. 5
( 12)


vate merchants exported alongside of¬cial traders. Most ¬nal sales of slaves to
European merchants were by coastal middlemen who strove to prevent either
138 africans: the history of a continent

white men penetrating the interior or inland traders reaching the sea “ per-
haps by telling each that the others were cannibals. In Senegambia and Upper
Guinea, these middlemen were often Afro-Portuguese. Elsewhere they were
usually Africans, the best-known group being the Ijaw traders of the Niger
Delta who employed an institution, the canoe house, which was a combination
of descent group, trading company, and political faction, the core lineage being
swollen by slaves and dependents who paddled huge canoes up the Niger to
collect slaves:

The Black Traders of Bonny and Calabar, who are very expert at reckoning
and talking the different Languages of their own Country and those of the
Europeans, come down about once a Fortnight with Slaves; Thursday or
Friday is generally their Trading Day. Twenty or Thirty Canoes, sometimes
more and sometimes less, come down at a Time. In each Canoe may be Twenty
or Thirty Slaves. The Arms of some of them are tied behind their Backs with
Twigs, Canes, Grass Rope, or other Ligaments of the Country; and if they
happen to be stronger than common, they are pinioned above the Knee also.
In this Situation they are thrown into the Bottom of the Canoe, where they lie
in great Pain, and often almost covered with Water. On their landing, they are
taken to the Traders Houses, where they are oiled, fed, and made up for Sale.9

The European merchants who now bought the slaves practised two trading
systems. One, known as the factory trade, was in effect a commercial dias-
pora on African lines where political authorities permitted Europeans to estab-
lish permanent coastal settlements to bulk slaves in readiness for ships. These
factories were expensive and were founded only by seventeenth-century char-
tered companies or where slaves were especially numerous, as at Dahomey.
Private traders, by contrast, negotiated with the African merchants at a single
post or, less often, cruised down the coast purchasing a few slaves at a time
until they had full cargoes. Both systems were under ultimate African control
and both operated by lengthy and skilful haggling, lubricated by hospitality,
bribery, political alliance, copious alcohol, and personal relations as well as
institutional mechanisms to secure credit and enforce ful¬lment of contracts.
Europeans have often asserted that Africans sold one another for ˜mere
baubles or the weapons of war™. Baubles were sometimes part of the deal, espe-
cially in the early days. Even in the 1680s, some 40 percent of Senegambian
imports were beads and semiprecious stones. Generally, however, Europeans
sold to Africans much the same kinds of goods as they sold to American
colonists. At least half of West Africa™s imports during the seventeenth and
eighteenth centuries were cloth, initially mostly from India or elsewhere in
Africa, later mostly from Europe. Raw iron and copper were also important, as
were cowrie shells (as currency) in the Bight of Benin. In the eighteenth century,
four items other than cloth each formed about 10 percent of imports: alcohol,
Atlantic slave trade 139

tobacco, miscellaneous manufactures (chie¬‚y metal goods), and ¬rearms and
gunpowder. North Europeans began to sell guns in quantity during the late
seventeenth century, when cheap and more reliable ¬‚intlock muskets led states
on the Gold Coast and the Bight of Benin to rearm their forces. A century later,
sub-Saharan Africa was importing nearly 200,000 muskets a year.
In confronting European traders, the eclecticism and competitiveness of
African societies made imported goods fatally attractive. None were essentials,
except, in a sense, ¬rearms, but most were consumption goods suf¬ciently val-
ued to entice African rulers and many ordinary people to sell other Africans
towards whom they felt no obligation, much as medieval Venetians and Genoese
had sold other Europeans to Muslims. Some Africans opposed this, not neces-
sarily on moral grounds. Several stateless peoples refused to trade in slaves,
Benin closed its slave market, King Afonso of Kongo bewailed the trade™s
effects, and there are accounts of ordinary people helping slaves to escape.
Given African concern to build up numbers, to sell people was uncongenial
and tragically ironic. Its logic lay in the divorce between collective and individ-
ual interest, for powerful men sold slaves to acquire goods with which to attract
still more personal followers. They sold people in order to acquire people.
The haggling was ended and the slave had passed to his new, European owner.
The ¬rst task was to brand him, as at every change of ownership. The second
was to load the slave on a ship for America before he died. There are no reliable
statistics of mortality before embarkation. Joseph Miller has estimated that of
every one hundred people enslaved for export from Angola in the last decades
of the eighteenth century, ten may have died during capture, twenty-two on the
way to the coast, ten in coastal towns, six at sea, and three in the Americas before
starting work, leaving fewer than half to work as slaves.10 Higher estimates could
be quoted for every stage: in the late seventeenth century, Gambia slaves cost
at least ¬ve times as much at the coast as at their inland place of enslavement.
Nothing more precise is possible, but time spent in coastal slave pens or aboard
ship waiting to sail was thought to carry high risks of disease, suicide, or
attempted escape:

When our slaves are aboard we shackle the men two and two, while we lie
in port, and in sight of their own country, for ™tis then they attempt to make
their escape, and mutiny . . . they are fed twice a day . . . which is the time
they are aptest to mutiny, being all upon deck; therefore all that time, what of
our men are not employ™d in distributing their victuals to them, and settling
them, stand to their arms; and some with lighted matches at the great guns
that yaun upon them, loaden with partridge, till they have done and gone
down to their kennels between decks.11

The moment of sailing was traumatic. ˜The slaves all night in a turmoil™, a
sailor™s diary recorded. ˜They felt the ship™s movement. A worse howling I
140 africans: the history of a continent

never did hear, like the poor mad souls in Bedlam Hospital. The men shook
their fetters which was deafening.™12 The anguish was in part because many West
Africans believed that Europeans were sea creatures, cannibals from the land
of the dead, whose black shoe-leather was African skin, whose red wine was
African blood, and whose gunpowder was burnt and ground African bones.
Similar fears existed in Mozambique and among those exposed to the Saharan
slave trade. Yet slaves owned by West African masters were also capable of
desperate violence, whether suicide or murder, born of offended honour and
love of freedom. Revolts may have taken place on some 10 percent of slave
voyages. An average of about twenty-¬ve slaves died in each known revolt.
The risk of death was perhaps four times as high as the chance of liberation,
for of 369 revolts where something is known of the outcome, in only 12 does
any slave appear to have returned to Africa as a free person. Taken as a whole,
probably fewer than one slave in a thousand of those exported regained freedom
before reaching America. The two most successful known revolts took place
on the Marlborough in 1752 and the Regina Coeli in 1858; in each case some 270
slaves escaped after seizing control of the ship while still close to their point
of embarkation. Revolt was most common on ships sailing from Senegambia,
Upper Guinea, and the Gold Coast “ all locations where slaves may have had
strong traditions of military honour “ and on those with large proportions
of female captives, possibly because women were commonly allowed greater
freedom of movement.13 Not that anyone had much freedom in a tumba, a
cof¬n, as the Portuguese called their aging slave ships. The average vessel in the
eighteenth-century French trade was twenty metres long, six metres wide, and
carried about three hundred slaves. In 104 ships measured between 1839 and
1852, the average deck space per slave was about 0.4 square metres. Mortality
depended chie¬‚y on place of embarkation, length of voyage “ averaging two
to three months in the eighteenth century but sometimes much more “ and
whether an epidemic broke out, usually dysentery, smallpox, or scurvy. Some
12 percent of slaves despatched to the Americas between 1519 and 1867 died at
sea.14 Sharks sometimes followed ships for a month.
Accounts by slaves who survived the Middle Passage generally stressed three
memories: the disgusting atmosphere in the slave quarters, where sometimes a
candle would not burn; the crew™s pervasive brutality; and especially the thirst,
for water was the crucial scarce resource: the normal ration was about one
litre per day. Olaudah Equiano, who claimed to have been kidnapped in Igbo
country at the age of 11 and sold to British slavers in 1756, wrote the most vivid

The closeness of the place, and the heat of the climate, added to the number in
the ship, which was so crowded that each had scarcely room to turn himself,
almost suffocated us. This produced copious perspirations, so that the air
Atlantic slave trade 141

soon became un¬t for respiration, from a variety of loathsome smells, and
brought on a sickness amongst the slaves, of which many died. . . . This
wretched situation was again aggravated by the galling of the chains, now
become insupportable; and the ¬lth of the necessary tubs, into which the
children often fell, and were almost suffocated. The shrieks of the women,
and the groans of the dying, rendered the whole a scene of horror almost

demographic consequences
Because the struggle to build up population had hitherto been African history™s
chief theme, the slave trade™s demographic impact was potentially its most
important. Unfortunately, it is also the most dif¬cult to investigate. Although
the number of slaves exported is now reasonably clear, there is no reliable way
of estimating loss of life before embarkation, nor do we know how large West
Africa™s population was when the trade began, whether and how fast it was
increasing, and whether and how fast it might have increased thereafter if the
slave trade had not happened. Historians can construct models of the demo-
graphic processes involved, as Patrick Manning has done, but many ¬gures fed
into the models must be guesses. Manning took census data for 1931, assumed
a natural (or intrinsic) population growth rate of 0.5 percent a year for most
of the previous centuries, allowed for the slave exports suggested by estimates
then current, and concluded that the area of western Africa supplying the
Atlantic slave trade contained twenty-¬ve million people in 1700. Using the
known age and sex composition of slaves exported, plus estimates of casualties
at earlier stages in the trade, he calculated that by 1850 the equivalent popula-
tion had fallen to about twenty million, with the worst losses in Angola and
the Bight of Benin. He also argued, however, that the true demographic cost
was to the likely population growth if there had been no slave trade. Using the
same assumptions, he reckoned that in 1850, but for the slave trade, the pop-
ulation of all sub-Saharan Africa might have been about 100 million but was
in fact about 50 million. This loss of potential population took place during
rapid demographic growth elsewhere “ China™s population doubled in the
eighteenth century alone “ so that Manning estimated that Africa™s proportion
of the combined population of Europe, Africa, the Middle East, and the New
World declined between 1600 and 1900 from about 30 percent to a little over
10 percent.16
Most historians would agree that Angola suffered especially severely, for
it was quite sparsely populated, its slave exports were continuously high for
three centuries, and there is much descriptive evidence of depopulation. Not all
would agree that the Bight of Benin suffered so badly, for Manning assumed that
most of its exported slaves came from close to the coast, which is disputed. There
142 africans: the history of a continent

is no consensus on whether western Africa™s population declined absolutely,
nor by how much, although most experts might think any decline to have been
relatively small. A few specialists believe that western Africa had little scope
for population growth before famine and epidemic would have checked it,
but more might point to its large areas of sparse population and agree with
Manning that the crucial question is how western Africa™s population might
have increased but for the slave trade. Unfortunately, two considerations make
this question virtually unanswerable. One is that Manning™s crucial assumption
of a natural growth rate of 0.5 percent a year has no evidential basis and is
much higher than normal growth rates in traditional societies. (Between 1550
and 1820, the population of England increased by 0.5 percent a year; that of
Western Europe, by 0.24 percent a year.) The second consideration is that two
other unquanti¬able consequences of European expansion in¬‚uenced western
Africa™s population history at this time.
One was the arrival of American crops, especially maize and cassava. In moist
savanna regions, maize produces nearly twice as many calories per hectare
as millet and 50 percent more than sorghum. Cassava produces 150 percent
more calories than maize and is less vulnerable to drought. Maize was easier
to integrate into established agricultural systems and spread more quickly. It
was a staple grain in the Kongo kingdom by 1640 and was especially success-
ful in forest-savanna borderlands like Asante, where it helped to feed a rapidly
expanding population and provided the army with easily transportable rations.
Cassava demanded new methods of cultivation and processing, so that it spread
more slowly, especially in West African forest areas, but it could be conserved
and transported as ¬‚our, so that it became the staple food of long-distance
traders in western equatorial Africa and was absorbed into agricultural sys-
tems along the trade routes as far eastwards as Kazembe™s kingdom in modern
Zambia, where a visitor in 1831 found ˜unending cassava gardens™. These new
crops almost certainly made more food available in a region of relatively poor
nutrition, although cassava “ widely considered a food of the poor “ was nutri-
tious only if eaten with a protein-rich accompaniment such as ¬sh. New crops
are a major reason for thinking that the potential for population growth at this
time was high.
Against this was the fact that Atlantic trade also exposed western Africa to
new diseases, although without the devastating effects they wrought in more
isolated America. These complaints may have included tuberculosis and bacil-
lary pneumonia, for West Africans show little resistance to these. They probably
included plague, from which the Sahara had hitherto protected West Africa;
epidemics appear to have affected Kongo and parts of Angola in 1655“60 and the
coasts of Senegal and Guinea around 1744. Venereal syphilis, possibly a Latin
American disease, was added to the long-established endemic syphilis and
yaws, although they were so closely related that early references are dif¬cult
Atlantic slave trade 143

to interpret. The major problem concerns smallpox, for although West Africa
probably had its own relatively mild strains, Europeans appear to have intro-
duced the virulent strains that devastated their own continent between the six-
teenth and eighteenth centuries. Coexistence of different strains might explain
the diverse responses to the disease that European observers reported, ranging
from indifference to panic-stricken witch-hunting. Equatorial regions appear
to have had least resistance. Smallpox was reported in the Kongo area in 1560
and a major epidemic took place there and in Angola in 1625“8, followed
by recurrent epidemics until the early twentieth century, often associated with
famine. But regions further north also suffered. The Bight of Benin, for example,
experienced several major epidemics from the seventeenth century onwards.
There the cult of the smallpox god, Sakpata, was allegedly introduced from
the north by the early eighteenth-century King Agaja of Dahomey. Certainly
West Africans practised inoculation against smallpox, teaching the skill to their
masters in America. In other continents, deathrates among those contracting
virulent smallpox averaged 25 percent or more. Accounts of western African
epidemics suggest mortality on that scale. Moreover, those in other continents
who recovered from smallpox were commonly left sterile. If the effects were
similar in western Africa, the disease must have cut deeply into any population
growth that American crops permitted.
In sum, we do not know how severely the slave trade affected western Africa™s
demographic history. Our best hope of assessing it will come from detailed
studies of the colonisation or abandonment of land. The most likely answer
at present is that the slave trade caused population decline in Angola and
severely retarded growth elsewhere, although the potential for growth was
substantially less than Manning™s model suggested. This happened during rapid
demographic expansion in other continents. Given the central importance of
underpopulation in African history, the slave trade was a demographic disaster,
but not a catastrophe. The people survived.

political consequences
Political consequences are better documented and perhaps easier to summarise.
Like merchant capital elsewhere, slave trading could coexist with almost any
political system. The Igbo, for example, supplied many slaves but experienced
little political change and remained predominantly stateless. Yet most trade was
conducted by citizens of major states, which often bene¬ted at the expense of
stateless peoples. The chief political consequence was to shape the character of
these states in a mercantilist direction, meaning that political and commercial
power fused, either by rulers controlling trade or by traders acquiring political
power. Such a fusion of power was not previously normal in this region. That
it now occurred was probably more a consequence of international trade than
144 africans: the history of a continent

speci¬cally of the slave trade, for similar changes happened along the Asian
coastline as European maritime trade concentrated wealth and power there at
the expense of land empires ruled by Mughals, Ottomans, or Safavis. Moreover,
in western Africa, it was the import and use of ¬rearms rather than the capture
and export of slaves that enabled small, well-armed minorities to dominate
larger populations. And foreign trade was only one of many forces shaping
western Africa™s political history at this time, not always the most important.
Three major states in western Africa disintegrated during the slave trade,
although not necessarily because of it. The play of forces can be seen in the ¬rst
important kingdom that the Portuguese contacted, Greater Jolof in Senegal.
This was a land empire, based in the inland savanna, ruled by horsemen, deeply
engaged in trans-Saharan trade, and exercising only loose suzerainty over its
four Wolof units “ nuclear Jolof, Waalo, Kajoor, Bawol “ and its Serer subjects.
By selling a few horses to the Wolof coastal states in return for slaves, Portuguese
traders encouraged centrifugal forces, but the northern trade remained more
important and Greater Jolof was probably more severely weakened by the cre-
ation of a pagan state in Futa Toro to the east during the 1490s, which inter-
rupted its inland trade. Forty years later, the other Wolof states withheld tribute
and Greater Jolof disintegrated. Now the Atlantic trade became an important
force shaping the successor states, especially when ¬rearms arrived in the sev-
enteenth century. The new Wolof kingdoms were dominated by armies of
royal slaves (ceddo), hard-drinking warriors with a code of military honour,
a deep investment in slaving, and a rough way with peasants. Against them,
however, was posed the continuing southward expansion of Islam, a histor-
ical process of more enduring signi¬cance than the slave trade. During the
next three centuries, Wolof politics centred on con¬‚ict between the forces of
mercantilism “ kings, ceddo, European traders “ and those of Islam, repre-
sented by rural-based marabouts (clerics) seeking to convert the peasantry
and create Islamic theocracies, defying the ancient West African tradition that
clerics prayed while warriors ruled. Among the Wolof, the victor was mer-
cantilism. The main revolt here, in 1673, arose from con¬‚ict in Mauritania
between Berber clerical tribes and their Arab conquerors; the leader, Nasir al-
Din, turned southwards and gained control of Waalo, Kajoor, and Futa Toro,
but after his death in 1674 the mercantilist forces regained power and slaugh-
tered marabouts. Further inland, however, mercantilism was less effective. In
c. 1698 a marabout named Malik Sy created an Islamic theocracy in Bundu, an
area formerly under Soninke rule but recently settled by sedentary Fulbe. In
1725 Fulbe clerics in Futa Jalon rebelled against their Mande-speaking rulers
and established a largely Fulbe theocracy, which for the ¬rst time in West Africa
translated the Koran into the vernacular. Half a century later, the clerical party
seized control of the Fulbe state of Futa Toro. Its ¬rst ruler banned the sale of
Muslim slaves, but the theocracies did not escape the lure of the Atlantic trade.
Atlantic slave trade 145

Futa Jalon, in particular, became a major slave exporter, with the most state-
controlled economy in West Africa and exceptional dependence on agricultural
slaves, who may have formed a majority of its population. Still further inland,
among the Mande-speaking Bambara who now dominated the old nucleus of
Mali, mercantilism rather than Islam prevailed. First a young hunter of low
birth, Biton Kulibali, expanded his age-set into a military force with which,
in precolonial Africa™s most dramatic generational revolt, he created the Segu
kingdom in c. 1712. Then slave generals overthrew his Muslim successor and
established a ceddo regime dependent on the slave trade and the use of slave
labour in agriculture and craft production.
The second important state to disintegrate during the slave trade was the
Kongo kingdom, where the European impact was more crucial because of the
proximity of the Portuguese colony in Angola. Yet Kongo™s collapse was long
delayed. After the crisis of 1526 when slaving threatened to escape royal con-
trol, Afonso I reestablished authority and con¬ned slave exports to foreigners
and convicts. His long reign (1506“43) secured his close kinsmen a monopoly
of provincial governorships. Adoption of Christianity as a state cult provided
literate subordinates to staff the administration and ritual resources to set
against the indigenous religion. By the seventeenth century, the state also had
a standing army of some ¬ve thousand, including ¬ve hundred mercenary
musketeers to whom the king sought to reserve ¬rearms. Aristocrats distin-
guished themselves from commoners by elements of European culture. They
clustered in the capital, renamed S˜o Salvador, which dominated the country-
side, where peasants and slaves gradually fused into a single subject population.
This reconstructed kingdom survived for nearly a century and a half, but it was
weakened by factionalism within the huge royal patrilineage. During one cri-
sis, in 1568, S˜o Salvador was destroyed by the militarised Imbangala. The king
needed Portuguese help to regain his throne. Meanwhile trade patterns shifted
to his disadvantage. Portuguese trade from Luanda after 1576 gave southern
provincial rulers independent access to ¬rearms and other imported goods,
while Dutch trade at Soyo from 1600 did the same in the north. A gifted king,
Garcia II, struggled to preserve the kingdom during the mid-seventeenth cen-
tury, but his successor died in 1665 at the Battle of Mbwila, precipitated by
Portuguese designs on Kongo™s copper deposits. Soyo sacked S˜o Salvador and
the Kongo kingdom disintegrated into its component provinces and villages.
One spectacular attempt at reuni¬cation was made in the early 1700s by a
young noblewoman named Beatrix Kimpa Vita, who, in a complex synthesis
of indigenous and Christian beliefs, declared herself possessed by St. Anthony
and was installed as national leader by her peasant followers at a rebuilt S˜o a
Salvador, only to be burnt at the stake in 1706.
The third major state in western Africa to collapse during the slave trade was
Oyo, the dominant Yoruba kingdom in the southwest of modern Nigeria. Here
146 africans: the history of a continent

too the interaction between indigenous processes and foreign commerce was
complex. Oyo was an inland savanna state with a cavalry elite and a political
system that dispersed power among structurally opposed groups and institu-
tions in a manner characteristic of ancient Yoruba towns. Power in the capital
was shared between the Ala¬n, a largely secluded ruler with ritual authority,
and the Oyo Mesi, a council of eight chiefs from the most important descent
groups. By the early seventeenth century, Oyo was an important supplier of
slaves to the Atlantic trade. In order to export them, it conquered a savanna
corridor to the coast through the Dahomey Gap, making Dahomey itself trib-
utary in 1726“7. Oyo also subjected many Yoruba towns and exerted some
predominance over Borgu and Nupe to the north. But the problem of control-
ling this empire (and not merely the slave trade) destabilised Oyo, just as it had
destabilised the Egyptian New Kingdom. Because power was widely dispersed
in Oyo, so were the pro¬ts of empire. The Ala¬n gained new administrative
functions exercised through royal slaves. The chiefs greatly increased their mil-
itary power. In the contest for supremacy, the senior chief, the Basorun Gaha,
pushed the Ala¬n aside and dominated the state from c. 1754 to 1774, until his
unpopularity enabled Ala¬n Abiodun to use military forces commanded by the
Kakanfo to overthrow Gaha and make himself supreme until 1789. Thereafter
con¬‚ict tore the political system apart, the subject peoples broke away, and in
1817 a dissident Kakanfo incited a revolt by Oyo™s numerous Muslims, which
ended with the overrunning of the capital. By c. 1835 it was deserted. Inter-
nal structural tensions, imperial expansion, and militant Islam had together
destroyed the state.
While old land empires collapsed, new mercantilist states arose, either by
merchants gaining political power or by rulers controlling commerce. The
most successful merchants were those of the Niger Delta, where the heads of
the most powerful canoe houses emerged in the eighteenth century as hered-
itary ˜kings™ in several small trading towns. In equatorial Africa, among the
Vili traders of Loango on the Gabon coast, a kingdom existed before foreign
trade became important, but when its ruler sought to supervise trade, tradi-
tionalists insisted that he avoid corruption by eschewing contact with white
people and their products; wealth and power therefore passed to merchants,
who ousted territorial chiefs from the royal council and eventually overshad-
owed the monarchy, which had no incumbent for a century after 1787. Away
from the coast, among the Tio traders of Malebo (Stanley) Pool on the River
Congo, the kingship became purely ceremonial and power passed to provincial
trader-chiefs. Leadership by Big Men had long predominated in this equato-
rial region and dovetailed neatly with the Atlantic trade. Among the Bobangi
merchants of the middle Congo, for example, canoe houses were as domi-
nant as in the Niger Delta, except that they had less continuity in this newly
commercialised and competitive region.
Atlantic slave trade 147

The most constructive political effects of the Atlantic trade took place among
the Akan people of the Gold Coast. By the seventeenth century, their wealth
in gold had bred a populous, commercialised, and strati¬ed coastal society
dominated by Big Men whom a European trader described as ˜wonderful proud
and haughty™. Greater power, however, lay with military rulers in the hinterland,
who used the new ¬‚intlock muskets of the late seventeenth century to create
citizen armies, enlarge their states, and control coastal ports in order to secure
their arms supply. In 1680 the ¬rst of these new states, Akwamu, captured Accra.
Eighteen years later, its rival, Denkyira, conquered Assin. But the eventual victor
was Asante, a dependency of Denkyira that threw off its control in 1701 under
the leadership of Osei Tutu, conquered Denkyira™s other dependencies, and
became the most powerful of the Akan states.
Asante™s underlying wealth lay in agriculture. With its capital at Kumasi,
some ¬fty kilometres south of the northern forest edge, it could draw on both
forest and savanna produce. Land was effectively held by lineages but was
freely available and was cultivated chie¬‚y by elementary peasant households,
whose villages of small, high-gabled, thatched huts in forest clearings contrasted
dramatically with the dangerous, competitive, domineering capital, housing
perhaps 20,000 to 25,000 people in the early nineteenth century and surrounded
by a twenty-kilometre belt of dense agricultural settlement and craft specialisa-
tion. Asante was also a major trading state, with four main roads radiating north
from Kumasi into the savanna and another four reaching south to the coast. The
northern trade, especially in kola nuts, was open to private merchants as well
as state agents, but the southern trade in gold, slaves, and ¬rearms was more
closely regulated. The main roads also facilitated control of Asante™s conquests,
made ¬rst to the south, between 1701 and 1720, and then northwards between
1730 and 1752. At its peak around 1820, the empire embraced over 250,000 square
kilometres divided into three broad regions: the six metropolitan chiefdoms
that had made up Osei Tutu™s military confederacy; an inner ring of conquered
peoples who were mainly Akan and paid tax each year to state of¬cials; and
the outlying non-Akan tributaries of Gonja and Dagomba from which Asante
residents demanded a thousand slaves a year, repressing frequent rebellions.
Asante always remained at root a military society with a citizen army, a harshly
militaristic ideology, and great brutality towards the weak.
Ruling an empire presented Asante with the same problems that destabilised
Oyo. Its success in meeting them was a measure and a cause of its political
sophistication. Like Ala¬ns of Oyo, Asantehenes “ especially Osei Kwadwo
(1764“77) “ created of¬cials to administer conquests, but unlike Ala¬ns they
drew them from free matrilineages, appointed them to chieftainships sup-
ported from attached grants of land and people, and permitted them to build
up administrative departments with specialised skills. Both the Exchequer and
the Chancery employed literate Muslims, a rare use of literacy in Asante and
148 africans: the history of a continent

an equally rare reliance on Muslims, who were otherwise kept at a distance.
This administration, like Buganda™s, was patrimonial: it grew from the royal
household, remained subject to the king™s favour, became in part hereditary,
and earned no regular salaries. Created to administer the empire, however,
it became also a means by which kings gradually asserted supremacy over the
military chiefs within the metropolitan provinces. Kings also exploited rivalries
between provinces, created an internal security force (the ankobea), attracted
cases into royal courts, established a state cult of the Golden Stool, elaborated
an annual Odwira festival dramatising royal power, and built up a richly com-
posite culture that accumulated dances, musical instruments, and medical and
other skills from conquered peoples. Yet the kingdom™s chief strength lay in its
political institutions, which did not counterpose king and chiefs in structured
opposition, as in Oyo, but integrated them into an annual national council, the
Asantemanhyiamu. Asantehenes were kings in council. They were chosen by
the Queen Mother and prominent chiefs from several matrilineal candidates,
a system that largely freed Asante from the succession disputes so destructive
to African kingdoms.
Asante was the only part of Africa where rich agricultural and mineral
resources coincided. Its gold bought ¬rearms and also, initially, slaves, until
the high slave prices of the eighteenth century led the state to export them in
return for munitions, husbanding its gold for the domestic economy, where
even a few bananas had their price in gold dust and every man of substance car-
ried scales and gold weights, often beautifully cast in brass. Gold gave Asante its
spectacular opulence. ˜There was gold everywhere™, a dazzled coastal emissary
reported; the Asantehene™s morning bath was accompanied by the rattling of
the treasury keys. Through loans of gold, the king bound ambitious followers to
him. When converted into slave labour, gold defended Asante against the encir-
cling forest. When converted into muskets, it defended the kingdom against its
enemies. Gold accumulated by a chief belonged not to his descendants but to
his chiefdom or the state, which (at least in the nineteenth century) extracted
heavy death duties from rich men. Private success was therefore public virtue.
Gold gave Asante a means, notably lacking to most Africans, of channelling
individual competitiveness into the service of the state, although only within
limits set by the established order of rank and royal power.
The second major coastal state created in response to the Atlantic trade was
Dahomey, but it had no gold and differed accordingly. In the late ¬fteenth
century the chief polities among the Aja-Ewe peoples of the Bight of Benin
were Allada and Whydah. Dahomey was a hinterland state apparently cre-
ated in the seventeenth century as an offshoot of Allada. When Allada tried
to control the intensive trade in slaves and ¬rearms that began at that time,
Dahomey conquered Allada in 1724 and became the dominant local power,
although tributary to Oyo. Its king was restrained by his major chiefs and by
Atlantic slave trade 149

the practical obstacles to absolutism in any pedestrian society, but Dahomey
was nevertheless a more ef¬ciently authoritarian state than its predecessors. Its
royal succession, largely by primogeniture, ensured that only ten kings reigned
between 1650 and 1889. Raiding its neighbours but never creating an empire,
it remained a small kingdom closely administered through commoner chiefs
and royal courts. The religious system was under strong royal control. The
army consisted largely of regular musketeers, renowned for their brutality and
accurate marksmanship. Its famous Amazon corps, probably in origin a palace
bodyguard, was one of several means by which the kingdom gave women an
important public role, perhaps because such a small and aggressive state needed
all its human resources. Rank and etiquette were strict, militarism was strident,
all captured slaves belonged to the king, and their treatment was exceptionally
Mercantilist states also came into being in Angola. The ¬rst, Ndongo, had
come into existence during the ¬fteenth century among the Mbundu people in
the hinterland of Luanda. When the Portuguese occupied Luanda in 1576 and
began to expand inland, Ndongo successfully withstood their attacks until the
Mbundu were also assaulted from the rear by Imbangala warbands. Ndongo
¬nally collapsed in 1671, but during the 1620s its queen, Njinga, had adopted
Imbangala militarism, retreated inland to Matamba, and established a new
state that became a focus of the long-distance slave trade. The most powerful
Imbangala chief, the Kasanje, created a similar kingdom over Mbundu subjects
during the 1630s. Matamba and Kasanje were Angola™s most important mer-
cantilist states and enjoyed a kind of stability after they evolved institutions to
contain the violence of the slave trade, enslavement by war giving way to kid-
napping and the distortion of judicial procedures. Yet the frontier of violence
only moved further inland. During the eighteenth century, the chief supplier
of slaves to Angola was Lunda, while the new growth point was further south,
where the Ovimbundu people, probably reorganised politically during the sev-
enteenth century by borrowing Mbundu and Imbangala innovations, had, by
the 1790s, crossed the continental divide to explore the slaving possibilities of
the upper Zambezi.

economic and social consequences
The slave trade™s economic impact was as complex as its other effects. Slaving
was only one sector in economies that remained overwhelmingly agricultural.
Specialists have estimated that in the mid-1780s, at the slave trade™s peak, the
average value of overseas trade per person in West Africa was only £0.10 per
year, compared with £2.30 in Britain and £5.70 in the British West Indies.17
Imported cloth then amounted to less than half a metre per West African per
year. The slave trade was growing much faster than international trade as a
150 africans: the history of a continent

whole and it was, of course, most unevenly distributed within West Africa. But
the main point about its economic impact was how little change it stimulated.
Western Africa traded with the Atlantic world for over three hundred years
without experiencing any signi¬cant economic development.
The impact of imported manufactures on West African domestic industries
illustrates the point. Only on the coast itself, in Angola, and along the River
Senegal “ where the slave trade cut deepest “ did imported cloth damage indige-
nous textile industries during the eighteenth century. Elsewhere an expanding
market absorbed both local and imported products. Igbo textile production
is thought to have increased in the eighteenth century, Yoruba cloth found a
market in Brazil, and Asante established a new weaving industry with imported
northern skills. Much the same was true of other crafts. On the coast and along
the Senegal, iron-smelting “ the least competitive African industry “ often
gave way to imported raw iron, but transport costs and consumer preferences
generally protected smelters elsewhere, while blacksmiths were, if anything,
encouraged. Imported brass similarly brought Benin™s craftsmanship to superb
levels and new kingdoms like Asante and Dahomey developed the whole range
of court industries. New specialities grew up among boatmen and professional
porters. But nothing in the Atlantic trade either encouraged change in the
structure of western African industries or improved the transport system that
was the crucial bottleneck. Moreover, western Africa exported scarcely any
agricultural products, in striking contrast to Caribbean plantations. The main
agricultural export was in fact food for slave ships, often grown largely by slave
villages; at Allada in 1663 it was reckoned that to feed every eight slaves at sea
cost the value of another slave. But plantation production of tropical crops
within western Africa, although often mooted, would have needed a coop-
eration between African authorities and European merchants that was never
achieved. One of the slave trade™s most destructive effects was to retard African
commodity production.
Another effect, with more ambiguous consequences, was to foster slav-
ery within western Africa, especially female slavery. A census of Portugal™s
Angolan territories in 1777“8 showed twice as many adult women as men
because so many young men had been exported. Women were valued, as
ever, for both their reproductive capacity and their labour. Food-producing
slave villages surrounded the Lunda capital and lined its trade routes. One
entrepreneur was said to have 140 plantation slaves at work in mid-eighteenth-
century Futa Jalon. In some highly commercialised coastal societies, owner-
ship of slaves became a criterion of full citizenship and most heavy physical
labour became ˜slave work™. A merchant on the Gold Coast estimated during
the 1770s, doubtless with exaggeration, that every Fante free man owned at
least one or two slaves. Where slaves were so numerous, some might gain
privileged status, like the Ala¬n™s administrators or the royal slaves in Wolof
Atlantic slave trade 151

states who supervised work parties of free peasants. More commonly, however,
the proliferation of slaves reduced their status from poor relations towards
mere labourers. Slaves probably seldom reproduced themselves. Visitors to
Asante and Dahomey noted that male slaves had dif¬culty in obtaining wives
where polygynous masters accumulated women, while research on the nine-
teenth century suggests that slave wives bore few children, perhaps deliberately.
Mass slavery also tended to reduce the status of free peasants, as in Kongo,
and this probably affected especially free women because so many slaves were
The tensions endemic to slave societies might ¬nd release in periods of
licensed disorder at festivals such as Asante™s Odwira, ˜the commonest mechan-
ics and slaves furiously declaiming on state palavers™. Other tensions evoked
more open violence. Some ninety-two shore-based attacks on slave ships are
recorded, many of them along the Senegal and Gambia Rivers. There were
localised slave insurrections against European traders in the Gambia in 1681 “2
and Senegal in 1698. Revolts against African slaveholders were particularly
common in Futa Jalon, where slaves were an exceptionally large proportion
of the population. One in 1756 established an independent slave settlement at
Kondeah. A second in 1785 decapitated numerous masters, burned rice ¬elds,
and created an autonomous slave community that survived for eleven years.
Shortly afterwards, some ¬fty slaves died in a third conspiracy. In general, how-
ever, rebellions were relatively rare before the ending of slave exports left large
concentrations of unfree men in coastal settlements. One reason for the rarity
was probably the ease of escape, whether to the slave™s home or to one of the
maroon communities fringing slave-owning regions. In seventeenth-century
Angola, for example, the Portuguese were untroubled by slave rebellion but
suffered massive slave ¬‚ight from Luanda and its environs, either to the nearby
forest refuge of Nsaka de Casanze, to the still unconquered southern thornbush
of Kisama “ probably West Africa™s most important focus of maroonage “ or to
join Queen Njinga™s resistance in Matamba, where slaves were promised land
and freedom.
The ethos of slaving societies was brutally inegalitarian and acquisitive.
Benin™s art became increasingly elaborate in its decoration, with much empha-
sis on symbols of Olokun, god of wealth. Legends told with symbolic truth that
the cowrie shells imported as currency had grown on the corpses of slaves cast
into the sea. Many held, with Bobangi traders, that wealth could be increased
only at others™ expense, commonly by witchcraft or sorcery, especially by sac-
ri¬cing a relative™s life, much as Asimini, the second king of Bonny in the Niger
Delta, was believed to have sacri¬ced his daughter to the sea gods in order
to deepen the estuary for Portuguese ships. Many oracles designed to test the
truth of witchcraft accusations now sentenced those convicted to enslavement
and sale. Among the Jola people on the southern coast of modern Senegal,
152 africans: the history of a continent

the spirit shrines that in the past had secured prosperity and warded off mis-
fortune were now supplemented both by shrines aiding success in capturing
slaves “ the rite required a priest who had himself captured a slave to pour
blood and palm wine over wooden slave-fetters “ and by shrines protecting
slaves excluded from the shrines consulted by free people. In the Anlo region of
the Gold Coast, similarly, a new cult enabled slaves to communicate with the
ancestors whom they could no longer venerate in their home areas. Perhaps
the most illuminating response to this social pathology took a medical form, as
was so characteristic of African thought. This was the Lemba cult of the middle
and lower Congo, created in the seventeenth century. Lemba (˜to calm™) was a
complaint, symptomatised by abdominal pains, dif¬cult breathing, and steril-
ity, which af¬‚icted the mercantilist elite of chiefs, traders, and successful slaves.
Perhaps the real disease was anxiety at the envy and sorcery they provoked. The
cure was to pay heavily to join the Lemba society of the powerful who could
protect one another and ensure, as their rhetoric claimed, that their lineages
did not die out. Lemba healers used drums and bracelets marked with cowrie
shells. They controlled markets, adjudicated disputes, and policed a trading
system traversing political borders across half of equatorial Africa. The cult
disappeared early in the twentieth century, along with the slave trade, but like
the consequences of that trade, it survived in the Caribbean.
Lemba illustrated the capacity of privileged Africans to create or adapt insti-
tutions in order to survive the slave trade. Successful Igbo traders formed similar
associations, while the Order of Christ, originally a Portuguese chivalric body,
took Lemba™s place in Angola. There, as in many regions, senior men also
reshaped family systems to give themselves greater control over the behaviour
and liberty of their dependants, whom they both exploited and protected. Given
the family™s central importance in African societies, its destruction was often
the slave™s most bitter experience. Olaudah Equiano™s memoirs, for example,
show him attempting throughout his life to create a surrogate kinship network.
But for those who remained in Africa, the family was probably the chief defence
against the effects of slaving, which may well have strengthened kinship systems.
Polygynous marriage, for example, provided for orphans and surplus women.
In this sense, their historical experience had equipped Africans to withstand
slaving better than other peoples might have managed, just as their codes of
honour and their training in the endurance of pain gave them fortitude to
withstand cruelty. Africa™s previous history helped to make the slave trade not
only possible but survivable.

the impact of abolition
In 1807 the British Parliament resolved to abolish the Atlantic slave trade.
Abolitionists believed that this would open a new chapter in West African
Atlantic slave trade 153

history. Some historians, too, have seen abolition as a major discontinuity
giving West Africa its modern place in the world economy as a supplier of
agricultural produce, with its attendant social and political consequences. There
is some truth in this, but the continuities between the eighteenth and nineteenth
centuries were equally striking, for parliamentary resolutions had little impact
in Africa. Britain could enforce abolition on its citizens, but its only weapon
against foreigners was to station naval vessels off the West African coast to
intercept slave ships. ˜They took off all the fetters from our feet, and threw
them into the water,™ a freed slave remembered, ˜and they gave us clothes that
we might cover our nakedness, they opened the watercasks, that we might
drink water to the full, and we also ate food, till we had enough.™18 In all, the
navy captured 1,635 ships and freed just over 160,000 slaves, landing many at the
colony created in Freetown in 1787. Yet no fewer than 3,446,800 slaves left Africa
for the Atlantic during the nineteenth century, or more than half as many as
during the previous century.
The British campaign did change the pattern of the trade. Most slaves now
went to Brazil or the Spanish sugar-producing colony in Cuba. The traders were
predominantly Brazilian or Cuban (i.e., Spanish) merchants, who established
permanent coastal factories ready to load slaves the moment a ship arrived, in
order to evade British patrols. One Brazilian ship arrived at Cabinda, loaded
450 slaves, and sailed within less than one hundred minutes. Resident foreign
traders gained political in¬‚uence, especially in the Bight of Benin where Felix
de Souza in the 1830s and Domingo Jos´ Martins in the 1840s became mer-
chant princes. The illegal trade was risky and pro¬table. Between 1856 and
1865, Cuban ships probably averaged over 90 percent pro¬ts on their outlay,
against perhaps an average of 10 percent during the eighteenth century, chie¬‚y
owing to increased selling prices in the Americas. It became more pro¬table to
export children, who now numbered half of those shipped from Angola. Slave
origins also changed. Civil war in Yorubaland from 1817 made the Bight of
Benin an important supplier to the last, but otherwise the trade shifted south-
wards to escape British patrols, concentrating on the Loango coast, Angola,
and Mozambique, which together provided over 80 percent of slaves shipped
after 1855.
West Africa™s foreign trade as a whole expanded dramatically during the
¬rst half of the nineteenth century. Trade with Britain and France multiplied
six or seven times between 1820 and 1850, while imports of European cotton
cloth multiplied about ¬fty times. The expansion was due chie¬‚y to lower
prices for European factory products, followed after 1850 by the impact of metal
steamships. The effects must not be exaggerated. The average per capita value of
West Africa™s overseas trade during the 1860s was only one-fortieth of Britain™s
and one-eighth of Brazil™s. Moreover, while Africa™s share of world trade
had increased during the eighteenth century, it declined thereafter. Domestic
154 africans: the history of a continent

production and consumption remained its economic core, although overseas
trade was nevertheless its growth point.
The consequences were not all bene¬cial. One was in¬‚ation, for after 1850
European ships imported vast quantities of cowrie shells, which depreciated
the currency. In Lagos cowries lost 87 percent of their value between 1850 and
1895. Cheap European cloth largely ousted local textiles from the mass market in
commercialised areas like southern Yorubaland and Asante, although wealthier
consumers still preferred African cloth. In Yorubaland in 1862, a traveller from
Osiele to Abeokuta passed 1,305 people of whom 1,100 wore only European
cloth. European competition damaged Igbo domestic industries for the ¬rst
time. Local iron-smelting declined very widely.
Yet the important point about nineteenth-century trade was that West Africa
increasingly exported agricultural and forest products in a pattern that was to
survive until the 1960s. Small quantities of these products had been exported
even in the eighteenth century. Old Calabar shipped palm oil during the 1770s,
gum (used in the textile industry) overtook slaves in Senegambia™s exports
between 1780 and 1820, and it was generally major slave-trading regions that
pioneered legitimate trade, often using existing commercial structures. But
European demand for vegetable oils multiplied in the early nineteenth cen-
tury because industrial populations did not produce their own oils and fats
and because technical innovations made oil-processing more pro¬table than
before. Between the 1820s and the 1850s, Britain™s imports of West African palm
oil multiplied more than six times and the price roughly doubled. By the 1830s,
West Africa™s palm-oil exports were worth more than its slaves. Oil reached
Britain chie¬‚y from the Niger Delta, which was supplied chie¬‚y by Igbo and
related hinterland peoples whose oil-palms grew virtually wild. Other impor-
tant suppliers were Yorubaland, where women were the main producers, and
Dahomey, where merchants and chiefs established slave-worked palm planta-
tions. Vegetable oil also came from groundnuts, a savanna crop ¬rst exported in
1834 from the Gambia and in 1839 from Senegal, where they were grown by local
peasants, migrant ˜strange farmers™ from the interior, and slaves. Between 1868
and 1877, Senegal exported an annual average of 27,000 tonnes of groundnuts,
the output of an estimated 70,000 producers. By then coffee (¬rst exported in
1844) and forest products (ivory, wax, and rubber) were earning even Angola
more than the slave trade had ever paid.
Many African leaders resisted the abolition of the slave trade. Kings of Asante,
Dahomey, and Lunda all warned that unsold captives and criminals would have
to be executed. Historians have suggested that the transition to agricultural
exports created a ˜crisis of the aristocracy™, because whereas slave-trading had
been a business for ˜kings, rich men, and prime merchants™, agricultural exports
pro¬ted small traders and producers. Aristocrats consequently intensi¬ed their
exploitation of commoners who were themselves better placed to resist, so that
Atlantic slave trade 155

con¬‚ict escalated. This thesis is stimulating but needs much quali¬cation. Not
only was there much political continuity even in regions active in the Atlantic
trade, but foreign commerce was not the most important in¬‚uence on most
West African political systems.
Asante, for example, had never relied entirely on slave exports and continued
for much of the nineteenth century to sell slaves to coastal peoples in return for
European goods purchased with palm oil. The kingdom also expanded both
its gold exports to the coast and its kola exports to the Islamic north. After
losing control of the coast to the British in 1826, Asante™s militarists also lost
internal power to advocates of peaceful trade. Kwaku Dua I (1833“67) presided
over a long period of peace, commercial prosperity, and authoritarianism.
The bureaucracy grew into a hereditary stratum with jealously guarded skills.
Power passed from the national council to an inner circle around the king. The
government tightened its control over both the provinces and private citizens,
especially rich traders whose prosperity and experience of the outside world
made them resentful of mercantilist restraints. Mid-nineteenth-century Asante
was not only wealthy and sophisticated but brutal and increasingly hostile
to innovation that might threaten the established order, rejecting missionary
teachers who offered literacy lest their teaching should undermine military
valour or slave obedience. Major discontinuity came only when Kwaku Dua™s
successor resumed an expansionist policy leading, in 1873“4, to a war in which
a British expedition destroyed Kumasi, all but the nuclear chiefdoms of the
empire seceded, and long-contained forces of change were released to create
half a century of turbulence. Much the same happened in Dahomey, which
survived abolition with little dif¬culty and reached its apogee under Gezo
(1818“58). Not only did Dahomey continue to export many slaves until the mid-
century, but it supplemented and eventually supplanted them by slave-grown
palm-oil exports. Although indigenous merchants gained greater in¬‚uence,
the military aristocracy remained dominant until the French invasion of 1892.
Thus the two major West African coastal kingdoms experienced much con-
tinuity during the transition from slave to legitimate trade. Senegambia also
experienced continuity, but a continuity of instability. It has been argued that
groundnuts enabled peasants to buy ¬rearms with which, under the leader-
ship of Islamic marabouts, they resisted oppression by ˜pagan™ (often eclectic)
rulers and their ceddo gunmen, themselves impoverished by abolition. The
result, supposedly, was ˜maraboutic revolutions™, the most important being
that of Maba Jaaxu on the northern bank of the lower Gambia in 1860“7. Yet
such con¬‚ict had racked the region since the seventeenth century and was
little altered when rulers had to tax groundnut exports or set unsold slaves
to groundnut production, especially when the important slave trade north-
wards to the Sahara continued unchecked. The main difference was perhaps
that some earlier maraboutic revolutions had succeeded whereas those of the
156 africans: the history of a continent

nineteenth century failed, partly because the French were increasingly prepared
to intervene against them. Only after European conquest did marabouts pre-
dominate in Senegambia. The main local consequence of abolition was rather
to encourage Futa Jalon, the major slaving state, to expand its power to the
Upper Guinea coast, partly to share in legitimate trade and partly to acquire
land for its expanding population, both slave and free.
Continuity largely characterised the nineteenth-century Niger Delta. The
transition from slaves to palm oil was gradual there. Delta ports produced
neither, but only marketed them, the trading machinery of canoe houses serving
equally well for both. These houses had long allowed mobility for talented men
of low birth, even slaves. They had long competed with one another and divided
when they grew too large. These processes continued during the nineteenth
century. In Bonny, for example, the monarchy collapsed amid rivalry and the
loser in the subsequent struggle among canoe houses, a talented slave named
Jaja, withdrew in 1869 to create a new town at Opobo and seek traditional
legitimacy. The exception to this pattern was in Old Calabar (on the Cross
River), which lacked canoe houses to provide mobility. Instead the commercial
elite used a secret society, Ekpe, to repress with great brutality the growing
numbers of slaves, who responded by combining together to defend themselves
as Bloodmen (who had sworn a blood oath), although they did not oppose
slavery as such and continued to support their masters in factional politics.
Continuity within change was especially clear in nineteenth-century Yoruba-
land. As Oyo collapsed after the Muslim revolt of 1817, perhaps half a million of
its people retreated southwards into the forest, where they carved out huge areas
of farmland, adopted cassava and maize, established major cities at Ibadan,
Ijaye, Abeokuta, and New Oyo, and initiated almost a century of warfare with
¬rearms bought with slaves and palm oil. Yoruba society was extensively mil-
itarised, war chiefs accumulated power and huge followings of ˜warboys™, and
even youths too young to ¬ght followed the armies in units called ˜Father said I
should not run away™. Ibadan emerged as the most powerful town, but it failed
to create a new Yoruba order, chie¬‚y because its character as a confederacy
of military chiefs contravened Yoruba political traditions. Rival towns with
ancient monarchies, including the Ala¬n at New Oyo, despised Ibadan as ˜a
people without a king or even a constitution™. Yorubaland ¬nally gained a new
order in 1893, but it was British.
Abolition had greater impact on political systems in equatorial Africa, where
slaving had been most destructive. The decline of Angola™s slave exports during
the 1850s created crisis among Imbangala rulers. Lunda, the major slave supplier
but still largely without ¬rearms, suffered especially severely from the expansion
of trade in forest products, which brought into the kingdom mobile, well-armed
bands of Chokwe hunters who intervened in succession disputes, killed the
king, sacked the capital, and destroyed the empire during the 1870s and 1880s.
Atlantic slave trade 157

Mobility was even more striking in Gabon and Cameroun, where imported
goods attracted the acquisitive Fang and Beti peoples to carve their way through
the equatorial forest towards the coast during the mid-nineteenth century,
setting off ˜a paroxysm of violence™. Yet the most important new state in this
region, the Bamoum kingdom in the Cameroun grass¬elds, owed much of
its early nineteenth-century expansion to the use of newly arrived ¬rearms to
conquer, enslave, and resettle thousands of captives, much like the creation of
Asante a century earlier.
As Bamoum demonstrates, one reason for continuity from the eighteenth
to the nineteenth century was that the prohibition of slave exports positively
expanded slavery within Africa, where slaves became more numerous than in
any other continent. The slaving frontier pressed ever deeper into the interior.
Around 1850 Ovimbundu slave traders from southern Angola ¬rst reached the
Luba of Katanga and the Ovambo of modern Namibia. In West Africa, the major
new slaving area was Mossi country (probably embracing a broad Voltaic area),
which provided 59 percent of the 605 slaves bought by one merchant in Kumasi
between 1837 and 1842. Forty years later, the slave trade in this region culminated
in the conquests of the Dyula state-builder Samori, who sold thousands of slaves
towards the coast for ¬rearms. Most of them probably went to Senegambian
groundnut farms, for there, as in Dahomey, Yorubaland, Old Calabar, and
many other regions, legitimate trade rested in part on expanded slavery.
Masters had always feared their slaves™ witchcraft. Now they feared also
rebellion as slave numbers grew and a larger proportion were men. Following
disturbances early in the nineteenth century, Asante broke up slave concen-
trations around the capital and dispersed them into the countryside. A major
revolt began in Futa Jalon in 1845 when slaves ¬‚ocked to a dissident marabout,
Alfa Mamadu Dyuhe. Known as Hubbu (from hubb, ˜love™, the key word in their
distinctive Arabic chant), they took Futa Jalon™s capital but later retreated into
a maroon community until exterminated by Samori in 1884. From 1848 to 1851,
rebellious slaves dominated Itsekiri in the Niger Delta. Shortly afterwards slaves
destroyed Ode Ondo in eastern Yorubaland and seized their freedom. Slaves in
Douala created such insecurity between 1858 and 1877 that their masters sought
European protection.
The need to terrorise numerous male slaves was probably one reason for
˜human sacri¬ce™, as nineteenth-century Europeans described it. Ritual killing
of retainers to accompany great men into the afterlife had been practised in early
dynastic Egypt, Kerma, Ghana, Igbo-Ukwu, Ife, and many other cultures. Both
Muslims and Christians had opposed the practice bravely and it survived in
the nineteenth century only where they had little in¬‚uence. Some killings were
daily rituals, others took place at annual festivals, others again at major funerals,
which perhaps became more common as aristocratic wealth increased. Most
killings had a religious purpose, but they also punished criminals, frightened
158 africans: the history of a continent

adversaries by executing captives, and terrorised slaves. ˜When a slave gets
very familiar, we take him to a funeral custom™, said an Asante proverb. ˜If
I were to abolish human sacri¬ces,™ Kwaku Dua told a missionary, ˜I should
deprive myself of one of the most effectual means of keeping the people in
subjection.™19 Long-established in Asante, Dahomey, and Benin, the practice
became increasingly common during the nineteenth century in other areas
with swollen slave populations, notably among the Bobangi, in Douala, and in
Old Calabar. It was ironic that abolition should have stimulated such cruelty.
Historians have argued that the abolition of the slave trade stimulated pop-
ulation growth in West Africa from the mid-nineteenth century. The best evi-
dence comes from Igboland, where, from about 1840, several observers reported
agricultural colonisation, the planting of cassava in poor soils, and growing local
land shortages. ˜The population is so great and it is so dif¬cult to live from their
farms™, wrote an early missionary, ˜that if they hear we shall want carriers they
come in great numbers begging to be used.™20 Yet the obstacles to population
growth remained strong. Igboland suffered severe smallpox epidemics in 1864,
1867, and 1873, as did many parts of West Africa. That of 1864 is said to have
killed one-quarter of Luanda™s population. Angola also suffered serious crop
failures in 1857, 1863“9, and 1876“84, the export of many ˜free™ labourers to
S˜o Tom´ , and unabated internal warfare and slaving, so that its demographic
a e
decline probably continued to the end of the century. Senegambia, similarly,
experienced the region™s only major cholera epidemic in 1868“9, in addition
to repeated violence and death. We do not know whether West Africa™s popu-
lation as a whole recovered during the later nineteenth century. We do know
that local experiences varied from expansion in Igboland to probable decline
in Angola. As elsewhere in nineteenth-century Africa, local factors dominated
demographic history.

Brutal though the European impact was, West African thought patterns and
institutions largely contained it until the late nineteenth century. Christian
missionary work provides the best illustration. Portugal despatched its ¬rst
missionaries to West Africa in 1458 and retained control until 1622, when the
Papacy created its own missionary agency, Propaganda Fide, and began to send
non-Iberian clergy. Missionary work declined during the eighteenth century
but revived after 1800, when it ¬rst involved Protestants as well as Roman
The chief mission ¬eld during the sixteenth and seventeenth centuries was
Kongo. Its newly established rulers lacked ritual legitimacy, were at odds with
indigenous religious institutions, and sought to make Christianity a state cult
focused especially on the spirits of dead kings “ the central church at S˜o a
Atlantic slave trade 159

Salvador was built in the royal cemetery. Afonso I saw Christianity as an ally
in the attack on ˜the great house of idols™ that he launched soon after seizing
the throne. He sent dozens of young noblemen to Portugal for education. His
son Henrique became a bishop, heading the Church from 1521 to perhaps 1530.
Several Kongo priests were ordained during the seventeenth century, literate
catechists and interpreters were recruited from the aristocracy, and heredi-
tary ˜slaves of the church™ acted as repositories of Christian ritual expertise,
although the missionaries never created a self-sustaining Kongo priesthood.
Although Christianity was at ¬rst chie¬‚y an aristocratic cult, ordinary people
also adopted Christian practices that served their needs, in the eclectic and
pragmatic manner of African religions. Baptism was widely desired as a pro-
tection against witchcraft. A new category of sky spirits, probably of Christian
inspiration, offered individuals personal advancement, healing, and protection,
again especially from witchcraft. The Virgin Mary was associated with fertility.
Continuing veneration of ancestors made All Souls Day and Holy Thursday
the most popular festivals. Sixteenth-century missionaries responded much
like early North African bishops or St. Takla Haymanot: they accepted Kongo
adaptations of Christianity, which had many parallels in rural Europe, but
treated all purely indigenous practices as works of the Devil, drawing no dis-
tinction between witchcraft and religion. These views hardened during the
seventeenth century. Ascetic Capuchin friars sent by the Vatican in 1645 were
so zealous in destroying ritual objects that they alienated Kongo aristocrats,
but in the province of Soyo, which was seeking independence, their impact
was suf¬ciently profound to in¬‚uence even aristocratic marriage practices.
Elsewhere, however, Kongo rejected monogamy and church marriage, while
missionaries treated polygyny not as an obstacle to baptism but as the sin of
Although the Kongo kingdom collapsed after 1665 and the attempted rein-
tegration by the syncretic Antonine movement was suppressed in 1706, an
Africanised Christianity survived in Kongo to bewilder nineteenth-century
missionaries. Eighteenth-century Jesuits had success in the Ambaca region of
Angola. Elsewhere in tropical Africa, however, missionary work before 1800
had little lasting impact. Benin™s rulers were hostile. Their Itsekiri rivals pro-
fessed Christianity from c. 1570 to 1733, but only at court. Rulers on the Bight
of Benin saw Christianity either as a potential ally in creating state cults or
as a source of magic but turned hostile when they learned that Christianity
involved abandoning rather than strengthening existing ritual techniques. The
Portuguese mission to Central Africa led only to Goncalo da Silveira™s murder
and three centuries of largely fruitless endeavour on the Zambezi. Approaches
to Ethiopia ended with the missionaries™ expulsion in 1632.
Active evangelisation in West Africa resumed after 1800. Roman Catholics
began work in Senegal and Gabon, expanding thereafter chie¬‚y among Yoruba
160 africans: the history of a continent

and Igbo. Protestants settled ¬rst in Sierra Leone in 1804 and extended their
work along the coast, especially in Liberia, the Gold Coast, and the south of
modern Nigeria. Coming from rapidly modernising societies, these missionar-
ies were less tolerant than their predecessors of African institutions like poly-
gyny, slavery, aristocracy, and kingship. If ruling classes opposed them, they
were eager, as the Holy Ghost Fathers™ constitution put it, ˜to concern them-
selves with the most abandoned souls™. Not only were they con¬dent of moral
and technological superiority, but they genuinely had more modern skills to
offer than barefoot Capuchins. They also had recaptured slaves as ready-made
African agents. Protestant missions, in particular, relied upon the Saro (from
˜Sierra Leone™) who returned as pastors or laymen to their homelands elsewhere
in West Africa, the most famous of these, Samuel Ajayi Crowther, becom-
ing Bishop of the Niger in 1864. Even with the aid of the Saro, however, no
nineteenth-century mission achieved anything approaching the general con-
version of an African society. Rather, Christianity was again largely absorbed
into a framework of functioning, though changing, African institutions.
Secular considerations usually shaped initial African responses to mission-
ary work. Presented with a New Testament, Asante™s state council declared, ˜It
is the Word of God, and had better remain unopened™, lest its contents threaten
the established order.21 By contrast, Akwapim, having allied with Britain to
throw off Asante™s control during the 1820s, welcomed missionaries in 1835
and became a centre of Christian innovation. Yet most early Christians there
were slaves, children, marginal adults, or princes who could not succeed to the
throne, and this became the normal pattern. ˜The bulk of the ¬rst Christian
converts™ in Igboland, Elizabeth Isichei has written, ˜were drawn from the poor,
the needy, and the rejected: the mothers of twins, women accused of witchcraft,
those suffering from diseases such as leprosy which were seen as abominable.™22
Among freemen, polygyny deterred many household heads, for, unlike their
predecessors in Kongo, most nineteenth-century missionaries believed that
Christian teaching prohibited the baptism of polygynists. Household heads, in
turn, often barred their dependents from attending Christian teaching, some-
times without success. Although many early congregations centred on one or
two large Christian households, most early Christians were women “ 80 percent
of Anglican communicants in Abeokuta in 1878 “ or the young, although the
generational protest later common among pioneer African Christians was less
pronounced in nineteenth-century West Africa, chie¬‚y because education and
literacy exerted surprisingly little appeal outside coastal colonies.
As in indigenous religions, Africans looked to Christianity chie¬‚y to meet
this-worldly needs. Many Yoruba converts, for example, joined the Church as
they might have venerated an orisa who had cured them: ˜As God has granted
her recovery through her prayers she resolves to serve Him.™23 The Christian
God was in effect added to the Yoruba pantheon, as was Allah. The Ifa cult
Atlantic slave trade 161

also absorbed many Christian and Muslim ideas. Most West African peoples
were remarkably tolerant of early Christian teaching, but it was the tolerance of
eclecticism. Persecution, unless for political reasons, came not when converts
adopted Christian practices but when they abandoned indigenous duties, as
with Nigeria™s ¬rst martyr, the slave Joshua Hart of Bonny. Missionaries, by
contrast, worshipped a jealous God, intolerant of eclecticism, and their African
agents often followed them in equating indigenous religions with witchcraft and
the Devil. As an evangelist told one enquirer, ˜So as you never in one day eating
with dog in dunghill, the same the Great God never eating with any Idols.™24
Moreover, not all of Christianity could be embraced within African thought
patterns, especially its ideas of heaven, hell, and resurrection, which had struck
Kongo villagers as frankly incredible and often provoked the same astonishment
in the nineteenth century, but could work powerfully on thoughtful minds.

coastal colonies
Partly through missionary work, an elite of western-educated Africans took
shape in coastal colonies. The oldest, Luanda, declined after 1850 as slave exports
ended, but Portuguese power expanded inland to share in the growing legiti-
mate commerce. Further north, a French fort constructed at Saint-Louis-du-
S´ n´ gal in 1659 formed the nucleus of a coastal colony inhabited chie¬‚y by an
Afro-French community trading up the Senegal River and aspiring to French
citizenship. A British trading post established on the Gambia in 1661 served
similar functions. Sierra Leone was created in 1787 as a depository, successively,
for the black poor of London, escaped slaves from North America, and slaves
recaptured by the British navy, some seventy-four thousand of whom landed in
the colony and dominated it. Libreville, on the coast of Gabon, was also settled
by recaptives in 1849, but Liberia was colonised from 1822 chie¬‚y by black freed-
men from the United States, so that its culture was more westernised when it
gained freedom in 1847. The other two colonies of the mid-nineteenth century
were the Gold Coast, which grew slowly out of trading forts, and Lagos, seized
by Britain in 1861 and settled by indigenous Yoruba, Saro recaptives from Sierra
Leone, and Amaro freedmen from Brazil.
Because either the law or their principles generally ruled out the slave
labour needed for large-scale farming, most settlers in coastal colonies sought
wealth through commerce. Many traded with goods taken on credit from
European merchants, although the most successful imported directly from
Europe. In 1875 Liberian merchants owned ¬fty-four ships. Prosperous and
often Christian, these traders had profound faith in education. By 1868 about
one-sixth of Sierra Leone™s population was at school, a higher proportion than
in Britain. The brightest pupils went on to Fourah Bay College, which was
af¬liated with Durham University in 1876, or even to Europe, where Bishop
162 africans: the history of a continent

Crowther had all his six children educated. By 1880 Africans had held posts
in the Lagos administration as treasurer, superintendent of police, and acting
colonial secretary, while James Bannerman, a successful merchant of mixed
parentage married to an Asantehene™s daughter, had held the Gold Coast™s high-
est of¬ce as lieutenant-governor. The ¬rst West African doctor with European
quali¬cations was probably John Macaulay Wilson, who practised in Sierra
Leone from 1817. The most successful professional man in a British colony was
Sir Samuel Lewis of Sierra Leone, called to the Bar in 1871 and knighted in 1896.
James Russwurm, graduate of an American university, launched the Liberia
Herald in 1830.
Merchants, of¬cials, and professionals made up the elite of coastal colonies,
the ˜aristos™ as they were known in Freetown. Until perhaps the 1880s, they
saw themselves as the agents through whom European enlightenment could
advance their continent. ˜With education . . . you will see the fall of all those
gross, not to say dishonourable, ways known unhappily as the custom of the
country™, wrote the Abb´ Boilat of Senegal.25 Yet his was an extreme position.
More would have agreed with Dr. James Africanus Horton of the Gold Coast in
seeking to blend the best of both cultures. They were in fact as eclectic as their
ancestors. The language they invented in Freetown, Krio, combined Yoruba
syntax with a vocabulary drawn from many African and European tongues.
Among the Saro of Lagos in the late nineteenth and early twentieth centuries,
99 of 113 elite men entered into Christian companionate marriages, but 54 of
them also made quasi-polygynous ˜outside unions™, a common pattern also in
Liberia. Even the most respectable clergymen prayed at their ancestors™ graves
and displayed a family sense verging on nepotism. Less acculturated men on
the frontiers of colonial society kept slaves, married into local families, and
˜went fantee™ like the prazeros of Mozambique. Many involved themselves in
homeland politics, aiding their ancestral towns in the Yoruba civil wars, sup-
porting Gold Coast chiefs in resisting direct taxation during the 1850s, allying
with them to form the Fante Confederation of 1868 from fear of Asante aggres-
sion, or creating the Egba United Board of Management in 1865 to spearhead
the modernisation of Abeokuta. Liberian settlers, by contrast, had no ances-
tral ties with surrounding peoples and possessed the political independence
to undertake inland expansion; by 1874 the republic™s territory extended eight
hundred kilometres along the coast and some two hundred kilometres inland,
at least on paper.
By the 1880s, elite eclecticism and acceptance of European control were
both under strain. One reason was commercial decline. This began in Sene-
gal, where the abolition of slavery in 1848 impoverished African and mulatto
traders, groundnut exports from the 1840s and steamship transport from the
1850s attracted French merchants, and military expansion during the 1850s
opened the River Senegal to European penetration, reducing many formerly
Atlantic slave trade 163

independent local businessmen to agents of European ¬rms. Lagos merchants
faced similar competition after 1861, but their main crisis came during the
international depression of 1880“92. By 1890 only the richest Saro merchant,
R. B. Blaize “ a millionaire in modern currency “ still exported on any scale. In
1880 some 57 percent of the Lagos elite were merchants; in 1902, only 38 per-
cent were. Even in Liberia, where legislation protected black traders, commerce
passed largely into European hands during the 1880s and 1890s. Political rivalry
between blacks and mulattos also escalated there, leading in 1878 to the victory
of the black settlers™ True Whig Party. In British colonies, much con¬‚ict took
place within Protestant churches. Contests for church control had led Wesleyan
secessionists in Sierra Leone to set up Africa™s ¬rst modern independent Church
in 1821. Later schisms in Lagos extended to doctrine and discipline. When the
African Church Organization broke away from the Anglican Church in 1901,
for example, it admitted polygynists, expressing a dissatisfaction with Victorian
marriage that was shared by women who resented the economic dependence
to which it reduced them. A minority of elite members interested themselves
in local history and culture, adopted African names and dress, and sought to
move beyond eclecticism to some cultural synthesis. More were embittered by
restraints on their advancement in government service, an issue in Sierra Leone
since at least 1829 but accentuated in British colonies after 1880 when govern-
ments acquiring power in the interior insisted on keeping it in white hands. In
1883 Africans held 9 of the 43 higher posts in the Gold Coast administration;
in 1908, they held only 5 of 274.
Only the Liberian elite generally opposed the penetration of European power
inland. Horton, for example, served in the British expedition that destroyed
Kumasi. Penetration was in any case halting and gradual, except in Senegal
where Governor Faidherbe launched, in 1855, a deliberate conquest of the Wolof
kingdoms. Determination to stop the slave trade led the British to abandon
their long deference to coastal rulers, impose anti-slave-trade treaties from the
1830s, appoint a consul in 1849, create joint courts of Europeans and Africans in
coastal towns from 1850, and interfere increasingly in African politics, especially
in Yorubaland and the Niger Delta. This strategy culminated in the seizure of
Lagos in 1861, the same year as slow expansion from Freetown into Sherbro
country began. Con¬‚ict with Asante led to the enlargement of the Gold Coast
colony in 1874, but the withdrawal after destroying Kumasi in that year signalled
Britain™s anxiety to avoid inland commitments. In 1876, when West Africa stood
on the brink of rapid partition, nothing in the local situation dictated such a
dramatic break with the past.

Regional diversity in the nineteenth century

even where the atlantic slave trade had not compounded the
dif¬culties, underpopulation had retarded Africa™s development and obstruc-
ted attempts to overcome political segmentation by creating enduring states.
Between the ¬fteenth and nineteenth centuries almost every part of the
continent was drawn into a world economy dominated by Europe and
a political order dominated by the growing use of ¬rearms. These both
threatened African peoples and gave them new techniques and opportunities
to overcome segmentation, techniques that supplemented ancient strategies
and new devices of African invention. Ultimately most attempts to enlarge
the scale of economic organisation and political loyalty in nineteenth-century
Africa failed, partly because European aggression overwhelmed them, but also
because they did not meet the underlying problem of underpopulation, often
rather compounding it by the demands they placed on existing populations.
Beneath the surface, however, more profound changes took place. For the
¬rst time, certain regions escaped ancient constraints and embarked on rapid
population growth. Others, by contrast, experienced demographic stagnation
or decline comparable to Angola™s. This regional diversity “ the lack of an
overall continental trend “ was a major feature of nineteenth-century Africa
and makes it necessary to treat each region in turn: ¬rst the north, then the
Islamic west, the south, and ¬nally the east.

northern africa
The incorporation of North Africa (excluding Morocco) into the Ottoman
empire began in 1517, when Turkish musketeers defeated Egypt™s outdated
Mamluk cavalry in twenty minutes. Further west, Turkish privateers con-
tested the Maghribian coastline with local rulers and Iberian invaders until
an Ottoman force took Tunis in 1574 and made it a provincial capital, along
with Tripoli and Algiers. During the next two centuries, control from Istanbul
weakened to the advantage of provincial forces. In Egypt the army so overawed
the governors that they turned for support to the surviving Mamluks, whose
leaders (known as beys) regained predominance in the eighteenth century. In
Regional diversity in the nineteenth century 165

9. Regional diversity in the nineteenth century.
166 africans: the history of a continent

Tunis and Tripoli, where Ottoman garrisons were recruited from the soldiers™
children by local women, military commanders founded semi-independent
dynasties in Tunis in 1705 and Tripoli in 1711. In the frontier province of Algiers,
by contrast, the soldiers remained more alien, electing an of¬cer as dey and gov-
erning the hinterland by using favoured tribes to extract taxes from the others.
All these North African societies under Ottoman rule were segmentary, either
in the narrow sense that nomads belonged to autonomous tribes subdivided
into clans and lineages, or in the broader sense that society divided into peasant
villages or specialised, self-regulating, corporate groups like the 240 guilds and
100 wards of late eighteenth-century Cairo. Below the military noblemen and
merchants who dominated capital cities, most townsmen were small traders
and craftsmen “ Cairo™s average workshop had only three or four producers “
while the lowest work fell to day-labourers and the black slaves who consti-
tuted 4 or 5 percent of urban populations. Most countrymen were peasant
farmers controlling their own land, using a technology little changed since the
early Islamic centuries, and supporting the ruling class through taxation. Rich
peasants produced rice or sugar in Egypt or olive oil in the prosperous Sahel
of eastern Tunisia. Poor sharecroppers “ khamanisa, keeping only one-¬fth of
their produce “ cultivated estates on the coastal plains. Nomadic pastoralists
in the hinterland paid taxes only to military expeditions.
The Ottoman states enjoyed their greatest prosperity during the seventeenth
century. Tunisia remained prosperous during the eighteenth century and Egyp-
tians were probably as rich on average as Frenchmen in 1800, but by then the
whole region was falling under European economic predominance. Industry
was constrained by guild control, which prevented concentration of production
into larger units permitting mechanisation. Although late eighteenth-century
Egypt exported almost as much cloth as it imported, the imports came from
Europe, whereas the exports went largely to the rest of North Africa. Egypt
also took almost all its metals from Europe as well as most of its currency,
the shipping that conducted its Mediterranean trade, and some of the foreign
merchants who increasingly dominated its external commerce. ˜In a certain
measure,™ Andr´ Raymond has written, ˜it is already a commerce of “colo-
nial” type.™ Tunisia™s increasing dependence on olive-oil exports had similar
Behind this relative decline lay the stagnation of North Africa™s population.
In 1800 Egypt may have contained between four million and ¬ve million peo-
ple, probably equal to its Ptolemaic numbers. Tunisia probably had only one or
one-and-a-half million. Most estimates for Algeria suggest three million in 1830,
although some claim ¬ve million.2 The chief reason for demographic stagna-
tion was plague, which had remained endemic throughout North Africa since
the Black Death (chie¬‚y as an urban disease, thanks to regular contacts with
the central Ottoman lands), whereas plague had disappeared from northern
Regional diversity in the nineteenth century 167

Europe by the eighteenth century. Algeria suffered ¬ve plague epidemics dur-
ing the eighteenth century alone; Egypt and Tunisia, three. Contemporary
mortality ¬gures are unreliable, but the epidemic of 1784“5 was said to have
killed one of every three or six people in Tunisia. This was added to endemic
disease, the recurrent slaughter of infants by smallpox, and famine, although
famine was largely absent during the eighteenth century until its disastrous last
This picture of relative economic decline and political instability rooted
in recurrent demographic crisis was equally true of Morocco, which escaped
Ottoman control by adopting Ottoman military innovations. First, Ahmad
al-Mansur (1578“1603) used mercenary musketeers to free the country from
Turkish and Christian threats. Then Mawlay Ismail (1672“1727) created an
army of black slaves that temporarily reunited most of Morocco. After his
death, however, the army only added to the kingdom™s powerful segmentary
tendencies. Declining towns and foreign trade weakened the monarchy™s chief
supports. More than half the population were Berber highlanders who despised
the state™s largely unpaid administration and respected only military force or the
king™s spiritual authority. The monarchy™s religious status denied it the freedom
of action enjoyed by more secular Ottoman rulers, exposing it to denunciation
by the ulama (clerics) if it demanded non-Koranic taxes. Behind this frailty of
authority lay demographic weakness. Between 1500 and 1800, Morocco suffered
at least ten plague epidemics, which were most destructive in the towns and
surrounding plains where royal power centred. Famine was especially serious
during the seventeenth century, that of 1661 “2 in Fes belonging to the rare
category in which, as a chronicler wrote, ˜Those who died there among the
great and the rich “ and they were numerous “ died of hunger like the poor.™3
Morocco™s population in 1800 may have been three million or four million,
barely more than in 1500.4
North Africa™s demographic crisis culminated during the ¬rst third of the
nineteenth century. In 1818“20 the Maghrib suffered its last great plague epi-
demic. Egypt experienced one more, in 1835, thought to have killed 200,000 peo-
ple. Thereafter plague virtually disappeared, as mysteriously as it had already
deserted Europe. But other disasters intervened. Tunisia suffered an acute agrar-
ian crisis, largely due to drought, which roughly halved the cultivated area of
state domain between the mid-eighteenth century and the 1820s. In 1831 Egypt
suffered its ¬rst attack of Asiatic cholera, carried by pilgrims using the faster
transport now available from Mecca. It allegedly killed 150,000 Egyptians. Four
years later the same pandemic reached the Maghrib, the ¬rst of ¬ve to af¬‚ict
nineteenth-century North Africa. Although a hideous disease that killed about
half of those contracting it, cholera caused fewer deaths than plague, except
during the 1830s or when associated with famine. It probably hindered popu-
lation recovery rather than causing absolute decline. But it came immediately
168 africans: the history of a continent

after plague, with no interval for demographic growth such as Europe and
China had enjoyed.
Paradoxically, however, it was during this period of disaster that Egypt took
the ¬rst steps towards the rapid demographic growth that has dominated mod-
ern African history. Elements of a new Egypt emerged in the later eighteenth
century. Mamluk beys gained predominance over Ottoman forces, especially
under Ali Bey (1760“72), who made himself almost independent ruler of an
increasingly Egyptian state. Expanded trade with Europe contributed to eco-
nomic crisis but also encouraged commercial agriculture, weakened guild insti-
tutions, and bred a coalition of artisans and radical clerics who became an
important political force in Cairo. French invasion in 1798, designed to secure
grain supplies and threaten Britain™s position in India, stimulated Egyptian
patriotism, destroyed the Mamluk-dominated military and governmental sys-
tem, and provided new models of military and administrative organisation to
supplement those which the Ottoman Empire itself had begun to adopt during
the 1780s. The ¬rst bene¬ciary was Muhammad Ali (1805“48), an of¬cer in the
Ottoman army that recaptured Egypt in 1801. In the ensuing power struggle, he
supplanted the Ottoman governor by winning popular support in Cairo, only
subsequently to exclude these supporters from power. Muhammad Ali was an
Ottoman autocrat seeking to create a dynasty with maximum independence
from Istanbul. Illiterate to the age of 47, suspicious and superstitious, his pen-
etrating mind and Ottoman receptivity to military innovation convinced him
that a modern army needed not merely guns but a supporting industrial and
technical infrastructure. He recruited his army initially from black slaves and,
after 1823, from Egyptian peasants conscripted for life. It numbered 200,000
at its peak and cost 60 percent of state expenditure. To ¬nance it he abolished
all existing claims to revenue from land, collected tax directly from peasant
villages, and multiplied the proceeds six times during his ¬rst sixteen years.
From 1821 he compelled peasants to grow long-staple cotton as a taxable crop.
Irrigation works increased the cultivable area by 37 percent between 1805 and
1863. The state bought and sold all cotton, craft products, and many other
commodities. Muhammad Ali established industrial enterprises “ especially
textiles, shipbuilding, and armaments “ employing European technology, usu-
ally driven by animal power. Egypt™s spinning output per head became the
¬fth largest in the world.5 The need to import iron, coal, technology, and
skills was an obstacle to industrialisation. Assets were cheap cotton and labour,
cheap and ample food, excellent transport, and relatively high levels of average
wealth. To administer his programme, Muhammad Ali created a patrimonial
bureaucracy staffed by Turks in the upper ranks and Egyptians in the lower, all
holding military ranks. He founded primary and technical schools for a peak
of ten thousand students, sent some ¬ve hundred Egyptians for training in
Europe, but restricted education to state needs. This attempt to create the ¬rst
Regional diversity in the nineteenth century 169

industrial state outside Europe may never have been feasible in a country of
only ¬ve million people, compared with sixteen million in the United Kingdom
in 1801 and perhaps thirty million in Japan when it began to modernise in 1868.
But the question was never tested, for the British so feared that Muhammad Ali
might threaten their power in Asia that in 1838“41 they compelled him to reduce
his army to eighteen thousand men and abolish the commercial monopolies
that excluded European manufactures from Egyptian markets. Unprotected,
Egypt™s industries collapsed. By 1849 only two factories remained. The country
became more exclusively agricultural and dependent than it had been in the
eighteenth century.
Yet one lasting achievement of Muhammad Ali™s reign was demographic
growth, which has continued to this day. In 1800 Egypt probably had between
four million and ¬ve million people. A census in 1897 showed 9,734,405.6 The
initial reason for growth was probably the disappearance of plague after 1835.
Cholera caused less mortality. The government™s main contributions were to
foster vaccination against smallpox “ by 1850 Egypt had more than 2,500 barber-
vaccinators immunising 80,000 children a year “ and to create, from 1836, a
provincial health service. Other contributions were Muhammad Ali™s success
in ending internal warfare, the expansion of permanent irrigation, and perhaps
the adoption of maize during the eighteenth century. Population growth was
due to lower death rates, with no known changes in fertility. It is noteworthy
that Africa™s modern population growth began in its ¬rst state with a modern
Under Muhammad Ali™s descendants, Egypt was wide open to European
economic penetration. By 1876 more than 100,000 Europeans lived there. They
pro¬ted from the cotton crop, which multiplied roughly ten times between
Muhammad Ali™s death and the early 1880s, occupying enough land to cause
Egypt to import grain from 1864. To transport the cotton, some 1,750 kilome-
tres of railway were built between 1852 and 1879. The Suez Canal opened to
shipping in 1869. A liberal economic strategy fostered private land-ownership.
By 1884 the royal family, notables, of¬cials, and Europeans owned some 48
percent of cultivated land, although even large estates were generally farmed in
small tenant plots. A rural elite of village sheikhs and notables expanded, as did
a landless class whose availability for urban labour ended slavery. Cairo™s pop-
ulation doubled between 1850 and 1900. Plans to transform the city on Parisian
lines were largely frustrated, but the new palaces screening urban poverty sym-
bolised the extravagant, elitist modernisation of the Khedive Ismail (1863“79).
The government of this Turkish autocrat was as patrimonial as his grandfather™s
and his Council of Deputies, created in 1866, was chie¬‚y designed to impress
Europeans. Beneath this surface, however, a political class of Arabic-speaking
landlords, clerics, and western-educated of¬cials and of¬cers gathered in¬‚u-
ence and made Cairo the centre of an Egyptian Enlightenment, as oppressive
170 africans: the history of a continent

to the peasantry as the Enlightenment in Europe but comparable in its in¬‚u-
ence on the Islamic world. Its chief embodiment was the Sala¬yya or Mod-
ernist Movement, centred on the Azhar mosque-university, with ten thousand
students and a teaching staff led by Muhammad Abduh (1849“1905), modern
Africa™s most important intellectual. Abduh taught that the way to revitalise the
Islamic world against Western aggression was to restore the pristine, supremely
rational Islam of the early Caliphate so that it could blend again harmoniously
with science and technology. Too elitist a doctrine to win mass enthusiasm, it
nevertheless inspired young intellectuals throughout North Africa. Listening to
Muhammad Abduh, one wrote, ˜We felt in our souls that any of us was capable
of reforming a province or a kingdom.™7
Unlike his grandfather, Ismail ¬nanced his modernisation partly by foreign
borrowing. By 1876 the of¬cial debt was £91 million and an international
commission took control of Egypt™s ¬nances. When Ismail mobilised his of¬cer
corps and Chamber of Deputies against foreign interference, the Ottoman
Sultan deposed him in 1879 at European behest. The attempt to modernise
a segmentary society had led to the verge of colonial invasion. But Egypt™s
example already reverberated throughout North Africa.
As always when Egypt was strong, the Sudan suffered. In 1820 Muhammad
Ali invaded it in search of slaves for his army. Egypt™s slaves had hitherto come
mainly from the southwest, where the cavalry of the Darfur kingdom raided
the savanna agriculturalists to the south and southwest in Dar Fertit (the land
of slaves), but Muhammad Ali judged this source inadequate. ˜The end of all
our effort and this expense is to procure Negroes™, he told his commander
in the Sudan.8 His troops overcame the warlords dominating Lower Nubia,
received the submission of the sedentary Arab states that had replaced ancient
Christian kingdoms, and took the Funj sultanate™s capital at Sennar on the Blue
Nile. Some thirty thousand Sudanese slaves were conscripted, but when they
died like ¬‚ies in Egypt, Muhammad Ali instead sought pro¬t by compelling
Sudanese to grow cotton. Sudanese traders, in turn, penetrated southwards
through the Shilluk kingdom to the Dinka of the Upper Nile, seeking ¬rst ivory
and then, from the 1860s, slaves for Egyptian cotton farms and the wider Islamic
world. The stateless pastoralists of southern Sudan had little interest in trade,
no chiefs to be seduced, and no guns to resist with. The traders therefore built
fortresses, raided for slaves with modern ¬rearms, and created anarchy and
depopulation remembered by Dinka as the time when ˜the earth was spoilt™.
By the mid-1860s Khartoum traders had intervened in succession wars as far
south as Bunyoro and the prosperous Mangbetu kingdom in the northeast
of modern Cougo, which distintegrated after its king died in battle against
Khartoum-backed enemies in 1873. Three years later, Egyptian troops reached
Lake Victoria. But in 1877 Ismail™s ¬nancial weakness enabled Europeans to
coerce him into appointing an Englishman, General Gordon, as governor of
Regional diversity in the nineteenth century 171

the Sudan to attack the slave trade. As in Egypt itself, European intervention
was deepening.
Muhammad Ali™s transformation of Egypt also affected Ethiopia. After the
repulse of Muslim invasion in 1543, the Christian kingdom was not fully restored
because warfare had allowed the stateless, Cushitic-speaking Oromo people
to in¬ltrate the Christian highlands from the south. Amhara settlement, in
response, edged northwards and westwards, a tendency crystallised in 1636 by
the establishment of a permanent capital at Gondar. In Ethiopia, segmentation
meant regionalism, especially during the ˜Era of the Judges™ between 1769 and
1855, when provincial warlords seeking to control powerless emperors reduced
the throne to ˜a worthless ¬‚ower that children pluck in the autumn rains™.
Border provinces pro¬ted especially at this time from long-distance trade in
slaves and ¬rearms. Tigray in the north acquired the largest arsenal. Shoa in
the south reconquered territory from the Oromo and created Ethiopia™s ¬rst
nascent bureaucracy. But the revival of central power was initiated by Tewodros,
a district governor in the western lowlands, who was defeated by Muhammad
Ali™s new army in 1848, sought to imitate its discipline and ¬repower, and
fought his way to the throne in 1855. The expense of his attempts to consolidate
central power by creating an armaments industry and replacing regional war-
lords by appointed governors led him in 1860 to con¬scate church lands. The
clergy responded by backing regionalism. In 1868, when his authority scarcely
extended beyond his fortress at Magdala, Tewodros shot himself as a British
expedition approached to punish him for maltreating their consul. Central
power passed to the ruler of Tigray, Yohannes IV (1872“89), who had aided
the British but now resumed the attempt to reunite Ethiopia, both by mili-
tary force and by such traditional methods of diplomacy as a marriage alliance
with Menelik of Shoa. When Yohannes died in battle with Muslim forces from
the Sudan, Menelik (1889“1913) succeeded peacefully and launched a cautious
programme of modernisation, introducing Shoa™s taxation system and bureau-


. 5
( 12)