AFRICAN POLITIES • COMPARATIVE HISTORICAL RESEARCH

AFRICA’S HISTORICAL WORLDS

A source-critical study of Aksum, Kush/Nubia, Ghana, Mali, Songhai, the Swahili city-states, and Great Zimbabwe

━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━

African oral and written traditions • Arabic texts • inscriptions and coins • archaeology • environmental evidence

Author & Researcher: Shubham Phougat

Research synthesis • October 2026

Shubham Phougat | 1

Executive summary

This report compares seven African historical worlds whose chronologies and evidence bases differ substantially: Aksum in the northern Horn of Africa; ancient Kush/Meroë and the later Christian kingdoms of Nubia along the Middle Nile; Ghana/Wagadu, Mali and Songhai in the western Sahel and Niger basin; the Swahili-speaking city-states on the East African coast; and Great Zimbabwe on the southern African plateau. The cases range from the first millennium BCE to the early modern period. They are not stages in a single civilizational sequence, and no one case can stand in for “Africa.”

The report’s central methodological conclusion is that no source category is sufficient on its own. Royal inscriptions tell us how rulers wanted military success and legitimacy remembered; coins reveal the changing public language of kingship and exchange; Arabic geographers and travellers illuminate trade and political reputation but often depend on informants and inherited conventions; oral epics preserve social memory and political argument, not a verbatim transcript of the distant past; buildings, ceramics, botanical remains, animal bones and settlement patterns recover ordinary production and domestic life largely absent from elite texts. Historical claims are strongest when independent evidence lines converge. This approach follows current African-history methodology, which stresses source criticism and triangulation rather than either dismissing external texts or accepting them as neutral eyewitness records [1–3].

• Aksum was an African Red Sea power. Its inscriptions in Ge’ez, Greek and South Arabian-related scripts, coinage, monumental stelae and the port of Adulis show a state that operated across highland and maritime environments. The conversion of King Ezana is securely associated with fourth-century inscriptions and coinage, but the pace and social depth of Christianisation remain open questions; new scholarship argues that royal religious messaging could be calibrated for different audiences [4–7].
• Kush was not simply an Egyptian appendage. Napatan and Meroitic rulers developed their own royal traditions, religious landscapes, writing, urban systems and political history. Kushite rulers conquered and governed Egypt as its Twenty-fifth Dynasty, while the later Meroitic kingdom built a distinctive capital and royal necropoleis. Meroitic script is readable but the language remains only partially understood. The later Christian kingdoms of Nobadia, Makuria and Alwa belong to a different historical phase and should not be folded into ancient Meroë [8–12].
• Ghana/Wagadu’s importance in trans-Saharan commerce is well supported by Arabic accounts, but the precise geography, capital, political structure and end of the polity remain uncertain. Koumbi Saleh is a leading archaeological candidate, not an indisputable identification. The claim that an Almoravid conquest in 1076 single-handedly destroyed Ghana is not established by the evidence [13–15].
• Mali’s fame rests partly on unusually visible connections: Mansa Musa’s pilgrimage in 1324–25, the reputation recorded by Arabic authors, and his image on the Catalan Atlas of 1375. Yet even for this empire, the location and form of the capital remain debated. The Sundiata epic is a living, multiple-version tradition shaped by griots (jèliw), not a single fixed chronicle [16–20].
• Songhai’s rise around Gao, the expansion under Sunni Ali and Askia Muhammad, the scholarly world of Timbuktu and the 1591 Moroccan invasion are supported by a mix of chronicles, architecture, manuscripts and archaeology. A major recent reassessment of the Timbuktu textual tradition warns that works previously conflated under the title Tarikh al-Fattash are distinct textual entities; editions and translations therefore affect historical interpretation [21–24].
• Swahili city-states were African urban societies whose Bantu-language communities built enduring Indian Ocean connections. Coral-stone mosques and palaces, local pottery, coins, imported ceramics and Arabic inscriptions attest to a cosmopolitan but locally rooted world. Elite genealogies of Persian or Arabian origins must be read as political memory and claims of status, not accepted automatically as demographic descriptions [25–29].
• Great Zimbabwe was built and occupied by local African communities, generally associated with Shona-speaking populations. Its stone architecture, settlement layouts, production systems, cattle economy and

long-distance objects reveal both local economic complexity and Indian Ocean links. Recent archaeological reanalysis rejects models that reduce the political economy to a central ruler redistributing exotic imports; occupation and activity were more varied than a simple “abandoned in 1450” story [30– 33]. A comparison across the seven cases suggests that long-distance trade created opportunities, but did not automatically create or destroy states. Political strength depended on how rulers and communities combined agricultural production, control of routes, military organisation, ritual legitimacy, negotiated local authority and adaptation to ecological conditions. “Decline” usually means a change in political centre, trade corridor or settlement pattern—not the disappearance of a people, culture or historical legacy.

How to read this report

The report uses bracketed reference numbers that point to the bibliography at the end. Dates are ranges where the evidence does not justify precision. “Empire,” “kingdom,” “city-state,” and “state” are analytical labels, not claims that the societies had identical institutions. Modern national borders are not used as direct proxies for premodern political boundaries. Where the evidence remains uncertain, that uncertainty is treated as a finding rather than filled with a plausible-sounding narrative.

Scope and limitations

This is a comparative research synthesis, not a critical edition of every primary text and not an exhaustive excavation catalogue. The source base is uneven: some courts left durable inscriptions and coins, while some regions are known mainly through later chronicles, oral traditions and archaeology. Several important archives and excavation contexts remain access-limited or incompletely published. The report therefore distinguishes high-confidence findings from hypotheses and identifies where additional research would change the picture.

Questions and method

Questions guiding the comparison

• What can local written traditions, Arabic accounts, inscriptions, coins and archaeology tell us—and what can each not tell us?
• How did political authority relate to agriculture, pastoralism, craft production, trade routes and urban settlements?
• How did religious institutions shape legitimacy without making every policy a “religious” policy?
• What is the difference between evidence for contact and evidence for control, migration or cultural replacement?
• What can responsibly be said about transition, fragmentation and abandonment, and what remains conjectural?

A source hierarchy without source chauvinism

There is no universal hierarchy in which written evidence always outranks oral evidence, or archaeology automatically settles every question. Instead, the method is claim-specific. A contemporary inscription is high-value evidence for what the issuing ruler publicly claimed; it is not automatically an objective account of the campaign. A traveller’s itinerary can anchor a date and route; it may still generalise about people he did not meet. A coin can establish a name, title, symbol, minting practice or economic connection; it cannot by itself reveal the full tax system. Oral traditions can preserve names, relationships, moral conflicts and community conceptions of the past; individual versions must be dated to their performance and recording contexts.

Excavated evidence can identify settlement, diet, exchange and technology, but its interpretation depends on sampling, preservation and context [1–3].

Chronological orientation

1. Aksum: a highland kingdom facing the Red Sea

1.1 Geography, formation and political economy

Aksum developed in the northern Horn of Africa, drawing on highland agriculture, craft production and access to the Red Sea through ports including Adulis. Its strength did not derive from “trade” in the abstract: it depended on connecting productive highland zones to coastal exchange, mobilising labour for monumental construction, and projecting authority over corridors that joined the Ethiopian-Eritrean interior with Arabia and the wider Indian Ocean. Aksumite political reach varied over time; a map of maximal influence should not be confused with routine administrative control of every region represented.

The first-century Greek sailing guide conventionally called the Periplus of the Erythraean Sea describes Adulis and the regional exchange in which it participated. Later, Cosmas Indicopleustes’ sixth-century Christian Topography preserves information about Red Sea trade and inscriptions he reported seeing. These sources are valuable but have different purposes: the Periplus is practical commercial geography; Cosmas writes as a Christian author with theological aims. Archaeology has confirmed long-distance contacts while also exposing uncertainties about the exact location and phases of the ancient harbour. Adulis stood inland from the modern shoreline because of coastal geomorphology; the identification of the harbour landscape and associated settlements has required fieldwork, not a literal reading of ancient distances alone [4–6].

Aksum exported or trans-shipped goods including ivory and other products drawn from African hinterlands, while imports helped connect it to Red Sea and Indian Ocean circuits. Foreign objects are evidence of circulation, not proof that the society was culturally or politically foreign. The state’s monumental stelae, elite tombs, palatial remains, inscriptions and coinage show a political centre with its own capacities and public idioms.

1.2 Inscriptions, languages and royal power

Aksumite royal inscriptions are unusually valuable because they survive in Ge’ez, Greek and South Arabian-related writing systems. The language choice is itself historical evidence. It shows a court operating across local, regional and international audiences, not simply a one-directional borrowing from an external civilisation. The fourth-century inscriptions attributed to Ezana record campaigns and claims to divine support. Their versions and languages must be read comparatively: small differences between surviving texts matter, and a royal victory inscription is a crafted statement of kingship rather than a full campaign diary [4, 7].

Coinage adds another line of evidence. Aksum produced gold, silver and copper coins, and the imagery changes over time. Earlier issues used symbols commonly associated with pre-Christian royal religion; later issues use Christian signs and formulas. Coins moved through practical exchanges and public messaging: they made the royal name and symbolic regime portable. A single coin design cannot establish how quickly ordinary households converted, what all provincial communities believed, or whether Christianity was imposed uniformly.

1.3 Ezana and Christianisation: a sequence, not a single switch

King Ezana’s fourth-century reign is closely associated with Aksum’s adoption of Christianity. Earlier inscriptions invoke different divine language and symbols; later texts use explicit Christian language, including formulations concerning the Father, Son and Holy Spirit. The evidence makes royal conversion much more secure than it makes any simple claim about total social conversion. A 2025 article by Aaron Michael Butts, published in 2026, argues that Ezana’s inscriptions and coins complicate the familiar story derived chiefly from later ecclesiastical narratives: royal religious messaging may have been strategically

adapted to different audiences, while Christianity’s spread could have involved social processes beyond a top- down decree [7]. This is a valuable revision, but an argument from inscriptions and coins is not a direct survey of belief across the kingdom.

Ethiopian Christian literary traditions, including later accounts of the kingdom’s origins and sacred history, are important evidence for how communities remembered Aksum and legitimised later institutions. They should not be read as contemporaneous documentary reports for every early event. The distinction is not between “true archaeology” and “false tradition”; it is between different genres, dates and purposes of evidence.

1.4 Transition and unresolved questions

Aksum’s political and commercial position changed across the later first millennium CE. A simplified story that attributes its decline to one invasion, a single environmental disaster or the rise of Islam is inadequate. Red Sea politics, shifts in maritime circuits, internal regional changes and ecological pressures are possible parts of a longer transition. The chronology and scale of these changes vary by site and evidence type. The later history of the Ethiopian highlands must be followed through post-Aksumite political and religious developments rather than treated as a vacuum after Aksum’s monumental phase.

Aksum: confidence assessment

2. Kush, Meroë and medieval Nubia: related landscapes, distinct histories

2.1 Kushite political history and Egyptian power

The Middle Nile supported a succession of African polities whose histories cannot be reduced to Egypt’s southern frontier. The kingdom of Kush developed its own royal centres, religious landscapes and systems of power. During the Napatan phase, Kushite kings conquered Egypt and ruled as the Twenty-fifth Dynasty in the eighth and seventh centuries BCE. They used Egyptian royal forms and religious language, but adoption does not imply that Kush had no political culture of its own. Kushite royal inscriptions articulate campaigns, restoration, causality and the king’s relationship with the gods in ways that should be treated as historical writing, not as a lesser substitute for Greek or Roman historiography [8–10].

The Piye Victory Stela is a particularly important example: it narrates a campaign in Egypt as both military event and legitimate restoration. Such inscriptions establish the political ambition and self-representation of rulers. They also need external checking for details of numbers, motives and outcomes. Kushite control of Egypt was neither a timeless state nor evidence that “Nubia” and “Egypt” were identical. It was a specific imperial episode within a much longer Nile Valley history.

2.2 Meroë as a city, capital and production landscape

Meroë, in the region between the Nile and Atbara systems, became a major capital of Kush. Its archaeological remains include royal cemeteries, palaces, temples, domestic areas and evidence of iron production. Royal pyramid fields at Meroë and other Kushite centres materialise kingship, family, elite hierarchy and mortuary religion; the pattern of large pyramids with smaller surrounding burials can illuminate relationships among rulers, kin and dependants [10–11]. The landscape was not simply a royal monument zone: agriculture, cattle, craft work and the maintenance of urban institutions sustained the court and population.

Ironworking at Meroë has long invited the claim that it was an exceptional “iron centre” or that metalworking alone explains Kushite power. The archaeological evidence confirms iron production, but the scale, commercial reach and causal importance of that industry are more difficult to quantify. Smelting installations are not the same as evidence for a monopoly over iron. The state’s longevity and regional reach must be understood through several linked systems: farming, pastoral resources, river corridors, craft specialisation, political control and religious authority.

2.3 Meroitic writing: deciphered script, incompletely understood language

Meroitic is central to the history of African writing. From the late centuries BCE the kingdom used its own script in monumental and cursive forms, with an alphabetic inventory derived ultimately from Egyptian scripts. The script was deciphered in the early twentieth century, but the language has not been fully understood because there is no extensive bilingual corpus comparable to the Rosetta Stone. Scholars can read many signs and identify formulaic elements and some words, yet interpretation of many texts remains limited [11–12]. This distinction matters: “the writing has been deciphered” does not mean every inscription can be translated with confidence.

The surviving corpus demonstrates local written practice, but the kinds of texts that survive are skewed toward funerary, royal and religious contexts. Everyday administrative and household writing may have existed in forms less likely to survive. We must avoid concluding that the society lacked complex historical or administrative thought because the language remains only partly understood.

2.4 Meroë’s end and the danger of single-cause collapse

Meroitic political power weakened over the later first millennium CE; Aksumite intervention in the fourth century is important, but it cannot automatically be treated as the sole cause of a sudden end to an entire civilisation. The causes, timing and geographic pattern of transition remain debated. Settlement shifts, changing regional exchange, political conflict and environmental history may have interacted differently across the kingdom. A royal centre’s decline is not the same thing as the disappearance of its inhabitants or all cultural practices associated with it.

2.5 Medieval Nubia: Christian kingdoms and Arabic documentary evidence

Several centuries after Meroë, Nubia contained the Christian kingdoms conventionally called Nobadia, Makuria and Alwa (Alodia). Their histories include connections to Byzantine and Coptic Christian institutions, the use of Old Nubian, Greek and Coptic, church construction and painting, and sustained diplomatic and commercial relations with Muslim-ruled Egypt. Makuria’s agreement with the early Islamic caliphate, conventionally known as the Baqt and associated with the unsuccessful siege of 651–652, shaped relations across the frontier. It should not be reduced to a simple statement that Nubia became a tributary in the modern sense: its terms, meanings and implementation have been debated, and trade in enslaved people was part of wider commercial realities [12–14].

The Qasr Ibrim archive has transformed understanding of medieval Nubia. Its legal and administrative texts show landholding, monetisation, private transactions, letters and the practical work of officials. Old Nubian was used in documents and literary/religious translation; this is a key reminder that local language and writing were not displaced in a simple, immediate process by Arabic. Documentary evidence challenges broad claims based solely on external geographers about Nubian society [13–14].

The later transition from Christian rule to Muslim rulers was uneven across communities and regions. The last known Christian Nubian ruler at Qasr Ibrim is attested in an Old Nubian inscription dated to 1483; Soba, capital of Alwa, was in ruins by the early sixteenth century. But this is not an instantaneous civilisational replacement. Languages, settlement communities, religious practices, family networks and political institutions changed at different speeds. A 2026 study of a newly excavated Arabic order linked to King Qashqash at Old Dongola demonstrates how one carefully contextualised document—combined with coins, radiocarbon and other texts—can illuminate administrative life in a period usually treated as poorly documented [15].

Kush/Nubia: confidence assessment

Claim Assessment Reason Royal inscriptions and Egyptian historical

Production evidence exists, but political

Connected geography, but distinct periods, Ancient Kush and Christian Nubia are one Incorrect institutions, languages and external continuous unchanged state relations

3. Ghana / Wagadu: powerful trade networks, uncertain political geography

3.1 The name and the polity

The Ghana Empire is among the earliest well-known political formations of the western Sahel, but it is also among the cases where confident popular narratives outrun the sources. Arabic-language writers mention Ghāna by the ninth century; in medieval texts the word can refer to a ruler, a court, a city or a political region, depending on context. The polity is often linked with Soninke-speaking populations and with the oral traditions of Wagadu. That link is historically useful, but it does not mean that each surviving version of the Wagadu tradition is a direct chronicle of every early medieval event [16–17].

The political landscape was a network of towns, farming communities, pastoral zones and trade routes. The rulers’ power likely depended less on digging all the gold themselves than on negotiating with the areas and networks that brought gold, salt and other goods into exchange. The wealth described by Arabic writers reflects both real commercial importance and literary conventions that often foreground the splendour of distant rulers. It is unwise to translate descriptions of “gold” into an exact modern map of mines or to assume a centralised bureaucracy identical to later empires.

3.2 What al-Bakri can—and cannot—tell us

Al-Bakri’s eleventh-century account is indispensable for the political world of Ghana. Yet he wrote in al-Andalus and relied on information transmitted through merchants and other informants; he was not a resident recording daily life firsthand. His description of the ruler, court, Muslim community, ritual practices and access to gold preserves exceptionally valuable detail, but each detail must be read in its textual context. The account’s distinction between a royal town and a Muslim town is often discussed, but it should not be converted into a rigid blueprint of a two-city state without archaeological and contextual support [16–18].

Arabic geographical writing often categorised polities through the political and commercial questions most important to its authors and audiences. It can tell us that Ghana was recognised as a major power and was connected to Muslim merchants. It is less direct evidence for the social experience of non-elite cultivators, the exact territorial border, or how authority was negotiated across communities.

3.3 Koumbi Saleh: important site, unsettled identification

Koumbi Saleh, in present-day Mauritania, has repeatedly been identified with Ghana’s capital. The archaeological site’s urban remains and material links make it a serious candidate for a major centre in the region. However, the equivalence “Koumbi Saleh = the capital of Ghana” should be stated as a hypothesis, not a settled fact. The Oxford Research Encyclopedia’s review highlights continuing uncertainty about location, state formation, territorial reach and the polity’s demise [16]. This does not mean archaeology is useless; it means that the archaeological site, the geographical names in Arabic sources and the remembered Soninke past have not yet been aligned beyond reasonable dispute.

3.4 The Almoravid story and the problem of dramatic collapse

A popular narrative says that Almoravids conquered Ghana in 1076 and destroyed the empire. This is too neat. North African religious and political expansion, conflict over trade corridors, internal political change and the rise of competing centres are relevant, but the evidence does not establish a single decisive conquest as the exclusive cause of Ghana’s political transformation. A polity can lose regional primacy gradually while

its people and political traditions continue under new arrangements. The fact that Ghana’s ruler was later described as subordinate to Mali does not prove the moment or mechanism of a clean fall [16–18].

3.5 What research could settle next?

The most valuable future work would link archaeological surveys across candidate urban centres and surrounding rural settlements; more secure dating of occupational layers; environmental studies of agricultural and pastoral changes; careful collection and contextualisation of Soninke oral traditions; and re- edition of Arabic sources in their manuscript and textual histories. The aim is not to “find the capital” at any cost but to explain how political authority, trade and settlement were distributed across space.

4. Mali: imperial scale, pilgrimage, oral epic and the problem of the capital

4.1 Rise and imperial structure

The Mali Empire emerged in the thirteenth century in the Mande-speaking region and developed into one of the most influential polities of medieval West Africa. Its reach expanded through relationships among rulers, local authorities, tributary communities, military elites, merchants and religious specialists. It should not be imagined as an evenly governed block of land shaded on a modern map. Imperial power could be strongest along corridors and around key political and commercial nodes, while the status of outer regions shifted with local alliances and dynastic competition [19–20].

Mali benefited from networks that linked gold-producing regions and agricultural zones south of the Sahel to salt supplies, caravan routes and markets north of the desert. Yet “control of trade” did not necessarily mean a ruler owned every mine, taxed every transaction or monopolised each route. Authority was negotiated. Local production, household activity, seasonal movement and regional markets mattered alongside elite exchange. This is one reason archaeology and oral history are essential complements to the Arabic accounts.

4.2 Sunjata/Sundiata and the oral epic

The Sunjata epic is a foundational Mande oral tradition, performed and developed by hereditary specialists known in Manding languages as jèliw. It provides narratives of origin, exile, alliance, struggle, social order and the moral basis of rule. It is simultaneously historical memory, literary performance and political argument. There is no single unchanging “original epic.” Versions recorded from the late nineteenth century into the twentieth century vary in episodes, names, emphasis and political interpretation; those differences are evidence about the communities and performances that produced them [20].

The epic can help recover what written texts often obscure: how a community explained the legitimacy of a dynasty, how power was imagined in relation to kinship and alliance, and how historical prestige was transmitted. But it cannot be used as a literal itinerary for every thirteenth-century event without comparison to other evidence. A responsible reading asks what each episode means in the performance tradition, when a particular version was recorded, and whether independent evidence supports a specific historical claim.

4.3 Mansa Musa and the Arabic evidence

Mansa Musa’s pilgrimage in 1324–25 made Mali visible to a wider Islamic world. The accounts associated with al-Umari, Ibn Battuta and Ibn Khaldun form the core written evidence for Mali at the height of its reputation. These are not interchangeable witnesses. Al-Umari’s 1340s account is based substantially on information gathered in Cairo from people who had encountered Musa and his entourage; Ibn Battuta travelled in Mali in 1352–53 and observed court practices at that later moment; Ibn Khaldun’s dynastic lists preserve a different kind of information and require scrutiny for chronology and transmission [18–19].

Ibn Battuta’s account is especially valuable for what a visitor could observe: court ceremony, roads and travel, hospitality, religious practice and the relationship between local authority and the mansa. But even an eyewitness observes a slice of a society, and the traveller’s own moral and cultural expectations shape what he chooses to record. A critique of his text should neither dismiss him as biased nor treat every statement as transparent fact.

Musa’s portrayal in the Catalan Atlas of 1375 is another form of evidence. It is not an African court painting or a direct portrait from life; it is a Mediterranean cartographic representation that combines geographic knowledge, political reputation and the conventions of manuscript imagery. The image proves that Mali’s

ruler had become a recognisable figure in an international knowledge network. It also shows how external audiences transformed African power into the visual language of their own political culture [22].

4.4 Capital, cities and the limits of certainty

The location and nature of Mali’s capital remain debated. Arabic writers refer to a royal centre, but place names and distances in medieval texts can be difficult to align with modern geography. Candidate sites, including Niani, have attracted substantial scholarship, but no simplistic formula should turn a hypothesis into certainty. The empire may also have used more than one court or seasonal centre, and the political centre need not match one archaeologically obvious city. This uncertainty is not a failure of scholarship: it identifies where future archaeology, environmental evidence, place-name study and source comparison can produce real advances [19].

4.5 Economy, social order and transformation

Mali’s prosperity did not rest on gold alone. It relied on agricultural production in river and savanna ecologies, pastoralism, artisan work, regional markets, salt and trans-Saharan commerce. Political authority involved tribute, appointments, military service and the management of alliances. Islamic scholarship and law had prestige at court and in urban centres, while local religious and social institutions continued to shape life. There is no basis for assuming a uniform Islamisation of every community simply because elite rulers and merchants were Muslim.

The empire fragmented over the late fifteenth and sixteenth centuries, but this was not the disappearance of Mande-speaking peoples or of Mali’s historical memory. Successor formations—including Kaabu and other regional powers—reworked older institutions. Mali remains politically and culturally consequential because oral narratives, dynastic claims, scholarly traditions and modern identities have continued to draw on its memory.

5. Songhai: Gao, Timbuktu, statecraft and the textual history of chronicles

5.1 Gao before imperial expansion

Songhai’s imperial phase grew out of an older political and urban world around Gao and the Niger bend. Gao-Saney and Gao Ancien offer archaeological evidence for settlement, Islamic burial and long-distance connections before the empire reached its later scale. The Niger was not merely a line on a map; it joined agricultural zones, fishing communities, pastoral routes, transport and markets. Political authority depended on access to people and productive landscapes as much as on control of a named trade route.

5.2 Sunni Ali, Askia Muhammad and imperial strategies

Sunni Ali (r. 1464–1492) expanded Songhai’s power through military campaigns and the incorporation of important urban and riverine centres. Askia Muhammad (r. 1493–1528) pursued a different but overlapping political strategy that linked government, Muslim legitimacy, scholarship and trans-Saharan diplomacy. The Tomb of Askia at Gao—traditionally dated to 1495—is a remarkable monumental expression of the empire and of the adaptation of local Sudano-Sahelian earthen architecture to Islamic needs. Its permanent projecting timber elements are not just decoration: they support the tradition of regularly replastering and maintaining earthen buildings [23].

It is misleading to divide Songhai into a “secular” Sunni Ali phase and a wholly “religious” Askia phase. In both periods, legitimacy, coercion, administration, commerce and religious institutions interacted. The rulers’ religious policies were important, but so were military organisation, control of river routes, relations with local elites and the practical demands of holding a large and diverse territory.

5.3 Timbuktu and the manuscript world

Timbuktu’s reputation rests not only on trade but on scholarship, teaching, legal practice, manuscript production and networks that connected local scholars with wider Islamic intellectual life. Arabic was both a transregional learned language and a language of locally produced histories and documents. Manuscripts reveal African scholarly agency: people in the Sahel wrote about their own rulers, cities, scholars, moral controversies and historical experiences within Islamic genres. Their use of Arabic does not make these records “foreign sources.” They are African-authored sources embedded in an international intellectual language.

The chronicles must nevertheless be read critically. Al-Sa‘di’s Tarikh al-Sudan was compiled in the seventeenth century, after many events it narrates, by an author with his own political and scholarly setting. It preserves indispensable traditions and biographies but also selects and evaluates the past from the vantage point of later Timbuktu and Songhai politics. The Tarikh al-Fattash textual tradition is even more important for lessons in editing: a new English-language edition and study published in 2025, reviewed in 2026, argues that two distinct works—identified as Tarikh Ibn al-Mukhtar and Tarikh al-Fattash—were merged in the 1913 French translation. Textual conflation can therefore produce historical claims that appear to come from one coherent chronicle but actually depend on editorial history [21, 24].

5.4 The 1591 invasion and political afterlives

The Moroccan invasion of 1591 severely disrupted Songhai’s imperial centre. Firearms were a factor, but a one-line explanation of “guns defeated Songhai” makes technology do too much work. The expedition faced distance, logistics and supply constraints; Songhai’s internal dynastic struggles and regional relationships

shaped the outcome; and the post-invasion political order did not exercise the same reach as the imperial system. The invasion changed the balance of power rather than wiping out the region’s societies, towns, languages or scholarly traditions.

Gao, Timbuktu and Djenné retained historical importance under altered political arrangements. Songhai political memory was re-used by later states and by manuscript authors. The “end” of an empire should therefore be distinguished from the end of an urban, intellectual or social world.

5.5 What remains uncertain

Important questions concern the reach of direct administration, the role of provincial authorities, the fiscal and labour systems of the empire, the relative autonomy of towns, and the economic effects of the invasion across different subregions. Arabic chronicles provide elite views; archaeology and manuscript studies can recover the spatial and documentary details that the narrative sources omit. The best new work will combine manuscript codicology, Arabic textual criticism, archaeometry, settlement survey and local oral memory rather than relying on one grand chronicle.

6. Swahili city-states: African urbanism in the Indian Ocean

6.1 Language, coast and hinterland

The Swahili coast was a chain of ports and urban communities, not a single empire. Settlement histories varied from island to island and from northern Somalia/Kenya through Tanzania and into Mozambique. Swahili is a Bantu language with substantial Arabic and other loanwords, shaped by long-term contact; describing it as a simple “mixture of Bantu and Arabic” misstates its linguistic structure and the local roots of its speakers. The city-states developed through the interaction of African coastal communities, farming and fishing, craft production, inland exchange, and maritime networks across the Red Sea and Indian Ocean [25– 28].

Imported ceramics, glass, beads and other goods reveal contact with Arabia, Persia, India and China, but they do not by themselves demonstrate demographic replacement or political control by foreigners. Local ceramics, house forms, food remains and settlement sequences show how towns developed from regional communities. The coast was connected, but it was not merely a line of foreign trading stations.

6.2 Kilwa: archaeology, coinage and maritime links

Kilwa Kisiwani and Songo Mnara offer unusually visible evidence of urbanism: coral-stone and lime-mortar buildings, mosques, palatial complexes, streets, residences, public spaces and burial grounds. Kilwa was occupied from the ninth to the nineteenth century and reached a major prosperity peak in the thirteenth and fourteenth centuries. Its trade connected inland gold and ivory to Indian Ocean goods; it minted its own currency during the eleventh to fourteenth centuries. The Great Mosque and Husuni Kubwa palace attest to local architectural skill and social differentiation [26].

Ibn Battuta visited Kilwa in 1331–32 and praised the city. His account is valuable for the urban atmosphere and political setting of one particular moment, but it is a traveller’s description rather than an archaeological inventory. Excavations and imported objects can test some aspects of the account, while also revealing daily life not described by the visitor. Kilwa’s relationship with the gold trade should be described as a network of coast-hinterland connections, not as evidence that the city directly ruled all gold-producing regions.

6.3 The Kilwa Chronicle and origin stories

The Kilwa Chronicle survives through Arabic and Portuguese versions with textual problems. Its dynastic narratives include claims about rulers’ origins, including traditions of Persian descent from Shiraz. These are significant statements of prestige, connection and political identity; they are not automatically a census of migration or proof that the city was founded by foreign settlers. Historians have debated the relationship between the versions and the transmission history of the Arabic text; one scholarly critique warns that treating the Portuguese version as a direct transcript of the surviving Arabic account creates unresolved problems in king lists [27].

The correct question is not whether Swahili societies were “African or foreign.” They were African societies deeply connected to an oceanic world, and identity could be locally grounded while drawing on transregional Islamic, commercial and genealogical language. Cosmopolitanism and local rootedness were compatible.

6.4 Islamic institutions, class and unequal exchange

Islamic practice linked Swahili elites to networks of law, learning, pilgrimage and commerce. Mosques, inscriptions and burial practices show that Islam became embedded within local urban life; the built environment also reflects local construction knowledge and adaptation to coastal materials. Social differences within cities likely involved lineage, access to trade, household labour, gender and political office. The exact nature and scale of slavery, dependent labour and servitude varied over time and place. Later large-scale patterns of the East African slave trade, especially under Omani/Zanzibari and nineteenth-century conditions, should not be projected unchanged across the whole medieval era.

6.5 Portuguese intervention and changing ports

Portuguese expansion into the Indian Ocean from the late fifteenth century altered the strategic environment. Armed interference, fortification, competition over routes and port-specific political choices changed the relative position of coastal towns. But “Portuguese arrival caused all Swahili decline” is too broad: cities had different trajectories, and local dynastic disputes, shifting commercial opportunities and hinterland relations also mattered. Kilwa’s sixteenth-century fort and subsequent history are part of that story, not a complete explanation for every city along the coast.

Swahili coast: confidence assessment

7. Great Zimbabwe: monumental stonework, cattle and regional political economy

7.1 A locally built African capital

Great Zimbabwe is a monumental settlement on the Zimbabwe plateau, built chiefly between the eleventh and fifteenth centuries in the conventional chronology. Its Hill Complex, Great Enclosure and Valley Ruins show sophisticated dry-stone construction, spatial differentiation and a long sequence of occupation. The city is widely associated with local Shona-speaking populations and with the wider Zimbabwe tradition of southern African societies. Its architecture should be understood in relation to the plateau’s local settlement systems, political institutions, cattle economy and ritual landscapes—not as an unexplained anomaly [30–32].

The false or colonial-era claims that Great Zimbabwe must have been built by outsiders, ancient Near Eastern colonists or an imagined lost foreign civilisation have been rejected by archaeological evidence. That correction does not mean every interpretation of the site is settled. The history of excavation matters: early antiquarian work removed objects, damaged contexts and imposed assumptions about gold and external origins. Later archaeological and African scholarly work has refined chronology and challenged models built from poorly contextualised finds [30–32].

7.2 The political economy beyond gold and porcelain

Gold and imported glass beads, ceramics and other objects demonstrate Great Zimbabwe’s participation in broader networks, including routes to the Indian Ocean. Kilwa coins and imported ceramics are part of the evidence. Yet low-frequency imported luxuries cannot alone explain why the city was large, powerful or architecturally monumental. Most people’s lives depended on more frequent activities: crop production, cattle, domestic craft, food preparation, labour, regional exchange and obligations to political authorities. Archaeological work on cattle and settlement has helped move the discussion away from an exclusively long-distance trade model [30–33].

Shadreck Chirikure’s 2020 review argues for a more mixed, household-centred and seasonally variable political economy in which leaders could claim shares of production and circulation but did not necessarily command a fixed redistribution system. Local and interregional exchange in grain, livestock, cloth, meat, skins and other perishable goods is harder to trace than imported porcelain, precisely because organic materials preserve unevenly. The visibility of exotic objects is therefore a preservation and sampling problem as well as an economic fact [30].

7.3 Architecture, sacred authority and social differentiation

The stone walls are not only defensive structures. They mark spaces, restrict and channel movement, display skilled labour and materialise rank and authority. The Great Enclosure and hilltop complexes were likely used and reworked by successive leaders rather than being a single planned complex finished all at once. Interpretations assigning every enclosure a fixed universal meaning—king’s house, priestly zone, wives’ quarters or a single ritual institution—have been criticised for exceeding what the architectural and stratigraphic evidence can demonstrate [31–32].

Local concepts of land, ancestors, sacred authority and custodianship can help frame interpretation, but they must be used carefully. Modern oral histories and ethnographic analogies are not unmediated time machines. Their value increases when they are tied to archaeological sequences, local community knowledge and transparent argument about gaps between present and past.

7.4 Abandonment, reoccupation and successor states

The conventional date around 1450 is a useful shorthand for the movement of political focus away from Great Zimbabwe, but it can be misleading if used to imply that everyone left at once or that the site never mattered again. Recent chronology and the study of legacy collections suggest cycles of expansion and contraction, with activity in surrounding unwalled settlements and some evidence of use into the seventeenth century. The scale of reoccupation and its relationship to later communities still require refinement [30].

The rise of Khami and other Zimbabwe tradition centres, and the development of states associated with Torwa and Mwene Mutapa/Mutapa, indicate a reconfiguration of power rather than cultural extinction. Portuguese documents from the sixteenth century and later are helpful for successor polities, but they do not straightforwardly describe Great Zimbabwe itself. Misidentifying references to a Mutapa capital as direct descriptions of Great Zimbabwe can create a false documentary certainty [30, 33].

7.5 Archaeological and conservation priorities

The priority is to publish and reassess excavation contexts, increase radiocarbon sampling of both walled and unwalled settlements, connect archaeobotanical and zooarchaeological data with settlement pattern, and work with local communities about the site’s continuing spiritual and national meaning. The stonework is not merely the relic of a vanished political order; it remains a place of living memory, tourism, conservation challenge and cultural significance.

8. Cross-case analysis: what comparison reveals

8.1 Trade mattered, but trade was not destiny

All seven cases were connected to networks larger than their immediate political cores. Aksum linked highland production to the Red Sea; Kush operated across the Nile and its adjoining ecologies; Ghana, Mali and Songhai worked through trans-Saharan corridors and Sahelian/Nigerian production zones; Swahili towns connected coastal and inland exchanges to the Indian Ocean; Great Zimbabwe linked regional plateau economies to routes that reached the coast. But the mechanism differed. Aksum minted coins and projected royal power across maritime interfaces; Sahelian rulers built relationships among local political units and long-distance merchants; Swahili cities were port polities with their own urban rivalries; Great Zimbabwe drew power from a combination of cattle, agriculture, labour, craft and regional exchange.

A crucial distinction is between proximity to trade and control of trade. A polity might profit from market access without owning the source of a commodity; the movement of gold does not prove that a ruler directly controlled every gold field. Likewise, imported ceramics at a settlement establish contact, but not necessarily foreign occupation, elite dependence or an economy dominated by imports. Trade routes are political opportunities; how rulers use them is a separate question.

8.2 State formation and uneven sovereignty

“Empire” can imply a uniform bureaucracy that the evidence often does not support. These polities may have used layered authority: royal households, provincial rulers, tribute relationships, ritual specialists, urban councils, military leaders and local communities. Control likely varied by ecological zone and distance from the capital. The evidence for Mali’s capital remains uncertain; Ghana’s political geography is disputed; Great Zimbabwe’s state boundaries are unknown; and the Swahili coast was never one unified polity. In each case, uncertainty about boundaries should not be replaced by overly precise maps.

8.3 Religion: legitimacy, knowledge and social practice

Christianity in Aksum, Kushite and Nubian royal religion, Islam in the Sahel and on the Swahili coast, and local religious/ancestral practices around Great Zimbabwe all shaped how authority was understood. Religious affiliation could connect rulers to transregional networks and give authority a language of justice, divine protection, pilgrimage or sacred custodianship. Yet religion does not mechanically explain political events. Conversion could be strategic, social, gradual, negotiated or regionally uneven. Religious monuments are evidence of institutions and public commitment; they cannot alone show the beliefs of every inhabitant.

8.4 Cities as social spaces, not just capitals

The archaeological evidence encourages us to move from capital-centred history to the households and work that sustain cities. Great Zimbabwe’s outer settlements, cattle and food production; the residential and port architecture of Kilwa and Songo Mnara; domestic and palatial remains at Meroë; manuscripts and trading communities around Timbuktu; and the urban settlement systems of Gao and Ghana’s candidate centres all show that political power was embedded in material life. The non-elite record is often less visible than the royal record, but absence from elite texts is not absence from history.

8.5 Gender, labour and hierarchy

Court accounts and royal monuments often foreground male rulers and military events. That is a bias of survival and genre, not a transparent picture of all social power. Kushite royal cemeteries include ruling queens and royal women; royal and household labour supported monumental architecture; markets and craft production depended on people rarely named in chronicles; and oral traditions may encode women’s roles in lineage, marriage, alliance and conflict. Slavery, dependent status, tribute and labour obligations must be treated as part of the history of these societies rather than ignored in a celebratory narrative. But practices should be dated and located rather than assumed identical across eras.

8.6 Ecology and climate: conditions, not automatic causes

Agriculture, rainfall variability, river dynamics, pasture and forest resources affected every case, but environmental explanations are strongest when connected to local evidence. A drought proxy by itself does not prove state collapse. Flooding, sedimentation, crop changes or deforestation may alter settlement opportunities, but these interact with trade, conflict, leadership and household choice. At Great Zimbabwe, deforestation and food supply have been proposed to explain the capital’s movement; newer evidence of varied occupation makes a single-cause abandonment story less secure. The same caution applies to Aksum, Meroë and Sahelian states: environmental change is a condition to investigate, not a universal answer.

8.7 Military power and technological determinism

Campaigns and invasions matter, but “one weapon defeated a civilisation” is rarely adequate. Songhai’s defeat in 1591 involved firearms as well as logistical distance, internal political dynamics and the limitations of occupation. Aksumite and Kushite inscriptions emphasise military victory because rulers wanted their power remembered; these records should be compared with political and material consequences rather than treated as complete histories. After conquest, many cultural and economic systems continued under new authorities.

8.8 The source-language problem: outside versus inside

“Arabic source” does not equal “foreign source.” Al-Bakri wrote from al-Andalus about Ghana; Ibn Battuta was a traveller; al-Sa‘di wrote in Timbuktu; Swahili intellectuals used Arabic for local dynastic and religious histories; Nubian communities wrote in Old Nubian as well as Greek and Coptic; Aksumite rulers used Ge’ez and Greek; Kushite rulers wrote monumental texts about their own past. The author’s relationship to the

society, the date, the audience and the genre are more informative than a binary native/foreign label. Similarly, oral tradition is not a weak substitute for writing; it is a source with its own methods, custodians and historical affordances [1–3].

9. Comparative source matrix: strongest evidence and main cautions

10. Conclusions and research priorities

The most defensible comparative conclusion is not that these polities all followed the same path, but that African state formation took multiple forms. Monumental writing and coins in Aksum, royal texts and pyramids in Kush, Christian Nubian documents, Sahelian Muslim chronicles, Swahili urban architecture, and Great Zimbabwe’s stone-built landscape are different material forms of political knowledge. They show that these worlds were connected to wider networks without being reducible to outside influence.

The methodological lesson is equally important. Where an Arabic chronicler, local oral tradition and archaeological evidence do not align, the disagreement should be investigated rather than erased. A late written chronicle may preserve an old memory; a monument may have more than one phase; imported porcelain may reflect long-term circulation but not the scale of everyday exchange; a royal inscription may be

persuasive political rhetoric rather than neutral report. A better history usually results from preserving these distinctions.

Priority questions for the next phase of research

• Aksum and Adulis: publish and compare inscriptional, numismatic and port-archaeology datasets; investigate the changing relation between the highland capital and maritime access.
• Kush/Meroë: improve contextual publication of settlement and production archaeology; strengthen linguistic interpretation without claiming full translation of Meroitic; compare environmental and political chronologies.
• Medieval Nubia: connect Old Nubian archives, Arabic legal documents, church inscriptions, coins and radiocarbon dating to map transition region by region rather than with one general “fall” date.
• Ghana/Wagadu: survey rural settlements and candidate centres; test the chronology and geography of Koumbi Saleh; record oral traditions with performer, community, language and collection context clearly documented.
• Mali: continue regional archaeology in candidate capital landscapes; compare multiple Sunjata epic versions; distinguish oral-political memory from claims that can be anchored to thirteenth-century events.
• Songhai: use critical Arabic editions, manuscript provenance and codicology; distinguish different chronicle works and redactions; pair urban archaeology at Gao and Timbuktu with the regional landscape.
• Swahili city-states: expand research on local food production and hinterland exchange; combine inscriptions, coins, household archaeology and oral histories; resist over-generalising from one city’s chronology.
• Great Zimbabwe: refine radiocarbon dates for walled and unwalled settlements; study livestock mobility, crops and craft beyond elite contexts; integrate community stewardship and oral knowledge in interpretation. The final aim is not to manufacture certainty. It is to make the uncertainty legible, show which evidence supports each claim, and demonstrate that the histories of African societies become richer—not weaker— when they are studied critically and comparatively.

Selected bibliography and research trail

The bibliography prioritises peer-reviewed research, university-press scholarship, primary-source editions, UNESCO documentation and open scholarly repositories. It is selective rather than exhaustive. Links are included where stable public landing pages or identifiers are available. Arabic titles are transliterated in commonly used scholarly forms; transliteration varies by edition. [1] Mobley, Christina. “Documentary Sources and Methods for Precolonial African History.” Oxford Research Encyclopedia of African History, 2018. DOI: https://doi.org/10.1093/acrefore/9780190277734.013.150. Methodological framework for critically triangulating documentary evidence with archaeology, oral history, linguistics and environmental research. [2] Schoenbrun, David. “Early African Pasts: Sources, Interpretations, and Meanings.” Oxford Research Encyclopedia of African History, 2018. DOI: https://doi.org/10.1093/acrefore/9780190277734.013.147. On writing early African history with archaeology, linguistics, palaeoenvironment and oral traditions. [3] Levtzion, Nehemia, and J. F. P. Hopkins, eds. Corpus of Early Arabic Sources for West African History. Revised edition. Princeton: Markus Wiener, 2000. Major translated collection for early Arabic accounts of West Africa. [4] Phillipson, David W. Foundations of an African Civilisation: Aksum and the Northern Horn, 1000 BC–AD 1300. Woodbridge: James Currey, 2012. Core synthesis integrating Aksumite inscriptions, coins and archaeology. [5] Peacock, David, and Lucy Blue, eds. The Ancient Red Sea Port of Adulis, Eritrea: Report of the Eritro-British Expedition, 2004–5. Oxford: Oxbow Books, 2007. DOI: https://doi.org/10.2307/j.ctvh1dpst. Fieldwork and reassessment of Adulis and its maritime landscape. [6] Casson, Lionel, trans. and ed. The Periplus Maris Erythraei: Text with Introduction, Translation, and Commentary. Princeton: Princeton University Press, 1989. Edition and commentary for the first-century Red Sea/Indian Ocean sailing guide. [7] Butts, Aaron Michael. “Ezana of Aksum: The First Christian African King.” Aethiopica 28 (2025 issue; online 2026), 7–38. DOI: https://doi.org/10.15460/gfj9t143. Recent reinterpretation of Ezana’s conversion using inscriptions and coins. [8] Emberling, Geoff, and Bruce Beyer Williams, eds. The Oxford Handbook of Ancient Nubia. Oxford: Oxford University Press,

2021. DOI: https://doi.org/10.1093/oxfordhb/9780190496272.001.0001. Recent multi-author synthesis of Nubian archaeology, history and epigraphy. [9] “History and the Kushite Royal Inscriptions.” In The Oxford Handbook of Ancient Nubia, 2021, pp. 395–410. Study of Kushite inscriptions as a form of historical writing and political memory. [10] Grzymski, Krzysztof. “The City of Meroe.” In The Oxford Handbook of Ancient Nubia, 2021, pp. 545–561. DOI: https://doi.org/10.1093/oxfordhb/9780190496272.013.26. Urban layout and development of Meroë. [11] Rilly, Claude. “Meroitic Writing.” UCLA Encyclopedia of Egyptology, 2022. https://escholarship.org/uc/item/2ts0n9p0. On script, language and interpretive limits. [12] Edwards, David N. “Meroe and the Sudanic Kingdoms.” Journal of African History 39, no. 2 (1998). Contextualises Kush and the Middle Nile in African history. [13] Ruffini, Giovanni R. Medieval Nubia: A Social and Economic History. Oxford: Oxford University Press, 2012. DOI: https://doi.org/10.1093/acprof:oso/9780199891634.001.0001. Uses Qasr Ibrim archives to reconstruct property, law, trade and social relations in medieval Nubia. [14] Khan, Geoffrey. Arabic Documents from Medieval Nubia. Cambridge: Cambridge University Press, 2024. Edition and discussion of Arabic documentary materials from Nubian archives. [15] “The King of Nubia at Work: Archaeological Context and Text Edition of a Sixteenth/Seventeenth-Century Arabic Document from Old Dongola.” Azania: Archaeological Research in Africa 61, no. 2 (online 2026), 349–374. DOI: https://doi.org/10.1080/0067270X.2026.2615518. New document contextualising the later political history of Dongola. [16] Gestrich, Nikolas. “The Empire of Ghana.” Oxford Research Encyclopedia of African History, 2019. DOI: https://doi.org/10.1093/acrefore/9780190277734.013.396. Key synthesis of uncertain geography, Wagadu traditions and trans-Saharan trade. [17] Gomez, Michael A. African Dominion: A New History of Empire in Early and Medieval West Africa. Princeton: Princeton University Press, 2018. Broad reinterpretation of West African empire and regional systems. [18] al-Bakri. Account of Ghana, in Levtzion and Hopkins, Corpus of Early Arabic Sources for West African History, revised ed.,

2022. Principal translated source for eleventh-century Ghana. [19] Canós-Donnay, Sirio. “The Empire of Mali.” Oxford Research Encyclopedia of African History, 2019. Open-access accepted manuscript via UCL Discovery: https://discovery.ucl.ac.uk/id/eprint/10070150/. DOI: https://doi.org/10.1093/acrefore/9780190277734.013.266. Critical synthesis of oral, Arabic, European and archaeological evidence, including uncertainty around the capital.

[20] Bulman, Stephen, Stephen Belcher, and Valentin Vydrin, eds. Corpus of Early Accounts of the Sunjata Epic, 1889–1959. Oxford: Oxford University Press / British Academy, 2023. DOI: https://doi.org/10.5871/bacad/9780197267387.001.0001. Multiple recorded versions of the Sunjata epic and their contexts. [21] Hunwick, John O. Timbuktu and the Songhay Empire: Al-Sa‘di’s Tarikh al-Sudan down to 1613 and Other Contemporary Documents. Leiden: Brill, 1999/2003. Standard English scholarly access to the Tarikh al-Sudan and related documents. [22] Smarthistory. “The Catalan Atlas.” https://smarthistory.org/catalan-atlas/. Contextual discussion of Mansa Musa in the 1375 atlas; the image is an external representation, not a direct African court portrait. [23] UNESCO World Heritage Centre. “Tomb of Askia.” https://whc.unesco.org/en/list/1139/. Monument, architecture, chronology and conservation context for Gao. [24] Nobili, Mauro, Zachary Wright, and Ali Diakite, eds./trans. The Chronicles of Two West African Kingdoms: The Tarikh Ibn Al-Mukhtar of the Songhay Empire and the Tarikh Al-Fattash of the Caliphate of Hamdallahi. The British Academy, 2025. Review by Rahmane Idrissa, African Affairs, published online 2026. DOI of review: https://doi.org/10.1093/afraf/adag025. Recent critical edition distinguishing textual traditions previously merged in a French translation. [25] Horton, Mark, and John Middleton. The Swahili: The Social Landscape of a Mercantile Society. Oxford: Blackwell, 2000. Broad synthesis of Swahili society, urbanism and commerce. [26] UNESCO World Heritage Centre. “Ruins of Kilwa Kisiwani and Ruins of Songo Mnara.” https://whc.unesco.org/en/list/144/. Archaeology, architecture, trade, chronology and conservation. [27] Saad, Elias. “Kilwa Dynastic Historiography: A Critical Study.” History in Africa 41 (2014). Critical reading of Arabic and Portuguese versions of the Kilwa Chronicle. [28] Horton, Mark C. Shanga: The Archaeology of a Muslim Trading Community on the Coast of East Africa. British Institute in Eastern Africa Memoir 14. London/Nairobi, 1996. Detailed archaeological study of a coastal Swahili settlement and its changing urban history. [29] Kusimba, Chapurukha M. The Rise and Fall of Swahili States. Walnut Creek, CA: AltaMira Press, 1999. Archaeological and historical synthesis of Swahili political and urban development. [30] Chirikure, Shadreck. “New Perspectives on the Political Economy of Great Zimbabwe.” Journal of Archaeological Research 28 (2020): 139–186. DOI: https://doi.org/10.1007/s10814-019-09133-w. Open-access review challenging narrow redistribution/exotic- import models and reassessing occupation chronology. [31] Chirikure, Shadreck, and Innocent Pikirayi. “Inside and Outside the Dry Stone Walls: Revisiting the Material Culture of Great Zimbabwe.” Antiquity 82, no. 318 (2008): 976–993. DOI: https://doi.org/10.1017/S0003598X00097726. Reassessment of architecture, stratigraphy and material evidence. [32] Pikirayi, Innocent. The Zimbabwe Culture: Origins and Decline of Southern Zambezian States. Walnut Creek, CA: AltaMira Press, 2001. Regional framework for Mapungubwe, Great Zimbabwe and successor states. [33] UNESCO World Heritage Centre. “Great Zimbabwe National Monument.” https://whc.unesco.org/en/list/364/. Site description, architecture, traditional significance and recognised chronology. [34] UNESCO World Heritage Centre. “Khami Ruins National Monument.” https://whc.unesco.org/en/list/365/. Successor political centre in the wider Zimbabwe tradition. [35] Niane, D. T., ed. General History of Africa, Volume IV: Africa from the Twelfth to the Sixteenth Century. UNESCO / Heinemann,

1984. Broad comparative reference; older synthesis should be read alongside newer regional work. [36] Mokhtar, G., ed. General History of Africa, Volume II: Ancient Civilizations of Africa. UNESCO / Heinemann, 1981. Wide-ranging synthesis of ancient African societies; supplement with recent excavation and epigraphy studies. [37] UNESCO World Heritage Centre. “Aksum.” https://whc.unesco.org/en/list/15/. Site overview of monuments, inscriptions and Aksumite connections. [38] Black, Stephanie L. “In the Power of God Christ”: Greek Inscriptional Evidence for the Anti-Arian Theology of Ethiopia’s First Christian King. Bulletin of the School of Oriental and African Studies 71, no. 1 (2008): 93–110. DOI: https://doi.org/10.1017/S0041977X08000062. Close reading of Ezana’s Greek inscription and theological language. [39] Uhlig, Siegbert. “Eine trilinguale ʿEzana-Inschrift.” Aethiopica 4 (2001): 7–31. DOI: https://doi.org/10.15460/aethiopica.4.1.488. Text-critical comparison of trilingual Ezana inscription witnesses. [40] Chirikure, Shadreck, et al. “Debating Great Zimbabwe.” Azania: Archaeological Research in Africa 46, no. 2 (2011): 221–231. DOI: https://doi.org/10.1080/0067270X.2011.580149. Debate about Portuguese sources, chronology and interpretive methods. [41] Rilly, Claude. “Meroitic Language.” Wiley / Encyclopedia of Ancient History, first published 2021. DOI: https://doi.org/10.1002/9781119399919.eahaa00443. Concise discussion of script and language status. [42] “Early Christianity in East Africa and Red Sea/Indian Ocean Commerce.” African Archaeological Review 31 (2014): 637–647. DOI: https://doi.org/10.1007/s10437-014-9172-5. Review of literary and archaeological evidence for Red Sea commerce and early Christianity.