In May 2012 the United Nations estimated that 18 million people in the Sahel region of West Africa were suffering from food shortages caused by drought and conflict and that nearly 1.5 million children were near starvation. The Sahel was defined broadly, although technically it is only the driest of the savanna zones south of the Sahara Desert. Nevertheless, how could it be that one of the areas of early urbanization and state formation considered in this book is situated within this apparently disastrous zone? Clearly, its occupants did not always face the bleak situation that now exists. Particularly this seems to have been the case immediately bordering the southern edge of the Sahara, which was the setting for important social and political developments during the first and second millennia AD. Some historical sources indicate conditions that were far from bleak. For example, in the early sixteenth century Leo Africanus said of Jenné, situated in the Inland Niger Delta on the edge of the Sahel zone: ‘This place exceedingly aboundeth with barlie, rice, cattell, fishes, and cotton’ (Africanus 1896: vol. 3, 822). This description was by an outsider, for Leo Africanus came from North Africa, but a local scholar, al-Sa'di, writing in about 1655, wrote: ‘This city is large, flourishing and prosperous; it is rich, blessed and favoured by the Almighty … The area around Jenné is fertile and well populated; with numerous markets held there on all the days of the week’ (al-Sa'di 1964: 22–4). Although it was pointed out that al-Sa'di might have been biased in favour of Jenné (McIntosh and McIntosh 1980: vol. 1, 49), Bovill (1968: 135) thought that this was a ‘convincing tribute’. Thus, it seems from historical sources, oral tradition, and archaeological data as if the West African savanna was in the past a zone that offered opportunities as well as constraints to the people who lived there.

Geographical location and environmental factors

Throughout the history of the human race in Africa, the most important ecosystem in the continent has probably been the savanna. Rich in faunal and floral resources, by the Holocene attractive for both cereal agriculture and livestock rearing, it offered conditions of relatively easy movement in which natural resources and manufactured products could be readily exchanged.

This book presents case studies of archaeological evidence for urbanization and state formation in precolonial Africa. The choice of areas has been dictated by the character of the archaeological evidence and by the extent of archaeological research, resulting in an incomplete picture. Not only might the individual case studies be unrepresentative of developments in the selected areas, but also there were precolonial cities and states in Africa, known from the evidence of ethnohistory or oral tradition, that are excluded. This omission is because the relevant archaeological evidence is either absent or severely limited.

The basic problem is archaeological visibility (Connah 2008). Some manifestations of early urbanization and state formation in Africa are of such a character that their archaeological investigation is peculiarly difficult. For instance, the capital city of Buganda (Gutkind 1963), a state near Lake Victoria, was described in 1889 as ‘one of the great capitals of Africa’ (Ashe 1889: 52), but it was constructed totally of grass, wood, and other organic materials and it moved frequently, particularly on the death of the Kabaka, the ruler of Buganda (Gutkind 1960: 29). The sites of these large settlements are, therefore, unlikely to have much depth of deposit or structural remains, and no archaeological investigation of them has been attempted (the last two are covered by modern Kampala). Thus there is no archaeological information about a settlement that John Hanning Speke, its first European visitor, described as ‘a magnificent sight. A whole hill was covered with gigantic huts such as I had never seen in Africa before’ (Speke 1863: 283).

There is, however, no reason why archaeologists should not be able to investigate such short-lived settlements of ephemeral materials, as was done, for example, at the South African site of Mgungundlovu (Parkington and Cronin 1979). Provided that subsequent activity or erosion has not destroyed the inevitably shallow evidence, survey and excavation techniques do exist that enable information to be obtained from sites of this type. Nevertheless, they have been seldom applied in Africa, and a lack of sophistication in research design has led to the excavation of many larger African settlement sites with structures of stone, fired brick, or mud, while sites with less substantial structural remains have been ignored.

For many observers it has been the isolation of this region, particularly that of the central highlands of Ethiopia and Eritrea, that has made the greatest impression. The core of the region is a great block of mountains, everywhere more than 1,000 metres in height, that reaches a general level of 2,300 metres and in places exceeds 4,200 metres above sea level. One could not imagine a less likely setting for state emergence and urbanization. Nevertheless, in the modern countries of Ethiopia and Eritrea there is evidence for such developments during the first few centuries AD, and indications of their origins during the first millennium BC (D.W. Phillipson 2009a; 2009c; 2012). Not only are these dates early for the attainment of social complexity in tropical Africa, but these achievements were highly sophisticated. Almost 2,000 years ago, the state of Aksum boasted urban centres; its own form of writing; coinage in gold, silver, and bronze; masonry buildings of a distinctive architectural style; unique monuments that indicate quarrying and engineering skills; extensive trading contacts both within and outside Africa; and a significant role in international politics. Indeed, Aksum seems to have been one of the first states to adopt Christianity (D.W. Phillipson 1998: 145 note 1). However, the antecedents of Aksumite culture can be discerned as early as the eighth century BC in the period formerly known as the Pre-Aksumite. Large settlements existed at Yeha, in Ethiopia, and at Matara and Sembel, in Eritrea, the last one of these belonging to the possibly earlier Ona Culture (Schmidt and Curtis 2001; Schmidt et al. 2008; Curtis 2009; Fattovich 2009; Schmidt 2009). In some places inscriptions appeared and there was sophisticated stone masonry, sculpture, and metallurgy, although stone artefacts were also still used. Furthermore, there were contacts with South Arabia and to a lesser extent with the Nile Valley. Far from being isolated, Ethiopia and Eritrea appear to have formed, at times, an important zone for cultural integration. For Edward Ullendorff (1960: 23) the region was ‘a bridge between Africa and Asia’.

Other than Pharaonic Egypt, some of the earliest evidence for urbanization and state formation in Africa was along the middle Nile, in the region called Nubia. One of the best-known examples was Meroë, located about 200 kilometres north-east of the modern city of Khartoum, in Sudan. Mentioned as early as the fifth century BC by Herodotus (Powell 1949: vol. 1, 121), its antecedents lay in Napata and Kerma, the latter dating back to before 2000 BC, and it was succeeded by the kingdoms of Christian Nubia, which survived until the early second millennium AD, and by Islamic states that existed until recent times (Welsby 1996; 2002; Edwards 2004; 2007; UCL 2013).

At the beginning of the twentieth century, scholars attributed such developments to the influence in turn of Pharaonic, Ptolemaic, Roman, Byzantine, and Islamic Egypt, located to the north of the area. Later the emphasis was changed to one that stressed the indigenous character of Nubian achievements but acknowledged the contributions made to them by northern cultures. William Adams, the author of a 1977 monograph on the archaeology of this area, described Nubia as ‘the transition zone, between the civilized world and Africa’ and entitled his book Nubia: Corridor to Africa. According to Adams, the importance of this narrow corridor through the barren land of Nubia arose from the fact that it was for long the only dependable route across the great barrier of the Sahara Desert, a major road into the heart of Africa. The African interior contained resources much coveted by the outside world – gold, ivory, and slaves – but also other mineral, animal, and vegetable products (Adams 1984: 40). These could be tapped via the Nubian corridor. However, with the development of Red Sea shipping from the first millennium BC onwards, and of trans-Saharan camel caravans during the first millennium AD, the middle Nile Valley gradually lost its significance as a major world trade route. That role was finally destroyed by the expansion of maritime trade around Africa's coasts in the sixteenth and seventeenth centuries AD.

Both archaeological and documentary evidence support the idea of Adams' trade corridor, and the growth of cities and states was centred on it. This suggests that this part of Africa can provide us with an example of trade as a major stimulus towards the development of social complexity.

In the second half of the fifteenth-century AD European sailors first set eyes on the southerly coast of West Africa. What they saw was hardly encouraging: from a distance, a vague grey line pencilled in between an immensity of sea and sky; from close-in, either a dangerous, surf-pounded, sandy beach or an uninviting network of mangrove swamp, creek, and river mouth. Whatever the character of the shoreline, however, behind it there was nearly always an impenetrable tangle of trees and other vegetation. Experience soon taught such visitors that this was a coast to be reckoned with: ships' crews died of fever, and shipworms (Teredo spp.) ate the bottoms out of their ships. Yet it was neither altruism nor curiosity that tempted most Europeans to such a region; it was profit. The very names they gave to different parts of this coast indicate their motives: ‘The Grain [pepper] Coast’, ‘The Ivory Coast’, ‘The Gold Coast’, ‘The Slave Coast’ (Bosman 1967: frontispiece map). Europeans quickly discovered that behind the coast lay a forested and savanna hinterland rich in resources, where the inhabitants were willing to trade on a considerable scale. Not only that, but those inhabitants lived in highly organized communities, some of which took on a size and density which left the visitors in no doubt about what they were dealing with. Within some parts of the West African forest and its fringes there were, indeed, hierarchical states, towns, and cities. Because of their conspicuousness, it was these large settlements that particularly attracted European attention. Thus, writing in AD 1507–8, the Portuguese Duarte Pacheco Pereira described Ijebu-Ode (now in Nigeria) as ‘a very large city’ (Bascom 1959: 38). In a similar vein, the Englishman Towerson, writing in 1557, could claim, perhaps with some exaggeration, that a town in what is now Ghana was ‘by the estimation of our men, as big in circuit as London’ (Blake 1942: vol. 2, 406).

These quotations are selected from early in the history of European West African contact, because state development and urbanization in the West African forest and its margins have sometimes been written about as if they were developments resulting from that contact rather than pre-dating it.

North Africa is situated between the Sahara Desert and the Mediterranean Sea, between an inhospitable environment discouraging travel and one facilitating extensive maritime contacts. Consequently, it has long tended to look to the north rather than to the south, constituting a frontier zone for the rest of the African continent. Excluding the lower Nile Valley, North Africa's habitable coast to the east is so narrow that in places the desert reaches the sea, but the coast broadens in the west into cultivable plains and still further west there are mountains separating it from the desert. Climatically, North Africa belongs with the Mediterranean, with hot dry summers and mild winters when most of the modest rainfall occurs. Culturally, the region has been a meeting place for people from the Levant, Greece, Rome, Europe, and Arabia and the indigenous Libyans. As a result, North Africa has played a role in the history of Europe as well as in that of Africa. The Mediterranean has been the link, providing relatively easy movement for traders and migrants and a frequent focus for armed conflict. Commencing early in the first millennium BC, a series of colonial and indigenous settlements developed that included urban communities and indications of state development. These were technologically sophisticated societies, at least partly literate, that constituted early examples of African social complexity, not as early as its development in the Nile Valley but still preceding that of most of the rest of Africa. As such, North Africa was eventually to stimulate changes in other parts of the continent, particularly in the Sahel beyond the Sahara and in adjacent regions.

Geographical location and environmental factors

North Africa extends from the lower Nile Valley to the Atlantic coast of Morocco. In terms of modern political geography, it includes western Egypt, Libya, Tunisia, Algeria, and Morocco. In Egypt and Libya the settled zone is mainly restricted to the coast by the Sahara, but in Tunisia, Algeria, and Morocco a variety of different environments, including those of the Atlas Mountains that rise to more than 4,000 metres, have provided greater opportunities for human settlement. Extending almost 3,700 kilometres from east to west, the geographical and environmental diversity of the region is striking.

Africa is huge; it is so big that you can put the United States and Australia into it and still have a bit left over. It extends from about 37° north to about 35° south and has an altitudinal range from depressions that are below sea level to mountain peaks that exceed 5,000 metres. As a result it has an incredible diversity of environments. It contains some of the driest deserts in the world, and yet has three of the world's major rivers: the Nile, the Niger, and the Congo. Some of the hottest places on earth are in Africa, and yet there are glaciers on its highest mountains. There are steaming rainforests and dry savanna grasslands, low-lying river valleys and high plateaux, extensive deserts and gigantic lakes, mangrove coasts and surf-pounded beaches. This is only an impressionistic picture of the very large number of differing environments in the African continent. In reality the major zones merge into one another, resulting in an even greater variety of conditions that have been further complicated by climatic variation through time.

For at least 2 million years, human beings have been learning how to get the best out of the kaleidoscope of African environments. Those environments have not determined what men and women could do, nor could the latter ignore the environments in which they have lived. Instead there has been a dynamic relationship between the two, in which people have sought to turn to their advantage the opportunities offered by each environment and to come to terms with its constraints. First as hunters, gatherers, and fishers who gradually intensified their exploitation of available resources, then as pastoralists and cultivators, and eventually as city-dwellers, artisans, and traders, men and women have continued to interact with their environment, retaining a remarkable variety of strategies for doing so.

Archaeology is a major source of information about Africa's past. Documentary sources for African history are limited: their coverage often chronologically patchy and geographically fragmented. For large areas of Africa, particularly tropical Africa, their time depth is restricted to the last century or two.

In this chapter, ‘South Africa’ describes the extremity of the continent south of a line from the Limpopo River, in the east, to Walvis Bay, in the west. It consists of the Republic of South Africa, together with southern parts of Namibia, Botswana, and Mozambique, as well as Lesotho and Swaziland, although attention will be limited to the first of these modern states. Two characteristics of the archaeology of this region are of relevance to the subject of urban and state origins in Africa. First, environments in which trees were rare or absent resulted in the use of drystone construction for parts of many settlements, rendering them highly visible archaeologically, particularly if subsequently abandoned during the turbulent history of recent centuries. This allows detailed analysis of the layout of settlements and their relationship to landscapes. Second, the sociopolitical developments that are the subject of this book took place in South Africa within the last five hundred years at most and in many cases more recently, so that both documentary records and oral traditions are copious and valuable sources of information. In addition, linguistic and genetic studies can provide information on the relationships and movements of ethnic groups, to which investigations of pottery and other aspects of material culture can contribute. Thus, in the case of South Africa, we can examine processes of change for which there is much less evidence in some other parts of the continent at earlier dates. Not only can this throw light on the development of complexity in Africa as a whole, but an investigation of the South African data also has global relevance.

Geographical location and environmental factors

Mixed farming seems to have been well established in South Africa by the middle of the second millennium AD. However, the western and south-western parts of this huge area were too dry to grow African cereal crops or had winter rainfall instead of the necessary summer rainfall. In those parts of southern Africa, livestock-herding, hunter-gathering, or a combination of the two remained the subsistence strategies until European colonization over the last three hundred years or so. In contrast, to the east and north-east both cultivation and livestock, particularly cattle, were important.

‘Kilwa is one of the most beautiful and well-constructed towns in the world.’ This is how the much-travelled ibn Battuta described, first-hand, ‘the principal town on the [East African] coast’ in 1331. Kilwa (properly called Kilwa Kisiwani), situated in what is now Tanzania, was no isolated phenomenon. On the same coast, in what is now Somalia, was Mogadishu, of which he wrote that it was ‘a very large town’ (Freeman-Grenville 1975: 27–31). Ibn Battuta represented the scholarly opinion of the fourteenth-century Islamic world. By the end of the following century there were less-scholarly visitors from the Christian world of Western Europe, but they also were impressed with settlements on the East African coast. Thus in 1498, the unknown author of the Journal of the First Voyage of Vasco da Gama, 1497–1499 compared ‘the town of Malindi’ (now in Kenya) to Alcouchette, a town near Lisbon in his native Portugal (Freeman-Grenville 1975: 55–6). Indeed, the account of Vasco da Gama's second voyage (1502) by Gaspar Correa (written about 1561) describes Kilwa as follows:

The East African settlements would be expected to make a favourable impression on sailors several months outward bound from Portugal, who had just endured the eastern Atlantic on a dull if not inadequate diet. Nevertheless, historical sources such as these demonstrate that settlements of considerable size had already developed on the East African coast before the middle of the second millennium AD. Archaeological and oral traditional evidence supports this conclusion. The problem has been to explain how such a development took place, a development that was limited to a narrow strip comprising 3,500 kilometres of coastline, from southern Somalia to southern Mozambique and including various offshore islands, as well as the Comoro Archipelago and parts of Madagascar.

I started to write this book in 1983, with previous editions appearing in 1987 and 2001 and a Japanese translation in 1993. Any book that originated so long ago and has remained in print for so long will become seriously out of date, but this is particularly the case for a publication about later African archaeology, which has seen an enormous increase in research activity in recent years. I am therefore very grateful to Cambridge University Press for the opportunity to provide a third edition. In doing so, I feel it essential to stress two points. First, people are sometimes unconvinced about the newness of new editions, but this really is a new edition and I hope that it will assist those who have ignored the second edition and continued to cite the outdated first edition. Second, preparing a new edition of a book will inevitably be constrained to some extent by the thinking that influenced its original content and form, in this instance many years ago; the only way to prevent this is to write a completely new book.

Failing this, what is new about this ‘new’ edition? First, it is larger and covers the whole continent, not just tropical Africa, like the two previous editions. This has meant the addition of three completely new chapters, on Egypt, North Africa, and South Africa. Although the book is consequently longer, the extra chapters could only be added by also shortening some of the chapters that already existed. In addition, those chapters have required considerable additions to include recent research, and as a result parts of some of them have been substantially rewritten or subjected to numerous smaller changes. Inevitably this has resulted in the deletion of older source material, wherever it could be replaced by newer information. This has included the replacement of some illustrations, as well as the addition of new ones. The whole process has taken more than three years but I remain concerned about recent publications that might have been missed; in spite of the Internet, accessing sources remains one of the main difficulties for the writers of syntheses. However, as far as possible, I have tried to include a representative sample of recently published material. I hope that not too many people will think that I have ignored their work.

Zambezia is the region between the Zambezi and Limpopo Rivers and between the Kalahari Desert and the Indian Ocean. The name was already in use by 1890 (Maund 1890) and has remained relevant (Pikirayi 2001b). This region comprises modern Zimbabwe, southern Mozambique, north-eastern Botswana, and the extreme north of South Africa. From the archaeological perspective, it is characterized by widespread evidence for the development of social complexity, commencing in the first millennium AD and continuing until the nineteenth century. Most famous of this evidence is the site of Great Zimbabwe, perhaps one of the most ill-used of Africa's archaeological sites. Its fame is such that it has given its name to the country in which it is situated, the country formerly known as Rhodesia and before that as Southern Rhodesia. Its ill usage has been both intellectual and physical, starting when it first became known to Europeans. The first such visitor was a German geologist, Carl Mauch, in 1871. After giving an account of the impressive stone ruins, Mauch felt it necessary to explain their presence deep in the African interior and did so by associating them with King Solomon and the Queen of Sheba. Perhaps it is understandable that a nineteenth-century European should grasp at such an unlikely biblical explanation, but unfortunately the myth of alien origin for the Great Zimbabwe structures was to survive for a long time, even surfacing as a political issue in the 1960s and 1970s (Garlake 1973: 209–10). This was probably because such a belief became psychologically essential for some European colonial settlers in this part of Africa. Perhaps the most damaging aspect of the African colonial experience was the attempted denial to African peoples of their own cultural heritage, of which the attribution of Great Zimbabwe to outside influence must be the classic example.

For those who understood the archaeological evidence, there was never any doubt about the African origins of Great Zimbabwe (e.g., Garlake 1973; 1978a). Nevertheless, the resulting controversy has influenced much of the research that has been conducted at this famous site. Intellectually, it has dictated the questions that have been asked by researchers; physically, it has sometimes occasioned excavations and restoration work that damaged or destroyed archaeological evidence.

The earliest example of social complexity in Africa was along the lower valley of the River Nile. Some of the first indications have been found at Hierakonpolis, on the west bank of the Nile in southern Egypt, where extensive habitation and cemetery sites date to the fourth millennium BC. This and other sites suggest that social changes were already taking place amongst the formerly dispersed agropastoral people of the region. Local elites were emerging, and by about 3000 BC there is evidence for the First Dynasty of Pharaonic rulers of ‘Ancient Egypt’. This polity has long occupied a special place in Western consciousness, both scholarly and public. Pharaonic Egypt has been regarded as one of the great ‘civilizations’, as an almost mythical construct and treated as separate from the overall study of Africa's past. However, archaeologists are now considering Egypt in its African context, for example, O'Connor and Reid (2003b) and Exell (2011), but for a long time this was not done. It was insisted that Egypt during the last three millennia BC demanded extensive archaeological investigations, but until the middle of the twentieth century the rest of the continent was given far less attention. Consequently, the capacity for Egyptology to shed light on the processes leading to social complexity, in Africa and elsewhere, has sometimes been overlooked. This is unfortunate because the Egyptian evidence has great comparative value.

That evidence is exceptional in three ways. First, the climate and environment of the Nile Valley and its surrounding deserts have resulted in the survival of a range of organic materials rarely encountered in archaeological contexts elsewhere in Africa. Second, the monumental expression of social order by ancient Egyptians has left extensive physical evidence in the form of pyramids, tombs, temples, statues, and other structures that have dominated scholarly attention. Third, the development of writing, of which there was already an early form by about 3300 BC (Davies and Friedman 1999: 36–7; O'Connor 2011: 143–7), means that there are substantial documentary sources, particularly for later periods, as well as archaeological evidence. These factors have resulted in a massive quantity of data that has affected much of its interpretation. The Ancient Egyptians recorded their history in terms of the reigns and dynasties of their rulers, the latter numbered from one to thirty.

The examples of African precolonial urbanization and state formation discussed in Chapters 2 to 11 come from a wide range of environments. These include river valley, such as Egypt and Nubia; mountain plateau, for example, Aksum and Zambezia; savanna plain, for instance, Kano; rainforest, such as Ife and Central Africa; maritime fringe, as with the Swahili coast; grassland, such as the South African Highveld; and desert margins, for example, North Africa. The geographical locations of these developments within the continent are equally diverse. Egypt and Nubia were so positioned that they had substantial and long-standing contacts with the eastern Mediterranean. Aksum was situated beside a major trade route between the Mediterranean world and the lands of the Indian Ocean. Zambezia also had contact with Indian Ocean trade, although more remotely. The West African savanna, on the other hand, was separated from the outside world by one of the greatest deserts on earth, the traversal of which weakened external cultural influences. This was even more the case in the West African forest, situated far to the south of the desert margins and backed by the empty Atlantic Ocean that for many centuries was a more effective barrier than the Sahara. Different again were the cities and towns of the East African coast, which looked out on the Indian Ocean that provided links with distant but technologically sophisticated cultures. More remote from such contacts was Central Africa, which had relatively little contact with the outside world until quite late, as was also the case for the South African Highveld. North Africa looked north to the heart of the ancient world but also south into the Sahara, with some remarkable adaptations to its extreme environments. Such environmental and geographical diversity would suggest that we must look elsewhere for any common factors that might explain the appearance of social complexity in some parts of the African continent and, equally, its failure to appear in other parts.

Nevertheless, each area had its own characteristic environmental and geographical opportunities. In every case there were also environmental constraints, but the ingenuity of human culture partly mitigated their impact. Nowhere is this more starkly illustrated than on the middle Nile; on the one hand, the river offered life, but on the other hand, parts of the surrounding environment were so dry as to be scarcely habitable.