Different cities: Jenne-jeno and African urbanism

RODERICK], MCINTOSH

Why study African cities?

If abundance of agricultural land, non-restrictive land ownership, lack of labor, and low population densities generally can be said to have character­ized Africa before the last days of colonialism, life in cities has been the increasing to dominant experience of Africans since the mid-decades of the twentieth century.

Certainly, the continent has been at best an afterthought to most global urban social histories. Strange to say, it was sub-Saharan Africa's very variability of the city experience that rendered - at least to the early European visitors' perceptions - even some massive conurbations as merely “‘a heap of huts' - clusters of scattered settlements interspersed with tracts of cultivation and pasture, seeming to lack order or formality to the visitor's unfamiliar eye.”^[334]

Why was African variability misunderstood and underappreciated? (And what does that tell us about how we have thought about the circumstances in which “exceptional” cities emerged elsewhere?) Why was sub-Saharan Africa for so long considered lacking in cities worthy of grand comparative theory? The answers to these questions are important to an understanding of how African variability will, in the future, influence how urban evolution­ary theory may develop. In the first instance the answer is disciplinary: the study of the African city has traditionally been the preserve of geographers, sociologists, and political scientists^[335] and not, until recently, of urban social historians or archaeologists.

We now recognize the shallowness and nonsense of this view. It is a legacy of a colonial period tradition of explanation that stripped Africans of their active agency. Dumped into the anomie of cities created by exotic, non-African interventions, as people poured into cities without an appre­ciable “indigenous” past, Africans became peoples without their own his­tory, motivated by external politics and economies, and thus of marginal interest to the deep histories of cities.

A second reason for the late recognition of Africa's urban past has everything to do with scholars' (and nineteenth-century explorers') expect­ations of what cities should look like. African cities were, emphatically, not merely “a heap of huts.” For our global understanding of the commonalities and differences in the urban experience in deep and in shallow time, African cities are pertinent examples, but also require new kinds of explanations.

Over and over, visitors to ancient African cities or to their ruins, whether Arab traders or European imperialists and mercantilists, saw organisms so different from those of their own experience, so alien from their expect­ations, that they were not even recognized as cities. Take this classic example from the late 1950s: a distinguished French historian-archaeologist, Raymond Mauny, walked over Jenne-jeno, a massive urban tell blanketed in potsherds in present-day Mali, and wrote of his bewilderment: “The archae­ologist is utterly at a loss... here we are in quicksand (en terrain mouvant), and we are left with the impression of being only at the very beginning of the pioneering stage of archaeology.”^[336] Lacking any prior mention of this town in the Arabic sources and lacking the expected urban signatures in its form and in monumental architecture, Mauny felt intellectually adrift and swallowed up - his own words^[337] - by the enormity of the site and by his own incomprehension due (as he himself acknowledged) to his expectations of what an early city should look like.

Clearly, the mere identification of an early African city presented scholars with a recognition problem and a comprehension problem. Pre-colonial urban forms in sub-Saharan Africa were different, one from another, and many from the scholarly expectation of what an ancient city should look like and how they functioned. It is a commonplace of Africanist historians and archaeologists to say that African cities defy attempts to classify and typo- logize them. That is to say that they don't fit normative urban models. But I will assert that (among many other causal factors) the expansive differences in African cities' form and function reflect the continent's “exuberance” in throwing up alternatives to the normative ideas of the distribution of power. I predict that the greatest contribution of the study of Africa's varied cities, when understood as the signature on the landscape of multiple expressions of power and authority, will be to encourage students of comparative early cities to take a more expansive view of power, its relations, its expressions, its abuses, and implications of resistance to it, in those urban landscapes.

African alternatives to the distribution of power in cities

I begin by posing the following question: Can we conceive of cities without citadels and palaces? In other words, is the expectation so engrained that there is a causal relationship between the appearance of cities and the emergence of new and areally expansive expressions of power, defined as the influence of a few over the many based on the threat of force, that it is impossible to think of cities as other than concentrations of political power? To be sure, as the archaeology of cities advances around the world, this simple causal equation of political power and urban form has been rethought and recast. However, it is fair to say that much of urban social thought is not commensurate with new prehistoric and early historic data. Cities without citadels and palaces - the imposing seat of power and unam­biguous signposts of force - are still considered anomalous or the “excep­tions that prove the rule” (unless they are just proof of archaeologists' incompetence, because surely the seat of power must be there, somewhere!).

As more early African cities are studied and as the archaeology of these sites develops and as the extreme subtlety of power and authority relations of the communities they sheltered is understood, the causal relationship of power and urbanism has been problematized. This is not to say that all cities lack citadels. Great Zimbabwe^[338] or Aksum,^[339] as we shall see, are examples of symbols of power in stone; Meroe^[340] with over 200 pyramids is a place of monumentality. Yet, there is widespread consensus among Africanist archaeologists that a useful working definition - one that will guide field work and aid interpretation - must reflect what cities do and the role of the city within the evolving landscape in which they evolved. Understanding the function of the city within a larger and interconnected and always changing field of settlements is now understood as critical in the investigation of the several alternatives to distribution of power that characterizes the African landscape. One working definition of a city is “a large and heterogeneous unit of settlement that provides a variety of services and manufactures to a larger hinterland.”^[341]

Using this working definition, archaeologists can document the power relations (or alternatives to hierarchical power relations) that bind together the inhabitants within the city and with inhabitants of rural settlements in the served hinterland. Those power relations may be found in massive wealth or status asymmetries; power can be studied in a variety of expres­sions including, as we shall see, forms of great persuasive authority.

Map 17.1 Jenne-jeno and other Middle Niger cities and regions. The numbers refer to the number of excavation units sunk into each mound.

problem? Not at all, now that we have a working definition of urbanism that allows us to comprehend what is, arguably, the best-understood African city that is a case study of an alternative to the hierarchical distribution of power in an urban place.

The clustered city: Jenne-jeno

Over the marshes, winding streams, and rice fields of Mali's Middle Niger floodplain rises a tell that would not be out of place in Mesopotamia (see Map 17.1). Jenne-jeno's¹° descendant town, Jenne, lies 3 kilometers away; there its present-day inhabitants walk about on 9 meters of ancient city deposits. Within 4 kilometers are seventy tells in total, apparently all occupied contemporaneously with Jenne-jeno, and most abandoned at the

¹⁰ Roderick J. McIntosh, The Peoples of the Middle Niger (Oxford: Blackwell, 1998); and McIntosh, Ancient Middle Niger.

same time, around 1400 ce. In fact, in the 55,000 square kilometers of the seasonally inundated Middle Niger floodplain there are many hundreds of similar mounds. In the roughly 170,000 square kilometers that were flooded yearly within the past four to five millennia (much now in the deepest Sahara) there are hundreds of tells more. Yet, as recently as the mid-1970s this was not recognized as an urban landscape.

When Raymond Mauny walked over Jenne-jeno in the late 1950s and cast a late-colonial eye over the jumble of potsherds, mudbrick houses, and eroding burials, he couldn't understand what he saw.

In the era when a city was denoted by its citadel and palace, it was incumbent on the archaeologist to excavate the king's palatial domain, the associated residences of the elites, and the splendid temples in which kingship and its relation to the gods were celebrated, then to roughly gauge the area enclosed by the (obligatory) city wall. From such work the city could be delimited. Now, Jenne-jeno does have a city wall (curious, now that we think there was no warfare during its 1,500-year occupation; perhaps it was to prevent flooding), but no elite residences, no public architecture nor monuments, much less a king's palace. What the site does provide in ample volume is evidence of multiple occupations, multiple manufacturing areas and multiple “identity groups” as indicated by a great diversity of contemporaneous burial practices. After some thirty-five years of controlled stratigraphic excavations at some twenty-two units spread over the 33-hectare area of the site we have a plot of how that occupational and identity diversity burgeoned from the relatively simple (yet quite large at 20 hectares) community at its founding in the third century bce. Note that sinking so many excavation units, some as large as 10 meters by 6 meters, into the 6-meter-(and more)-deep deposits is enormously time consuming, but absolutely necessary. One is, after all, investigating the emergence and evolution of complex society, that intricate mix and dynamics of interactions of multiple corporate groups (defined as self-identified groups that hold real or symbolic property in common).

A complication to the archaeologist's tasks is the tight clustering of nearby settlements in a density that roughly decreases with distance from the principal site. What does clustering mean? One way to answer that question is to surface collect artifacts and surface features (burials, house foundations, etc.) from all the sites in order to ask the question, are there anomalous concentrations of certain artifacts or features, compared to the range and relative proportions at Jenne-jeno that reveal special activities, concentrated at different locales? Take the case of iron smelting. This is a dirty, hot, and noisy activity, one that is today in Mande society associated with the production of highly dangerous occult power. Not something one wants next door. Through the Jenne-jeno sequence we see the migration of smelting from the focus site to multiple satellites. Adding to the tasks of sampling the activities at the main site, the urban archaeologist also must study the activities in each of the satellite sites. Already in the first season of surface collection in the Jenne-jeno Urban Complex we detected a pattern of mutual exclusivity, or near exclusivity, of activities at satellites.

Thus, as excavations began at Jenne-jeno, we began a randomized collec­tion of sites within a larger ι,ιoo-kilometer-square region - just to get a feel for how the distribution of artifacts and the pottery assemblages compared to those at Jenne-jeno. On the basis of that season's analysis, it was clear that site density dropped off radically after a c. 4-kilometer radius and that within that cluster some sites had an “overabundance” of certain classes (or of a single class) of features or artifacts. The next stage (simultaneous with expansion of the excavations at Jenne-jeno) was surface collection of 50 percent of the satellite sites and, ultimately, of 100 percent. The last stage has included the sinking of excavation units into a sample of satellite sites. A considerable degree of concentration of activities (especially, iron- working, various styles of fishing, hunting aquatic mammals and snakes, weaving) continues to be the best explanation for the non-uniform, non­random distribution of surface and stratified artifacts in the Jenne-jeno Urban Complex. What does this clustering mean and what does this landscape say, potentially, about the distribution of power and authority within the larger community?

Some three and a half decades of research in the Jenne region allow us to say two things, with considerable confidence, about how this society was structured. First, clustering appears to have been a solution to the problem of how different specialist corporations might maintain and display their distinct identities, while at the same time having immediate access to other providers of needed goods and services and access to those to whom they themselves provided materials. All well and good, but one needs at the same time to answer the question, who monitors or controls the exchange of those manufactures and services? An older theory of the nature of cities would have answered, “the king and his administration, backed up by force, of course.” However, in the case of the Jenne-jeno Urban Cluster we appear to have an exchange system that functioned organically through the rela­tions of reciprocity forged among corporate groups. If corporate groups as well as expectations and rules were defined unambiguously, including consequences if the rules were transgressed, the urban system would not have required a vertically hierarchical control structure - which in fact we do not see. Rather, corporate ownership of individual mounds by members of an occupationally defined/kin group is indicated by the archaeological data indicating near exclusivity of occupational debris at each of the satellite sites.

Second, we have no hard evidence of a state-like, top-down, elite-driven political engine powering this kind of urbanism through time. We find no indications of kings, citadels, palaces, or, indeed, any obvious elites. The political and economic organization of late first millennium bce and later urbanism in the Middle Niger seems heterarchical. That is, one identifies separate, if sometimes overlapping, domains of authority, all functioning in an interactive field, not a vertical hierarchy of kings and subjects and unidirectional flows of information.

Emergent urbanism in the Middle Niger region

Recent research reveals cities even earlier than Jenne-jeno and especially a “pre-urban” landscape that was potentially several millennia in the making. This “pre-urban” landscape includes the creation of specialized sites and activities in a regional economy (a self-organizing landscape in the language of complex systems). This landscape displayed great resilience in a challen­ging environment as the Sahara transformed from a well-watered savannah (with vast lakes) as late as the fourth millennium bce to its present condition by the late third or early second millennium.

Our search for antecedent settlements, and possibly causal conditions, starts in the then better-watered Sahara, some ten centuries or more before the founding of Jenne-jeno. In a curvilinear arc spanning hundreds of kilometers in the central Mauritanian wasteland are many score Late Stone Age stone-built ruins - urban in proportions. Just how urban were they? This is difficult to answer since decades after the first (and only) systematic survey and dating of hundreds of these distinctive stone ruins in the Dhar Tichitt region of Mauritania, we still speak of a generalized “sedentary pastoral” adaptation (see Figure 17.ι). However, remote sensing reveals hundreds of multi-cellular Late Stone Age ruins (that is, multiple, essentially identical, enclosures with an exterior kraal walling, which we presume were cattle enclosures, with the residences in the interior). Unfortunately, the first systematic survey and test excavations of the 1960s have not been repeated since. Therefore, we are uncertain about the dating and contemporaneity of sites and whether there were smaller or less exotic settlements, which were observed in the initial survey. As we will see below, the same kinds of large stone enclosures are indeed associated slightly later with a variety of other sites in what we can begin to call an emerging urban landscape in which clustering is very much in evidence.

The Dhar Tichitt sequence occupies the latter half of a long, drawn-out movement of peoples beginning c. 4500 bce as highly arid episodes increas­ingly disrupt the Sahara's second Holocene wet period. The instability of the climate intensifies after c. 2300 bce, while the overall drying trend intensifies. These peoples, and those occupying the once inundated Azawad Basin of the ancient Middle Niger (stretching hundreds of kilometers north of Tim­buktu), found refuge centuries later in the as yet inundated part of the southern Middle Niger. MacDonald hypothesizes that many peoples, with many emergent subsistence and perhaps even craft specializations, were thrown together in these more southerly refuges.¹¹

One might imagine an initial situation of short contact and exchange among different groups of slightly different subsistence preferences, slightly different territorial ranges, which led to a seasonal, predictable co­occupation of a region. In this model of incipient complexity, “rules of engagement” for sustained interaction of a few communities of proto­specialists (some that favored fishing, some herding, and even some slightly differentiated as gatherers of various local grains) were developed. If regional exchange networks were successful, a movement toward true specialization ensued.

Kevin C. MacDonald, “Dhar Nema: From Early Agriculture to Metallurgy in South­eastern Mauritania,” Azania 46 (2011), ι and 49-69.

Different cities: Jenne-jeno and African urbanism

Figure 17.1 Top: Late Stone Age pastoral site of Akreijit of the Dhar Tichitt cliffs of central Mauritania (modified for publication with permission of Robert Vernet). Bottom: general plan of Kerma, Republic of Sudan (by permission of Ch. Bonnet).

Starting as early as perhaps 2000 bce, we hypothesize a movement of some of the Tichitt herders and very likely the generalized, aquatic-oriented hunters/gatherers/fisherfolk of the great playalands to the south, attracted by the still-vibrant lakes and waterways of Timbuktu's hinterland and Lake Region further to the south. On present evidence, these peoples would have moved down palaeochannels, such as the Wadi El-Ahmar (visible just east of Timbuktu). On satellite imagery we see familiar-looking stone architectural sites on outcrops and escarpments of southeast Mauritania. Here, stone Tichitt-like sites lie near “tell” sites of varying size, some of stone, others of wattle-and-daub or dry mud walls. These are not yet urban in size, but clustering of apparently contemporaneous settlements of very different styles suggests intensive relations of very different peoples. Our corporate groups of Jenne-jeno would be their direct legacies.

The seasonal co-evolution in which increasing numbers and kinds of specialists spent more months of each year in proximity (in the early and middle first millennium bce) would only work, of course, if the sites are contemporaneous. This we cannot yet verify for the Lake Region. But we do have initial evidence that this is the case for the Mema, on the basis of deep stratigraphic excavations at several mounds comprising the Akumbu Urban Cluster. Here streams and shallow lakes persisted into the first millennium ce, and aquatically oriented hunters/fisherfolk (present at least as early as 2000 bce) lived with semi-specialized herders sharing lake-beach and levee localities. And, 400-500 years later, they were joined by proto-cultivators in clusters of sites on the landscape. This pattern implies that not only were the “rules of engagement” for peaceful interaction among these groups maturing, but also that the sense of security these networks of reciprocity provided encouraged certain communities to become even more specialized (with, perhaps, the appearance of specialist artisans, such as potters).

Our best evidence for the next step of the progression comes from extensive survey and excavation of several large and small tells of the Mema and, especially, of the Timbuktu region, dating to the Late Stone Age-Iron Age transition (or, about 2500 bce).^[343] Clustered sites in preferred locales of the Mema transmuted into clustered Iron Age mounds (including, eventu­ally, several new satellites devoted predominantly to the craft of the iron smelters). If the inferred social contract is successful, then, the stage is set for the appearance of cities. There is a contemporaneous explosion of site size in the Timbuktu region. Tells are of urban proportions (50-100 hectares) during the mid-first millennium bce.

To bring the story full circle, as these cities were expanding, population also streamed into the previously unexploited far southern basins of the southern Middle Niger. In the last centuries before the Common Era, Jenne- jeno exploded on the scene, already large (>20 hectares) at its foundation. But, as opposed to the northern cities, the main tell remained at a “mere” 33 hectares throughout its maximal prosperity during the mid- to later first millennium ce. The urban population was dispersed among scores of specialist-oriented satellites of the Jenne-jeno Urban Cluster.

Although the evolution of urban centers in the region does not include the appearance of kings and urban elites, the trajectory to urbanism in other African cities is quite different (see Map 17.2).

Map 17.2 Cities and states mentioned in the text.

A spectrum of early African cities (outside Egypt)

The understanding of the evolution and nature of east African cities has similarly changed greatly in light of new archaeological field work.^[344] For example, the early Nubian city of Kerma was once believed to have flourished during the Egyptian Middle Kingdom (c. 2000-1600 bce) in the Nile's Dongola Reach, a product of trade with Egypt (see Figure 17.1). The city looked Egyptian, imitative in monuments (including huge brick “pyra­mids” that were religious structures called defuffas) and clearly derivative of Egyptian cities.

However, after thirty years of extensive excavation by the Swiss Arch­aeological Mission to the Sudan (Universite de Neuchatel), our knowledge of Kerma has changed the picture considerably. Although trade and cultural contacts down the Nile were important, the city's role as focus of a political and cultic capital owes much to local developments. The importance of manufactures and of crafts, especially of a spectacular ceramic production and perhaps even of the invention of faience, are clear. Also, from the earliest days, transhumant cattle-herding groups swelled the population seasonally. The symbolic importance of cattle is evident even in the tombs and tumuli of the kings and elite in a vast nearby cemetery. Anything but derivative, Kerma was urban by c. 3000 bce and occupied a locale sacred to the worship of cattle that is traceable at least four millennia before that.

Kerma is one of the southern-most of a string of late fourth millennium “proto-towns,” each with religious significance, which were part of an exchange-oriented and areally extensive Nilotic culture that only later coalesced in various forms, including the elite culture of ancient Egypt in which distinctive Egyptian cities developed (see Baines, Chapter 2, this volume). Kerma was eventually succeeded by Meroe, the southern capital of the Kingdom of Kush from approximately 800 to 350 bce. Although Meroe was also once thought to be a derivative of Egypt, mainly because of the hundreds of distinctively Nubian pyramids nearby, we now appreciate the local jewelry, gold and iron industries, and the great importance of Red Sea trade (as opposed to north-south Nile trade) (see Figure 17.2).

Other African cities have formerly been regarded as secondary develop­ments from more civilized corners of the ancient world. Such was the case with Aksum, long presumed to have begun as a Sabaean Arabian outpost,

Figure 17.2 Right: Plan of Meroe (from P. L. Shinnie and J. R. Anderson, The Capital of Kush 2: Meroe Excavations 1973-1984 [Wiesbaden: Harrassowitz Verlag, 2004], with permission). Left: Great Zimbabwe - Hill Complex above; Great Enclosure below (modified for publication with permission of Innocent Pikirayi).

but now understood as a purely African response to the potential of its location on the Horn of Africa. Indeed, Aksum participated in trading relations with lands as distant as India. Long controlling the lower Red Sea, and to a lesser extent the Persian Gulf from the first century ce until the latter half of the first millennium, the Aksumite Kingdom through its main port of Adulis was a major player in the pan-Indian Ocean trade linking Southeast Asia, India, the Gulf, and the Swahili coast of eastern Africa. Archaeology is only now revealing the extent to which the importance of this Indian Ocean network rivaled that of the Silk Route.

We know relatively little about the pre-Aksumite origins of urbanism on the eastern Horn (which made it far too easy to argue that cities were imported). Excavations have perhaps not surprisingly concentrated on the famous underground tombs of kings and nobles, on the foundations of the massive (some over 30 meters tall) commemorative stelae, and to a lesser extent upon the residences of the elite. Cosmopolitan and outward looking, Aksum welcomed traders from the Arabian Peninsula, Rome, Egypt, India, and Nubia to its markets featuring goods from the African interior - ivory and rhino horn, live wild animals, slaves, and grains. Aksum, the capital, was also a classic primate city, much larger and more highly stratified than the kingdom's other ports and interior towns.

Other classically primate capitals of African kingdoms, dating to later times than this volume's cities, such as Njimi of eleventh-fifteenth-century Kanem-Bornu (Chad and northeast Nigeria), and Kano (capital of the Hausa state in central Nigeria, which may date as early as the tenth century), are known more from ethnohistorical sources than from archaeology, and their origins are obscure. New field work by African archaeologists will remedy this lack of knowledge.

For hundreds of kilometers along the eastern coast of Africa one finds a string of important city-states: the trading cities of the Swahili coast.^[345] United by a common language (Bantu at base), by belief in Islam, and by a long­distance mercantile identity, many joined in temporary alliances, but none achieved dominance over the others. Cities such as Kilwa, Sofala, Lamu, Manda, Zanzibar, or Malindi varied as much by size and by elegance of their mosques and elite residences, as by the advantages of their harbor or as by the relations with affiliated villages and hamlets in the interior. Trade was in slaves, gold, ivory, sandalwood, and spices. In a strange, but not uncom­mon, mismarriage of history and archaeology, the towns' origins were thought not to have pre-dated the tenth century, the time of colonization by Shirazi Persian immigrants - and, hence, archaeologists tended to stop digging at places like Kilwa when “pre-Shirazi” levels were reached.

More recent excavations of towns and villages at Sanga, Pemba Island, and the Comoros Islands, and coastal and other surveys outside of the historically documented towns reveal a very different origin of “Swahili” urbanism. The towns and cities were built upon layer upon layer of older trade centers that were oriented to trade with China, Southeast Asia, and South Asia. New research is just beginning to bring these urban origins to light.

These long-distance interactions from the interior and then through ocean-going trade are the backdrop for state formation and urbanism in the interior. The best-known city is Great Zimbabwe (see Figure 17.2).^[346] Dominating the interior-coast ivory routes and set near important gold reefs, Great Zimbabwe was architecturally impressive (so much so that early European colonists refused to credit an indigenous origin) and symbolically highly charged. European mining interests despoiled gold artifacts and cult objects such that no undisturbed deposits remained for the first legitimate excavations in the 1920s. We rely mainly upon oral traditions of the des­cendant Shona peoples, and increasingly on excavations at other, earlier settlements (such as Mapungubwe, see below) to understand the trade, cultic, and political importance of Great Zimbabwe.

Conclusion

To summarize, some African cities developed in the context of interregional trade, others were politically dominant in their regions, and still others were “clustered cities,” showing little political or economic hierarchy. African urbanism encompasses many kinds of cities and many kinds of power. There is much work, clearly, to be done to understand African cities. There is also a certain quest for African cities that can serve as the endpoint and prospective of this chapter.

There is first the pre-fourteenth-century capital of the Kongo Empire, which is known only from oral histories of Kongo people. In southern Mali and northern Guinea, archaeologists still search for the capital of the Mali Empire (Dakajalan, in the oral traditions), while respecting the warnings of traditionalists among the local Mande peoples of places not to go. Some cities are too dangerous for prying eyes. And in the highlands of Ethiopia, we know from oral accounts that there were wandering royal capitals that left no trace. All these cases, of course, have to be added to the calculus of what makes African urbanism so enthusiastically original.

Furthermore, extensive archaeological survey and excavation have failed to find undisputed urbanism at Mapungubwe on the banks of the Limpopo in South Africa. The eleventh-century site of that name did provide impres­sive elite burials as well as gold objects and glass beads from the Indian Ocean trade (dominance over which it relinquished to Great Zimbabwe in the mid-thirteenth century). But the site cannot be claimed to have been a city, however areally extensive the polity for which it served as political and ritual focus. What kind of power is this? Even more curious is the case of the Middle Senegal River Valley, on the other side of the continent. The very first state mentioned by the earliest of the Arabic Chronicles of West Africa is Takrur, which extended far north into the western Sahara and from the Atlantic to the interior. Its gold workshops provided currencies for the Mediterranean world of late antiquity and medieval periods. To date, however, survey in the region of the Middle Senegal Valley, guided by oral traditions and the Arabic Chronicles, has yielded only clusters of hamlets or small village mounds that date to the early centuries of the Common Era. These clusters, however, cannot be called in any sense urban.

Early African cities and the distribution of power in them were neither cut to a normative pattern, nor did they develop from any single cause. Neither did every city need a state, nor every state a city.

further readings

Bonnet, Charles, The Nubian Pharaohs, New York: The American University in Cairo Press, 2003.

Connah, Graham, African Civilizations, Cambridge: Cambridge University Press, 1981. Garlake, Peter, Early Art and Architecture of Africa, Oxford: Oxford University Press, 2002.

Great Zimbabwe, London: Thames & Hudson, Ltd., 1973.

Insoll, Timothy, The Archaeology of Islam in Sub-Saharan Africa, Cambridge: Cambridge University Press, 2003.

Kusimba, Chapurukha M., “The Collapse of Coastal City-states of East Africa,” in Akinwumi Ogundiran and Toyin Falola (eds.), Archaeology of Atlantic Africa and the African Diaspora, Bloomington: Indiana University Press, 2007, pp. 160-84.

“Early African Cities: Their Role in the Shaping of Urban and Rural Interaction Spheres,” in Joyce Marcus and Jeremy A. Sabloff (eds.), The Ancient City: New Perspectives on Urbanism in the Old and New World, Santa Fe, NM: School of Advanced Research Press, 2008, pp. 229-46.

McIntosh, Susan Keech, ed., Beyond Chiefdoms: Pathways to Complexity in Africa, Cam­bridge: Cambridge University Press, 1999.

Shaw, Thurstan T., Nigeria: Its Archaeology and Early History, London: Thames & Hudson, Ltd., 1978.