature and the colonisation of land, to which
Christian merit attached. Settlement concentrated on the relatively warm and
moist plateau between about 1,800 and 2,500 metres, avoiding arid lowlands,
bleak mountain slopes, and densely wooded valleys. On the plateau, the settler
surrounded his homestead with concentric rings of gradually less intensive
cultivation and defended his ¬elds against the natural forces beyond them. The
hagiography of the Shoan abbot St. Takla Haymanot (traditional dates 1215“
1313) describes his monks carving their ¬elds from the bush, while nearby ˜the
mountain land was waste and uncultivated™. When animals raided the crops,
the saint counselled patience: ˜Let them alone, for it is we who have invaded
their habitation, and not they who have invaded ours.™ But when a huge ape
robbed a poor widow, the saint exerted his authority: ˜By the Word of God
Whom I serve, be ye kept in restraint, O all ye beasts of the desert, for ye have
overrun the boundaries which have been appointed to you.™18 To maintain
those boundaries was at the heart of Ethiopian culture. Holy men like St. Takla
Haymanot often protected people miraculously against wild animals. Satan,
when cast out from the sick or sinful, usually appeared as an ape. Only holy men
58 africans: the history of a continent

could safely cross the boundary between culture and nature to live as hermits
among animals and eat wild produce.
The cultivator had other enemies. Rainfall was probably more generous than
it is today, if lake levels are indicative, and famine was less common than it later
became, but it was still a constant threat. The Islamic principality in Shoa
suffered three famines during the later thirteenth century. In 1520 a Portuguese
missionary, Francisco Alvares,

travelled ¬ve days through country entirely depopulated, and through millet
stalks as thick as canes for propping vines; it cannot be told how they were all
cut and bitten, as if bitten by asses, all done by the locusts. . . . The people were
going away from this country, and we found the roads full of men, women,
and children on foot, and some in their arms, with their little bundles on their
heads, removing to a country where they might ¬nd provisions.19

The sources also mention many epidemics, but in terms too general to identify,
although they probably included smallpox, which was attributed to an Aksum-
ite army in 569“70. Ethiopia™s exceptional range of environments supported a
variety of endemic diseases, ranging from leprosy (especially in remote rural
areas) and malaria (which Ethiopians associated with mosquito bites) to the
intestinal parasites that later European doctors found almost universal. Drastic
folk-remedies were supplemented by herbalism, the quasi-magical techniques
of the Church™s debtera (deacons), and miraculous healing at shrines. Dust
from St. Takla Haymanot™s grave ˜gave children to barren women, and he gave
relief unto women who suffered pain at the time of childbirth, and he gave
seed to eunuchs, and he healed the sick, and he destroyed the wild beasts of the
desert, and the wild beasts of the belly.™20
The cultivator™s art was to minimise his vulnerability to disaster. ˜We sow
so much,™ they told Alvares, ˜with the hope that even if each of those said
plagues [locusts and hail] should come, some would be spoiled, and some
would remain, and if all is spoiled the year before has been so plentiful that we
have no scarcity.™21 Self-reliance was vital, for the tropical highland location,
avoidance of river valleys, and the absence of even a single bridge prevented
anything but local transport of food. Wheat, barley, and teff were staples on the
plateau; ensete (false banana), in the well-watered south. Crops were rotated
and ¬elds permanently cultivated with the plough, uniquely in sub-Saharan
Africa, but it was a scratch-plough drawn by one or two beasts, so that no
manorial structure or serfdom came into being. Only men handled the plough;
women did much other agricultural work but had less economic independence
that in many African regions. Alvares noted the fertility and populousness
of long-settled Tigray. Other highland areas had population concentrations,
but much land was still pasture or bush. Families mentioned in Solomonic
hagiographies generally had few children, giving point to St. Takla Haymanot™s
Christianity and Islam 59

grave-dust. It was probably already true, as in the nineteenth century, that
not only women but men married young, which was rare in Africa and was
probably related to ecclesiastical penalties for polygyny “ although great men
de¬ed them “ and a bilateral kinship system in which young men inherited
land rights from both parents and moved away with their brides to establish
independent households. Generational con¬‚ict was consequently muted in
Ethiopia, where Christian commoners had neither corporate lineages nor even
family names. Small hamlets existed in some regions, but elsewhere elementary
families formed dispersed homesteads whose chief institutional nexus was the
parish and its church.
Scattered through this mobile, colonising society were noble households,
which over time gained greater permanence. Their wealth came from estates
cleared by ancestral pioneers and from royal grants of the right to collect tribute
in kind and labour from surrounding peasants. Such grants were revocable
in theory but often hereditary in practice, so that royal power depended on
constant territorial expansion in order to reward followers. Holders of tribute-
collecting rights were charged to maintain law and order and supply ¬ghting
men, who at this period were not normally peasants but their enemies:
Whose face have you not dis¬gured?
Whose wife and child have you not captured?22
War-horses were probably more important than ploughs in enabling the ruling
class to extract up to an estimated 30 percent of the peasant™s crop. But the
peasant (gabbar, tribute-payer) was no serf, for his multiple land rights enabled
him to quit an unpopular lord. The nobleman, too, was a Big Man (tellek
saw) whose status was earned by talent and favour in a ¬‚uid and competitive
military society whose strongly localised rulers had little corporate identity
or distinctive culture. They displayed their rank by a surfeit of servants and
by ostentatious largess to the incapacitated poor who thronged public places.
Popular insurrection on class lines appeared only in the seventeenth century,
and then under leaders claiming to be messiahs or rightful kings.
The government of this dispersed and mobile society was necessarily loose
and personalised. Yikunno Amlak and his successors ruled partly by right
of Solomonic blood but chie¬‚y by force of arms. They were generally suc-
ceeded by their sons, usually their eldest sons, but only after con¬‚ict among
them. The king ruled an agglomeration of principalities whose chiefs valued his
recognition but resented his control, which depended on appointed regional
governors with garrisons of royal troops from elsewhere in the kingdom. To
exert their authority, Solomonic kings abandoned permanent capitals until the
mid-¬fteenth century in favour of huge itinerant camps. ˜They have no written
Laws,™ a seventeenth-century European wrote, ˜Justice and Right is determined
by Custom, and the Example of their Ancestors: and most differences are ended
60 africans: the history of a continent

by the Will of the Judge.™23 Punishments were usually physical and brutal, as in
other societies where offenders were seldom caught. Un¬‚inching endurance of
pain was a point of honour for all classes, while noblemen observed a heroic
code personi¬ed by Amda Siyon:

Some among them said to the king, ˜Let us go inside the defences of the camp
and ¬ght there.™ But the king said, ˜No, I will not die in my wife™s embrace, but
I will die the death of a man in battle.™ . . . So saying, he bounded like a leopard
and leapt like a lion, and mounted his horse whose name was Harab Asfare. . . .
They surrounded him with their swords and he, his face set hard like stone
and his spirit undaunted by death, clove the ranks of the rebels and struck
so hard that he trans¬xed two men as one with a blow of his spear, through
the strength of God. Thereupon the rebels scattered and took to ¬‚ight, being
unable to hold their ground in his presence.24

Amda Siyon™s conquests created a vast mission ¬eld for the Ethiopian Church.
Its evangelists were the spiritual counterparts of military heroes: holy men like
St. Takla Haymanot, usually of gentle birth, who created pioneer monasteries in
non-Christian areas, practised extreme self-morti¬cation, waged epic struggles
against indigenous nature religions, and attracted the people to Christianity
by their power, their sanctity, their miracles, and the services they could per-
form in the new Christian order. Monasticism had existed in Tigray since the
¬fth century, according to tradition. Iyasus Mo™a extended it southwards to
Amhara in c. 1248. His pupil, St. Takla Haymanot, created the great monastery
of Debra Libanos in Shoa in c. 1286. Other pioneers built smaller institutions
throughout the south during the next two centuries, while Tigrayan monasti-
cism was revitalised by Ewostatewos (c. 1273“1352), both monastic movements
expressing regional hostility to royal centralisation. Indigenous Cushitic reli-
gion centred on nature spirits, which could possess and speak through their
priests or ordinary people. Christian holy men accepted the reality of nature
spirits but identi¬ed them as demons or manifestations of Satan and waged
personal warfare against them. On one missionary journey, for example, St.
Takla Haymanot had the people cut down the tree housing the spirit they ven-
erated, ˜and that tree was by itself suf¬cient to provide all the wood which was
required in the church™.25 In retaliation the indigenous priests had the holy man
¬‚ogged and tortured, while the pagan king twice hurled him from a precipice,
only for St. Michael to save him. Yet other local rulers allied with holy men and
were the ¬rst to accept Christianity, perhaps to free themselves from indigenous
priestly control. This could bring violent persecution of the old religion, but
more frequently the conquered peoples appear to have added aspects of Chris-
tianity to indigenous practices in an eclectic way, worshipping at the church
built from the sacred tree, celebrating the Maskal feast of the Cross supplanting
the festival at the end of the rains, and perhaps even becoming possessed by
Christianity and Islam 61

St. Michael or St. Gabriel. Indigenous, Christian, and Islamic spirits gradually
fused into a possession (zar) cult providing psychological relief for the marginal
and unfortunate.
Missionary adaptations reinforced the distinctiveness of Ethiopian Chris-
tianity. The kingdom™s retreat southwards into the highlands, together with the
simultaneous expansion of Islam, had accentuated Ethiopia™s partial isolation
from the Eurasian core of Christianity. The Bible “ as much the Old Testament
as the New “ therefore came to dominate Christian imaginations. Ethiopia was
Zion, a nation de¬ned by religion, a second Israel defending its faith against
surrounding enemies. That faith stressed the majesty of Jehovah and the divin-
ity rather than the humanity of Christ. Judaic practices “ dietary restrictions,
ritual dancing, use of the tabot or holy ark “ were emphasised, while polygyny
was hard to eradicate and eschatology and mysticism were less prominent than
in European Christianity. Yet New Testament practices also shaped behaviour,
as in the emphasis on charity, miracles, and spiritual healing. The only bishop,
sent irregularly from Alexandria, concentrated on ordaining numerous priests,
often barely literate and very young, lest the kingdom should long be without
a bishop. These secular clergy were almost a hereditary caste, married and
engaged in agriculture. Monks generally had more education, but few noble-
men could read. The result was a colourful, symbolic, largely oral, village Chris-
tianity with little hierarchical structure but a clear distinction between the laity
and a spiritual elite “ a pattern remarkably akin to secular society. Ethiopian
Christianity expressed a heroic culture: the spectacular ˜contendings™ of holy
men, the self-morti¬cation of fasting, the symbolic role of St. George and the
Archangels, all headed by a South Arabian priest-king renowned for his vio-
lence in war, notorious for his polygyny, claiming in the Byzantine manner to
preside over declarations of doctrine.
The king who ¬lled this role most completely was Zara Yaqob (1434“68). He
imposed control over the church during its outburst of monastic evangelisation,
codi¬ed its distinctive practices, strengthened the parochial system, and strove
to uproot eclecticism and enforce orthodoxy. In secular affairs, similarly, Zara
Yaqob sought to consolidate his predecessors™ conquests into a stable kingdom,
creating a ¬xed capital at Debra Berhan in Shoa and reviving the ancient custom
of coronation at Aksum. Yet he only partially succeeded and acted with such
authoritarian brutality that his death was followed by particularistic reactions
in all directions. The new capital was abandoned and centralisation was relaxed.
Between 1478 and 1527, the average age of kings at their accession was 11.
The bene¬ciary was the Sultanate of Harar, where zealous Muslims had taken
refuge from Amda Siyon. Reinforced by the Islamisation of the neighbouring
Somali and by Turkish and Arab adventurers, Harar™s forces invaded the Chris-
tian highlands in 1529 under the leadership of the Imam Ahmad ibn Ibrahim.
His astounding success was not only because Christian forces were divided
62 africans: the history of a continent

and ill-led but because newly conquered, Cushitic-speaking subjects joined
the invaders in hope of regaining independence. Muslim forces devastated the
highlands for fourteen years, destroying Debra Libanos and leaving damage
still visible on the rock churches at Lalibela. The Imam appointed governors in
each province, but in 1543 he was killed in battle with a Christian army which
included a body of Portuguese musketeers. His forces dissolved back to Harar,
leaving the Ethiopian Church, alone in Africa, to survive in independence into
the modern world.

Colonising society in western Africa

e q u i p p e d w i t h ag r i c u lt u re a n d i r o n , t h e p e o p l e s o f
western Africa sought to build up their numbers, humanise the land, fer-
tilise it with their dead, consolidate their societies, and send out more colonists
to extend the struggle with nature. These were tasks so compelling that they
gave social organisation and culture a character that still underlies African
behaviour today. This chapter describes the evolution of colonising societies in
the savanna and forest of West and West-Central Africa between the eleventh
and mid-seventeenth centuries, before the Atlantic slave trade made its most
widespread impact. But some evidence is also taken from later centuries when
it illuminates long-standing social patterns.

colonisation and agriculture
From Senegal to Angola, most western Africans of the forest and the imme-
diately adjoining savanna spoke Niger-Congo languages. North of them, also
in the savanna, were survivors of groups probably driven southwards by the
desiccation of the Sahara, speaking either Nilo-Saharan languages (possibly
including the Songhay people of the middle Niger) or Afroasiatic tongues (the
Hausa of modern northern Nigeria). Desert peoples “ Berbers, Moors, Tuareg “
also spoke Afroasiatic languages. Further desiccation in the north and labo-
rious forest clearance in the south bred a continuing southward population
This drift was not the only pattern of colonisation. The West African savanna
had no single moving frontier like North America or Siberia. Rather, clusters
of pioneer agriculturalists were scattered through the region at favoured and
defensible locations like the early settlements along the middle Niger or on
mounds above the ¬‚oodplain south of Lake Chad. By the early second mil-
lennium ad, such areas of intensive crop production and rich culture had
multiplied, often in river valleys or defensible highland outcrops where the hoe
and digging-stick were the only practicable tools. During the eleventh century,
for example, a people known to their successors as Tellem settled on the edge of
the Bandiagara escarpment in modern Mali, cultivating the plateau margins,

64 africans: the history of a continent

storing their grain and interring their dead in inaccessible caverns in the cliffs,
and making some of the earliest cloth and the oldest wooden objects “ hoes,
statuettes, musical instruments, neck-rests for the dead “ yet found in sub-
Saharan Africa. From the ¬fteenth century, they were joined and eventually
supplanted by diverse immigrants known as Dogon who practised an excep-
tionally intensive agriculture designed to utilise every scarce drop of water,
besides creating some of Africa™s ¬nest wood-carvings and its most colourful
masquerades. The staple crops in this dry savanna region were millet and fonio
(a tiny grain). Further south, where annual rainfall exceeded seven hundred
millimetres, sorghum prevailed, while rice was grown in favoured areas like the
internal delta of the Niger. The grains of the period recovered by archaeologists
are often much smaller than modern varieties, suggesting that to wrest a secure
subsistence during a short growing season needed the skill and energy that
cultivators later displayed so eagerly in public hoeing competitions.
The open plains of the West African savanna also had their population clus-
ters, drawn together by the need for defence, the advantages of low internal
transport costs and life in society, or the exercise of political power. Each nucleus
was generally surrounded by frontier settlements and separated from the next
nucleus by a tract of wilderness. Within the nucleus, each village or looser
grouping of homesteads was similarly surrounded by concentric rings of per-
manent cultivation, temporary ¬elds, and outlying woodland “ karkara, saura,
and daji, in the Hausa language “ before entering the next village territory.
In this exceptionally uneven pattern of population distribution, each cluster
had its own frontier, expanding in good times and contracting in bad. But if
numbers increased too greatly, if drought or witches or enemies attacked the
nucleus, if dissent or ambition or thirst for adventure grew beyond control,
young men might carve a new nucleus from untamed land:

Bagauda made the ¬rst clearing in the Kano bush,
It was then uninhabited jungle;
A vast forest with nothing save antelope,
Waterbuck, buffalo and elephant.
Bagauda, he had his home back at Gaya;
He was a mighty hunter, a slayer.1

Village names caught the pioneering ethos: New Village, Do™s Village, Hard
Soil, Water Wood, Hyena “ to quote a cluster from northern Cote d™Ivoire.
Traditions of migration oversimplify the process, suggesting concerted pop-
ulation movements from one location to another, whereas colonisation was
normally a gradual diffusion of families and small groups, often to settle along-
side people of quite different origin. The colonists who became known as Dogon
preserved traditions of migration from many directions and spoke languages
so diverse as to be unintelligible to villages only a few hundred metres away.
Colonisation in western Africa 65

6. Colonising society in western Africa.
66 africans: the history of a continent

To reconstruct this history of colonisation will be almost as laborious as the
operation itself. But even more surely than the peopling of North America or
Siberia, it created a mobile society responding to pressure on resources by yet
further movement.
To the south, in and around the West African forest, colonisation was espe-
cially laborious. From Senegambia to Cote d™Ivoire cultivators enjoyed only one
annual peak of rainfall and specialised in growing rice, either extensively in the
interior uplands or intensively in arti¬cial coastal polders whose sophistica-
tion impressed ¬fteenth-century Europeans. From Cote d™Ivoire eastwards,
by contrast, rainfall peaked twice each year and the staple crop was yam,
whose great productivity on virgin soil rewarded even the clearing of tropical
forest supporting up to 1,250 tonnes of vegetation per hectare. Yam-growers
were therefore compulsive but very gradual colonists. During perhaps three
millennia or more they had cleared most of the forest from the present grass-
¬elds of Cameroun. The ancestral Yoruba and Igbo of modern Nigeria had
probably colonised southwards into the forest edge for much the same period,
perfecting cultures that exploited both savanna and forest environments.
Related Edo-speaking people had penetrated the forest to the west of the Niger
in pre-Christian times, but at the end of the ¬rst millennium ad, new pioneers
pushed southwards into the region, building some ten thousand kilometres
of earth boundaries to enclose the villages and kinship territories they carved
from the bush. At that period, the northern forest edge may generally have
been some 160 kilometres north of its present position, but ¬ve hundred years
later most forest regions supported agricultural communities, although few but
hunters yet penetrated the deepest jungles of modern Ghana, Cˆ te d™Ivoire, and
The laborious colonisation of the West African forest created an even stronger
pattern than in the savanna of settled clearings surrounded by circles of pro-
gressively wilder vegetation. Later Igbo villagers, for example, focused their
communities on central meeting- and market-places surrounded by rings
of residential compounds, then belts of oil-palms (which ¬‚ourished close to
human settlements), then village farmlands, and ¬nally ˜bad bush™ frequented
by evil spirits, heroic hunters, and herbalists. The Edo-speakers™ earth bound-
aries reveal a core of small and complex enclosures surrounded by a penum-
bra of larger enclosures and wasteland, indicating a gradual outward thrust
of colonisation. From the later ¬rst millennium ad, village clusters in such
core territories were coalescing into the ¬rst microstates which were to be the
building-blocks of political development.
We know more about the colonisation of western equatorial Africa, thanks to
Jan Vansina™s skill in eliciting historical information from surviving languages.2
Here Bantu-speakers had entered an immensely complex environment. The
equatorial forest, containing little to eat or hunt, was hard to penetrate and
Colonisation in western Africa 67

harder to clear. Bantu cultivators left it mostly to Pygmy bands with whom
they established ties of exchange and patronage. Interpenetrated with the
forest, however, were more favourable microenvironments: forest-savanna
edges, swamps and rivers rich in ¬sh, riverside toe-holds that farmers could
enlarge into cultivable land. Following the rivers, pioneers could expand more
swiftly than their ancestors in West Africa. The ¬rst Bantu colonists used stone
axes and digging-sticks to cultivate yams, oil-palms, and possibly plantains.
Their descendants acquired iron tools and gradually expanded their num-
bers, penetrating almost the entire region by ad 1000. Thereafter small groups
no longer sought new land each year for their crops. Instead stable popu-
lations consolidated around semipermanent plantain gardens and sent out
colonising offshoots when densities grew too great, although they retained an
utterly unsentimental, instrumental attitude to the exploitation of nature. As
groups specialised to distinctive local environments, their cultures and lan-
guages differentiated and ethnic groups took shape. On the northeastern edge
of the equatorial forest, Bantu-speaking forest cultivators met and interacted
with grain farmers speaking Nilo-Saharan languages to produce a rich com-
posite culture. To the southwest, beyond the forest in the savanna of modern
Angola, farming peoples acquired cereal crops and cattle from the east, mingled
with earlier forager-hunters, created population concentrations and emergent
ethnic groups in river valleys, and expanded to more arid lands as far south-
wards as the purely pastoral regions of modern Namibia. Yet widely as the Bantu
spread, they left vast areas almost unoccupied. Much of the eastern uplands
of Kivu Province was still uninhabited in the nineteenth century. Even more
than elsewhere in western Africa, equatorial agriculture required collaborative
effort, for it needed a group of at least twenty men to clear equatorial forest and
humanise a local environment. Colonists therefore lived in nucleated villages
forming clusters separated by vast empty wastelands. Most clusters were on the
forest-savanna edges where clearing was easiest and men could exploit multi-
ple environments. Here the ¬rst substantial polities would take shape during
the second millennium ad.
In colonising the land and building up their numbers, western Africans
struggled to establish an equilibrium with their exceptionally hostile disease
environment. Disease was probably very common, as is suggested by the many
complaints and deformities represented in early terracotta ¬gures from Nok
and from the Yoruba town of Ife. But many conditions may have been chronic
rather than fatal, precisely because parasites had had so long to adapt them-
selves to human hosts in Africa. Malaria was probably the biggest killer, espe-
cially of infants, in all but the coolest and driest regions; its absence from the
high grass¬elds of Cameroun was a reason for their intensive settlement. But
western Africans had evolved a relatively high level of resistance, just as they
possessed much resistance to hookworm anaemia and suffered two childhood
68 africans: the history of a continent

complaints “ yaws in equatorial regions and endemic syphilis in the savanna “
less acute than the related venereal syphilis from which the region was spared
until the sixteenth century.3 Leprosy was common when Europeans penetrated
beyond the coast in the nineteenth century, especially in equatorial regions
and Igboland, but there it, too, generally took a milder form than in other
continents and only the most severe cases were ostracised. Tsetse ¬‚ies transmit-
ting trypanosomiasis infested many wooded areas, especially along waterways,
causing Gambian sleeping sickness; its victims included the mid-fourteenth-
century King Diata II of Mali, but West Africans generally had much resistance
and the disease took a protracted form. Similarly, modern research has shown
that West and East Africa had a distinct, relatively mild strain of smallpox.4 Long
familiarity had also contributed to medical skills. The ancestral Bantu language
had a root for medicine, -ti-, which also meant a tree, indicating the herbal
basis of African medical practice. Many Bantu languages also had a common
word for the cupping-horn with which doctors bled patients. This practice was
reported by sixteenth-century missionaries to the Kongo kingdom in modern
Angola, in addition to the use of herbs, ointments, purgatives, and magical
remedies. Hausa specialists included herbalists, bone-setters, midwives, and
barber-surgeons, as well as exorcists using spiritual procedures. Anthropologi-
cal research has generally stressed the rational, experimental character of West
African medical systems and the widespread knowledge of folk-medicine. Yet
disease was common and debilitating, especially when compounded by diets
de¬cient in animal protein and vitamins “ slaves taken to the Americas were to
grow markedly taller than their African ancestors “ and when supplemented
by the ˜head-aches, bloody-¬‚uxes, fevers . . . cholicks, pains in the stomach™
noted on the seventeenth-century Gold Coast. These maladies were due chie¬‚y
to drinking bad water, as was the agonising complaint of Guinea worm, ˜the
misery™ as it was known in Borno, which disabled great numbers throughout
West Africa, especially among the poor. Yet the region was protected by the
Sahara against Old World epidemics. The Black Death appears to have spared
West Africa. Several unspeci¬ed epidemics affected savanna towns during the
sixteenth century, but not until the 1740s was ˜plague™ reported simultaneously
there and in North Africa.
Famine was a second obstacle to population growth in all but the best-
watered regions. Both oral traditions and the Islamic chronicles of savanna
towns stressed its devastating effects. Portuguese records of Angola from the
sixteenth century show that a great famine occurred on average every seventy
years; accompanied by epidemic disease, it might kill one-third or one-half
of the population, destroying the demographic growth of a generation and
forcing colonists back into the river valleys. Whether famines were so devas-
tating before Europeans brought their acute strains of smallpox is uncertain,
but they were destructive enough. They might be due to locusts (which Ibn
Colonisation in western Africa 69

Battuta reported in Mali in 1352), unseasonably heavy rains, abuses of power,
or warfare that prevented people from practising survival skills, but the most
common reason was drought. From about ad 300 to 1100, West Africa enjoyed
an interval of relatively good rainfall, as the prosperity of the Niger Valley sug-
gests. Lake Chad, too, was high for most of the period. The next four centuries
experienced renewed desiccation. Desert conditions spread southwards into
former savanna, making al-Idrisi in 1154 the ¬rst of many Jeremiahs to warn
that the Sahara was advancing. Rulers of Kanem left their ˜land of famine and
austerity™ for a more southerly location in Borno. By 1400 Old Jenne was aban-
doned after a thousand years of prosperity. Meanwhile savanna conditions in
turn ate into the northern forest edge, enabling horsemen and cattle-owners to
establish a new dominance over agriculturalists. The sixteenth century saw a
brief improvement in rainfall, but soon after 1600 desiccation resumed. During
the next 250 years, the western Sahara expanded two hundred to three hundred
kilometres southwards. The deterioration was signalled by crop failure in the
Niger Valley in 1639“43, when New Jenne™s townsmen sacked their ruler™s store-
houses. The worst crises were in the 1680s, when famine extended from the
Senegambian coast to the Upper Nile and ˜many sold themselves for slaves,
only to get a sustenance™, and especially in 1738“56, when West Africa™s great-
est recorded subsistence crisis, due to drought and locusts, reportedly killed
half the population of Timbuktu. ˜The most distinguished people ate nothing
but . . . seeds of grasses . . . or of any other grain which ordinarily were eaten only
by the most vile and impoverished people™, the chronicler recorded,5 adding
that the poor were reduced to cannibalism, the standard African metaphor
for the collapse of civilisation. Famine deaths on this scale were possible,
for three well-documented famines in Cape Verde between 1773 and 1866
each killed roughly 40 percent of the population. But such mortality was
rare. Famine was generally only one among several obstacles to demographic
Surrounded by these obstacles, western Africans attached supreme impor-
tance to the production of children. ˜Without children you are naked™, said a
Yoruba proverb. Virility was vital to a man™s honour; a Kuba village on the
southern edge of the equatorial forest might have a celibates™ quarter known
as ˜the street of small children.™ Childlessness was even more bitter for women.
˜The fruitful Woman is highly valued, whilst the Barren is despised™, wrote an
early visitor to Benin. Children were essential to parents™ social standing, to
their welfare in old age, to their survival as ancestors, and to the group™s very


. 2
( 12)