Show Summary Details

Page of

PRINTED FROM OXFORD HANDBOOKS ONLINE ( © Oxford University Press, 2018. All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a title in Oxford Handbooks Online for personal use (for details see Privacy Policy and Legal Notice).

Subscriber: null; date: 13 August 2020


Abstract and Keywords

This article examines early urban societies in Africa. It emphasizes three key issues: the strikingly wide geographical range and structural variety of urban forms; the apparent dichotomy between more hierarchical and more heterarchical urban societies; the contrasting functions of towns in the service of state formation or inter-regional exchange. The earliest cities in Africa are linked to the great rivers of the continent, in particular the Nile and the Niger. There have also been significant urban expressions along the Mediterranean seaboard, or on the Red Sea and East African coast, where contact with neighbouring civilizations was part of the context. Yet, African urban forms assume a dazzling array of expressions, confounding traditional expectations of normative Old World archetypes of what defines ‘urban’.

Keywords: urban societies, urban forms, towns, state formation, ancient African cities, urbanization

With the exception of the Mediterranean and Nilotic civilizations, there has been little acknowledgement of pre-Islamic, or pristine urbanization in Africa. We believe this to be a notable omission in studies of global cities. In this chapter, we review first some thematic and definitional questions and then present brief surveys of several of the early urban societies in Africa. Three key issues are stressed from the outset: the strikingly wide geographical range and structural variety of urban forms; the apparent dichotomy between more hierarchical and more heterarchical urban societies; the contrasting functions of towns in the service of state formation or inter-regional exchange.

Regarding the localization and concentration of urbanism, it is not surprising to find that the earliest cities in Africa (see Regional Map i.3) are linked to the great rivers of the continent, in particular the Nile and the Niger. There have also been significant urban expressions along the Mediterranean seaboard, or on the Red Sea and East African coast, where contact with neighbouring civilizations was part of the context.1 Yet, African urban forms take on a dazzling array of expressions, confounding traditional expectations of normative Old World archetypes of what defines ‘urban’.

Definitions and Methodological Problems

Early urbanization in Africa has been treated as exceptional (Egypt and the Nile), as due to Mediterranean civilizations (Phoenicians, Greeks, and Romans) or to have begun only as a result of external trade stimulus in Islamic times (most others). This chapter presents the now substantial evidence for indigenously developed, pre-Islamic urban centres in Africa. The emergence of urban societies on the Upper Nile in Sudan has long been recognized (Kerma, kingdom of Kush/Meroe), but we now know that there were also towns in the Western Desert of Egypt and in the Central Sahara as well as in West Africa around the Inland Niger Delta and the Niger Bend from the late 1st millennium (p. 67) bce. What is clear, however, is that these early urbanizations in Africa were not part of a homogeneous process and there are significant local particularities that set these early urban societies off from one another and from contemporaneous Mediterranean urbanization to the north. Current definitions of early urban societies cannot be easily applied to all these African examples.2 By ‘pre-Islamic’ we intend not a specific calendar date, but rather the epochs prior to the first influence of the Islamic (or European) worlds. As such, relatively pristine and isolated polities of early 2nd millennium ce date, such as Ife are referred to in this chapter, though the bulk of our examples relate to the late 1st millennium bce and the 1st millennium ce.3


figure 4.1 Comparative plans at two different scales of a number of African urban and proto-urban sites, with settlement areas shaded. The largest enceintes at Ife, Kano, and Ibadan are of post-Islamic date, but are included for comparison with the extended settlement complexes around Dia and Jenné-jeno (compiled from various sources, montage D. Mattingly).

A full survey of all examples of early urbanism in Africa is beyond the scope of this chapter, but we have selected a range of examples of the varied forms of urban settlement that are to be found. Coming up with a definition of urbanisation that neatly encompasses all of our case studies is impossible, because of their geographical and cultural (p. 68) diversity. The urban footprint is not a reliable measure as very different constraints operated on the size of settlements and localized populations, for instance, in the central Sahara in comparison to the sub-Saharan zones (Fig. 4.1). For some of the societies under review there is literary evidence to support the urban ascription (Egypt, the Maghreb), but many of our examples relate to sites whose urban function must be derived solely from archaeological evidence. For these reasons it is impractical to make distinctions between towns and cities. We define ‘urban’ in broad terms: nucleated settlements, larger in terms of their extent and population in comparison with other contemporary sites, possessing allied hinterlands for their subsistence support, often with evidence of specialized manufacturing, trade, political, and/or religious activity. The sites may have a monumental aspect, again contrasting with lower echelon settlements. In the next section, we also discuss a particular distinction that can be made between the communities behind some of these early urban experiments: hierarchical and heterarchical societies.

Hierarchy and Heterarchy

Many instances of Africa urbanization seem to be linked to state formation processes, though there was considerable variability in the sort of society that emerged. Some might be termed city-states (micro-states) and others as forming elements of territorial (or macro-) states.4 At first glance African urbanization may be regionally divided between societies where pronounced and coercive hierarchies were emergent (Egypt, Garamantes), and others where more heterarchical processes were at work, sometimes encapsulated within wider hierarchical structures (the Middle Niger). Heterarchy may be defined as a mode of organization whereby power is spread horizontally, rather than vertically, with lineages and specialist groups (ideological and technological) negotiating corporate decisions.5 However, upon closer examination, degrees of hierarchy across early urban sites probably were historically and politically contingent. For example, recent historical investigations in the Segou region of Mali have revealed a strong folk taxonomy dividing urban centres into long-lived, semi-autonomous mercantile centres (Markadugu) and those created by the power of the state (Fadugu and Dendugu).6 Many heterarchically organized Middle Niger cities, such as Jenné, survived multiple periods of immersion in state territories while retaining relatively robust self-governance. The granting of such autonomy was linked equally to the fear of ‘killing the golden goose’—the economic prosperity generated by the long-established trade networks of such cities—and the reputed supernatural (‘eternal’) reputations of the places themselves. Large settlements whose raison d’être was wedded only to state power were comparatively ephemeral, and thus are more difficult to document archaeologically. Comparative studies between different African urbanizations are at an early stage, but the archaeological definition of hierarchy and heterarchy in such circumstances will be important poles for future debate.7

(p. 69) Functions and Connectivity

The pre-Islamic urbanization of the Sahara and sub-Saharan zones was often linked to state formation as well as long-range contacts and trade. Commercial factors at local, regional, and inter-regional scales often appear as significant drivers for African urbanization, with geographic nexus points creating important foci for exchange. In Ethiopia, for instance, the kingdom of Aksum in the 1st millennium ce combined a highland Ethiopian trade web with longer-range trade contacts to the Nile Valley and the Red Sea. However, it is important to realize, especially when historical data are to hand, that ideological factors also played an important role in African urbanism, with centres playing critical roles as foci of religion and (sometimes) royal cult. In the central Saharan oases, towns and villages were often visibly fortified, perhaps projecting localized power as much as defence. South of the Sahara urbanism allowed ranges of specialists in metallurgy and sculpture to flourish, whether at Dia, Jenné-jeno or Ife. Despite the diversity of urban trajectories exhibited across Africa and the profound socio-economic and structural differences that existed between the societies served by these early towns, the new evidence requires us to consider carefully the potential for interconnectivity between urban networks. While there are factors that suggest a strong local and independent impulse in urban developments, the examples we shall describe are no longer as isolated and insulated from other urban networks as was once thought. For example, impressive developments in West Africa were contemporaneous with significant changes in the central Sahara. While it is far from the case that we need to return to long discarded diffusionist views, it is also clear that early Saharan and Sahelian urban networks were in contact with each other. The nature and scale of such interactions merit greater investigation. Long-range trade, whether across the desert or by sea (as in the case of the Red Sea and Indian Ocean) appears to be an important feature of several of the early African urban stories.

The First Urbanization in the Nile Valley

The process of settlement nucleation and state formation from the 4th millennium bce has been well documented in the Nile Valley. However, Pharaonic Egypt and, to an extent, the story of Kerma and the later kingdoms of Kush and Meroe in Sudan have tended to be seen as exceptional in an African context, linked to the life-enhancing potential of the great river. This is a well exposed urban story, although in fact the number of extensively excavated urban sites in the Nilotic region is comparatively small, with most attention having focused on their temple and funerary complexes. However, the existence of extensive nucleated settlements around temples is now demonstrated sufficiently to counter the older view that Egypt was a state without towns.8 There is also (p. 70) increasing evidence for the existence of nucleated towns in the pre-dynastic period, though the overall size, layout, and organization of these early towns are much less clear than those of Mesopotamia. Where more extensive remains survive of a particular phase of an urban centre it is generally due to the town being of short duration on a virgin site (as at el-‘Amarna or el-Lahun/Kahun), and the typicality of such sites must be doubted.

Nonetheless, the rich iconographic and textual record from Dynastic Egypt provides substantial contextual data on the function and organization of these urban settlements within a monarchical and highly hierarchical state. Some of the early phases of development have similarities with other urban civilizations of the Bronze Age Near East, with which Egypt was in certain periods closely connected, but Egyptian society was also uniquely structured in many respects.9 Some towns and large villages had highly specific functions—providing housing for tomb and pyramid constructors or serving as regional administrative centres—many focused on monumental temple complexes or centres of royal power.

The continuation of urban forms of settlement through the Persian, Hellenistic, and Roman periods added complexity and external influences to the ancient towns of Egypt. For the Classical world, the survival of extensive documentary records at sites like Oxyrhynchus provide a fascinating level of detail of life in such settlements, only rarely attainable for ancient urban centres. There is growing evidence to suggest that urbanization was transmitted from the Nile to the emerging oases of the Western Desert, as early as the 1st millennium bce and that there was eventually a chain of such oasis settlements stretching into the Central Sahara. The oasis towns of the Egyptian Western Desert often betray Egyptian cultural influences in architecture, religion, and burial rituals and were politically aligned with Egypt.10 As we shall see below, the oasis towns further out into central Sahara are less obviously modelled on Egyptian/Nilotic ideas.

Early urban development in the Upper Nile in Sudan related to the kingdoms of Kerma (2500–1500 bce) and Kush/Meroe (c.800 bce–350 ce). While these states were at one time seen as a pale imitation of Pharaonic Egypt to the north, it is now clear that they were sophisticated civilizations in their own rights.11

The Red Sea Trade and Kingdom of Aksum

The discovery of the secret of the monsoon winds in the second half of the 1st millennium bce led to the development of trade routes between the Red Sea and India and East Africa, gaining in intensity in the Roman period. Recent research has shown these commercial contacts to have been much larger scale and longer lived than previously imagined. One aspect of this trade was the creation of small mercantile towns on both sides of the Red Sea coast, as at Myos Hormos, Berenike, and Ardulis, and at Kané opposite the Horn of Africa.12 Although these port towns were generally of quite small scale, they were notable centres of population on a desert coastline, with cosmopolitan populations and material culture. Berenike, for instance, is thought to have had a maximum (p. 71) population of only c.1000, but was an important centre for transhipment and regulation of cargo. By the late Roman period, some of these ports were in decline, though this may in part have been connected by the increasing power of an independent kingdom in Ethiopia, in control of the port of Ardulis.

The kingdom of Aksum had its origins in the 1st millennium bce, with its capital an 8–15 day journey inland to the Ethiopian uplands at Aksum. The kingdom reached its apogee in the period between the 3rd and 7th centuries ce, when its political authority extended to southern Arabia, over a large part of the Red Sea and towards the Nile in Sudan. The coins issued by a succession of kings, Christian from the 4th century ce onwards, were widely distributed. The Red Sea outlet at Ardulis was highly significant to the economy of Aksum at all periods and linked the kingdom not only with the Indian and East African trade, but with southern Arabia, Egypt, and the Mediterranean. Overland contact also existed with the kingdom of Meroe on the Upper Nile.13 The monumental centre of Aksum, its initial capital, appears to have been founded in the early 1st century ce. Though it was eventually very extensive, its urban core has been less explored than its cemeteries and stelae field. These monolithic funerary stelae, the largest of which weighed 500 tonnes and stood 33 m tall, and the bases of more than two dozen stone ‘thrones’ spread around the capital, strongly evoke the political and sacred character of this city at its mid-1st millennium ce peak. Archaeological evidence reveals that Aksum would have housed a range of craft specialists including stone carvers, stone tool makers, rope makers, and glass makers (or re-workers) amongst others. To support this centre, particularly for grain and firewood, a strong network of supply would have been required from a hinterland which archaeological research has only recently begun to assess.14

The Mediterranean and Its Hinterland

Classical Mediterranean urbanization is covered elsewhere in this volume and a few comments on the African specificities will suffice here. The North African littoral was affected at various times and places by Greek, Phoenician, and Roman colonization/imperialism. The Phoenician/Punic contribution to North African urbanization has become clearer in recent decades as a result of extensive excavations at Carthage and a number of smaller emporia, such as Kerkouane.15 The Roman province of Africa Proconsularis (roughly modern Tunisia) was one of the most heavily urbanized regions in the Roman empire, with more than 300 towns. Most previous studies have approached these settlements in terms of their conformity with Punic or Greco-Roman models. Less attention has been paid to their divergence from the Mediterranean norms and the extent to which they may be seen as African adaptations of the urban form.

Some early urban centres were colonies or emporia of external polities: Cyrene and Euesperides in Greek Cyrenaica; Carthage and Lepcis Magna in Punic and Roman Africa. However, the massive expansion of urban centres into the interior of Rome's (p. 72) African provinces went far beyond colonial foundations, building on urban experiments relating to the Numidian and Mauretanian kingdoms. While Greece and Rome provided some of the most obvious architectural trappings of these towns, the African character of many sites is apparent in matters of layout (the winding lanes of Numidian Thugga in contrast to the rigid checkerboard of the Roman military colony of Thamugadi), religious practices, and nomenclature. Nonetheless, in general the urbanization of Mediterranean Africa connected with the world to north and east and was a rather separate phenomenon to Saharan and sub-Saharan urbanization.16 There is substantial evidence to show that by late Roman times the nature of the North African city was undergoing profound transformations.17 While such changes are paralleled elsewhere in the Mediterranean, the possibility that there are African peculiarities in what emerged has not been much considered hitherto.

The Central Sahara

The Garamantes of southern Libya are the best example of a pre-Islamic urban civilization in the Central Sahara. South-west Libya (Fazzan) appears blank on many historical maps of Africa, though the region contains one of the densest clusters of oases.18 Modern perceptions have been framed around accounts of early European travellers of impoverished desert oases and the inhumanity of the trans-Saharan 19th-century slave trade. We can challenge such limited perceptions. In particular, it is now established that by the early centuries ce a people called the Garamantes controlled a powerful kingdom covering c.250,000 km2 of the Libyan Sahara. Far from being the stereotypical ‘nomadic barbarians’ suggested by the classical sources, the Garamantes lived in towns and villages and practised irrigated agriculture of considerable sophistication.19 A convincing case can be made for identifying them as an early state, founded on the twin pillars of irrigated agriculture and urbanized settlement networks. Settlement in this region is concentrated today in three roughly parallel bands of oases (each of these 100–150 km in length), the Wadi ash-Shati, the Wadi al-Ajal, and the Zuwila-Murzuq-Barjuj depression. Archaeological work in the central oasis belt, the Wadi al-Ajal, has produced dramatic evidence of technical accomplishment and rich material culture in their heartland territory. The proto-urban origins of Garamantian towns have been traced in defended escarpment settlements (hillforts) such as Zinkekra, where nucleated settlements of simple oval buildings were transformed by the later 1st millennium bce into large population centres utilizing complex, multi-roomed mudbrick dwellings. The Garamantian capital at Jarma (ancient Garama) was evidently founded c.300 bce and typifies later sites, being located in the centre of the valley, away from the escarpment, but adjacent to the oasis cultivation zone. Studies around the Taqallit headland west of Jarma have revealed a pioneer phase of colonization of the landscape, with the simultaneous creation in the last centuries bce of monumental cemeteries, irrigation systems called foggaras (equivalent to the Persian qanat), and numerous nucleated settlements.


plate 4.1 Fragments of architectural elements recovered from the Garamantian town of Garama (Old Jarma, Libya) in the central Sahara (Photomontage © D. Mattingly).


figure 4.2 Comparative plans of Garamantian towns and fortified villages: (a) Old Jarma, (b) Qasr ash-Sharraba, (c) Qasr bin Dughba, (d) HH1, (e) HH 6–8 (montage © D. Mattingly and M. Sterry).

(p. 73) There are clear indications of monumental developments in the architectural repertoire at Jarma—where major public buildings with ashlar footings and columnar screens were erected (see Plate 4.1, Fig. 4.2). There is evidence that this was linked to large-scale trading connections to north and south of the Sahara and to technological sophistication in a range of manufacturing processes.20 That this was a centralized political and military power is indicated by the disparity in material culture between the core and the periphery of Garamantian territory, by the large-scale erection of forts and urban defences, and by the simple consideration of the amount of work required just to (p. 74) (p. 75) construct the irrigation systems (estimated at c.77,000 man years labour). The evidence now available supports the identification of the Garamantian Kingdom at its height as a macro-state, rather than one among a series of city-states, though it seems plausible that the rise and decline of the Garamantes involved first the incorporation and then re-emergence of smaller micro-states.

Most work on Garamantian settlement to date has focused on the heartland area of the Wadi al-Ajal near their capital Jarma, with some reconnaissance work on Garamantian and Islamic settlement in the northern and southern oases zones. These preliminary studies indicate that similar large-scale development of agricultural and village-based societies occurred in both the northern and southern oasis belts. The Murzuq area is of particular importance for understanding the transition between the Garamantes and the Islamic Fazzan, because, as the Garamantian Kingdom declined in late antiquity, the locus of power in the region shifted south-eastwards from Jarma in the Wadi al-Ajal to Eastern Fazzan. The successive capitals of medieval and early modern Fazzan were all located in this area: Zuwila, Traghan, Murzuq, and Sabha. The most direct route between Tripoli and the sub-Saharan kingdoms in the Lake Chad area (Bornu, Kanem) passed through Eastern Fazzan. The latest archaeological research has revealed a densely colonized Garamantian landscape of fortified villages, exhibiting planned layouts and with the largest sites verging on an urban scale (Fig. 4.2). The massive expansion of the oasis landscape in the Murzuq area appears to date to the early centuries ce, somewhat later than the similar process in the Wadi al-Ajal.21

Another Libyan desert site with huge potential to illuminate the theme of early urbanization is Ghadames (ancient Cidamus). Survey work has identified extensive pre-Islamic cemeteries, including monumental tombs of late Roman date, and the likely location of the early oasis town. There is a high probability that some other central Saharan oases that utilized the foggara irrigation system, for instance in southern Algeria, also originated in the pre-Islamic period.22

The Middle Niger

In the 1970s and 1980s Roderick and Susan McIntosh produced West Africa's first archaeological evidence of pre-Islamic urbanism, c.400–800 ce, at the tell complex of Jenné-jeno in the Inland Niger Delta of Mali. The definition of its urban status relied in part on concepts borrowed from ‘New Geography’, including the ‘Rank Size Rule’ and ‘Central Place Theory’. Yet, Jenné-jeno's role as a regional hub of commerce and interaction, together with the size of its core mound (33 ha), and evidence for the presence of many satellite specialist communities (an additional c.36 ha) have all become part of defining what constitutes an early African city.23 This definition may be summarized as comprising a localized and concentrated economic diversity of population, the presence of craft and ideological specialists, evidence for a range of trade networks, and a size of central settlement which renders it dependent upon its hinterland for assured subsistence.

(p. 76) More recently, Roderick McIntosh has emphasized the ‘self-generated’—rather than state or hierarchically generated—nature of Middle Niger urbanism.24 Deriving from his earlier ‘Pulse Theory’ for the genesis of economic specialization and symbiosis in the West African Late Stone Age, McIntosh proposes that Middle Niger urban centres were generated organically and gradually from localized networks of subsistence and occupational specialists who found means of maintaining diversified communities through economic symbiosis and heterarchical political organization. The most recent work at Dia, Jenné-jeno's notional ‘mother city’, has provided evidence for a large-scale (c.23 ha) occupation with satellite sites, banco architecture, and iron metallurgy between 800 and 400 bce; notionally pushing back dates for the advent of Middle Niger urbanism.25


plate 4.2 Excavations at the early urban site of Dia (Mali) in 1998.

However, there remain questions as to whether Middle Niger tell sites such as Dia and Jenné-jeno are really the points of origin for urbanized settlement organization in West Africa, and whether heterarchical self-generation is the only viable model for early Middle Niger urbanism. The Late Stone Age or ‘Neolithic’ polity of Tichitt (c.1900–400 bce) occupied a vast landscape, stretching across the south-eastern quarter of Mauritania and to the edge of the Middle Niger. Indeed, its distinctive pottery occurs in the earliest occupational layers of Dia (Plate 4.2).26 Many of its sites comprise large dry-stone walled compounds, with traces of more ephemeral internal structures, including granaries. Its largest site, Dakhlet el Atrous, covers a remarkable 80.5 ha and comprises 540 compounds. Although words such as ‘urban’ or ‘proto-urban’ have not been mooted about Tichitt's larger settlements, we could usefully ask why this is the case. Like Jenné-jeno, Dakhlet el Atrous was a large centre within a regional settlement hierarchy. Yet, such massive Tichitt settlements lack evidence for subsistence and occupational (p. 77) specialization (e.g. metallurgy) until late in their existence and there is little concrete evidence for long distance commerce beyond a few handfuls of carnelian and amazonite beads.27 Nor has the seasonality and contemporaneity of Dakhlet el Atrous’ many compounds been satisfactorily resolved. Nevertheless, large-scale occupations ancestral to the Middle Niger's first generally agreed cities are worthy of more attention than they have received, and perhaps form part of a long process of demobilization of mobile, hierarchical pastoral societies at the advent of generalized aridity in the Holocene Sahara (3rd millennium bce onwards).28 In other words, living together in large settlements was not—in the sense of the longue durée—a new thing to the 1st millennium bce populations of the Middle Niger.

Regarding the self-generation or political generation of large-scale settlements, one must also not lose sight of the fact that West Africa has a long history of statehood. Pre-existing polities such as Ghana and Kawkaw were recognized by the first Arabic visitors to the Sahel in the 8th and 9th centuries.29 The origins of the earliest of these—the empire of Ghana (or Wagadu)—might extend back into the 4th century ce, with the possibility of still earlier antecedents such as Tichitt. Written histories attest that states such as Ghana, Kawkaw, and Mali had ‘capitals’—although this term should not be understood in the classic sense of such localities acting in an administrative or economic sense for the entire state. Rather, they would have been places of royal courts, points of embassy, areas of sacred ritual spaces, and garrisons for core elements of the army. They do not appear to have served important mercantile functions. We can infer this both from primary sources and by the close study of more recent successors such as Segou (discussed above regarding issues of hierarchy and heterarchy). Such capitals may have been occupied for only a single reign and are very likely to have been abandoned at any point of dynastic rupture.30 Subsequent to the shift of a capital there is normally either a marked reduction in size of the site (with merely a vestigal village-sized population) or a complete abandonment. Lifespan is thus a key factor separating large Middle Niger ‘political’ settlements from more stable mercantile centres. The organization of mercantile centres may well have been locally heterarchical, but the political structure of the larger polities which surrounded them, if we are to believe early Arabic sources and some archaeological traces, are likely to have had strong coercive elements, including slavery.31

The West African Forest

Although the beginnings of urbanism in the West African forest were, comparatively speaking, rather recent—c.900–1300 ce—their individualistic and pristine nature makes them relevant to the present chapter. Ife, with its sacred groves, sophisticated art corpus, and concentric bands of bank and ditch fortifications covering over 1,000 ha, makes for a remarkable example. The city of Ife is one of several ancient walled Nigerian cities linked to the people today known as the Yoruba. Indeed, according to tradition it is not only Nigeria's first city, but the point from which the world was created.32

(p. 78) The origins of Ife may be traced to an enclosure within the current city known as Enuwa (meaning ‘we see eye to eye’). Radiocarbon dates from excavations within the area fall between 600 and 1000 ce and probably relate to an initial agglomeration of local populations behind a bank and ditch wall.33 Ozanne has effectively argued that Ife grew up as a series of roughly concentric walled rings around the sacred palace of its ruler, the Oni, whose descendants still reign in the city today.34 Ife's vast enclosures were only partially filled with housing and would have included large open areas with farmsteads, making any estimation of its ancient population difficult and low population density likely.

From early in its existence Ife would have participated in localized trade in agricultural resources, particularly yams, vegetable oils, and palm wine; however long-distance commerce was probably ongoing from c.1000 ce with exports including ivory from forest elephants, kola nuts, peppers, and slaves. In terms of imports the most visible is a large-scale trade for copper alloys, notably brass bars of trans-Saharan origin which went to form the majority of cast objects manufactured at the city. Ife's metallurgical traditions grew out of a long legacy of sophisticated metalworking in Nigeria, dating from the 1st millennium bce Nok culture, and the local copper alloy working traditions of Igbo Ukwu (c.900 ce). There is no doubt that the concentration of metalworkers in the city was a key factor in the legitimation of its royal cult. The beginnings of Ife's naturalistic tradition of terracotta sculptural portraiture, which ultimately developed into famous examples in cast copper alloy, began sometime between 900 and 1100 ce. By the end of this same period potsherd pavement shrine areas are documented, probably dedicated to both ancestral and regal cults, and utilizing terracotta or copper alloy portrait busts as part of their focus.

The enormous scale of public labour evidenced by Ife's vast networks of bank and ditch boundaries is evocative of enslavement and coercion. Historically, we know that slavery existed in the region, and early Ife artistic renderings of gagged captives or sacrificial victims in terracotta sculpture exist. The social system of Ife was ultimately highly coercive, though it probably did not begin that way.

In searching for the impetus behind the foundation of Ife, we are tempted to appeal to ideological factors over economic ones. The sacred nature of Ife kingship appears to form part of a wider trend across the West African forest. As Asombang writes, ‘it is yet to be demonstrated historically or archaeologically that…cities like Ife and Benin were military or administrative centers more than they were sacred centers.’35 Using the historical example of Bafut in western Cameroon, Asombang describes a situation very much like that of Ife: a palace housing a king who is also chief priest of the cults, surrounded by quarters of lineages competing for influence, with further settlements radiating out from this sacred core. It is further argued that, as an alternative to economic incentives, the supernatural abilities of individuals to control natural forces (e.g. rainmaking, as in Southall's segmentary state),36 or occult forces, can form the core around which major settlement centres and polities can grow. Ultimately the trajectories of such centres and polities can evolve to include commerce and, with increasing power of the sacred ruler and closely allied lineages, elements of coercion. Ife thus potentially constitutes a marked alternative pathway to urbanism: from the centre of sacred cult to a city.

(p. 79) Conclusions

Across early African cities, regionality is particularly marked in the degrees of mercantile, sacred, and political/military strategies mixed into each urban milieu. For the African forest the impetus of urbanism appears to have initially grown out of the gravitational power of a sacred cult (cult-generated urbanism?), eventually developing into a sacred kingship fed by long-range commerce in forest resources. Elsewhere in the sub-Saharan zone there are potentially profound contrasts. For instance, the citadel-focused character of many central Saharan sites looks very different to the unfortified ‘self-generated’ mercantile settlements of the West African Sahel in general and the Middle Niger in particular. However such trends are not absolute, as even in earlier (1st millennium bce) periods, fortified towns are now known from the Lake Chad Basin (e.g. Zilum).37 Indeed, by the time Arabic geographers were first writing about the states of the Sahel in the late 1st millennium ce, it is apparent that they possessed substantial ‘capitals’ or state-generated urban centres.

Yet, despite the uniqueness of early African urban phenomena, there is clear evidence to indicate that many early urban societies were in contact and interacting with each other. This is particularly the case with the African Sahel and the central Sahara, where pre-Islamic commerce and the exchange of ideas and technologies appears to be increasingly likely.38 The implications of such cross-regional contacts for the development of African urbanization remain to be explored.


Connah, G., African Civilizations. An Archaeological Perspective (Cambridge: Cambridge University Press, 2001, 2nd edn.).Find this resource:

    Dowler, A., and Galvin, E. R., eds., Money, Trade and Trade Routes in Pre-Islamic North Africa (London: British Museum Press, 2011).Find this resource:

      Edwards, D. N., The Nubian Past (London: Routledge, 2004).Find this resource:

        MacDonald, K. C., ‘Before the Empire of Ghana: Pastoralism and the Origins of Cultural Complexity in the Sahel’, in G. Connah, ed., Transformations in Africa: Essays on Africa's Later Past (London: Leicester University Press, 1998), 71–103.Find this resource:

          McIntosh, R. J., Ancient Middle Niger. Urbanism and the Self-Organising Landscape (Cambridge: Cambridge University Press, 2005).Find this resource:

            McIntosh, S. K., ed., Excavations at Jenné-Jeno, Hambarketolo, and Kaniana (Inner Niger Delta, Mali), the 1981 Season (Berkeley: University of California Publications in Anthropology, vol. 20, 1995).Find this resource:

              ——— ed., Beyond Chiefdoms: Pathways to Complexity in Africa (Cambridge: Cambridge University Press, 1999).Find this resource:

                ———  and McIntosh, R. J., The Early City in West Africa: Towards an Understanding. African Archaeological Review, 2 (1984), 73–98.Find this resource:

                  Mattingly, D. J., ed., The Archaeology of Fazzan. Vol. 1, Synthesis (London: Society for Libyan Studies, 2003).Find this resource:

                    Phillipson, D. W., Ancient Ethiopia. Aksum: Its Antecedents and Successors (London, British Museum Press, 1998).Find this resource:

                      ——— African Archaeology (Cambridge: Cambridge University Press, 2005, 3rd edn.).Find this resource:

                        Sears, G., The Cities of Roman Africa (Stroud: The History Press, 2011).Find this resource:

                          Shaw, T., Sinclair, P., Andah, B., and Okpoko, A., eds., The Archaeology of Africa. Food, Metals and Towns (London: Routledge, 1993).Find this resource:

                            Sidebotham, S., Berenike and the Ancient Maritime Spice Route (Berkeley: University of California Press, 2011).Find this resource:

                              Stahl, A. B., ed., African Archaeology (Oxford: Blackwell, 2005).Find this resource:

                                Trigger, B. G., Kemp, B. J., O’Connor, D., and Lloyd, A. B., Ancient Egypt. A Social History (Cambridge: Cambridge University Press, 1983).Find this resource:


                                  (1.) G. Connah, African Civilizations. An Archaeological Perspective (Cambridge: Cambridge University Press, 2001), 13.

                                  (2.) See A. LaViolette and J. Fleisher, ‘The Archaeology of Sub-Saharan Urbanism: Cities and Their Countrysides’, in A. B. Stahl, ed., African Archaeology (Oxford: Blackwell, 2005), 327–52; R. J. McIntosh, Ancient Middle Niger. Urbanism and the Self-Organising Landscape (Cambridge: Cambridge University Press, 2005).

                                  (3.) There is inevitably some overlap between our chapter and Bill Freund's (Ch. 33) later in this volume.

                                  (4.) See M. Hansen, ed., A Comparative Study of 30 City-State Cultures (Copenhagen: Historisk-filosofiske Skrifter 21, 2000), 11–34, for an extended discussion.

                                  (5.) R. J. McIntosh, The Peoples of the Middle Niger (Malden, Mass.: Blackwell, 1998).

                                  (6.) K. C. MacDonald and S. Camara, ‘Segou, Slavery, and Sifinso’, in J. C. Monroe and A. Ogundiran, eds., The Politics of Landscape in Atlantic West Africa (Cambridge: Cambridge University Press, 2012 in press).

                                  (7.) S. K. McIntosh, ed., Beyond Chiefdoms: Pathways to Complexity in Africa (Cambridge: Cambridge University Press, 1999).

                                  (8.) J. Baines and J. Málek, Atlas of Ancient Egypt (Oxford: Phaidon, 1980), on temple/funerary focus; Hansen, Comparative Study of 30 City-State Cultures, 14, on debate about Egypt as an urbanized state.

                                  (9.) See B. G. Trigger et al., Ancient Egypt. A Social History (Cambridge: Cambridge University Press, 1983) and papers by Hassan and O’Connor in T. Shaw et al., eds., The Archaeology of Africa. Food, Metals and Towns (London: Routledge, 1993), 551–86; see McMahon, Ch. 2, for Near Eastern comparanda.

                                  (10.) On Classical cities in Egypt, see P. Parsons, City of the Sharp-Nosed Fish: Greek Lives in Roman Egypt (London: Weidenfeld and Nicholson, 2006), 9; cf. also Osborne and Wallace-Hadrill, Ch. 3, this volume. On the desert routes and oases west from the Nile, M. Liverani, ‘The Libyan Caravan Road in Herodotus IV.181–4’, Journal of the Economic and Social History of the Orient, 43.4 (2000), 496–520; for the important recent work at Dakhleh oasis, see inter alia A. Boozer, ‘New Excavations from a Domestic Context in Roman Amheida, Egypt’, in M. Bommas, ed., Cultural Memory and Identity in Ancient Societies (London and New York: Continuum, 2011), 109–26.

                                  (11.) D. N. Edwards, The Nubian Past (London: Routledge, 2004); D. A. Welsby, The Kingdom of Kush. The Napatan and Meroitic Empires (London: British Museum Press, 1996).

                                  (12.) The recent research is best summarized in S. Sidebotham, Berenike and the Ancient Maritime Spice Route (Berkeley: University of California Press, 2011); D. Peacock and L. Blue, eds., Myos Hormos—Quseir al-Qadim: Roman and Islamic Ports on the Red Sea. Vol 1. Survey and Excavations 1999–2003 (Oxford: Oxbow, 2006).

                                  (13.) D. W. Phillipson, Ancient Ethiopia. Aksum: Its Antecedents and Successors (London: British Museum Press, 1998); D. Peacock and L. Blue, The Ancient Red Sea Port of Ardulis, Eritrea (Oxford: Oxbow, 2007).

                                  (14.) J. W. Michels, ‘Regional Political Organisation in the Axum-Yeha Area during the Pre-Axumite and Axumite Eras’, Etudes éthiopiennes, 1 (1994), 61–80.

                                  (15.) Various contributions in A. Ennabli, Pour sauver Carthage. Exploration et conservation de la cité punique, romain et byzantine (Paris: UNESCO, 1992); M. H. Fantar, Kerkouane, une cité punique au Cap Bon (Tunis: Maison tunisienne de l’édition, vols. I–III, 1983–6).

                                  (16.) R. Laurence, S. Esmonde Cleary, and G. Sears, The City in the Roman West 250 bc to 250 ad (Cambridge: Cambridge University Press, 2011); G. Sears, The Cities of Roman Africa (Stroud: The History Press, 2011).

                                  (17.) A. Leone, Changing Townscapes in North Africa from Late Antiquity to the Arab Conquest (Bari: Edipuglia, 2007); cf. D. Stone, D. J. Mattingly, N. Ben Lazreg, eds., Leptiminus (Lamta): A Roman Port City in Tunisia, Report no. 3, the Urban Survey (Ann Arbor: JRA Supplement, 2011) for an attempt to trace the long-term urban biography of a coastal port city.

                                  (18.) See e.g. C. McEvedy, The Penguin Atlas of African History (London: Penguin, 1995, revised edn.).

                                  (19.) The Garamantes have been overlooked or underestimated in most modern accounts of African civilizations (e.g. Connah, African Civilizations (2001), D. W. Phillipson, African Archaeology (Cambridge: Cambridge University Press, 2005, 3rd edn.), and Stahl, African Archaeology contain scarcely a reference to them). For modern research, see D. J. Mattingly, ed., The Archaeology of Fazzan. Vol. 1, Synthesis (London: Society for Libyan Studies, 2003); Vol. 2, Site Gazetteer, Pottery and Other Finds (2007); Vol. 3, Excavations of C. M. Daniels (2010). Cf. M. Liverani, ed., Aghram Nadarif. A Garamantian Citadel in the Wadi Tannezzuft (Firenze: All’Insegna del Giglio, 2006).

                                  (20.) D. J. Mattingly, ‘The Garamantes of Fazzan: An Early Libyan State with Trans-Saharan Connections’, in A. Dowler and E. R. Galvin, eds., Money, Trade and Trade Routes in Pre-Islamic North Africa (London: British Museum Press, 2011), 49–60.

                                  (21.) M. Sterry and D. J. Mattingly, ‘DMP XIII: Reconnaissance Survey of Archaeological Sites in the Murzuq Area’, Libyan Studies, 42 (2011), 103–16.

                                  (22.) A. I. Wilson, ‘The Spread of Foggara-based Irrigation in the Ancient Sahara’, in D. Mattingly, et al., eds., The Libyan Desert. Natural Resources and Cultural Heritage (London: Society for Libyan Studies, 2006), 205–16.

                                  (23.) S. K. McIntosh, ed., Excavations at Jenné-Jeno, Hambarketolo, and Kaniana (Inner Niger Delta, Mali), the 1981 Season (Berkeley: University of California Publications in Anthropology, vol. 20, 1995); S. K. McIntosh and R. J. McIntosh, ‘The Early City in West Africa: Towards an Understanding’, African Archaeological Review, 2 (1984), 73–98; id., ‘Cities without Citadels: Understanding Urban Origins along the Middle Niger’, in T. Shaw et al., The Archaeology of Africa, 622–41.

                                  (24.) R. J. McIntosh, Peoples of the Middle Niger; id., Ancient Middle Niger.

                                  (25.) R. Bedaux et al., ‘The Dia Archaeological Project: Rescuing Cultural Heritage in the Inland Niger Delta (Mali)’, Antiquity, 75 (2001), 837–48; id., ‘Conclusions: Une histoire de l’occupation humaine de Dia et de sa place régionale par la méthode archéologique est-elle déjà possible?’, in Recherches archéologiques à Dia dans le Delta intérieur du Niger (Mali): bilan des saisons de fouilles 1998–2003 (Leiden: CNWS Publications, 2005), 445–55.

                                  (26.) Tichitt: K. C. MacDonald et al., ‘Dhar Néma: From Early Agriculture, to Metallurgy in Southeastern Mauritania’, Azania: Archaeological Research in Africa, 44 (2009), 3–48; pottery: K. C. MacDonald, ‘Betwixt Tichitt and the IND: The Pottery of the Faïta Facies, Tichitt Tradition’, Azania: Archaeological Research in Africa, 46 (2011), 49–69.

                                  (27.) Dakhlet el Atrous: A. Holl, ‘Late Neolithic Cultural Landscape in Southeastern Mauritania: An Essay in Spatiometrics’, in A. Holl and T. E. Levy, eds., Spatial Boundaries and Social Dynamics: Case Studies from Food-Producing Societies (Ann Arbor: International Monographs in Prehistory, Ethnoarchaeological Series 2, 1993), 95–133; beads: K. C. MacDonald, ‘A View from the South: Sub-Saharan Evidence for Contacts between North Africa, Mauritania and the Niger, 1000 bcad 700’, in Dowler and Galvin, Money, Trade and Trade Routes, 71–81.

                                  (28.) K. C. MacDonald, ‘Before the Empire of Ghana: Pastoralism and the Origins of Cultural Complexity in the Sahel’, in G. Connah, ed., Transformations in Africa: Essays on Africa's Later Past (London: Leicester University Press, 1998), 71–103.

                                  (29.) N. Levtzion and J. F. P. Hopkins, eds., Corpus of Early Arabic Sources for West African History (Princeton: Markus Wiener Publishers, 2nd edn, 2001); cf. also P. J. Munson, ‘Archaeology and the Prehistoric Origins of the Ghana Empire’, Journal of African History, 21 (1980), 457–66.

                                  (30.) D. C. Conrad, ‘A Town Called Dakalajan: The Sunjata Tradition and the Question of Ancient Mali's Capital’, Journal of African History, 35 (1994), 355–77. For example, over a c.160-year existence Segou had six documented capitals, averaging displacements every 26 years, see MacDonald and Camara, ‘Segou, Slavery, and Sifinso’ (2011).

                                  (31.) P. J. Lane and K. C. MacDonald, eds., Slavery in Africa: Archaeology and Memory (Oxford: Oxford University Press, 2011).

                                  (32) T. Shaw, Nigeria: Its Archaeology and Early History (London: Thames & Hudson, 1978); cf. Connah, African Civilizations.

                                  (33.) H. J. Drewal and E. Schildkrout, Kingdom of Ife: Sculptures from West Africa (London: British Museum Press, 2010).

                                  (34.) P. Ozanne, ‘A New Archaeological Survey of Ife’, Odu, ns 1 (1969), 28–45.

                                  (35.) R. Asombang, ‘Sacred Centres and Urbanization in West Central Africa’, in S. K. McIntosh, ed., Beyond Chiefdoms: Pathways to Complexity in Africa (Cambridge: Cambridge University Press, 1999), 80–7 (quote from 80).

                                  (36.) A. Southall, ‘The Segmentary State and the Ritual Phase in Political Economy’, in S. K. McIntosh, ed., Beyond Chiefdoms, 31–8.

                                  (37.) On the Middle Niger: R. J. McIntosh, Ancient Middle Niger; McIntosh and McIntosh, ‘Cities without Citadels’; Lake Chad basin: C. Magnavita et al., ‘Zilum: A Mid-first Millennium bc Fortified Settlement near Lake Chad’, Journal of African Archaeology, 4 (2006), 153–69.

                                  (38.) Dowler and Galvin, eds., Money, Trade and Trade Routes; P. Mitchell, African Connections: Archaeological Perspectives on Africa and the Wider World (Walnut Creek: AltaMira Press, 2005), 140–6.