Sub-Saharan Africa — empires of the Sahel and the Swahili coast

Intuition Beginner

Long before any European ship reached African waters, two of the richest trade systems on Earth met across the Sahara and the Indian Ocean. In the west, camel caravans carried salt north and gold south across the desert, and the kings who taxed that flow built empires. In the east, wooden dhows rode the monsoon winds between African ports and Arabia, India, and China.

Three empires rose in turn on the same grassland strip called the Sahel: Ghana, then Mali, then Songhai. Each controlled the gold fields and the desert crossings. On the coast, stone-built cities like Kilwa and Mombasa brokered Indian Ocean trade. Far inland, Great Zimbabwe fed gold and ivory toward those ports. Underneath all of it, the slow Bantu migrations had already spread farming, iron, and languages across most of the continent.

These were not thin societies touched from outside by trade. They had standing armies, tax systems, mosques, universities, courts of law, and their own written and oral records. The reason they are under-told in many school curricula is a problem of historiography, of who wrote, preserved, and destroyed the records, not a problem of evidence. The evidence is abundant.

Visual Beginner

Figure: Two trade systems, one continent. The trans-Saharan caravans (gold and salt, north-south) sustained the Sahelian empires; the Indian Ocean dhows (monsoon-driven, east-west) sustained the Swahili city-states. Great Zimbabwe linked the inland gold fields to the coast.

Feature	Trans-Saharan system	Indian Ocean system
Vehicle	Camel caravan	Dhow, lateen-rigged
Driving force	Desert crossings	Monsoon winds
African export	Gold, ivory, enslaved people	Gold, ivory, enslaved people
African import	Salt, cloth, copper, manuscripts	Cloth, porcelain, glass beads
Political form	Territorial empires	Mercantile city-states
Key emporia	Timbuktu, Djenné, Walata	Kilwa, Mombasa, Sofala
Decline trigger	1591 Moroccan invasion	1498 Portuguese disruption

Worked example Beginner

In 1324 the Mali emperor Mansa Musa set out from his capital toward Mecca. The Egyptian official al-Umari, writing in Cairo about fifteen years later from the testimony of eyewitnesses, recorded the scale. His figures are the ones historians still cite, with the caution that they reach us through several layers of retelling.

Al-Umari put the retinue at roughly 60,000 people, including some 12,000 enslaved attendants. The emperor carried gold both to fund the pilgrimage and to give as gifts. Later chroniclers give totals on the order of 12 tons of gold. The metal was spent and donated in Cairo along the route.

The market could not absorb it. Al-Umari and the later historian al-Maqrizi report that the gold dinar's exchange rate in Cairo fell from about 25 silver dirhams to about 18. That is a drop of 7 dirhams out of 25, or 28 percent. The rate took well over a decade to recover.

Whatever the exact tonnage, the episode shows the mechanism. Mali taxed gold production and the desert crossings, stored the metal as state wealth, and then released it abroad as diplomacy. The hajj was a political act: it advertised Mali's power to the Islamic world and drew scholars and architects, among them Abu Ishaq al-Sahili, back to Timbuktu and to the building of the Djinguereber Mosque.

Check your understanding Beginner

Formal definition Intermediate+

The civilisations of this unit operated inside two distinct commercial geographies, and precise terminology is needed before any comparison.

Trans-Saharan trade denotes the network of caravan routes crossing the Sahara, linking the Sahel and the Niger bend to the Maghreb, Egypt, and through them the Mediterranean and the Islamic world. Gold, ivory, and enslaved people moved north; salt, textiles, copper, horses, and manuscripts moved south. The system was made commercially viable by the camel, in regular use across the desert from roughly the 3rd century CE, and by a chain of oases and market towns that staged the crossing.

Indian Ocean trade denotes the maritime network linking the East African coast to Arabia, the Persian Gulf, India, Southeast Asia, and China. It was sustained by the monsoon reversal: northeast winds from November to March carried vessels from India and Arabia to Africa, and southwest winds from April to October carried them back. This climatic regularity permitted scheduled long-distance trade by wooden dhows without advanced navigational instruments.

The Sahel (Arabic sahil, "shore" or "coast," meaning the shore of the desert) is the ecological transition zone, roughly 5,400 km long, between the Sahara and the humid savanna to the south. Ghana, Mali, and Songhai were Sahelian polities, positioned to tax trade moving between the desert and the tropical forest and gold fields.

Silent barter, sometimes called dumb barter, designates the exchange mechanism described by Arabic geographers for the gold trade between North African merchants and producers of the south. Merchants left goods at a boundary and withdrew; producers left gold beside the goods and withdrew; if both sides were satisfied the exchange stood. Whether the description is literal or stylised, it captures two real features: the absence of a shared language or currency, and the producers' interest in concealing the gold sources.

A Sahelian empire is best modelled as a territorial-tax state: its revenue came not from owning a single mine but from taxing the movement of gold, salt, and other goods across a controlled corridor, and from tribute from provincial rulers. Successive empires rose on the same territory because the corridor was a durable geographic fact.

A Swahili city-state denotes a coastal or island Bantu-speaking urban polity, Muslim and mercantile, trading across the Indian Ocean. Kilwa, Mombasa, Zanzibar, Malindi, Mafia, and Sofala formed a chain rather than a single state; each was politically independent and linked to the others by kinship, language, religion, and commerce.

The Bantu expansion is a linguistic and demographic concept, not an ethnic one. It denotes the multi-millennial spread of Bantu languages (a branch of the Niger-Congo family) from a homeland in present-day Cameroon and southeastern Nigeria across central, eastern, and southern Africa, carrying ironworking and agricultural practices with it. Reconstructed primarily by historical linguistics and corroborated by archaeology and genetics, it is the deep context for every later state in this unit.

Comparative framework Intermediate+

The three Sahelian empires are best understood as one political-economic succession on a single corridor, not as three unrelated states. They inherited the same resource base and the same problem: how to convert control of a trade chokepoint into a durable administrative structure.

Ghana (flourishing roughly 6th-13th century) governed the earliest well-attested Sahelian state. Al-Bakri, writing in 1067-68 from traveller accounts, describes a dual capital at Kumbi Saleh, where the king's town with its court and the Muslim trading town with its mosques were spatially separated but economically linked. The state's mechanism was a regulated gold trade: the king claimed a monopoly on gold nuggets while allowing gold dust to circulate, a two-tier system that captured revenue without glutting the market. Administration was tribute-based and personal, organised around the king's household and provincial chiefs rather than a standing bureaucracy.

Mali (13th-15th century) scaled this model into an imperial-provincial structure. Founded through the wars recounted in the Epic of Sundiata, Mali extended control from the Atlantic to the middle Niger and institutionalised rule through appointed provincial governors, a professional army, and regularised tax collection. Mansa Musa's 1324 hajj projected this imperial capacity abroad. Mali's distinctive contribution was the deliberate patronage of Islamic scholarship: Timbuktu and Djenné became centres of learning under imperial protection, and the state invested in mosques, libraries, and the scholars who staffed them.

Songhai (15th-16th century) added a bureaucratic-Islamic layer. Under Askia Muhammad I (1493-1528), the empire was divided into provinces governed by appointed officials, taxation was standardised, and the legal system was aligned with Maliki jurisprudence. The chronicles Tarikh al-Sudan and Tarikh al-Fattash, composed in Arabic by Timbuktu scholars in the 17th century, are themselves products of this administrative-literary culture. The succession ended at the Battle of Tondibi in 1591, when a Moroccan force equipped with firearms defeated the Songhai army.

The structural pattern across all three is the same: each empire's revenue depended on taxing the gold-salt flow, and each in turn lost coherence when that flow was disrupted or when succession crises and external shocks broke the administrative machine. The Sahelian cycle is therefore one of institutional layering on a constant economic base.

The two trade systems stand in instructive contrast. The trans-Saharan system ran through a small number of long, risky desert crossings, funnelled through chokepoints that a territorial state could seize and tax; this is why it produced large land empires. The Indian Ocean system ran along a coastline of many harbours, each defensible but none individually dominant; this is why it produced a chain of independent city-states rather than a single littoral empire. The trans-Saharan system was a monopoly machine, the Indian Ocean system a competitive network. The political forms follow from the economic geography.

The scholarly centres, Timbuktu and Djenné, belonged structurally to the trans-Saharan system. Both sat at the southern end of desert routes and north of the gold fields, and both drew their scholarly culture from the Islamic world through those routes. Djenné, the older commercial centre, was the upriver partner of Timbuktu and the site of the great earthen mosque rebuilt in 1907 on much older foundations. The manuscript libraries that survive in both cities, an estimated 700,000 across the region though counts are uncertain and many collections remain uncatalogued, document a learned tradition in law, theology, astronomy, mathematics, and correspondence that was continuous from the 14th century to the present.

Bridge. The structural comparison of the Sahelian and Swahili systems builds toward a single thesis: state power in medieval sub-Saharan Africa rested on controlling chokepoints in long-distance exchange, whether caravan stages across the Sahara or harbour entrepots on the Indian Ocean. This is exactly the pattern that appears again in the Atlantic period, when the same gold-and-people flows were rerouted toward European shipping; the central insight generalises to the inland Shona state at Great Zimbabwe, which controlled the southern gold source feeding the Swahili coast. The foundational reason the Sahel could sustain three successive empires on the same territory is that the gold-salt chokepoint was a durable geographic fact; putting these together, the rise and fall of Ghana, Mali, and Songhai tracks who taxed that flow, and the bridge is a chokepoint theory of African state formation.

Case study: Great Zimbabwe and the Shona state Intermediate+

Great Zimbabwe, in the hills of the southeastern Zimbabwean plateau, is the largest pre-colonial dry-stone complex south of the Egyptian pyramids, and the most archaeologically visible capital of an inland Shona trading state. Its development is the test case for African state formation that was neither Sahelian nor coastal.

The site has three architectural components laid out across a granite landscape. The Hill Complex, the earliest, was occupied from roughly the 11th century and served as a royal and ritual enclosure. The Great Enclosure, built in the 13th and 14th centuries at the peak of the state, is the largest single pre-colonial structure south of the Sahara: a curved wall up to 11 metres high and 250 metres in circumference, rising from a 5-metre base, built of precisely fitted granite blocks without mortar, enclosing a solid conical tower whose function remains debated. The Valley Ruins housed the bulk of the population. At its peak the city held an estimated 10,000 to 20,000 people.

The architectural sequence has a documented predecessor. The K2 and Mapungubwe settlements in the Limpopo valley, flourishing from roughly 1075 to 1220, show the transition from earlier farming villages to a class-stratified, stone-building polity. Mapungubwo's abandonment around 1220, plausibly linked to climatic stress, coincides with the rise of Great Zimbabwe, and the continuity of building techniques and material culture between them is the archaeological basis for treating Great Zimbabwe as an indigenous development rather than an imported one.

The state's economy rested on three bases. Cattle were the primary store of wealth and the medium of political tribute; the archaeology of cattle byres at the centre of the site indicates the ritual and economic centrality of pastoral wealth. Agriculture produced sorghum and millet for a large urban population. Gold, mined and panned in the surrounding greenstone belts, was worked locally and traded east, together with ivory, toward the coast. Cotton spinning and weaving are attested in the archaeological record. Great Zimbabwe's prosperity came from its position as the inland node that funnelled gold and ivory to the port of Sofala and, through Sofala, to Kilwa.

The trade connection is material, not inferred. Excavation has recovered Chinese celadon and Ming porcelain, Near Eastern glass beads, cowrie shells from the Indian Ocean, and coins minted at Kilwa. These are the same categories of import found at Swahili coast sites, confirming that Great Zimbabwe was integrated into the Indian Ocean network through the Sofala trade. The inland state and the coastal city-states were the two ends of a single gold-for-goods system.

Authorship was contested for political rather than evidential reasons. Colonial-era writers attributed the stonework to Phoenicians, Sabaeans, or the biblical Queen of Sheba, attributions consistent with the colonial assumption that sub-Saharan Africans were incapable of such construction. Gertrude Caton-Thompson's 1929 excavation, presented to the British Association, established that the ruins were of African origin and relatively recent date, a conclusion built on by subsequent excavation. The name itself records the builders: dzimba dza mabwe, "great houses of stone," in Shona. The state declined through the 15th century, probably under combined environmental and political stress, and its population dispersed to successor settlements such as Khami and Mutapa.

Case study: the Bantu expansion as deep context Intermediate+

Every state treated in this unit stands on a deeper demographic base: the Bantu expansion, the multi-millennial spread of Bantu languages and associated technologies across sub-Saharan Africa. It is the context that makes the Sahelian empires, the Swahili coast, and Great Zimbabwe legible as African developments rather than external implants.

The reconstruction is primarily linguistic. The approximately 500 Bantu languages spoken today descend from a reconstructable common ancestor, Proto-Bantu. The comparative method recovers features of that ancestor, and the geographical distribution of its daughter branches, ordered by shared innovations, maps the pathways of dispersal. The inferred homeland lies in the Benue-Cross river region on the present Cameroon-Nigeria border. From it, two principal streams diverged: a western stream through the equatorial rainforest toward the Atlantic and south-central Africa, and an eastern stream through the Great Lakes to the Indian Ocean coast and thence southward to Mozambique and KwaZulu-Natal.

The chronology is convergent. Historical linguistics places the initial breakup of Proto-Bantu in the early first millennium BCE, broadly from about 1000 BCE. Archaeology supplies an independent sequence: the appearance of Early Iron Age assemblages, characterised by Urewe and related wares and by iron smelting, across the Great Lakes region in the early centuries CE, and further south in the mid-first millennium. Bantu-speaking populations reached the East African coast by roughly 500 CE and the southern African interior over the following centuries. The full span, from the initial dispersal to settlement of the southern third of the continent, runs from roughly 1000 BCE to 1000 CE.

The expansion carried a technological and economic package: iron smelting and forging; agriculture based on yams, oil palm, and sorghum and millet where rainfall allowed; cattle pastoralism in suitable zones; and, in the Great Lakes, bananas adopted from the Indian Ocean contacts that reached the coast in the early centuries CE. The package was not uniform; ironworking reached some regions centuries before others, and the mix of crops and stock varied with ecology. What was shared was a set of social institutions and a language family dense enough to make the expansion visible in the linguistic record.

Two clarifications are necessary. First, the Bantu expansion was not a coordinated military conquest; it was a frontier movement of population growth, agricultural colonisation, and cultural diffusion operating over many centuries. Second, it did not occur in empty land. Bantu-speaking populations encountered Khoekhoe and San hunter-gatherer communities in the south and rainforest forager communities in the centre, and the outcome was interaction, trade, intermarriage, and in many cases assimilation or displacement, rather than uniform replacement. Genetic evidence indicates substantial admixture, consistent with a mixture of demic movement and cultural diffusion rather than simple population replacement.

The significance for this unit is structural. The Bantu expansion explains why the Sahel, the Swahili coast, and the Zimbabwe plateau shared broadly related agricultural and metallurgical foundations, and why the languages of the entire eastern and southern half of the continent are related. The states that arose from the first millennium CE did so on a base that the expansion had prepared, and a unit that treated those states without the expansion would mistake their foundations for borrowings.

Exercises Intermediate+

Advanced results Master

The under-telling of African history: two positions

The relative absence of sub-Saharan Africa from many Western curricula is a documented feature of 19th- and 20th-century education, and its causes are themselves a contested historical question. At least two positions are live in the literature, and a third seeks to mediate them methodologically.

The first position, articulated most fully in the UNESCO General History of Africa project (1964-1993) and in the work of Cheikh Anta Diop and Joseph Ki-Zerbo, treats the omission as a political artifact of colonialism. On this view, the claim that Africa had "no history" was an ideology that served the slave trade and colonial rule, formalised in Hegel's 1837 Philosophy of History declaration that Africa had "no historical part of the World." The corrective is not merely additive (inserting a few chapters) but structural: African history must be reconstructed from African sources, by African scholars, with periodisations derived from African experience rather than from the arrival of Europeans ^{[UNESCO 1981]}. The eight UNESCO volumes, the majority of their contributors African, were explicitly conceived as this corrective.

The second position, associated with more conservative historiography and with defenders of existing curricular organisation, attributes the under-representation chiefly to source availability and disciplinary inertia rather than to ideology. On this view, the documentary density for medieval Europe or China is simply higher than for oral-tradition-based societies, and history departments organised themselves around documentary abundance; the omission, while regrettable, reflected archival asymmetry rather than malicious intent. This position is increasingly untenable as archaeological, linguistic, and genetic evidence has closed the source gap, but it persists in practice in curricula slow to revise.

A third, methodological position, identified with Jan Vansina's oral-tradition work and with John Iliffe's synthetic histories, holds that the under-telling is corrected by method rather than by ideology in either direction ^{[Vansina 1985]}. The remedy is to treat oral tradition, archaeology, historical linguistics, and documentary sources as complementary evidence streams, each with dated provenance and stated limits. This position is dominant in professional Africanist historiography and is the one adopted in this unit.

The slave-trade African-agency debate: two positions

A second contested question concerns the Atlantic slave trade and the degree of African political agency within it. The literature contains at least two well-defined positions.

The African-agency thesis, advanced by John Thornton in Africa and Africans in the Making of the Atlantic World (1992, revised 1998), holds that Africans were not passive victims but active political and economic actors whose institutions shaped the trade ^{[Thornton 1998]}. Thornton emphasises that pre-existing African slaving institutions, African demand for imported goods, and African legal and political frameworks determined the terms on which captives were sold, and that Europeans depended on African intermediaries to operate at all.

The coercive-demand critique, associated with Joseph Inikori, Paul Lovejoy, and more recent work by Marixa Lasso and Stephanie Smallwood, holds that the "agency" framing understates the structural violence of the Atlantic system ^{[Iliffe 1995]}. In this view, plantation capital and European demand pulled captives across the ocean at volumes and under conditions that no African political actor could have sustained unaided, and "agency" exercised under the threat of firearms and military pressure is not free action. The trade enlarged and distorted African slaving institutions rather than merely channelling pre-existing ones.

These positions are not symmetric in their moral or evidential weight, and a unit following the humanities addendum must mark them as contested rather than resolving them. What distinguishes responsible presentation is that each position is attributed, dated, and stated in its strongest form before evaluation, and that the asymmetry of violence is not flattened into a neutral "debate" between equivalent parties.

Synthesis. Putting these together, the contested historiography of sub-Saharan Africa turns on the same structural fact: the surviving sources are filtered by who recorded, destroyed, or ignored them, so every quantitative estimate, caravan tonnages, manuscript counts, captive numbers, gold rates, carries a dated source and a stated margin. The central insight of the UNESCO project, that African voices must lead the reconstruction of the African past, is dual to the demand of the African-agency debate, that African political actors be treated as principals rather than as passive terrain; this is exactly why convergent evidence such as the Mansa Musa hajj or the Kilwa coinage matters methodologically, since no single tradition could fabricate a convergence across Arabic, African, and later European records. The foundational reason such convergences are decisive is that they are overdetermined by independent evidence streams, and the framework generalises to every colonial-era archive, where the bridge is a cross-check of oral tradition, archaeology, linguistics, and genetics against a documentary record thinned by conquest.

Connections Master

• The Atlantic slave trade 32.16.01. The gold-and-labour flows that the Sahelian empires taxed across the Sahara were, after the late 15th century, progressively rerouted toward the Atlantic. The same African political institutions that organised caravan trade also mediated the sale of captives to European shippers, so the African-agency debate treated above cannot be read without the Atlantic context developed in the slave-trade unit.

• Colonialism and imperialism 32.15.01. The systematic bias in the source record, the destruction of archives, the disruption of oral transmission, and the looting of objects such as the Benin Bronzes are colonial processes. The historiographical problem this unit foregrounds is a direct effect of the events analysed in the colonialism unit, and decolonising the curriculum is a response to that causal chain.

• Ancient Egypt and Nubia 32.03.01. The Sudanic state tradition, of which Ghana, Mali, and Songhai are late examples, has deep precedents in the Nile valley and the Meroitic state of Nubia. The relationship between Egyptian, Nubian, and Sudanic political cultures, and Diop's contested claim about the African character of pharaonic civilisation, connects this unit to the Egypt and Nubia material.

• The Mongol empire and Eurasian trade 32.13.01. The Indian Ocean network that sustained Kilwa and Mombasa was the southern arm of a Eurasian commercial system whose northern, overland arm was reorganised under the Mongols. The Ming voyages of Zheng He, which reached the Swahili coast in the early 15th century, belong to the same Indian Ocean world, and the comparison clarifies what was specifically African about the coastal city-states.

• Sub-Saharan African kingdoms (survey) 32.12.01. This unit deepens the companion survey, which covers the full set of kingdoms from Aksum to Kongo and Benin at introductory depth. The present unit treats the Sahelian empires, the Swahili coast, Great Zimbabwe, and the Bantu expansion at greater analytical depth, with the survey as its prerequisite.

Historical & philosophical context Master

The academic study of sub-Saharan African history is a young discipline. Before the 1950s most European-language scholarship on Africa was produced by colonial administrators, missionaries, and travellers whose assumptions were shaped by imperial administration, and African societies were classified by the absence of written records rather than by the presence of oral and material ones.

Cheikh Anta Diop's 1954 Nations negres et culture argued that ancient Egyptian civilisation was fundamentally African and that sub-Saharan Africa had produced complex civilisations systematically denied by European scholarship. His specific claims about the racial identity of the ancient Egyptians remain contested, but his insistence that African civilisations be studied without the lens of European racial hierarchy was influential across the decolonising field.

Jan Vansina's De la tradition orale (1961), revised in English as Oral Tradition (1965) and again as Oral Tradition as History (1985), established the methodology for treating oral traditions as historical sources ^{[Vansina 1985]}. Vansina demonstrated that the source criticism applied to written documents could be adapted to oral materials, with attention to genre, transmission, and the social position of the custodian. The work was initially resisted by historians trained exclusively in the documentary tradition.

The UNESCO General History of Africa, commissioned in 1964 and published in eight volumes between 1981 and 1993 under the scientific direction of an international committee the majority of whose members were African, was the first comprehensive multi-volume history of the continent produced by such a team ^{[UNESCO 1981]}. It was conceived explicitly as a corrective to the Eurocentric histories that had marginalised African civilisations, and its editorial decisions, including the use of internal African periodisations and the integration of oral and archaeological evidence, shaped the field's subsequent standards.

Christopher Ehret's linguistic reconstruction of African population and economic history, synthesised in The Civilizations of Africa (2002), and Roderick McIntosh's work on Middle Niger urbanism, have shown the depth of evidence available when historical linguistics, archaeology, and genetics are combined ^{[Ehret 2002]}. Thomas Spear's 2000 reassessment of Swahili origins is representative of the wider revision that has restored African authorship to coastal and Sahelian urbanism alike ^{[Spear 2000]}. The historiographical dispute between ideological-corrective and source-availability accounts of the under-telling, treated above, remains productive precisely because it keeps the field attentive to its own evidential and political assumptions.

Bibliography Master

@book{ehret2002,
  author = {Ehret, Christopher},
  title = {The Civilizations of Africa: A History to 1800},
  publisher = {University Press of Virginia},
  year = {2002}
}

@book{iliffe1995,
  author = {Iliffe, John},
  title = {Africans: The History of a Continent},
  publisher = {Cambridge University Press},
  year = {1995}
}

@book{reader1997,
  author = {Reader, John},
  title = {Africa: A Biography of the Continent},
  publisher = {Alfred A. Knopf},
  year = {1997}
}

@book{vansina1985,
  author = {Vansina, Jan},
  title = {Oral Tradition as History},
  publisher = {University of Wisconsin Press},
  year = {1985}
}

@book{thornton1998,
  author = {Thornton, John},
  title = {Africa and Africans in the Making of the Atlantic World, 1400--1800},
  edition = {2nd},
  publisher = {Cambridge University Press},
  year = {1998}
}

@book{levtzion2000,
  editor = {Levtzion, Nehemia and Hopkins, J. F. P.},
  title = {Corpus of Early Arabic Sources for West African History},
  publisher = {Markus Wiener Publishers},
  year = {2000}
}

@book{mcintosh1999,
  editor = {McIntosh, Susan Keech},
  title = {Beyond Chiefdoms: Pathways to Complexity in Africa},
  publisher = {Cambridge University Press},
  year = {1999}
}

@article{spear2000,
  author = {Spear, Thomas},
  title = {Early Swahili History Reconsidered},
  journal = {International Journal of African Historical Studies},
  volume = {33},
  number = {2},
  pages = {257--290},
  year = {2000}
}

@book{hunwick1999,
  author = {Hunwick, John O.},
  title = {Timbuktu and the Songhay Empire: Al-Sa'di's Tarikh al-Sudan down to 1613 and other contemporary documents},
  publisher = {Brill},
  year = {1999}
}

@incollection{kizerbo1990,
  author = {Ki-Zerbo, Joseph},
  title = {Introduction},
  booktitle = {UNESCO General History of Africa, Vol. I: Methodology and African Prehistory},
  publisher = {Heinemann/UNESCO},
  year = {1990}
}

@book{connah2001,
  author = {Connah, Graham},
  title = {African Civilizations: An Archaeological Perspective},
  edition = {2nd},
  publisher = {Cambridge University Press},
  year = {2001}
}

@book{inikori1992,
  editor = {Inikori, Joseph E. and Engerman, Stanley L.},
  title = {The Atlantic Slave Trade: Effects on Economies, Societies, and Peoples in Africa, the Americas, and Europe},
  publisher = {Duke University Press},
  year = {1992}
}

@book{lovejoy2011,
  author = {Lovejoy, Paul E.},
  title = {Transformations in Slavery: A History of Slavery in Africa},
  edition = {3rd},
  publisher = {Cambridge University Press},
  year = {2011}
}