. 7
( 12)


direct taxation was unnecessary and administration could be con¬ned to a
handful of white of¬cials seeking to guide African rulers gradually towards
European notions of good government, as had long been the approach of the
Cape administration in South Africa.
More commonly, however, early colonial administrators were military of¬-
cers who saw Africans as security risks. This was especially the tradition of
French of¬cers trained in Algeria. When Colonel Archinard took Segu in 1890,
for example, he deported 20,000 of its Tukulor rulers back to Senegal, installed
a chief from an indigenous and friendly Bambara dynasty, distrusted his loy-
alty, summarily executed him, appointed a rival Bambara candidate, and ¬nally
abolished the chieftainship and established direct French administration, all
within three years. From this security perspective, any powerful African was
dangerous. ˜We must look on all these chiefs as people to be ruined™, Archi-
nard™s mentor advised him. British of¬cers took the same view in Sudan, where
their initial hostility to vestiges of Mahdism verged on paranoia, and white
settler regimes took it everywhere. As this style of administration settled into
civilian routine around the turn of the century, it became known as direct rule
and was practised especially by French, Belgian, and Portuguese of¬cials who
believed in centralisation and saw hereditary rulers as ˜usually nothing but par-
asites™, in the words of William Ponty, governor-general of French West Africa.
French and Belgian of¬cials were more numerous than British; in 1926 Cˆ te o
d™Ivoire had one white of¬cial for every eighteen thousand people, Southern
Nigeria one for every seventy thousand. Frenchmen headed the cercles and
subdivisions into which the two federations of West and Equatorial Africa
Colonial invasion 207

were divided. Below them African chiefs administered cantons (often the old
kafu units) and villages. Direct rule forced African political systems into this
framework. Monarchs gave way to chefs de canton in Dahomey and Futa Jalon,
while the Mossi Mogho Naba (Lord of the World) was deliberately deprived of
power. Stateless peoples were subordinated to appointed chefs de canton, drawn
by preference from the local population but chosen mainly for their loyalty,
literacy, and ef¬ciency “ retired soldiers were often favoured. Chefs de canton,
in turn, relied on village chiefs, who were normally local men. The early Bel-
gian administration also levelled chiefs downwards and upwards in this way to
form the base of a bureaucratic pyramid. Thus even ˜direct rule™ was in practice
rule through Africans; the question was the level in the indigenous society at
which the link with the colonial bureaucracy was made and the contradictions
of colonial rule were therefore most acute.
That was the originality of ˜Indirect Rule™, which the British devised in the
Sokoto Caliphate (Northern Nigeria) before the First World War and then
extended to other colonies. Frederick Lugard, the soldier who conquered the
Caliphate and devised the system, brought with him the Indian Army™s hatred
of ˜the politically-minded Indian™, which he translated into a loathing of the
˜Europeanised Africans™ of Southern Nigeria. He wanted a more authoritarian
administration and he realised that the Fulbe emirs and their relatively sophisti-
cated institutions could serve his purpose, ˜for they are born rulers, and incom-
parably above the negroid tribes in ability™. Lugard™s unusually strong military
forces enabled him to defeat and replace ruling emirs without destroying their
administrations. ˜Every Sultan and Emir™, he proclaimed after taking Sokoto, ˜. . .
will rule over the people as of old time . . . but will obey the laws of the Governor
and will act in accordance with the advice of the Resident.™9 The Caliphate was
abolished as a political unit and each emir headed a distinct native adminis-
tration with powers of subordinate legislation, jurisdiction, and tax collection,
remitting part to the British authorities. Unlike the Kabaka of Buganda under
the Uganda Agreement of 1900, emirs had no entrenched position but ruled
purely by British favour. Many were replaced, but the Fulbe ruling class retained
power, at the cost of much oppression “ the early colonial period was known
to Hausa as ˜the tearing asunder™ “ and eventually of much stagnation.
Lugard devised Indirect Rule for the unique circumstances of Northern Nige-
ria, but he convinced himself and others that its principles could generally give
more orderly administration than the ad hoc arrangements of conquest. Yet
few African societies had possessed administrative institutions like Sokoto™s,
while conquest had often destroyed such institutions as had existed. Outside
Northern Nigeria, therefore, Indirect Rule meant rediscovering or inventing
institutions to ¬t the structure of native administrations, courts, and treasuries.
It thereby became a new idiom for African political competition. When Lugard
himself took over Southern Nigeria in 1912 in order to amalgamate it with the
208 africans: the history of a continent

north, for example, his of¬cials invented an imaginary eighteenth century by
subordinating Ibadan to a restored ˜Oyo Empire™ in which an ambitious Ala¬n
dominated a purely ceremonial Oyo Mesi and exercised novel powers of direct
taxation. In southeastern Nigeria™s stateless societies, the consequences were
even more disruptive, culminating in the ˜Women™s War™ of 1929 when Igbo
women who believed that they were to be taxed attacked chiefs, courts, and
European trading posts until repression cost ¬fty-three lives. Tanganyika (for-
merly German East Africa) adopted Indirect Rule in 1925 and applied it to state-
less peoples by creating councils of headmen, but it had dif¬culty in discovering
what institutions had existed before German conquest. There, as everywhere,
the units placed under native administrations came to be seen and to see them-
selves as ˜tribes™. During the 1930s, Indirect Rule spread to Nyasaland and North-
ern Rhodesia, where it replaced more direct administration through headmen,
and to Basutoland, Bechuanaland, and Swaziland, where the British used it to
reduce chie¬‚y power. The policy™s conservative thrust was strong. In Sudan,
for example, the Egyptian and Sudanese elites, who were initially employed
for their anti-Mahdist sympathies, were abandoned after 1924 when an army
mutiny revealed the ¬rst glimpses of Sudanese nationalism. Instead, the British
adopted ˜Indirect Rule™ and rehabilitated ˜tribal chiefs™ in a policy described by
the governor as ˜making the Sudan safe for autocracy™, one step being to exclude
all northern Islamic in¬‚uence from the non-Islamic south. In the Gold Coast,
similarly, the restored Asante Confederacy of 1935 lacked the bureaucracy that
had balanced hereditary chieftainship. The confederacy gave early priority to
abolishing ˜youngmen™s associations™ for their ˜unwarranted militancy™.
Not all British territories adopted Indirect Rule. White settlers thought it
made chiefs too powerful and obstructed the labour supply. Southern Rhode-
sia used chiefs purely as government agents and refused demands to restore
the Ndebele kingship. Kenya created administrative chiefs but from 1924 also
established district-level local native councils whose partly elected composi-
tion stimulated the liveliest rural politics in tropical Africa. Generally, however,
Indirect Rule not only became the distinctive pattern of British administration
but even in¬‚uenced other colonial governments, despite their suspicion that
it was a typical piece of British indolence. When the Belgians took Rwanda
and Burundi from German control after the First World War, they governed
through Tutsi monarchies, although rationalising them into unrecognisably
neat administrative pyramids. The Mogho Naba, hitherto ignored, gradually
became a key French ally after demonstrating his usefulness in recruiting troops
during the First World War and making his people grow cotton. In 1917 the
governor-general of French West Africa urged the recruitment of chiefs pos-
sessing true authority over their peoples, but adding that the chief must remain
˜our instrument™.10 That remained French policy even in Morocco and Tunisia,
where of¬cials protected indigenous rulers in theory but blatantly exploited
Colonial invasion 209

their prestige in practice. ˜There is in every society™, Morocco™s ¬rst French
governor had declared, ˜a ruling class born to rule. . . . Get it on our side.™11
His compatriots had been slow to heed, but by the 1930s of¬cials everywhere
in Africa shared his approach.

early colonial economies
A crucial issue for each colony was whether its economy was to rest chie¬‚y
on peasant production, European farms or plantations, mining, or some com-
bination of these. Although there were a few later changes of direction, most
colonies retained throughout their history the economic trajectory acquired
before the First World War. It was seldom due to deliberate planning, for
most European governments left economic development to private enterprise,
themselves contributing only infrastructure, a legal system, and an appetite for
taxation that drove their subjects into the cash economy. The outcome was that
each colony was integrated into the international economy as a specialised pro-
ducer of commodities for which it had some natural advantage, at a generally
prosperous period for commodity producers. The consequences then rami¬ed
through the rest of the economy. The amount of restructuring involved and its
human costs varied from one colony to another.
The process was least traumatic in colonies already integrated into interna-
tional trade. One was Egypt, where in 1879, three years before British invasion,
the main export crop, cotton, already occupied 12 percent of the cultivated area.
By 1913 it occupied 22 percent, for British rule only accentuated Egypt™s previ-
ous development trajectory, resulting in marked prosperity “ per capita income
rose by nearly 50 percent12 “ but also marked differentiation, for by 1917 the
proportion of rural families owning no land was 53 percent in Upper Egypt and
36 percent in the Delta. In tropical Africa, the West African coast was the region
hitherto most fully integrated into the world economy, giving it an advantage
throughout the colonial period, especially those areas exporting tropical crops,
which sometimes enjoyed spectacular expansion. Senegal™s groundnut exports
multiplied ten times between the 1880s and the First World War. Production,
hitherto con¬ned to the coast, spread inland along a railway built towards the
Niger, which reduced freight costs by over 95 percent. The main new produc-
ers were Wolof, transferring their energies from warfare to colonising the dry
inland plains. Their ruthless exploitation of virgin bush was often organised
by marabouts of the Mouride brotherhood, a local branch of the Qadiriyya
created in the 1880s by Amadou Bamba, which reintegrated the military ethos
of the ceddo and the materialism of the peasantry into disciplined pioneering
communities. As Governor-General Ponty remarked, ˜The blacks . . . make
perfect settlers.™ Even more spectacular was the colonisation of the eastern,
yam-producing sector of the West African forest to produce cocoa, a new tree
210 africans: the history of a continent

crop, which took ¬ve years to bear fruit. The pioneers were the Akwapim people
of the southern Gold Coast, who ¬rst learned of cocoa from the enterprising
Basel Mission and its West Indian agents. From the 1880s, Akwapim migrated
into the neighbouring Densu Forest to buy and clear land for cocoa farms,
sometimes packing their produce into casks and rolling them along the paths
to the coast. Many entrepreneurs were traders, teachers, or clergymen, employ-
ing paid labourers. As a railway from the coast reached Kumasi in 1903, cocoa
production also spread rapidly in Asante, where many pioneers were chiefs
who controlled stool wealth and the labour of slaves and other dependents. By
1911 there may have been twenty-¬ve hundred square kilometres of cocoa in
the Gold Coast and it was the world™s leading exporter. Another pioneer area
was Yorubaland, where unsuccessful African traders from Lagos established
cocoa plantations on the mainland, whence they spread inland. The success of
African commodity production in this region ensured, as the governor of Lagos
put it in 1901, ˜that the future development of this country must be by its own
people, through its own people, and for its own people™.13 Six years later the
Southern Nigerian government rejected an application from Lever Brothers to
establish palm-oil plantations. Instead the ¬rm obtained land in the Belgian
Equatorial Africa had not developed agricultural exports before colonial rule.
Its population was sparse and its physical and climatic environment made rail-
way construction especially dif¬cult. Its main ruler, King Leopold II, lacked the
¬nancial resources available to a state, while the neighbouring French regime
was notoriously short of men and money. For all these reasons, the equatorial
region experienced the most brutal exploitation of the early colonial period.
Leopold™s solution to his ¬nancial dilemma was for the Congo Independent
State itself to trade, with all the violence its military force permitted. By chance,
this decision coincided with increased international demand for rubber, as bicy-
cles and then motor cars multiplied. Between 1890 and 1910, the world price
of rubber more than trebled and tropical Africans in the thousands scoured
forests for wild sources. This ˜time of hot money™, as it was known in Yorubaland,
pro¬ted many, but not in the Congo, as a British Consul was told in 1899:

His method of procedure was to arrive in canoes at a village, the inhabitants
of which invariably bolted on their arrival; the soldiers were then landed, and
commenced looting, taking all the chickens, grain, etc., out of the houses; after
this they attacked the natives until able to seize their women; these women
were kept as hostages until the Chief of the district brought in the required
number of kilogrammes of rubber.14

In 1908, following international outcry, Belgium took over the territory and
established a less brutal but still authoritarian regime. Leopold™s other fund-
raising technique “ to lease vast regions to private companies “ was extended
Colonial invasion 211

in the late 1890s to French Equatorial Africa, where forty concessionary ¬rms
pillaged the region for a generation. Similar concessions occupied much of
Cameroun and Mozambique.
In contrast to this Raubwirtschaft or the expansion of coastal commodity pro-
duction, the key to economic development elsewhere was railway-building, the
period™s main achievement and a crucial means of escape from Africa™s vicious
circle of underpopulation and inadequate transport. By the 1880s Egypt, Alge-
ria, and the Cape Colony already had railway networks. New lines reached
Bulawayo (from South Africa) in 1897, Lake Victoria (from Mombasa) in 1901,
the Niger (from Senegal) in 1905, Kano (from Lagos) in 1912, and Lake Tan-
ganyika (from Dar es Salaam) in 1914. Governments built these new lines chie¬‚y
for strategic reasons, but their economic impact was even more profound, for
they frequently cut transport costs by 90 to 95 percent, restructured trading sys-
tems, released labour, and provided outlets for inland commodity production,
thereby creating distinctively colonial economies.
Many ancient trading systems collapsed when faced with railway competi-
tion. The trans-Saharan trade from Hausaland to Tripoli, which had ¬‚ourished
throughout the nineteenth century, declined as the railhead approached Kano
from the south, impelling the Tuareg desert traders into their desperate revolt
of 1916. As construction neared Lake Tanganyika, the caravan porters leaving
Bagamoyo for the East African interior declined between 1900 and 1912 from
43,880 to 193. Dar es Salaam, Mombasa, Dakar, and Conakry ¬‚ourished while
ports bypassed by railways sank into insigni¬cance. Railway towns like Bouak´ e
and Bamako supplanted the more romantic Kong and Timbuktu, although
other old centres like Kano and Kumasi drew new life from steam. In west-
ern and central Africa, railway transport ¬rst enabled European trading ¬rms
and their African or Lebanese agents to penetrate deeply into the continent,
establishing the ´conomie de traite “ the exchange of imported manufactures for
locally grown commodities “ as the dominant economic pattern. In East Africa,
the new traders were Asians; the most successful, Alidina Visram, owned some
240 stores when he died in 1916. African traders were not necessarily impov-
erished. In West Africa, they were gradually squeezed out of export“import
trade but found new opportunities in inland commerce, much stimulated by
colonial abolition of internal tolls and other restraints. ˜We turned ourselves
from warriors into Merchants, Traders, Christians and men of properties, keep
moneys in the Banks under British Protection and began to build huge houses™,
Asante™s nouveaux riches (akonkofo) recalled.15 In equatorial and eastern Africa,
by contrast, little of the nineteenth-century trading system except its lines of
communication survived into the colonial period.
Railways ¬rst made possible the large-scale exploitation of Africa™s chief
economic asset, its minerals. The arrival of railways at the turn of the cen-
tury made possible large-scale goldmining in Southern Rhodesia and Asante,
212 africans: the history of a continent

although neither became a second Witwatersrand. Diamond discoveries gave
new vitality to Southwest Africa from 1908 and Angola from 1912. Less spectac-
ularly, Tunisia became Europe™s chief supplier of phosphates, until challenged
by Morocco. Railway development enabled central Nigeria to become a signif-
icant tin producer during the First World War. But the main development was
the beginning of copper production in Katanga in 1911, following the arrival
of a railway from the south the year before. As an indication of mining™s pre-
dominance, of the estimated £1,222 million invested in sub-Saharan Africa by
1938, £555 million had gone to South Africa, £143 million to the Belgian Congo,
£102 million to the Rhodesias, and only £422 million to the remainder, much
of that as railway investment.
Railway transport was also vital to the establishment of white agriculture
in the highlands of eastern and southern Africa, which proved as suitable for
Europeans as for cows. In Southern Rhodesia, the white settlers of Rhodes™s Pio-
neer Column conquered the colony and seized one-sixth of its land during the
1890s, mostly on the central highveld and including almost the entire Ndebele
kingdom, together with most of its cattle. From about 1908, when hopes of gold
dimmed, settlers began serious grain farming and experimented with tobacco.
Eventually they controlled half the colony™s land. Extensive white settlement in
Northern Rhodesia began only when a railway was built through its centre to
Katanga and land along the line was reserved for Europeans. Swaziland, south-
ern Nyasaland, and Southwest Africa all became regions of European settlement
at this time. White colonisation of the Kenya highlands, designed to make the
strategic railway to Lake Victoria pay, eventually took some 18 percent of the
colony™s best agricultural land. Sisal plantations and private settlers in German
East Africa clustered along the railway in the northeast. North Africa was the
other region of white settlement. European landholdings in Algeria doubled
between 1881 and 1921, driving Muslims on to marginal land. In Tunisia in 1914,
Europeans owned 920,000 hectares, about half of it in the ancient grain-growing
areas of the north. Morocco, too, was seen as a potential granary and Euro-
pean farming was encouraged on the Atlantic Plains. Yet these ¬gures were
partly misleading, for in several settler colonies before the First World War,
African and Arab farming also ¬‚ourished as white settlement and transport
improvements created markets. For the Shona of Southern Rhodesia and the
Kikuyu of Kenya, the 1900s were a decade of prosperity and expansion. In 1913
Africans produced at least three-quarters of Kenya™s exports. Even the output
from European estates was often grown by African or Arab smallholders, for as
in Roman times many white farmers accumulated capital by exploiting indige-
nous sharecroppers or labour tenants, whether khamanisa in North Africa or
˜squatters™ in the east.
Peasant farming had victories even outside West Africa, for Luba cultivators
defeated Katanga™s white settlers in competition for the food market in copper
Colonial invasion 213

towns, while peasant cotton growers drove European planters out of business
in Nyasaland and Uganda. Like Nigeria or the Gold Coast, Uganda remained a
black man™s country because its peasants became commodity producers sup-
plying European merchants and supporting the colonial state through taxation,
growing cotton from 1903 in response to the railway™s arrival at Lake Victoria
two years earlier. Similarly, when the railhead reached Kano in 1912, it stimu-
lated a ¬‚urry of groundnut production and marketing among Hausa peasants.
Here the entrepreneurs were mainly former kola traders. Elsewhere they were
often educated Christians, as among early African coffee-growers in German
East Africa, or political leaders, notably the Ganda chiefs who coerced their
tenants into growing cotton in order to pay rent and tax.
Such control of labour was crucial in early colonial Africa, not least because
Africans had long relied heavily on slavery, which all colonial regimes were
committed to abolish. In practice they approached this cautiously. They knew
that rapid emancipation, as in the Cape Colony and Senegal, had been expen-
sive in compensation and had caused temporary economic crises followed by
new forms of dependency. In India, however, the British had abolished slavery
gradually, without compensation or dislocation (although also without abol-
ishing dependency), by ¬rst banning the slave trade and then declaring that
slavery had no legal status, leaving the slave to assert his freedom if and when
he wished. This policy was adopted in the Gold Coast in 1874 and generally
thereafter in Africa, although with much local variation. In Zanzibar, for exam-
ple, the whole economy depended on slavery and British rule rested on alliance
with Arab slave-owners; when the legal status of slavery was abolished in 1897,
therefore, vagrancy legislation obliged most freed slaves to remain on the plan-
tations, where they became labour-tenants. In Northern Nigeria, a major slave
society, Lugard™s anxiety to preserve the indigenous hierarchy led him to qualify
the abolition of the legal status by requiring male slaves to purchase their free-
dom, preventing them from acquiring independent land rights, and returning
them to their masters if they ¬‚ed, so that slavery only gradually disappeared
and did not become illegal until 1936. The British authorities in Sudan were
equally cautious. The military regime in French West Africa itself relied ini-
tially on slave labour, but France abolished the legal status in 1903 and slaves
quickly seized their freedom, perhaps because many were recent captives, often
Samori™s victims. Tens and perhaps hundreds of thousands quit their masters,
despite French attempts to stop them. Some colonial regimes initially freed their
enemies™ slaves, as in Ijebu and Dahomey. In other regions, slavery or something
similar lingered into the later colonial period and beyond, notably in Mauri-
tania and Botswana where master classes retained political power. Elsewhere
emancipation threw some African aristocracies into crisis. Arab planters on the
Kenya coast were ruined, some Gola owners in Liberia are said to have killed
themselves, and literati complained that ˜a mob of serfs and servants vaunt
214 africans: the history of a continent

themselves above their masters™. Yet generally emancipation went remarkably
smoothly. One reason was that it was carried out by alien and largely disin-
terested rulers. Another was that many slaves remained with their masters and
gradually acquired higher but still dependent status, in effect accelerating a
process indigenous to lineage slavery. This applied especially to the majority of
slaves who were women, for they often had few alternative means of livelihood
and were treated by male colonial authorities as concubines: ˜rather a question
of divorce than of slavery™, wrote Lugard. Slavery often gave way to pawnship,
an ancient form of dependence in which the pawn™s labour passed to a creditor
in lieu of interest until a debt was repaid; in 1909 Ibadan City alone was said to
contain 10,000 pawns and the institution survived far into the colonial period.
A further reason for smooth emancipation was that, except in Northern Nige-
ria, land was generally available for slaves to become peasants; Mourides, for
example, accepted slaves into colonising communities as genuine equals. And
the eager demand for male labour gave many slaves wage employment.
Land alienation created one source of labour, emancipation another, coer-
cion a third. The fourth, and perhaps the most rapidly increasing, was labour
migration, often compelled by taxation, from regions with no access to railway
transport by which to export produce. Migrant labour was no novelty. Much
unskilled work throughout early colonial West Africa was performed by Kru
from the Liberian coast who had served Europeans as seamen and dockwork-
ers since the eighteenth century, while most labourers on German plantations
in East Africa came from regions that had supplied porters for nineteenth-
century caravans. But many peoples entered the market for the ¬rst time.
Mossi migration to ˜England™, as they called the Gold Coast, began with tax-
ation in 1896, although, like other migrants, they also sought wages to buy
imported goods and invest in their domestic economy. Lozi migrants from the
Zambezi ¬‚oodplain, often former slaves, sought work shortly afterwards and
for the same reasons. Both regions had been famed in the nineteenth century
for agricultural prosperity but now decayed into rural backwaters, especially
Bulozi whose irrigation system collapsed without slave labour to maintain it.
This experience illustrated the unevenness of the early colonial impact. Just
as nineteenth-century East Africa had been restructured around Zanzibar and
its trade, so colonial Africa (except perhaps Egypt) was restructured around
new growth points: towns, mines, European estates, and African cash-crop
farms. Encircling these were zones supplying food, themselves surrounded by
remote regions exporting migrant workers or livestock. Most growth points
were in forested or highland areas, while labour reservoirs were mainly in
the savanna, reversing the main trends of earlier African history. For growth
points, the voracious colonisation of land had future costs, but the early colo-
nial period was prosperous. Labour reservoirs, by contrast, suffered crippling
Colonial invasion 215

environment and demography
Early colonial Africa did not experience a wholesale demographic catastrophe
on the scale that conquest and imported diseases brought to Latin America
and the Paci¬c. Not only were Africans already adapted to the world™s most
hostile disease environment, but as Old World populations they, like Asians,
already had some resistance to European maladies. Smallpox, for example, was
probably more destructive in West Africa during the slave trade than in the
early colonial period. Yet certain regions had experienced less contact with
Europe, while others were especially vulnerable owing to the nature of their
environments or of the colonial intrusion. Such regional diversity made the
early colonial period a demographic crisis, but a muted crisis. The African
peoples once more survived.
˜Wars, drought, famine, pestilence, locusts, cattle-plague! Why so many
calamities in succession? Why?™16 The missionary Francois Coillard™s anguish
in Bulozi in 1896 was widely shared. Military conquest itself was probably
not the greatest killer, but it had devastating regional effects. The twenty-one
years of intermittent warfare through which the Italians conquered Libya may
have killed one-third of its people.17 In 1904 the Germans suppressed a Herero
rebellion in Southwest Africa by driving the people into the Omaheke desert;
a census in 1911 showed only 15,130 surviving Herero, of perhaps 80,000.18 The
repression and famine that defeated the Maji Maji rebellion not only killed
up to one-third of the region™s population but ˜reduced the average fertility of
the surviving women by over 25 per cent™, according to a professional study
thirty years later.19 The ¬ghting between British, Belgians, Germans, and Por-
tuguese in East Africa during the First World War was a comparable disaster,
for it exposed well over 100,000 African troops and perhaps a million porters
and labourers to appalling deathrates from disease and exhaustion. These were
exceptional horrors, but they indicate that colonial violence could have sig-
ni¬cant demographic effects, although colonial conquest also ended much
violence. When the British took control of Nyasaland™s Shire Valley in 1891,
men sang:

Child of the baboon
Come down from the hills
The country is on its feet again.20

On a continental scale, violence was less destructive than famine. Through-
out the tropical African savanna, the favourable rains of the mid-nineteenth
century faltered during the 1880s, inaugurating forty years of relative arid-
ity before rainfall recovered during the 1920s. (This is supported by evidence
of lake levels but neglects much local variation.) Frequently interacting with
other aspects of the early colonial crisis, drought bred a sequence of famines
216 africans: the history of a continent

throughout the savanna. It began in East Africa during the 1880s and reached
its ¬rst crisis in Ethiopia in the famine of 1888“92 known as Awful Day, when
cattle plague killed the plough-oxen and dislocated the agricultural system.
Exacerbated by drought, locusts, violence, and human disease, this crisis was
said to have killed one-third of Ethiopia™s people, although the ¬gure was con-
ventional. Many areas of Sudan suffered equally. In 1896 the northern Transvaal
experienced South Africa™s last major famine that killed. Eastern Africa™s ˜Great
Famine™ of 1898“1900, as Kikuyu called it, was exacerbated by colonial food
purchases and was later found to have killed two-thirds of the people in one
Kikuyu mbari. The West African savanna experienced its crisis in 1913“14,
when an exceptional drought coincided with new tax systems, crop exports,
declining trans-Saharan commerce, and labour migration. The First World War
campaign created widespread famine in East Africa, especially the devastating
Rumanura famine of 1916“18 in Rwanda. Finally, French Equatorial Africa
experienced the worst famine in its history from 1918 to 1926, owing chie¬‚y to
excessive colonial demands for food and labour. As usual, these famines killed
chie¬‚y through disease, especially smallpox among those clustering together
for food and water. East Africa suffered a terrible epidemic in the late 1890s,
when a new, possibly Asian strain may have entered the region. Smallpox, like
other diseases, affected the increased numbers of people moving into unfa-
miliar environments, expecially as labourers, and it remained a regular killer
of children. A doctor in Nyasaland recorded that 93 percent of adults and
68 percent of children he examined in 1913 had suffered the disease.
The worst medical crisis was the sleeping sickness epidemic, which shared
many features with AIDS a century later. Sleeping sickness was caused by try-
panosomes, microorganisms that attacked the central nervous system, lead-
ing from fever and lassitude to coma and death. In West Africa the disease
was ancient, chronic, and chie¬‚y transmitted from man to man by a tsetse
¬‚y, Glossina palpalis, living close to water. Whether human sleeping sickness
also existed in eastern Africa before the 1890s is uncertain; no contemporary
observer recorded it, but Africans questioned after 1900 reported one ear-
lier epidemic on the southeastern shore of Lake Victoria and occasional cases
elsewhere,21 possibly transmitted by the dryland tsetse ¬‚y, Glossina morsitans,
which certainly infected game and cattle. The early colonial epidemic appears
to have been caused by two processes whose connections are unclear. The ¬rst
was rapid expansion of sleeping sickness in western Africa, possibly because
increased mobility exposed more people to it. Local epidemics took place on
the coast between Senegal and Angola from the 1860s. That on the lower Congo
spread up the equatorial river system along with waterborne traf¬c, probably
linking together existing local pockets of the disease. The epidemic reached
Lake Tanganyika by 1901 and killed up to 90 percent of the population in the
worst-affected localities of equatorial Africa.22 Some think it was also carried
Colonial invasion 217

across the Congo“Nile watershed to Lake Victoria, where epidemic sleeping
sickness appeared during the mid-1890s. Others think this epidemic was of
local origin. By 1905 it had killed over 200,000 people on the northern lakeshore.
The epidemic also spread down Lake Tanganyika into Central Africa, reach-
ing Northern Rhodesia by 1907. Here a second process intervened. The cattle
plague that had caused the Ethiopian famine of 1888 had spread throughout
sub-Saharan Africa during the 1890s, often killing 90 percent of cattle, pau-
perising their owners, and slaughtering game on which Glossina morsitans had
fed. This allowed bush harbouring tsetse ¬‚ies to reconquer pasture and threaten
cultivation, especially where violence, famine, land alienation, or labour migra-
tion reduced human populations. When game numbers recovered before the
First World War, therefore, they occupied larger bush areas and tsetse ¬‚ies
transmitted sleeping sickness more extensively to human beings. Numerous
local epidemics in savanna areas resulted, while vast areas hitherto occupied by
men and cattle were closed to human settlement. One estimate is that the early
colonial epidemic as a whole may have killed nearly a million people.23 The
only remedies for sleeping sickness before the First World War were almost as
deadly as the disease.
Colonial doctors were also alarmed by an apparent spread of venereal dis-
eases, especially syphilis, which was reported in 1908 to infect 80 percent of
Ganda. Although venereal syphilis was indeed spreading, the doctors™ tests
probably indicated the prevalence of yaws and endemic syphilis, related dis-
eases long present in tropical Africa. But the alarm obscured the spread in
equatorial Africa, probably since the 1860s, of another venereal disease, gon-
orrhoea, which had less acute symptoms but caused infertility, especially in
women, and was to become the chief reason for the region™s exceptionally low
Some Africans saw these diseases as European biological weapons. Others
blamed Arab slave traders, their own sins, or neglect of ancient gods. Nor was the
catalogue of disasters ended. Sand ¬‚ies that burrowed into the feet and caused
terrible sores reached Angola from Brazil around 1850 and had spread across
the continent to the Mozambique coast by 1895. The third plague pandemic
began in China in 1893 and affected African coastal cities from Mombasa to
Dakar. Epidemics of cholera, yellow fever, and cerebrospinal meningitis also
occurred. But the greatest disaster was the Spanish In¬‚uenza pandemic of 1918,
which killed between 2 and 5 percent of the population in almost every colony.
That it spread along new transport routes suggested that increased mobility and
external contacts without increased medical capacity underlay the clustering
of early colonial disasters.
Although reliable demographic data are scarce, the early colonial period
was probably most destructive in equatorial Africa, where violence, famine,
smallpox, sleeping sickness, venereal diseases, and in¬‚uenza coincided. A
218 africans: the history of a continent

knowledgeable guess is that between 1880 and 1920 the population of the Belgian
Congo fell by one-third or one-half. In 1914 it was probably declining at about
0.25 percent a year. Losses in French Equatorial Africa were probably at least as
great, especially in the riverain areas of Ubangi-Chari (Central African Repub-
lic) and the forest of Gabon, where 20 percent of women born before 1890 and
interviewed in the 1960s had never borne a living child, as against 13 to 14 percent
(itself a very high proportion) of similar women elsewhere in the region. Infer-
tility was due chie¬‚y to gonorrhoea, spread in Gabon by migration to work in
the timber industry and relaxed attitudes towards extramarital childbearing.24
Uganda, Burundi, northern Angola, and southern Sudan almost certainly lost
population. Most historians believe that savanna regions of eastern Africa also
suffered loss, although less severely. Violence, famine, sleeping sickness, and
in¬‚uenza also attacked the West African savanna, but population data there
are especially scarce.
By contrast, the early colonial crisis was less acute in the West African for-
est, except perhaps in Asante, which was disrupted by defeat in 1874. Earlier
exposure to European diseases may have given forest peoples greater immu-
nity. They also escaped famine and their prosperity from cash crops may have
improved nutrition and medical care. But the main areas of demographic
growth were probably northern and southern Africa. Apart from violence in
Libya and Morocco and the universal in¬‚uenza epidemic, North Africa escaped
the early colonial crisis. Egypt™s population continued to increase at over 1 per-
cent a year and Algeria™s somewhat more slowly, chie¬‚y owing to declining
deathrates, while both Morocco and Tunisia experienced modest demographic
growth.25 South Africa, too, escaped most catastrophes; the Cape Colony™s
census of 1904 suggested that its African population was growing as fast as
2 percent per year.26 Southern Rhodesia also experienced surprisingly rapid
demographic growth after 1900, perhaps because it escaped major famine or
disease (except in¬‚uenza).27
Whether Africa™s total population grew or declined between the 1880s and
the 1920s is unclear, but there was certainly no single pattern. To establish one
was to be a major consequence of the colonial period.

Colonial change, 1918“1950

a f r i c a ™s l e a d i n g h i s to r i a n s d i s ag re e p r o f o u n d ly a b o u t
the colonial period. For one among them it was merely ˜one episode in the con-
tinuous ¬‚ow of African history™. For another it destroyed an ancient political tra-
dition that had survived even the slave trade.1 They disagree partly because one
was thinking of western Nigeria and the other of the Belgian Congo, for the colo-
nial impact varied dramatically from place to place. But they differ also because
colonial change was contradictory and subtle. New did not simply replace
old, but blended with it, sometimes revitalised it, and produced novel and
distinctively African syntheses. Capitalism, urbanisation, Christianity, Islam,
political organisation, ethnicity, and family relationships “ central themes of
this chapter “ all took particular forms when Africans reshaped them to meet
their needs and traditions. To see colonialism as destroying tradition is to
underestimate African resilience. To see it as merely an episode is to underes-
timate how much industrial civilisation offered twentieth-century Africans “
far more than colonialism had offered sixteenth-century Latin Americans or
eighteenth-century Indians. Africa™s colonial period was as traumatic as it was
brief. Its major consequence, refuting any notion of mere continuity, was rapid
population growth, which by 1950 had become the new dynamic of African

economic change
If railways vitalised early colonial economies, the main innovation of the mid-
colonial period was motor transport. The ¬rst ˜pleasure cars™ (in the pidgin
term) appeared in French West Africa at the turn of the century. By 1927 ˜the
Ala¬n™s car, a Daimler-de-luxe in aluminium with sky ventilator and nine
dazzling head-lights, was the cynosure of all eyes™.2 More functional was the
lorry, which became common in the 1920s, the great period of road-building.
Lorries halved the cost of transporting Senegal™s groundnuts to the railhead
between 1925 and 1935 and then reduced it by another 80 percent during the
next thirty years. Lorries also released labour and provided opportunities for
Africans to move from farming and local trade into large-scale enterprise.
220 africans: the history of a continent

12. Colonial change and independent Africa.
Colonial change, 1918“1950 221

Lorries carried the ´conomie de traite into remote villages, replacing the
camels and donkeys on which Moors had hitherto transported groundnuts,
along with many other trading networks. The chief bene¬ciaries were the great
European ¬rms like the United Africa Company (UAC), which was created
during the 1920s by amalgamation and conducted nearly half of West Africa™s
foreign trade by 1930. Yet because the real value of West Africa™s overseas trade
multiplied about ¬fteen times between 1906“10 and 1955“9, there were many
opportunities for African traders. They retained control of traditional com-
merce in cattle and kola and indigenous cloth, moved into new products like
Nigerian groundnuts or Tanganyikan coffee, supplied growing towns with food
and fuel and building materials, and thronged West African markets, ˜thou-
sands of people . . . buying and selling minute quantities of the same things™.
Tropical Africa™s ¬rst successful indigenous bank, the National Bank of Nigeria,
opened in 1933. The depression of that period and the Second World War were
bad times, but after 1945 economies boomed and new trading communities
supplanted old in several cities, notably Douala, where enterprising Bamileke
became dominant, and Tunis, where rural immigrants submerged ancient mer-
chant families. Even UAC and its counterparts trained a new breed of African
managers; in 1951 some 22 percent of the members of Western Nigeria™s ¬rst
House of Assembly had such training.
The decline of some old trades, the survival of others, and the emergence
of new ones also characterised craft industries. Luxury trades often suffered
¬rst, especially where they had supplied aristocracies whose decay ravaged local
economies. Kairwan had twenty-three tanners™ shops before 1914 but none in
1940. Crafts with mass markets were also threatened. Iron-smelting disappeared
everywhere. Kano city had sixty-four blacksmiths in 1926 but only thirty-seven
in 1971. Textile industries had varied experiences. Most Ethiopians continued to
wear hand-woven cloth, but Tunisia™s famous cap-makers gradually lost their
markets to manufactured competition and changing Islamic fashions. Kano™s
cloth industry declined overall, but domestic weaving survived in the Hausa
countryside, as in many richer parts of West Africa, where cash-crop wealth
expanded markets for those making high-quality cloth or using innovations
like synthetic dyes and machine-spun thread. The most numerous artisans
in colonial cities were tailors, who pro¬ted from urban growth and imported
sewing machines. Construction trades also expanded, as did new crafts like tin-
smithing, bicycle- and motor-repair, and the manufacture of cheap household
goods from industrial waste. Egypt™s cigarette-makers pioneered mass produc-
tion for a global market, but generally there was little continuity from old to
new trades. West Africa™s stigmatised groups (other than praise-singers) lost
their craft monopolies, although stigma continued to obstruct intermarriage
and social advancement. Most guilds collapsed, except in Tunisia™s cap-making
industry and, for unknown reasons, in Yorubaland, where new trades adopted
222 africans: the history of a continent

them. Apprenticeship, however, remained widespread in West Africa and spread
into the east. In the early 1960s, Nigeria had some two million apprentices, four
times its labour force in large enterprises.
Motor transport enabled African cultivators to colonise further land for
cash crops, especially when world prices were generally high during the earlier
1920s, late 1940s, and 1950s. Pioneers created new cash-crop areas: cocoa in
Cameroun and Gabon; cocoa and coffee in Cˆ d™Ivoire; coffee in highland
areas of Tanganyika; tobacco in Nyasaland. The most important new enterprise
was the Gezira irrigation scheme on the Nile south of Khartoum, established
in 1925, where tenants cultivated over 400,000 hectares during the mid-1950s
and produced one-third of the world™s long-staple cotton. By contrast, a par-
allel French scheme near Segu, the Of¬ce du Niger, absorbed 48 percent of all
public investment in the French Soudan (Mali) during the colonial period and
was consistently unpro¬table, as were other cash-crop initiatives dependent
on of¬cial compulsion. Unrestrained private enterprise also had its costs. By
the 1940s, yields on Senegal™s older groundnut lands were falling, little virgin
forest remained in the Gold Coast™s pioneer cocoa areas, and swollen-shoot
disease had begun to kill cocoa trees. But cash-crop areas nevertheless enjoyed
unprecedented prosperity embodied in schools, churches, mosques, dispen-
saries, ˜storey-houses™, corrugated-iron roofs, shops, lorries, and lower infant
Agricultural change was not con¬ned to export crops. Africa had long con-
sisted of core areas of settlement surrounded by sparsely populated border-
lands. Pressure on core areas grew as human populations increased between
the wars (except in equatorial regions), herds recovered from cattle plague,
and many savanna peoples shifted emphasis from pastoralism towards agri-
culture. Cultivation often intensi¬ed in these areas. Igbo re¬ned their methods
of intercropping, Mossi manured in-¬elds and exploited valley-bottoms, and
many peoples replaced wooden or local iron hoes and even digging-sticks with
better imported tools. When rinderpest robbed Burundi™s intensive agricul-
ture of manure and colonial demands robbed it of labour, scarcity and disease
became common until adoption of cassava, sweet potatoes, and bananas during
the later colonial period restored viability. Many peoples diversi¬ed their crops
in this way. Maize continued to spread at the expense of millet and sorghum,
imported seeds often replacing older varieties. Cassava expanded even more
quickly, especially in densely settled areas where its productivity economised
on land and labour while its de¬ciency in protein could be supplemented from
cash-crop earnings. Potatoes spread faster in Rwanda and Burundi than they
ever had in Europe. Generally, however, it was only through better seeds, tools,
and transport that European innovations rivalled either exchange of crops
among Africans “ returning migrant labourers often took home unfamiliar
seeds “ or eager experimentation with local varieties by individual cultivators.
Colonial change, 1918“1950 223

Yet the pressure on core areas grew. Most Africans still had ample land, but
there were local scarcities in northern Ethiopia, Igboland, favoured highland
environments, areas of extensive alienation, a few cash-crop regions, and espe-
cially North Africa where many centuries of grain exports to Europe ended
during the 1930s. Fallow periods were shortening in many colonies.
Cultivators frequently responded by the traditional means of colonisation
into frontier zones. One common pattern was for crowded highlanders to
spread out into neighbouring lowlands from which insecurity had hitherto
barred them. In 1929 a ritual specialist led a pioneer column of Cushitic-
speaking Iraqw from their homeland in northern Tanzania to colonise the broad
plains of Karatu to the north. By 1937 more than half the Dogon had left the
Bandiagara cliffs for the better-watered plains, each clan creating ˜a string of new
villages, the more distant being the more recent™, their inhabitants returning to
the cliffs for burials and festivals. Savanna peoples, too, took advantage of colo-
nial peace. The Tiv of Nigeria dispersed from their stockaded villages, breaking
through the ˜wall™ by which the British sought to restrain them. The equally
individualistic Lobi expanded irresistibly across the borders of Upper Volta,
Cˆ d™Ivoire, and the Gold Coast, each family moving an average of one kilo-
metre a year. Such pioneer settlements were often culturally barren: ˜reduced
hospitality, impoverished language, anxiety and disarray in face of sickness,
boredom resented especially by the women and rendering households unsta-
ble, neglected housing.™3 Many were multiethnic. Bugerere, conquered territory
on Buganda™s northern border, had 10,302 inhabitants in 1931 and over 130,000
in the late 1960s, of whom only 38 percent were Ganda. With virtually no con-
straints on exploitation, it was Kampala™s main source of bananas. Commercial
food production was often the main activity in pioneer settlements but was not
con¬ned to them. By 1936 Tonga farmers in Northern Rhodesia, owning some
forty-three hundred ploughs, supplied maize to Katanga and the Copperbelt,
much as Kenyan plough-farmers supplied Nairobi. Yet technological change
was limited. Most Africans ˜went into colonialism with a hoe and came out
with a hoe™,4 although it was often a better hoe.
Agricultural entrepreneurship bred social ferment. ˜Cocoa is spoiling every-
thing™, a missionary in Akwapim complained in 1907. ˜There is internal strife . . .
dissatisfaction, fomentation, irregular living . . . parasites, corruption, extor-
tion, perjury, lies, drinking, laziness, pride and conceit.™5 A hybrid society was
taking shape, partly peasant, in that most members farmed their own land
with family labour and produced both for home consumption and the market,
and partly capitalist, in that a minority employed wage labourers, produced
chie¬‚y for the market, and reinvested pro¬ts. A sample in Yorubaland in 1951 “2
suggested that the 18 percent of biggest cocoa-growers marketed 53 percent of
the crop. Such big producers often pioneered cash crops but could not monop-
olise them or proletarianise their neighbours. One reason was that even the
224 africans: the history of a continent

poor generally retained access to land and therefore enjoyed much indepen-
dence and bargaining power. Even migrant labourers from the impoverished
savanna who cultivated cocoa for local entrepreneurs in the Gold Coast in the
1930s could afford to demand one-third of the crop in payment, while their
counterparts in Buganda and southern Cˆ te d™Ivoire had to be paid with access
to land, making labour migration a form of colonisation. Another obstacle to
aspiring capitalists was of¬cial hostility. African entrepreneurs were eager for
individual property rights. Some, including the Kikuyu, acquired them, but
colonial governments, like their African predecessors, saw wealthy property-
owners as not only politically dangerous but likely to create an equally threaten-
ing propertyless class. The British initially recognised freehold landownership
in West Africa but reversed their views in 1907“8 and insisted thereafter on
communal property as the basis for a ˜thriving peasantry™. Belgian authorities,
too, aimed at ˜a peasant class . . . attached by tradition and interest to social
peace™. European trading ¬rms similarly preferred to deal with peasants rather
than capitalist farmers with greater market strength; in the Gold Coast during
the late 1930s, cocoa-buying ¬rms combined with government to destroy the
African farmer-brokers who had hitherto acted as middlemen. These obstacles
were reinforced by a third impediment to capitalism: the survival of precapi-
talist obligations and attitudes. Men often continued to divide their property
equally among their heirs, of whom a wealthy polygynist might have many.
Wealth was still displayed and distributed to win dependants and power: one
early cotton-grower in Nyasaland made his workers watch over his banknotes
spread out in the sun, although most men of wealth contented themselves with
a corrugated-iron roof or a lorry. Many invested not in further production but
in education, believing correctly that white-collar employment brought easier
wealth and status; in Africa the ˜treason of the bourgeoisie™ was to invest out of
land into learning. Social obligations, especially bridewealth and ceremonies,
absorbed much capital, so that some successful men became Muslims, joined
an exclusive sect like Jehovah™s Witnesses, or otherwise sought to limit their
commitments. As one told an anthropologist, ˜I am a Christian; I don™t do
things for nothing.™ Yet clientship remained for many a powerful constraint on
capitalist relations. Ganda landowners, for example, chose the most populous
areas for their estates and behaved as seigneurs rather than capitalist farmers.
The growth of capitalism was more dramatic, but still ambiguous, among
European farmers. Here the chief interwar innovation was the Firestone
Company™s creation in Liberia of the largest rubber plantation in the world.
Libya became an important colony of settlement, while land alienation contin-
ued in the Maghrib and in East and Central Africa. Algeria™s 984,031 Europeans
(in 1954) dominated it politically, although the 250,000 in Tunisia and 363,000 in
Morocco did not exercise such power. Southern Rhodesia™s Europeans, number-
ing 136,017 in 1951, gained internal self-government in 1923, but Kenya™s settlers
Colonial change, 1918“1950 225

(38,600 in 1951) were frustrated by Asian opposition. Settlers in Portuguese
colonies (88,163 in Angola and 52,008 in Mozambique in 1951) had no more
political freedom than other Portuguese. Power or in¬‚uence enabled Europeans
to reshape colonial economies to their advantage. Railways and roads ran
through settled areas. Land banks gave Europeans credit. In 1942 Kenya™s
chief native commissioner described its maize marketing system as ˜the most
barefaced and thorough-going attempt at exploitation the people of Africa
have ever known since Joseph cornered all the corn in Egypt™.6 European agri-
culture largely monopolised export production in these colonies and shifted
from smallholder grain-farming to plantation crops: wine in Algeria, fruit in
Morocco, coffee and tea in Kenya, tobacco in Southern Rhodesia “ a trend rein-
forced by mechanisation, which became signi¬cant in North Africa between
the wars and in tropical Africa after 1945. By the 1930s over half Algeria™s exports
were wine and half of it came from 5 percent of producers. Most Europeans,
by contrast, became townsmen: 58 percent in Kenya and 78 percent in Algeria
in the late 1940s.
European farming did not simply proletarianise Africans or Arabs but differ-
entiated among them. Mechanisation in North Africa drove khamanisa share-
croppers from the land, reducing their numbers in Algeria from 350,715 in 1901
to 60,300 in 1954. By the 1950s at least one-quarter of the Maghrib™s Muslims were
landless. But Tunisia still had prosperous Muslim olive-growers in the Sahel,
while Morocco had both wealthy landlords and modern Muslim farmers in the
Atlantic Plains. In Southern Rhodesia, similarly, white farming and population
growth reduced many African reserves to impoverished labour reservoirs, but
entrepreneurial farming survived in the Native Purchase Areas created in 1930.
The Kikuyu of Kenya retained most of their valuable land close to Nairobi but
could not expand it, so that land sales created both a prosperous ˜gentry™ and a
large landless class, the latter swollen by eviction of labour-tenants from white
farms in response to mechanisation.
Most African farm labourers were not proletarians but, like cocoa workers,
were migrants with land rights at home. This was also true in the mining indus-
try. In 1935 it still supplied 57 percent of Africa™s exports. The main expansion
and pro¬t were in Katanga and on the Northern Rhodesian Copperbelt, which
produced fully from 1932. These mining companies sought to stabilise their
workers in order to ensure supply, improve skills, and silence critics, but they
still encouraged miners to retire to the countryside. Between 1921 “5 and 1931 “5
the proportion of Katanga™s African copper workers recruited annually fell from
96 to 7 percent. Yet less sophisticated industries continued to rely on short-term
migrants, now suf¬ciently available “ owing to population growth, taxation,
and other cash needs “ to make direct compulsion seldom necessary. Lack of
alternative cash sources was the key to migration. In Nyasaland in 1934, for
example, over 60 percent of men were absent from remote northern districts,
226 africans: the history of a continent

but only 10 percent from southern districts with local earning opportunities.
Most migrants were poor men like the former slaves of the West African savanna
or the subject peoples of the Ndebele kingdom, but in the most remote areas
almost all men might have to migrate, sometimes repeatedly. Governments
encouraged them to return periodically to rural homes, which they proba-
bly preferred. Oscillating migration was therefore another consequence of the
continuing control of land that distinguished the African rural poor from their
counterparts in Europe and Latin America.
Migrants generally organised their departure to minimise rural disruption.
They often travelled as parties under experienced headmen, either walking
well-trodden routes “ that from Southern Rhodesia to the Witwatersrand
was punctuated by sleeping-places built in trees “ or reluctantly accepting
a recruiter™s contract, which provided rail or lorry transport but left no free-
dom to choose employment. Those with fewest alternative cash opportunities
at home had to accept the worst jobs. All were likely to suffer contempt as man-
amba (numbers), as Europeans called them in East Africa, or even as slaves,
with whom some African employers equated them. One major ¬‚ow took West
African savanna peoples like the Mossi southwards to prosperous forest regions,
where up to 200,000 worked on Gold Coast cocoa farms in the early 1950s, or
westwards as nav´tanes (winterers) for the Senegambian groundnut season. The
Gezira relied on West Africans working their way to or from Mecca. Central
Africans headed for the copper and gold mines. Algerian workers became
numerous in France during the First World War and West Africans after 1950.
Urban studies in the 1970s suggested that 60“80 percent of migrants sent a pro-
portion (often 15“20 percent) of their earnings home regularly, plus savings
and purchases taken in person.
During the depressed 1930s, observers stressed labour migration™s destructive
effects on rural society. These were probably worst in savanna regions where
loss of labour allowed the natural ecosystem of bush, game, and tsetse to gain
over cultivation. But later analyses, in the prosperous 1950s, suggested rather
that many peoples used migration to preserve their social order, whether it be
Mambwe from Northern Rhodesia ˜raiding the cash-economy for goods™ or
Swazi employing cash earnings to rebuild their herds and homesteads in the
wake of cattle plague and land alienation. Kabylia, the chief source of Algerian
migrants, was famed for its stable social order. Migrants could nevertheless
be innovators, returning with new crops, religions, and ideas. In some remote
areas like Ovamboland in Southwest Africa they took the lead in revolt against
white oppression. But labour-exporting societies switched eagerly to commod-
ity production whenever transport improvements made it possible.
Of the growth points attracting migrants, the most spectacular were the
towns. Some ancient cities “ Cairo, Tunis, Kano “ expanded under colonial rule,
often through acute overcrowding, while others, deprived of railway transport,
Colonial change, 1918“1950 227

decayed. Governments seldom attempted to restructure old cities, rather build-
ing new suburbs for themselves and rural immigrants, much like the dual capital
of ancient Ghana. New colonial cities in Central Africa were built on segregated
South African models, but elsewhere they generally grew up more haphazardly,
with city centres ¬‚anked by European and African quarters that were engulfed
by later building, leaving of¬cials struggling to impose order by violently
unpopular ˜slum clearance™. By 1945 most towns were roughly segregated but
further rapid growth then surrounded them with belts of bidonvilles (tin-can
towns). Growth could be astonishing, especially from the 1930s. Casablanca,
Morocco™s main commercial and industrial city, grew between 1912 and 1951 “2
from 20,000 to 682,000 people.
Young countrypeople often had high expectations of urban life. ˜Make I go
Freetown,™ they said, ˜make I go free.™ The reality shocked them:

To my eyes the city [Cairo™s City of the Dead] was worse than a desert. It
was just as ugly and barren, but crammed with people. Everywhere as you
looked there were crowds of poor people who were dirty, ill-mannered and
ill-dressed. Everyone shouted and yelled at each other, there was no politeness
and no sign of modest behaviour. It seemed that people in the city had become

In 1910 the deathrate in African locations at Elisabethville (Lubumbashi) was
24 percent a year. Early immigrants sought accommodation, employment,
and help to survive from kinsmen or ˜home-boys™, recreating rural institutions
like the Nyau society, which ¬‚ourished among Chewa in Salisbury (Harare).
These gradually gave way to speci¬cally urban associations like the Beni dance
societies, imitating military bands and drill, which demobilised First World
War soldiers spread from Somalia to the Congo estuary, or the football clubs,
¬rst created by mission schoolboys, which became surrogates for violent rural
youth groups. ˜Tribal associations™ performed burials, supplied mutual aid, and
worked for rural advancement. In 1938 the Ibibio Welfare Union in southern
Nigeria sent eight students to Britain and America in a single day.
The ¬rst trade unions were generally small artisan societies like the Union
Mechanics™ Association of 1859 in Monrovia. The most strategic workers in
colonial economies were in transport and government employment. Egypt™s
¬rst major trade union organised Cairo™s tramway workers, while railway
employees were pioneers in Sudan and the Gold Coast. Dockworkers in East
Africa were also among the ¬rst to unionise, although government servants had
already formed associations there and schoolteachers were long Nigeria™s best-
organised employees. Agricultural labourers were not unionised at this period
and industrial workers were rarely so, except in Egypt. Mineworkers were slow
to combine “ the ¬rst major African union on the Copperbelt was formed in
1949 “ but they were quick to strike, owing to their concentration in compounds
228 africans: the history of a continent

and their often brutal working conditions. Much early industrial protest was
˜desertion™ or sabotage, but widespread strikes followed both world wars, owing
to in¬‚ation, and occurred also during the late 1930s.
This turbulence before the Second World War was a consequence of the
international depression, which both revealed and deepened Africa™s depen-
dence on metropolitan economies. Between the late 1920s and the early 1930s,
average export prices declined by about 60 percent and producer prices fell
more drastically. In 1932 even Union Mini` re made a loss, but the chief
sufferers were remote areas supplying food and labour to former growth points,
indicating how strong the chains of dependency had become. The proportion of
the Belgian Congo™s public revenue drawn from African taxes doubled between
1929 and 1932. Tax protests, rural revolts, and millenarian movements followed,
while cash-crop producers organised ˜hold-ups™ directed against low prices
and collusive European companies. Urban protest was less common, for wages
generally fell less than food prices and the chief victims were the unemployed
who had little redress, although both African and European copper work-
ers rioted. The major industrial protests came in the late 1930s when wages
recovered more slowly than prices. Governmental responses to depression
varied. The Belgians attempted to preserve the Congo™s economy by regu-
lating crops, production targets, and prices for Africans in each agricultural
region. Portugal™s African colonies supplanted Brazil as its major trading part-
ner. The British subsidised white settlers and pressed Africans to expand output
but otherwise practised retrenchment. The French state, by contrast, quadru-
pled its colonial investment (mainly through loans) so that colonial exports
could conserve foreign exchange and colonial markets could absorb surplus
French manufactures. ˜Urban Dakar and its rural outskirts have become vast
building-sites™, its administrator reported in 1932.8 With swelling cities, increas-
ing export production, and rising public debt, French colonies entered a new
phase of underdevelopment, but the pattern was different elsewhere. Between
1929 and 1933 Nairobi™s African population fell by 28 percent.
The economic strains of depression were followed immediately by war.
Liberian rubber and Southern Rhodesian tobacco pro¬ted from wartime
demand, but French colonies suffered extreme exploitation by Vichy and
Gaullist regimes, while the economic controls established during the 1930s
tightened everywhere. By July 1946, Britain owed its East and West African
colonies £209 million in unpaid wartime debts. Some 374,000 Africans were
serving in the British forces in May 1945, others were conscripted for private
employers, food was rationed in many cities, imported goods were scarce, and
in¬‚ation was high. Real wages in Douala halved during the war and protest was
sternly repressed. North Africa suffered special pressures, for Libya and Tunisia
were battle¬elds and even Egypt suffered in¬‚ation that reduced real industrial
wages by 41 percent.
Colonial change, 1918“1950 229

After 1945 European Powers used their empires to aid metropolitan recon-
struction. Britain, the world™s main debtor, extracted some £140 million from
its colonies between 1945 and 1951, in addition to sums withheld from producers
by colonial marketing boards for local investment. In the same period, only
some £40 million of metropolitan funds were invested under the Colonial
Development and Welfare Act. France was more generous, investing public
funds heavily in infrastructure and primary production. The Belgian Congo™s
¬rst development plan of 1948 concentrated on infrastructure and the sta-
bilisation of peasantries and urban classes, while Portugal invested chie¬‚y in
infrastructure and white settlement. Widespread impatience with peasant agri-
culture led to expensive absurdities like the Groundnut Scheme in Tanganyika,
but by 1952, when Europe™s postwar crisis ended, Africa was enjoying its ¬rst
prosperity for twenty-¬ve years, thanks to the Korean War commodity boom.
In 1949 cocoa already provided half of the Gold Coast™s national income;
during the next ¬ve years, its world price rose by 162 percent. Rapidly falling
real petrol prices encouraged motor transport, African entrepreneurship, and
exhilarating growth during the early 1950s. There was even structural change.
Hitherto, modern industries existed only in Egypt, where import substitution
had begun during the 1930s and accelerated during the war; in the settler terri-
tories of North Africa, Kenya, and southern Africa; around Katanga™s mining
industry; and on a small scale in Dakar. Elsewhere governments had often dis-
couraged industry. From the late 1940s, however, trading ¬rms built ˜market-
protecting factories™ (mainly brewing, agricultural processing, textiles, soap,
and cigarettes) in major colonies, using tariff barriers against rival importers.
A few African and Asian entrepreneurs followed their lead. In 1954 Kenya™s
manufacturing output ¬rst exceeded the value of its European agricultural
production. The classical colonial economy was changing.

education and religion
Alongside economic development, education provided the chief dynamic of
colonial change, not only as a reservoir of skills but as a source of social differ-
entiation and political con¬‚ict. Compared with wealth, education was easier
both to obtain and to transmit to the next generation, so that it became Africa™s
chief generator of both mobility and strati¬cation. It also bred individual lib-
eration and con¬‚ict, obliging educated men to create the personal syntheses of
inherited values and new ideas that gave the colonial period much of its vitality.
Fang parents in Gabon, who in the past had inserted a spear into the water for
an infant boy™s ¬rst bath, replaced it in the twentieth century with a pencil.
Christian missionaries pioneered Western education in order to create intel-
ligent Christians, but they found limited African interest until they reached
Buganda in 1877. There competition for advancement at court expanded into
230 africans: the history of a continent

a mass demand for literacy once educated ˜Readers™ seized power in 1889 and
British rule followed them. By 1900 at least one-quarter of all Ganda, adults and
children, were receiving instruction. Elsewhere colonial conquest frequently
stimulated young people to ˜marry the alphabet™ at a mission station, seeing
education as a way of escaping from agricultural labour into rewarding employ-
ment. The most eager demands came from competitive and often stateless
peoples like the Igbo of Nigeria, Beti of Cameroun, and Ewe of Togo. Quantity
rather than quality was the mission priority. In 1905 the Dutch Reformed
Church had fourteen thousand pupils among the Chewa of Nyasaland but
spent only £100 on them, teaching them to write in the dust. The development
of mission ˜bush schools™ into primary education systems in British colonies
came with state subsidies and inspection, generally in the 1920s. In French
West Africa, by contrast, anticlerical authorities refused such aid and primary
schooling lagged. In 1949“50 only 6 percent of its children of primary school
age were at school, compared to 16 percent in Nigeria, 26 percent in Kenya, and
33 percent in the Belgian Congo.
Secondary school enrolment in 1950, by contrast, was only 1 to 2 percent in
tropical colonies, compared to 7 percent in Egypt. Until the Second World War,
governments and missions stressed primary schooling, chie¬‚y as a preparation
for village life or a practical craft, whereas Africans increasingly demanded
secondary education as an escape from those occupations. Yet colonial regimes
needed trained African subordinates. The result was a few secondary schools
of high quality and prestige: the secular Sadiki Academy in Tunis, founded by
Khayr ed-Din in 1875; the Overtoun Institution in northern Nyasaland, whose
initial syllabus ranged from engineering principles to the Greek classics but
declined sadly during the 1920s; the Ecole William Ponty, founded near Dakar
in 1903, which formed a distinctive French West African elite; and Gordon
Memorial College in Khartoum, Achimota in the Gold Coast, and Alliance
High School in Kenya, each of which produced a generation of nationalist
leaders. Such schools were notably lacking in the Rhodesias and in the Belgian
Congo, where ˜No elites, no problems™ was a popular tag. North Africa™s ¬rst
Western-style university opened in Cairo in 1909; by the early 1950s, Egypt had
proportionally twice as many university students as Britain. Among tropical
Africans, only a tiny minority graduated, mostly in Europe or the United States,
until after the Second World War, when new universities opened in Nigeria,
the Gold Coast, Uganda, Sudan, and Ethiopia in 1948“51 and at Salisbury in
1956 and Dakar in 1957.
In contrast to India, African education at this time did more to foster social
mobility than to entrench old privileged classes, largely because tropical Africa
had no long-established literate elites except Muslim clerics. Egypt was one
exception to this generalisation, while colonial regimes gave privileged educa-
tion to sons of chiefs in many colonies and some ruling classes were quick to
Colonial change, 1918“1950 231

appropriate schooling, notably in Bulozi and Swaziland. Generally, however,
education expanded so quickly that even where privilege was transmitted, most
pupils were nevertheless of relatively low status. Of William Ponty™s students
in 1940“1, 38 percent were sons of farmers, herders, or ¬shermen, while 23
percent had clerical or professional fathers. The equivalent proportions at its
female counterpart at Ru¬sque, however, were 7 and 54 percent, because girls™
education was less widely available. Regional differentials were large and endur-
ing. Buganda provided some 40 percent of all East Africans entering Makerere
College at Kampala before 1953, whereas two years earlier the whole of Northern
Nigeria had only a single graduate (a Christian). Dissatisfaction with of¬cial
schooling led some Africans to establish independent school systems, notably
among the Kikuyu who founded some three hundred to four hundred inde-
pendent schools between 1929 and 1952.
Literacy fascinated young people:

At the pastures one of them began teaching another to write, and I watched
them idly . . . and said, ˜I too am going to try to put something down on
paper.™ . . . By-and-by I wrote many letters on paper. And at last my teacher
said, ˜These are all the letters. You have ¬nished.™ That day my love of learning

Writing still possessed the magic it had enjoyed when monopolised by Muslims,
but it also had material advantages “ the highest-paid African in Nairobi in
1927 was a linotype operator “ and it gave access to enlightenment and power
from which the illiterate knew themselves to be excluded. They had become
abatasoma, ˜they who do not read™, Buganda™s contemptuous of¬cial term for
pagans. Literacy could be a weapon for the elite, as in Ancient Egypt. Sir Apolo
Kaggwa acquired his own printing press to publish his version of Buganda™s
history and customs, while Njoya, the remarkable Bamoum ruler in early colo-
nial Cameroun, invented a language and script in which to write his chiefdom™s
history, record customary law, compile a local pharmacopeia, and codify a
new religion from elements of Islam, Christianity, and indigenous practices.
Yet because twentieth-century literacy was unrestricted, alphabetic, and rela-
tively easy to learn, it could equally be a weapon of dissent. John Chilembwe™s
followers read their scriptures and rebelled, as, less violently, did many early
Christians, for the Bible was commonly the ¬rst book printed in an African
language. Print also expanded the availability of Islamic texts and stimulated
vigorous literary activity in Egypt and Tunisia. Elsewhere the indigenous know-
ledge it popularised was mainly local history, custom, and folklore.
The ¬rst important African newspapers south of the Sahara, outside Liberia,
were Imvo Zabantsundu (Native Opinion), published in South Africa in 1884,
and the Lagos Weekly Record of 1891. The latter™s circulation in 1914 was only
seven hundred, compared with the ten thousand sold by Cairo™s leading
232 africans: the history of a continent

nationalist journal of the time, but by the mid-1940s the West African Pilot,
edited in Lagos by the pioneer nationalist Nnamdi Azikiwe, distributed twenty
thousand copies. Even in 1953 only about 10 percent of Nigerians were literate
in Roman script, but by then a popular readership existed. The ¬rst major trop-
ical African novel, D. O. Fagunwa™s Forest of a thousand daemons, appeared in
Yoruba in 1936 and retold folktales of dazzling vitality. Its nearest counterpart in
East Africa was Mr. Myombekere and his wife Bugonoka, written in the Kerebe
language during the Second World War by the childless Aniceti Kitereza ˜to
preserve the Customs and Way of Life of our ancestors . . . by writing the story
of a married couple whose marriage starts off being tested by a long period of
barrenness.™10 By contrast, the popular novelettes published in Onitsha after
1945 met the interests of urban primary school-leavers, stressing individual-
ism, material success, generational relationships, and romantic love, as did
popular Amharic literature. Onitsha pamphlets produced Cyprien Ekwensi,
whose novel of 1954, People of the city, introduced a generation of realistic nov-
elists who rediscovered their continent in print at the time of independence
and anatomised its dilemmas thereafter. The most gifted writer, however, was
the playwright Wole Soyinka, whose synthesis of Yoruba cultural forms and
Western dramatic traditions was to win him the Nobel Prize for Literature in
Christianity™s association with literacy and education in the early colonial
period gave it a more compelling attraction and a new character. For the ¬rst
time, the African Church became a church of the young, and it enjoyed the
success that has attended every movement in modern Africa that has associated
itself with the young. They often had the strongest material motives for seeking
education. They were attracted by novelty and not hindered by polygyny or
indigenous religious duties. Adolescence is generally the optimum age for reli-
gious conversion. And Africa™s polygynous societies had repressed the young,
for whom Christianity was a generational revolt much as communism was in
twentieth-century Asia. In 1912 every member of the ¬rst Anglican Church
Council in western Kenya was still at school. ˜Nearly all of the adherents are
young people™, an Anglican bishop wrote from Ijebu in Yorubaland in 1898. ˜Very
few exceed 40 years, the majority are in the very prime of life™.11 Ijebu™s mass
conversion of the young was repeated among the Igbo, the Beti of Cameroun
(90 percent of whom became Christians within thirty years), and less dra-
matically elsewhere. These school-based movements were not only rapid but
generally lasting, unlike more millennial assaults on disease or evil. Many mis-
sionaries quickly realised, as a Roman Catholic in Igboland explained, that ˜all
our concentration must be on schools otherwise our enemy the Protestants
will capture all the youth™.12 Benedictine monks, Salvation Army of¬cers, and
Moravian quietists found themselves abandoning their European traditions
and organising almost identical networks of schools and catechists. Often their
Colonial change, 1918“1950 233

pupils came from poor families, not least because African rulers often sus-
pected missionary work. The King of Rwanda, for example, initially allowed
White Fathers to settle only among his Hutu subjects, threatening to kill any
Tutsi who joined them. Many converts found refuge in paternalistic mission
settlements like the Jesuit ˜chapel farms™ of the Belgian Congo, where cate-
chists founded Christian villages alongside existing settlements to attract and
teach their young. Women, too, might seek protection in Christian communi-
ties. This strong element of self-emancipation by the young, poor, and female,
rather than a purely intellectual response to colonial enlargement of the world
in which Africans lived, was the chief motive force behind Christian growth
during the ¬rst half of the twentieth century.
By 1910 there were probably more than ten thousand European missionaries
in Africa. Their chief role was to build the church, whereas pioneer evange-
lisation was mainly done by Africans. Some of these African evangelists had
no speci¬c mandate or training, like the migrant labourers who carried the
teachings of the Salvation Army or Jehovah™s Witnesses throughout south-
ern and eastern Africa. Others were catechist-schoolteachers, often with bare
literacy, sent out, as one put it, to ˜the school so small in a remote village full of
witches and wizards and guineaworm™.13 By 1931 the Holy Ghost Fathers alone
employed 8,399 catechists. Missionaries were generally eager to train African
clergy, knowing, as Kongo had demonstrated, that this was vital to an enduring
church, but they were also cautious lest that clergy should prove unworthy.
Roman Catholics insisted on celibacy and the full rigour of seminary training,
ordaining their ¬rst African priests before the First World War and consecrat-
ing their ¬rst two bishops, Joseph Kiwanuka of Buganda and Joseph Faye of
Senegal, in 1939. In 1950 there were some eight hundred African priests of the
Roman Catholic Church. Protestant missions, not requiring celibacy or such
lengthy training, ordained their ¬rst clergy in the mid-nineteenth century and
subsequently trained larger numbers but were slow to give them church con-
trol, except during the First World War when several former German mission
¬elds became autonomous African churches. In 1910 there were perhaps seven
million Christians in Africa; in 1930, some sixteen million; in 1950, thirty-four
This rapid evangelisation was generally confrontational. Early twentieth-
century missionaries had little sympathy for the gradual Christianisation of
African customs. They might appropriate indigenous symbols, as the White
Fathers used the god Mukasa™s drum-call to summon people to church, but
the crucial test for Christian converts was to ˜throw away their idols™. Yet Africa
was no tabula rasa. In practice, adaptation took place, but it was done by the
converts themselves while experimenting with the new religion and reconciling
it with inherited beliefs and practices. As, at best, newly literate people, they
did this in an eclectic manner, but eclecticism could point in two directions, as
234 africans: the history of a continent

at earlier periods of African Christian history. Some Christians continued to
believe fervently in the reality of their old gods but saw them now as evil forces.
In Buganda, embittered by civil war, victorious Christians burned indigenous
shrines wherever they found them. ˜Go forward,™ their leaders urged, ˜detest
the gods of the past and love the Living God.™15 The ancient Chwezi cult of the
Great Lakes region resisted vigorously. ˜Their assemblies™, a catechist recalled,
˜were merely clubs of malcontents and reactionaries, just as my house was the
meeting-place of the young, of eager and simple hopes for a better future.™16
Yet even assaults on indigenous practices had indigenous precedents in pop-
ular movements to destroy sorcery implements and protective charms. The
most remarkable mass conversions of the early colonial period were led by
African prophets working in this tradition, notably William Wade Harris in
Cˆ te d™Ivoire during 1912“14, Garrick Braide in the Niger Delta during 1914“
16, and Simon Kimbangu among the Kongo in 1921. ˜Fetishes dey in town, in
bush, in wattah™, Harris was said to have proclaimed. ˜God send me for come
burn. . . . If you believe God, all be nutting. Everyting be ¬t do you.™17 Here
an indigenous dualism provided the basis for a Christian dualism in which
witchcraft and the spirits were identi¬ed with Satan without losing their real-
ity. Some communities held ceremonies to inform the spirits that they would
no longer be venerated.
Yet the eclecticism that permitted belief in both sets of spiritual forces led
other people to seek bene¬ts from both, thereby infusing Christianity with an
indigenous tolerance to which missionaries were strangers. ˜Christians simply
chose in the Christianity presented to them the elements that were meaning-
ful and helpful to them, and ignored the rest™, a student of Nyasaland has
written.18 They took literacy, the clearer Christian perception of the High
God, and elements of Christian eschatology, which was so much more com-
pelling than indigenous beliefs in the survival of spirits. ˜Do you want to
burn?™ Christian villagers on Lake Malawi demanded of their neighbours.
Yet Christians generally ignored missionary teachings about marriage, as had
their predecessors in Carthage, Ethiopia, and Kongo. Two generations after
evangelisation, only one-quarter of Anglicans in Buganda or Roman Catholics
in Southern Rhodesia married in church. Many Christians employed what-
ever spiritual resource was most likely to remedy their particular misfortune,
consulting diviners, using protective charms and medicines, and interpreting
Christian practices in magical terms. Over time, this eclecticism bred patterns
of rural spirituality that were both Christian and African, although anathema to
As missionaries and converts gained understanding, they moved in oppo-
site directions. After 1918 missionaries learned more respect for African reli-
gions and several attempted adaptations such as Christian initiation rites or
ritual use of African music. But these generally met African hostility, for many
Colonial change, 1918“1950 235

Africans identi¬ed indigenous practices with the Devil and sought models
rather in the Bible, now increasingly available in African languages. These
tendencies underlay many of the independent churches, outside missionary
control, which became a distinctive feature of African Christianity. The earliest
generally resulted from disputes over church control, ordination of African
clergy, and political tensions underlying these grievances. This was true of
independent churches founded in Lagos from the 1880s and of the ¬rst such
churches in South Africa, the Thembu Church of 1884 and the Ethiopian Church
of 1892 that gave its name to the type of organisation that retained the parent
mission™s doctrines and ritual but rejected its authority. After 1918 such protest
generally took secular political form and independent churches were created
for more spiritual reasons. These Zionist churches, ¬rst founded by white peo-
ple in South Africa in the 1890s, were more innovative in ritual and doctrine,
appealed to less educated people, and were especially concerned with healing
and spiritual experience. The major South African churches, Isaiah Shembe™s
Church of the Nazarites (founded in 1911) and Enginasi Lekhanyane™s Zion
Christian Church (1925), both followed this pattern, as did the Aladura churches
founded in Nigeria in response to the in¬‚uenza epidemic of 1918. In their belief
that religion must heal body as well as spirit, independents restored the holism
of African religious traditions and sought to use Christian resources to meet
the whole range of African needs. By studying the Bible and seeking to rid
Christianity of European accretions, they also sought a more spiritual church,
as did many who remained in the mission churches but, for example, joined
the Revival Movement that revitalised East Africa™s Protestant churches after
1929. Others read not only the Bible but the pamphlets of the Watchtower Bible
and Tract Society, the predecessor of Jehovah™s Witnesses, which from about
1908 inspired southern Africans (often migrant labourers) to create separated
˜Watchtower™ communities awaiting the imminent millennium and denounc-
ing both the indigenous and colonial orders as satanic. Zionists created similar
supportive communities, which gave satisfying roles to women, the old, the
illiterate, and the polygynous, but they were seldom politically active. Inde-
pendency was not simply resistance to colonialism, for churches were not most
common in the most oppressed areas “ there were many in prosperous, lightly
governed Yorubaland “ and they continued to proliferate after independence.
Rather, they were attempts to meet African needs in strongly Christian regions
where churches provided institutional models for people who had few relevant
indigenous institutions. Religious initiatives that were barely Christian, such as
D´ ima in Cˆ d™Ivoire, might ignore the Bible but imitate the church. In 1950
e ote
there were probably between one and two thousand independent churches in
Africa with some two million members.
European rule might seem a less propitious environment for Islam. Muslims
certainly thought so. After defeat, some withdrew as far as Mecca, while the
236 africans: the history of a continent

majority treated European governments with the outward submission and
inward reserve that the faith prescribed. Many Muslim regimes lost power and
their educational systems were marginalised or, as in Algeria, almost destroyed.
British rulers in Sudan and early French of¬cials everywhere viewed Islam
with deep suspicion. Yet Lugard extended Fulbe Islamic rule over non-Islamic
peoples, the Germans in East Africa relied on coastal agents, and even the French
sought Muslim allies in West Africa between the wars, ¬nding them espe-
cially among Senegal™s increasingly powerful brotherhoods. Moreover, colo-
nial change encouraged Islamic expansion. The need for social reconstruction
after conquest popularised the Mouride brotherhood and bred mass conver-
sions among the Yao of Central Africa and the Ijebu Yoruba. Many freed slaves
appear to have become Muslims in order to assert a new identity. Labour
migration exposed many countrymen to Islamic urban culture. For the Jola of
Senegal, Islam came with groundnuts and motor-lorries, but it was brought by
Muslim teachers, for deliberate proselytisation was the major engine of expan-
sion, as by the Qadiriyya brotherhood active in East Africa. Because Islam was a
complete culture, its initial adoption generally took an eclectic form. The High
God could be understood afresh as Allah, while lesser spirits were demonised
as jinn. Whereas Christians stressed healing, Muslims offered divination and
protective magic. The Yoruba ifa oracle, for example, found a substitute in
the Islamic hati, with the difference that the Muslim diviner consulted a book
rather than memorised texts and counselled charity rather than sacri¬ce. Yet
Islam™s scripturalism eroded eclecticism and encouraged deepening of belief.
The main agency in North Africa was the Sala¬yya modernist movement from
Cairo, which created independent schools throughout the Maghrib between
the wars and educated a generation of future nationalists. It reached West Africa
after 1945, but earlier initiatives there came from local educational associations
and from brotherhoods like the Reformed Tijaniyya.
Most people in Africa before 1950 probably relied on indigenous religious
resources. Because these lacked institutions and scriptures, their resistance to
the colonial impact can often only be glimpsed in Kinjikitile™s inspiration of
the Maji Maji rebellion, the Poro society™s coordination of the Hut Tax War,
or the Nyau society™s harrassment of village catechists and schoolchildren. In
1931 Dahomey™s priests preserved their functions little changed, a generation
after the destruction of the kingdom. Yet indigenous religions did not merely
resist. They had long been among the most adaptable elements in African cul-
tures. Many invented prophecies of white men, which made conquest easier to
bear because foreseen. The territorial cult of Mwari accommodated Southern
Rhodesia™s white settlers as Mwari™s sister™s children. Mbona the martyr took on
features of Jesus. Cults of af¬‚iction domesticated the forces of change by incor-
porating them as possessing spirits. Protective shrines and medicines prolifer-
ated with the greater mobility and new stresses of the time. Many cults adopted
Colonial change, 1918“1950 237

European elements. Kongo, for example, used both Holy Communion and
surgery as ordeals to test accusations of witchcraft, while the Fang of Gabon, at
a time of massive depopulation before the First World War, created the Bwiti
cult to revitalise their society by synthesising its beliefs and symbols within an
institutional framework modelled on a Christian church.
Despite this creative vitality, however, indigenous religions were generally
in retreat. By 1913 a Christian was planting cocoa in once-sacred land at
Arochukwu. Decline owed something to of¬cial disfavour “ in 1933 all chiefs
in Burundi practising indigenous rites were dismissed “ and more to Islamic
and Christian competition, but owed most to personal mobility and com-
mercial change, for indigenous religions were communal religions that could
¬‚ourish only within local environments. Characteristically, therefore, it was
communal rituals that decayed, whereas belief in a High God might even be
strengthened, as might belief in individualistic magic and witchcraft. Africans
probably always thought witchcraft was increasing, just as the weather was
deteriorating, but alarm during the colonial period was fuelled by the decline
of indigenous religious defences, increased mobility, new social strains, and
especially European governments™ refusal to take witchcraft seriously and their
insistence on punishing not only supposed witches but those seeking to iden-
tify them. Africans evaded these colonial laws by multiplying the hitherto rare
witch-cleansing movements that administered medicines to whole communi-
ties in order to make them invulnerable to witchcraft and kill any recipient
subsequently practising it. These movements, diverse in form and often with
millennial and anticolonial overtones, became as characteristic of the colonial
period as was Christianity itself.

political change
Western education transformed Africa™s politics. The political growth point in
1918 was the north. Egypt™s nationalists, driven underground in 1882, reemerged
before the First World War and in 1918 demanded a delegation to press the
Peace Conference for independence. Rebuffed, landlords and lawyers led by
Saad Zaghlul founded a party, the Wafd (Delegation), and mobilised support
through village sheikhs. When the British deported Zaghlul, widespread rural
violence followed. Unwilling to undertake prolonged repression, the British
sought to transfer responsibility to conservative Egyptians by declaring Egypt
independent in 1922, while reserving ˜rights™ of intervention to secure the Suez
Canal, defend Egypt, protect foreign interests there, and maintain the Sudan™s
integrity. As in subsequent decolonisations, the British failed to choose their
successors, for the Wafd won the ¬rst election in 1924. When it tried to force
complete British withdrawal by popular agitation, however, the British obliged
the King to dismiss the government, only for the Wafd to win the next election.
238 africans: the history of a continent

This debilitating cycle occurred three times between 1922 and 1952. Both Wafd
and monarchy grew ever more conservative and corrupt. Radical politics took
to the streets in the Communist Party and the Muslim Brotherhood, founded
by Hasan al-Banna in 1928 as North Africa™s ¬rst fundamentalist movement.
Population growth and accumulation of land by the wealthy drove country-
men into urban unemployment. Despite some industrial development, Egypt™s
national income per head fell by perhaps 20 percent between the 1900s and 1945.
Egypt was nevertheless a model for the Maghrib™s ¬rst nationalists. In Tunisia
a modernising elite had existed before French occupation and reemerged in
1907 as the Young Tunisians, seeking assimilation into the French political
system. Refused a voice in the peace settlement, some created the Destour
(Constitution) Party in 1920, seeking to restrain the bey™s formal autocracy
through which the French ruled. The party™s refusal to exploit popular unrest
exasperated younger men with European education and a Jacobin political
perspective from lower social strata and small towns. Led by Habib Bourguiba,
they created the Neo-Destour in 1934 as colonial Africa™s ¬rst true nationalist
party, with the goal of independence. In 1937 it had some 70,000 members in
over 400 cells. Frustrated by repression and settler opposition, it was moving
in the early 1950s towards terrorist violence.
Morocco and Algeria were more fragmented territories lacking pre-colonial
elites with modern education. In Morocco former student groups from the
urban bourgeoisie created the Istiqlal (Independence) Party in 1943, but it
was overshadowed politically by the king, Muhammad V, whose exile by the
French in 1953 made him a martyr, and by Berber chieftains and guerrilla
groups in the mountains. In Algeria, similarly, constitutional nationalism was
only one political tendency. Its exponents, the Young Algerians, were denied
full assimilation by settler opposition before the First World War. Muslim
politics then fragmented. The Westernised elite took a minority role in electoral
politics. Islamic modernists concentrated on creating an independent school
system. A populist organisation, the North African Star, was formed among
Algerian workers in France in 1926 and introduced to Algeria in 1934. When
electoral rigging by the authorities aborted constitutional reforms of 1947,
which promised Muslims one-half of the seats in an Algerian Assembly, young
militants launched in 1954 the insurrection that was to liberate Algeria.
For Italian colonies, by contrast, the Second World War brought indepen-
dence. Unable to agree on who should control Libya, the United Nations gave
it freedom in 1951 under the hereditary leader of the Sanusi brotherhood, King
Idris, although Western in¬‚uence remained strong. Somalia returned to Italian
trusteeship for only a ten-year period in 1949, but Eritrea was federated with
Ethiopia in 1952 against the wishes of many Eritreans. In other ways, too, the
Second World War was important to Ethiopia. Despite Menelik™s modernising
initiatives, his death in 1913 had sparked a succession war. The eventual victor,
Colonial change, 1918“1950 239

Haile Selassie (1916“74), had continued Menelik™s centralisation and moderni-
sation, but his forces had been too lightly armed and divided by aristocratic
factionalism to prevent conquest in 1935 by 500,000 Italian troops, although
guerrilla resistance had kept large areas of countryside beyond Italian control.
When British forces restored Haile Selassie in 1941 he used their presence to oust
his enemies who had collaborated with the Italians, disband provincial armies,
defeat regional revolt, create a salaried bureaucracy, and introduce direct indi-
vidual taxation. It was a crucial period in the creation of a modern Ethiopian
Because most tropical African colonies were purely European creations, they
lacked North Africa™s pre-colonial elites with colony-wide perspectives. Instead,
their modern politics generally began at two other levels. Most Africans con-
centrated on local issues: to defend their locality against European intrusion,
to increase its prosperity, and to advance their personal position within it.
During the early twentieth century, teachers, clerks, clergymen, traders, and
commercial farmers formed innumerable local associations ˜for the develop-
ment of our country and for the seeking of a system for the simple way to
civilization to our mutual advantage™.19 These associations often fostered tribal
identities. Pre-colonial Africans had possessed several social identities. They
might belong to lineages, clans, villages, towns, chiefdoms, language groups,
states, and almost any combination of these, the relevant identity depending
on the situation. Identities shaded into one another, for people speaking the
same language might belong to different chiefdoms, while one chiefdom might
embrace people speaking several languages. It was an immensely complex social
order. Individuals or social processes periodically simpli¬ed it by stressing one
identity over others. Ewuare the Great of Benin decreed that his subjects should
bear distinctive Benin facial markings. East African coastmen gave traders and
porters from the west a new identity as ˜Nyamwezi™ (People of the Moon). Colo-
nial circumstances, similarly, often accentuated one existing identity, which
Europeans called tribal, or occasionally created such an identity. Some gov-
ernments used tribal identities to divide their subjects, notably the British in
southern Sudan and the French in Morocco. More often they tried to demar-
cate tribes for administrative purposes, especially under Indirect Rule. Ethnic
differences were sometimes interpreted in racial terms, especially in Rwanda
and Burundi where Europeans regarded the Tutsi as a superior immigrant race
and reinforced their domination over the more numerous Hutu, while simul-
taneously destroying the military basis of Tutsi prestige. Also in¬‚uential were
missionaries who reduced Africa™s innumerable dialects to fewer written lan-
guages, each supposedly de¬ning a tribe. Yoruba, Igbo, Ewe, Shona, and many
other ˜tribes™ were formed in this way. This linguistic work relied on African
intellectuals who made many of the translations, staffed the churches and pri-
mary schools propagating tribal languages, recorded the traditions composing
240 africans: the history of a continent

tribal histories, and expounded the customs forming tribal laws. Some intel-
lectuals invented entirely new tribes such as the Abaluyia of western Kenya.
Others campaigned for the election of paramount chiefs to foster unity and
progress among previously divided groups, or advocated the reuni¬cation of
Ewe, Kongo, Somali, and other people divided by colonial borders. Yet tribal-
ism was not the work of intellectuals alone. ˜Modern tribes were often born on
the way to work™,20 for migrant labourers needed group solidarity and a means
of categorising other peoples in towns that highlighted differential access to
colonial change. ˜They used to say that we were cannibals, and that we did
not even make up a tribe™, one townsman recalled. ˜Therefore, the few other
intellectuals and myself organised a sort of federation in Abidjan to include


. 7
( 12)