Pathways to Social Complexity and State Formation

Afr Archaeol Rev
DOI 10.1007/s10437-008-9031-3
O R I G I N A L A RT I C L E
Pathways to Social Complexity and State Formation
in the Southern Zambezian Region
Nam C. Kim & Chapurukha M. Kusimba
# Springer Science + Business Media, LLC 2008
Abstract Theorists have put forth various anthropological perspectives on the
variables leading to social complexity and the emergence of state-level polities. This
paper incorporates data from the Zambezian region of Southern Africa in order to
contribute to the literature on social evolutionary theory. It traces the cultural
trajectories of communities that flourished during the region’s Iron Age within the
Shashi-Limpopo Basin, leading to the emergence of the Great Zimbabwe polity. In
examining the archaeological record, the authors discuss the emergence of state-like
societies, offering a review of current interpretations and explanations for the
emergent complexity.
Les théoriciens ont émis plusieurs points de vue anthropologiques sur les
variables qui conduisent à la complexité sociale et à l’émergence des états. Cet
article est une contribution à la littérature sur la théorie de l’évolution des sociétés
réalisée à partir des données de la région du Zambèze au sud de l’Afrique. Il retrace
les trajectoires culturelles de communautés qui sont apparues au cours de l’age du fer
dans le Bassin du Shashi-Limpopo et qui ont conduit à l’émergence de l’état du
Grand Zimbabwe. En s’appuyant sur les données archéologiques, les auteurs
discutent l’émergence de sociétés-états, offrant une revue des dernières interprétations et explications au sujet des complexités émergentes.
Keywords Great Zimbabwe . Zimbabwe culture . Mapungubwe . Fortification .
Social and political complexity . States . Urbanism . Zambezia
N. C. Kim
Department of Anthropology, University of Illinois at Chicago, 1007 W. Harrison Street, Chicago,
IL 60607, USA
C. M. Kusimba (*)
Department of Anthropology, Field Museum of Natural History, 1400 S. Lake Shore Drive, Chicago,
IL 60605-2496, USA
e-mail: [email protected]
Afr Archaeol Rev
Introduction
The origins of social complexity, urbanism, and archaic states are of profound
interest for social scientists. Investigating social evolution is an important aspect of
elucidating culture change and human organizational behavior, and this is one of the
core missions of anthropology. Studying the evolution of social complexity offers
insights into understanding strategies that segments of humankind have employed to
accumulate power (Earle 1997; Haas 2001; Holl 2000; Robb 1999), or in the causes
of social change and societal inequality (Blanton et al. 1996; Blanton 1998; Carneiro
1970, 1990; Feinman 2001; Johnson and Earle 2000; Trigger 2003). Furthermore, to
truly understand human grouping patterns and political organization within our
modern-day contexts, it is essential that we explore their historical dimensions and
manifestations. Archaeology thus contributes to the social evolutionary debate by
highlighting the origins of stratified societies as a precursor to more complex
conglomerations such as the city-state and territorial state. At the same time,
archaeology helps us understand how and why people living in similar ecosystems
may embrace different strategies, ranging from egalitarian or acephalous, heterarchical, or hierarchical, in creating communities (Ehrenreich et al 1995; Holl 2000,
2003; McIntosh 1999). This article contributes to the ongoing debate on the
emergence of social complexity worldwide; a debate that we believe has not yet
incorporated the rich data from Sub-Saharan Africa (cf. Holl 2000, 2003; Kusimba
1999; McIntosh 1999; Pikirayi 2001; Stahl 2001; Trigger 2003). In doing so, we
reexamine published archaeological records primarily from three extensively studied
sites of the southern Zambezian region, namely Bambandyanalo, Mapungubwe, and
Great Zimbabwe, to address factors that contributed to the rise and development of
social complexity and statehood in southern Africa. We discuss the main
archaeological indicators of social complexity, current interpretations of these data
and how they fit into the general debate of archaeology of social complexity
worldwide.
The Shashi-Limpopo Basin experienced major technological and sociopolitical
transformations towards the end of the first millennium AD (Manyanga 2006:138;
Pwiti 2005). These transformations included rapid demographic growth, due in part
to migration and natural growth, an increase in societal inequality evidenced by
differential household size, wealth and status, and the emergence of site hierarchies
(Manyanga 2006:138). To what can we attribute the transformation of acephalous
Iron Age societies of southern Africa into state societies? To answer these questions,
we first revisit the history of social formations traceable to the late first millennium
BC when pastoral and agrarian communities inhabited the region (Pikirayi 2001:34).
Second, we look at the underpinnings of societal inequalities using an intersite and
regional perspective that highlights the efflorescence of Bambandyanalo (also
known as K2), Mapungubwe (Huffman 2005, 2007), and Great Zimbabwe (Pikirayi
2006a). We believe this historical approach provides clues towards understanding
the prevailing conditions under which certain household heads could have
accumulated unequal wealth and legitimacy, enabling them to become leaders of
their various polities (Haas 2001; Kusimba 1999; Kusimba and Oka 2008). How and
in what ways did these polities function? How did these incipient leaders centralize
authority and consolidate their power? What was the nature of relationships amongst
Afr Archaeol Rev
Bambandyanalo, Mapungubwe, and Great Zimbabwe? And what was the relationship between these larger settlements with small, more peripheral and frontier states?
As in many regions around the world, larger and more built-up settlements that
were presumably capitals have received more scholarly attention. In contrast smaller
more regional/rural chiefdoms that may or may not have been allied to the state
capitals have received little or no attention (e.g., Usman 2001). It is important to note
how smaller-scale albeit peripheral communities may have interacted with their more
powerful neighbors. Was the nature of their relations based on tributary and
subordinate ties? Was it heterarchical? How may archaeology operationalize these
relationships? Understanding these relationships has implications for knowing how
and by what means incipient leaders emerge and consolidate power.
The emergence of social complexity and the state in southern Zambezia has been
attributed to a variety of factors. These include external trade (Huffman 1972, 1982,
1986b, 1996, 2005), accumulation of cattle (Beach 1998; Denbow 1984), religious
ideology (Beach 1980, 1998) and climate change (Pikirayi 2001). Some researchers
(e.g., Hall 1990; Manyanga 2006) have proposed the importance of warfare and
coercion as one of the means used to accumulate and keep power but little
archaeological evidence has been gathered thus far to support this hypothesis, at
least, for the initial stages of social complexity. While the material evidence for
organized violence and warfare in precolonial southeast Africa is abundant for the
second half for the second millennium (e.g. Pikirayi 1993; Huffman 2007),
archaeologists must intensify efforts to identify and develop the chronological
context of warfare and its aftermath during the nascent stages of state formation.
The lines of evidence for operationalizing the archaeological recognition of
warfare pertain to tensions over resources, societal attempts to monopolize resources,
and fortification and aggregation at resource-rich areas in proximity to permanent
sources of water. Forms of evidence include oral testimonies, written records,
iconography, weapons, fortifications, paleopathology, and violent destruction,
disruption of cultural patterns, and the relocation of settlements (Holl 1985, 1997,
2003; Webster 1998:315). The currently available material evidence hints at the
possibility that Iron Age communities of the Shashi-Limpopo Basin not only had
military defense on their minds, but that they may have also engaged in organized
violence in order to achieve objectives and political agendas.1
We propose that Mapungubwe and Great Zimbabwe achieved and maintained
economic, political, and ideological hegemony through a combination of both
peaceful and coercive means. Certain material indicators combine to support this
proposition, material signatures that relate to shifting patterns of settlement location,
fortification, mortuary practices, territoriality, and migration, all of which combine to
support warfare and violence hypotheses. We hope that this paper will initiate the
reevaluation and reinterpretation of current patterns in the archaeological landscape,
contribute to the global literature on state formation, and open discussions within the
African archaeology community on the pathways to complexity along lines now
common place in other regions (Trigger 2003).
1
We explore the role of coercion and warfare in the rise and fall of state societies in Southern Zambezia in
a separate paper to be published soon (Kim and Kusimba 2008)
Afr Archaeol Rev
Defining the State
Archaeology tells us that societal change is a universal cultural dynamic which all of
humankind has embraced, and they continue to embrace, albeit at different moments
and paces in history. Not all societies evolved into states but many of those that did
generally exhibit salient features that have guided archaeological theorizing on the
emergence of the state (Bilman 2002; Carneiro 1981; Flannery 1998; Fried 1967;
Johnson and Earle 2000; Keeley 1996; Marcus 1998; Service 1975; Spencer and
Redmond 2004; Trigger 2003; Yoffee 2005). There are countless definitions of the
state but for our purposes here, we will use Bruce Trigger (2003) since we believe it
is more relevant to the African context we discuss in this article.
In Understanding Early Civilizations, Trigger (2003:92) defines the state as: “a
politically organized society that is regarded by those who live in it as sovereign or
politically independent and has leaders who control its social, political, legal,
economic, and cultural activities” (2003:92). In this view, centralized control over
many aspects of everyday life in a society is a defining feature for the state. Power can be
manifest in economic, political, physical, and ideological forms, and one would expect
to see material indications for some or all of these forms within an emergent or
established state. Furthermore, due to the aggregating nature of states and the labor
resources required to administratively run the requisite operations, significant numbers
of human resources are necessarily present. In describing early states, Trigger (2003:47)
emphasizes the importance of kinship ties and how rulers regularly used force to
maintain their authority. His definition makes clear that early states were socially
stratified, and that individuals occupying the highest strata possessed the most wealth,
status, and political power, while concurrently possessing the means to maintain these
distributions of wealth and power. “The core of such an early state (or complex
chiefdom) was an ethnic group, tribe, or ruler’s kindred to which other groups
willingly or unwillingly paid tribute” (Trigger 2003:47). The means through which
these members of the upper strata were able to gain and maintain their power likely
included a mixture of physical as well as ideological power.
When discussing state-level societies, Trigger makes an important and necessary
distinction between “city-state” and “territorial state”. This dichotomy is germane to
our discussion of Iron Age southern Zambezia, as there are differences in size, scale,
and territorial extent for certain polities. Trigger (2003:92) echoes Akin Mabogunje’s
(1962) notion that city-states were relatively small polities, consisting of an urban
core surrounded by farmland containing smaller units of settlement. In contrast,
territorial states possess a ruler who governed a larger region through a multileveled
hierarchy of provincial and local administrators in a corresponding hierarchy of
administrative centers (Huffman 1986c, 2007). In terms of archaeological markers,
boundaries for early city-states were sometimes clearly demarcated by natural
features, artificial markers, or walled fortifications, with populations ranging
anywhere from 5,000 to 20,000 (Trigger 2003:94). To illustrate case studies of
early city-state civilizations, Trigger (2003:94) outlines evidence from the Uruk
period (3500 to 3200 BC) in southern Mesopotamia. In conjunction with the
emergence of agricultural practices, the region also experienced increasing social
stratification, organized warfare, urban centers, large public buildings, and a citystate system.
Afr Archaeol Rev
Given these features, there are important inferences to be made. First, the societies
within this region were fairly sedentary and committed to their locations. This is
evident in the amount of agricultural and architectural investment put into the cities.
Second, populations had grown sufficiently in these urban centers, enough to
accommodate all of the labor needs required for the urban construction. Finally, it is
clear that power and control were centralized, as large-scale production and
architecture could only efficiently occur through coordinated planning and
operational implementation.
It is thus evident that emergent and established prehistoric states were highly
populous, socially stratified, complex societies in which asymmetries of power and
wealth existed. The leaders and elites within these societies were able to achieve
their status and garner their power through a variety of means. In sum, states can be
characterized by centralized political authority resting in the hands of a small elite
group, wherein the power is oftentimes physical, economic, and ideological by
nature (Mann 1986; Morris 1998:98). Furthermore, an urban or semi-urban center
often served as the central nodal point of interaction, serving as the capital and
destination for agricultural and material tribute from the surrounding hinterland
(Yoffee 2005:91).
How may archaeologists operationalize the state archaeologically? As they rise,
decline and eventually collapse, states inevitably leave certain clues about their
history: technological, social, economic, and political. In addition to markers
indicating social stratification and wealth differentiation, these societies will also
exhibit signs of religious and ceremonial functions, higher populations and
centralized control over administrative functions (Flannery 1998:54, 55). As such,
archaeologists can reconstruct daily life by studying the structural and symbolic
edifices of an archaeological site. This can include the architecture, temples and
shrines, elite and non-elite residential quarters.
Pathways to Social Complexity and State Formation
Having reviewed the defining characteristics of state-level societies and some of
their material signatures, we now turn to the archaeology of Later Iron Age societies
in southern Zambezia, where important transformations from acephalous to state
societies occurred towards the end of the first millennium AD. The question we
address here is the form and degree to which states existed in prehistoric southern
Zambezia. What is the archaeological evidence there to indicate the presence or
absence of state-like societies?
Zambezia generally refers to all those regions drained by the Zambezi River
and the Zimbabwe plateau (Pikirayi 2001:3). It covered five countries in
Southern and Southeast Africa including Zambia, Zimbabwe, Botswana, South
Africa, and Mozambique. The earliest recorded manifestations of social complexity and state formation in central and southern Africa are in Southern Zambezia.
Both in terms of coverage, prominence and controversy, research in the Zimbabwe
plateau centered on Great Zimbabwe has dominated the region’s archaeology (e.g.,
Bent 1892; Beach 1980, 1998; Caton-Thompson 1970; Hall 1990; Huffman 1996;
Ndoro 2001; Pikirayi 2001; Soper 2006). Within this region, commonly referred to by
Afr Archaeol Rev
archaeologists as the Zimbabwe Culture, archaeologists have defined a number of
chiefdoms and states.
The Zimbabwe Culture is characterized by the presence of massive stone walls
built in a variety of styles (Pikirayi 2001:3, see also Huffman 1996). Dating from
approximately the eleventh to the late nineteenth centuries, the Zimbabwe Culture
can be divided into three main cultural periods. The first phase, the Mapungubwe
phase, date from the mid-eleventh century until the late thirteenth century AD. The
site of Mapungubwe, the type-site for this first phase, attained regional prominence
during the thirteenth century, managing the resources of a territory equivalent to a
state in both political and economic terms (Pikirayi 2001:3). The second phase is the
Great Zimbabwe phase dating from approximately 1270 to 1550 AD. The second
phase is highlighted by the city of Great Zimbabwe. The third phase dated from the
sixteenth century and was based at the Mutapa state and the Torwa polity (Pikirayi
2001:3). Although for purposes of this paper, we focus on the first two phases—
Mapungubwe and Great Zimbabwe—we cannot effectively assess the emergence of
these states regional dominance without first revisiting the region’s cultural history
prior to Mapungubwe’s ascension.
From Pastoralism to Farming
The earliest known inhabitants of the fertile highlands and open savanna grasslands
of Zambezia were hunter-gatherers (Huffman 2007; Mitchell 2002; Walker 1995).
These hunter-gatherer communities managed the land exploiting its resources,
relying primarily on stone-tool technology until the late Holocene. At about 150 BC,
the region witnessed a major transformation. The recovery of domestic fauna of Bos
indicas and ovicaprids and pottery at a number of sites suggests the appearance and
initiation of pastoralism to complement foraging as the primary means of
subsistence. Most archaeologists have attributed these initial transformations to
migration. However, we believe that the dynamics were subtler than merely one
group moving into the region, with a relatively superior way of making a living, and
completely overwhelming original inhabitants (Pikirayi 2001:73). The early first
millennium subsistence strategies in southern Zambezia were thus foraging and
pastoralism and the archaeological evidence suggests that the latter increasingly
became the more dominant subsistence strategy as more people acquired cows.
Pastoralism ushered in the germ of inequality as those with more cows carried with
them higher economic and social status. Although relatively few pastoral sites are
archaeologically known, the landscape would have been dotted with semi-sedentary
pastoral camps, hamlets, and villages of Khoisan speakers (Pikirayi 2001:77).
The first reported evidence of what would appear to be agrarian communities
dates to the first millennium AD (Pikirayi 2001:80). Interestingly the majority of
southern Africanists attribute these changes to Bantu speaking migrants from eastern
and north-central Africa. The evidence is drawn from the appearance in the
archaeological record of a complete tool kit of iron technology, distinctive ceramics,
and new crops (Huffman 1982; Mitchell 2002:259; Pikirayi 2001:80). Farming, put
more accurately gardening, herding, and foraging, all combined to create a more
diversified economy resulting in demographic growth (Huffman 1996). By the
Afr Archaeol Rev
fourth or fifth century AD the farming technology had crossed the Limpopo into
northern South Africa and areas further south (Hall 1990; Mitchell 2002; Pikirayi
2001:79; Van der Merwe 1969; see Mitchell 2002 for more citations).
The Zambezia landscape of the first millennium AD was dotted with temporary
rockshelter settlements, semi-sedentary camps and villages, and permanent settlements. Farming and pastoralism were extremely significant in transforming both the
cultural mindset and the landscape. Local and interregional trade flourished amongst
foragers, herders, and farmers. Iron smelters, blacksmiths, and potters prospered.
Change was in the air and it was rapid. Demographic changes were regional. Many
areas previously uninhabited were now settled. Residential areas became larger as
did gardens and farmsteads. Owning cattle signified wealth and status but also
required labor to maintain. Mitchell (2002:288) proposes that “domestic animals and
crops imply private property; long-term storage of cereals suggests this must have
been controlled within families or perhaps centrally within villages.” Presumably
property rights and an adherence to territorially affiliated beliefs and ancestor cults
were established at this time (Pikirayi 2001:79).
Towards the mid/late first millennium the interaction sphere in Zambezia had
extended to other areas. Long-distance trade and exchange with the coast had been
established and increasingly became one of the chief means of accumulating wealth
in addition to pastoralism and farming (Mitchell 2002:288, 2005). The hunting and
processing of ivory for export evidenced at larger sites indicates the complexity of
coastal-Zambezian relationships, which promoted investment in the procurement of
trade items in high demand. Such demand required investment in the infrastructure
necessary for producing these items. Specialized craft activities, such as ivory
working, recovered on larger first millennium sites also served as catalysts for
development of related crafts including iron working and gold mining. Ivory
procurement created a group of highly specialized hunters who would have sought to
restrict the specialized knowledge to their group. Trade with the coast thus indirectly
helped diversify the local and regional political economy of Zambezia. In Mitchell’s
words, “such specialisation…facilitated the expansion of trade and the possibilities
for individuals or groups to benefit…at the expense of others, including
opportunities for accumulating larger herds of cattle” (2002:289). In this regard
rapid demographic growth, food security, favorable and stable climate, investment in
highly specialized craft activities, and long-distance trade combined to lay the
foundations for the development of the larger chiefdoms to become states in
southern Zambezia (Manyanga 2006:21).
In sum, it is evident that by the mid to late first millennium AD greater amounts
of wealth and status were being conferred upon certain segments of societies in the
region (Manyanga 2006:139). By the late first millennium AD, ivory and skins were
already being exported overseas, with sites like Chibuene interfacing between
interior and transoceanic trade routes (Mitchell 2002:300; Pwiti 2005; Sinclair 1982;
Sinclair and Hakansson 2000). The presence of craft specialization and material
symbols of high status are telling, and we can infer the onset of social stratification
and private property on a level heretofore unseen in the region. The germ for
hierarchical relationships, social complexity and state development had been
planted. By the early centuries of the second millennium, these farming villages
and their lifeways had become well established. “Here and there small-scale
Afr Archaeol Rev
confederations of subsistence farmers lived in stable agricultural regimens well
adapted to their natural surroundings, adjusting to these changes in their social
environment” (Pikirayi 2001:95). These farming communities represented politically
autonomous villages that would soon became states.
Precursor to State Development: Bambandyanalo
The Iron Age sites known as Mapungubwe and Bambandyanalo on the farm
Greefswald, west of the modern town of Musina in the Limpopo Province of South
Africa, have aroused significant speculation ever since their discovery in the early
1930s (Fagan 1964:337). The area consists of a number of red sandstone hills, which
dominate the geology of the area. The main activities seem to have been focused on
Bambandyanalo Hill and its adjacent valley, where extensive mounds are located,
and Mapungubwe Hill just over a kilometer to the east, where extensive settlement
was found on the hill, its adjacent slope/terrace and base. Bambandyanalo (also
known as K2 after the mounds found in the Near East) is located at the Greefswald
farm, a few kilometers south of the Shashi-Limpopo confluence. K2 has massive
middens mixed with desiccated dung, measuring 200 m across (Huffman 2005,
2007; Robinson 1966; Pikirayi 2001:107; Summers 1967). Stratigraphic profiles
have revealed that the occupants settled in the Bambandyanalo valley by successive
or continuous occupations, and accumulated the mound of habitation debris that
survives today (Fagan 1964:338). Possessing cultural remains related to the
Leopard’s Kopje tradition, the site dates from the tenth to eleventh centuries
(Pikirayi 2001:107). It was a substantial village, part of a settlement system that
included Schroda, 6 km to the northeast, and other Zhizho and Leopard’s Kopje
villages (Pikirayi 2001:107). Compared to other periods, the K2/Mapungubwe
period represents a time of intense occupation and expansion in the region, with sites
oriented toward the rivers and floodplains (Manyanga 2006:80).
The large, circular mound of occupation debris of Bambandyanalo measures
some 182 m in diameter and up to 6m deep at its highest point (Fagan 1964:338).
The main midden at Bambandyanalo, dating between 1030 and 1220 AD, stands out
above the surrounding occupation area, reaching more than 6 m deep in some places
(Mitchell 2002:300). Covering more than 8 ha and possibly housing as many as
2,000 people, the settlement consisted of pole-and-daga houses with gravel floors
focused around and to the west of a large byre (Mitchell 2002:300–301).
The people at Bambandyanalo were cattle herders and elephant hunters who
worked extensively in ivory and bone (Fagan 1964:10, 343; Pikirayi 2001:107).
They were also agriculturalists. Ivory was carved into bangles and bracelets, while
bone was made into points. While these products were traded locally at first, they
were exchanged out of the region for glass beads obtained from as the coast and
towns like Chibuene (Pikirayi 2001:109; Sinclair 1982). The town had thus gained
prominence by the early eleventh century. Radiocarbon dates from test excavations
indicate successive house floors and great amounts of domestic refuse all
accumulated within a short period of time, maybe a single generation (Pikirayi
2001:109). Though researchers disagree as to exactly when Bambandyanalo was
abandoned, it is clear that occupation was short-lived and abandonment was
Afr Archaeol Rev
somewhat sudden, with a transition of power to nearby Mapungubwe Hill (Mitchell
2002:302; Pikirayi 2001:109). According to Pikirayi (2001:107), the town of
Bambandyanalo represents a significant step in the development of social
complexity in the region.
Seventy-four skeletons have been excavated there, buried with pottery and some
jaws of cattle. The bodies of the 74 skeletons were normally lying on their sides in a
flexed position, surrounded by pots (Fagan 1964:339, Meyer 1998; Steyn 1997;
Steyn and Henneberg 1995a, b, 1996, 1997). A study of at least 40 of the individuals
by Galloway (1959) proposed that the remains were likely to be of pre-Bantu
populations of the “Boskop-Bush” in physical type and “Hottentot” in culture.
Galloway also argued that the remains could not have been Bantu since they
exhibited non-Bantu characteristics. Later research indicated that these individuals
were Bantu (Rightmire 1970); the jury is still out on the ethnic identity of these
people. More research, including genetic studies, would be necessary to settle the
question of the ethnic identity of the people of Bambandyanalo.
According to Fagan (1964), the first occupants of Mapungubwe Hill were the
people who occupied Bambandyanalo in its latest stages, and their occupation is
sealed from the later levels by a layer of black ash. There is a complete break in the
pottery sequence at this stage, and spindle whorls and abundant iron tools make their
appearance. In addition, complicated daga structures appear at Mapungubwe, along
with gold and copper ornaments. Taken in sum, there are indications that by the
emergence of Mapungubwe’s occupation and Bambandyanalo’s abandonment in the
early thirteenth century, profound social changes were occurring in the area during
this transitionary phase (Fagan 1964:339; Mitchell 2002:302).
What was the nature of these profound social changes, marked by Bambandyanalo’s abandonment, and what caused them? The incursions of Bantu migrants into
the area and their incorporation into the cultural milieu of their hosts seem plausible
(Fagan 1964:352). Rapid population expansion due to natural growth, intermarriage,
migrations, and interethnic tensions may explain the rapid expansion and
abandonment of Bambandyanalo in favor of Mapungubwe. This view is bolstered
by the archaeological evidence indicating the interruption of continuous occupation
at Bambandyanalo by the arrival of new groups in the Greefswald region in the 11th
century. Brian Fagan (1964:340) believed that it was these immigrants along with
the citizens of Bambandyanalo, who were the first inhabitants of Mapungubwe.
As stated earlier, the archaeological record shows that the second half of the first
millennium witnessed profound changes in land use from primary subsistence based
upon pastoralism to a combination of agrarian and pastoral and foraging strategies.
The migration theme is still a dominant way of seeing change in southern Africa, but
we believe more archaeological research is necessary to fully parse through the range
of possible interpretations regarding how the newcomers interacted with the preexisting populations at Bambandyanalo, and how these changes relate to the site’s
abandonment and the occupation of Mapungubwe. We posit that the ethnic mosaic in
Africa and elsewhere has often been characterized by both cooperation and conflict.
Cooperation refers to the means through which people of different ethnic affiliations
inhabiting different but complementary ecological zones amicably find a way of
sharing resources. In eastern Africa the institutionalization of blood brotherhood and
sisterhood enabled those who would otherwise have been foes to freely trade and
Afr Archaeol Rev
trespass into each other’s territorial domains; for example, some groups gave passage
to elephant hunters in exchange for sharing the ivory. Conflict refers to competition
over resources, which would have included good pasture and arable land, and coveted
trade items in high demand regionally and extra-regionally (Herlehy 1984; Kusimba
and Kusimba 2005; Stahl 2005). It is thus quite plausible that demographic stress
carried with it undercurrents of ethnic and territorial tensions. Whether these were
sporadic outbreaks of violence or full-scale warfare needs to be revisited. For
example, could the layer of black ash at Mapungubwe Hill be indicative of conflict
especially when we consider that the cultural pattern sealed below the layer harkens
back to the pre-Iron Age culture at Bambandyanalo, whereas patterns above the
layer belong to the ironworking culture of the Bantu?
In sum, despite its size and the diversity and intensity of its material culture,
Bambandyanalo was not yet a state by the time of its abandonment. The settlement
had become a central location for an aggregating population, the signaled initial
manifestation of urbanism. However, the archaeological record strongly indicates
that Bambandyanalo lacked a high level of political centralization and control over
various aspects of cultural life. Huffman (2005, 2007) defines it as a level 3 town,
synonymous with regional chiefdoms. Diversity in craft specialization and mortuary
patterns point to social inequalities, ethnic, gender, and class differentiation, which
are all typical in complex societies. Southern Zambezia was on its way toward
statehood.
Emerging Statehood: Mapungubwe
The hill known as Mapungubwe (‘the hill of the Jackals’) stands out conspicuously
amongst those that surround it, both because of its precipitous cliffs, over 61 m high
in places, and because of the wide valley around it (Fagan 1964:338). Located just
over a kilometer east of Bambandyanalo Hill, Mapungubwe is a flat-topped, steep
sided hill measuring 350 m long and 80–100 m across. On the southern side of
Mapungubwe is a well-defined terrace of occupation debris, occupied at the same
time as the hill. The hill’s top is flat and contains remains of elite residences (Pikirayi
2001:115). The town appears to have developed beneath from the southwest in the
middle of the eleventh century and extended towards the hill top reaching the
summit in the early twelfth century (Hall 1990:77; Pikirayi 2001:115). Mapungubwe, like many contemporary settlements in the region, was spatially organized to
reflect status differences. Elite residences were located on the hilltop, and the steepsided nature of the hill and its limited access to the summit afforded ruling elites
with much needed security (Pikirayi 2008, personal communication). Ordinary
citizens homes were often built a respectable distance in the valleys. Both residential
types are characterized by what archaeologists refer to as central cattle pattern or
Zimbabwe Pattern (Hall 1990:82; Huffman 2005, 2007; Manyanga 2006:140). Elite
residences were constructed of dry stone architecture and have come to symbolize a
departure from more heterarchical forms of social organization to hierarchical ones
in southern Africa, which Huffman (1986c, 2007) defines as the Zimbabwe Pattern.
Residential structures on the hilltop were clearly constructed for elites. For
example, at Mapungubwe and its major satellite towns, livestock and the majority of
Afr Archaeol Rev
the human population lived beneath the hill, the top of which was reserved for only a
small part of the community (Hall 1990:82; Huffman 1996, 2005, 2007).
Archaeological finds recovered at the elite homes include highly polished pottery
in diverse forms and styles, gold beads and wire bangles, iron and copper objects,
and trade beads (Pikirayi 2001:115). Elite burials at the summit of the hill were
richly adorned (Hall 1990:77). The hilltop was the most desirable piece of real
estate. Old and decrepit homes would be leveled and new ones built instead of
relocating. In contrast, commoner residences were characterized by mud-on-wooden
frame homes that were often surrounded by a wooden and sometimes stone wall.
Homes were often located adjacent to cattle corals to the point where it is
archaeologically difficult to tease out the precise locations of homes and cattle
corals. Archaeological finds at these locations are dominated by utilitarian pottery
that is qualitatively inferior to that recovered at hilltop residences (Manyanga 2006).
The political economy of Mapungubwe was based on agropastoralism and longdistance trade. Regional trade involving the eastern Kalahari sandveld of Botswana,
south-western Zimbabwe, northern South Africa and southern Mozambique
flourished between Mapungubwe and surrounding towns. The circulations of
agricultural produce, metallurgical objects (copper and iron), and cattle formed the
backbone of regional trade (Manyanga 2006:140). The recovery of huge collections
of trade items including ceramics and glass beads at many sites in the ShashiLimpopo Valley and beyond points to trade with the coast (Mitchell 2005; Pikirayi
1993; Pwiti 2005; Sinclair 1987). Recently, archaeologists have emphasized trade
with the coast, sometimes minimizing the role regional trade had on the region’s
economy. We believe that coastal trade, while important, merely interlocked on an
economy that was already highly developed and quite complexly organized (Pikirayi
2001:116; Pwiti 2005; Wood 2000).
The quality of finished bone, ivory, pottery, and iron tools leaves little doubt that
this was the work of well-financed and highly skilled craftspeople. Investment in
time and training points to the possible presence of full-time time specialists engaged
in craft activities, and further suggests the robustness of the regional economy and
the ability and willingness to invest in quality products by elites. Hall (1990:80) is
convinced that Mapungubwe had gained state status. Clear distinctions between
rulers and commoners were being made in all areas of daily life, consumption, and
mortuary practices. For instance, the recovery of numerous items of gold in three
burials betrays the high status of those individuals (Mitchell 2002:303). The gold
rhinoceros, bowl, and scepter recovered from a male burial at Mapungubwe Hill
suggests that gold had become a symbol of royalty and its use and circulation
restricted to the ruling class elite. This evidence stands in stark contrast with an elite
child’s burial from Bambandyanalo that was accompanied by several hundred glass
beads, and seven large turquoise examples of probable Chinese origin (Mitchell
2002:303). As beads became more common in the region, elites shifted to a rarer
prestige item—gold—to maintain social distance from commoners.
At the zenith in the late eleventh century, Mapungubwe had attained the status of
state capital. Its leaders boasted control of a vast region held together by a complex
network of economic and social interaction and a tributary system (Pikirayi
2001:116). The town served as the primary manufacturing center for craft items
including iron and copper objects, copper wire, carved ivory, and elite pottery. At the
Afr Archaeol Rev
same time, the town monopolized the distribution of exotica from the coast including
glass beads (Davison 1972; Popelka et al. 2005). The presence of external trade
across the Indian Ocean also indicates the incorporation of southern Africa into the
global interaction sphere that enabled local elites to further consolidate their position
as they extended their reach to frontier regions through alliance building and
coercion or threat of coercion (Kusimba 2007).
The rapid rise of Mapungubwe to prominence is in part closely allied with the
developments at Bambandyanalo. However, in the second half of the twelfth century,
cracks began to develop in a once well-articulated hierarchy. The period between
1100 and 1220 AD was characterized by civil unrest,2 decline in quality of life,
demographic decline, and severe reduction in size of farmland. The majority of the
town’s residents moved to other regions, creating a smaller albeit impoverished
community. The hill was reoccupied after 1250 AD, when some of the stone walls
were erected as well. The reasons for Mapungubwe’s decline and abandonment have
not yet been satisfactorily explained. Mitchell (2002:302) believes that Mapungubwe’s decline and eventual abandonment was due to a complex series of social
and political events. Droughts are common in the area and it is plausible that shifts in
environmental conditions destabilized the status quo and weakened elite control over
social, ritual, economic and political power. Leaders who failed to fulfill their
citizens’ needs often lost legitimacy and found themselves isolated. In most
instances in Africa and elsewhere, people elected to vote with their feet. This was
very likely the chief reason for the rapid depopulation of Mapungubwe. By 1300
AD, Mapungubwe was no longer inhabited and the political power based shifted
north on the Zimbabwe Plateau (Pikirayi 2001:116). Apparently, this collapse had
nothing to do with climate change as was once believed (see for example Tyson and
Lindesay 1992; Tyson et al. 2000, 2002) but may have been due to political factors
(Smith et al. 2007).
In sum, despite its short-lived history, Mapungubwe functioned as a city,
satisfactorily fitting Trigger’s (2003:92) criteria of a state: a politically organized
society that is regarded by those who live in it as sovereign or politically
independent and has leaders who control its social, political, legal, economic, and
cultural activities. Mapungubwe was a socially stratified society. It had an urban core
and significant wealth in the hands of a few of its citizens and a large commoner
population. The town’s elite had developed networks of alliance and economic ties
to regions beyond, including the Swahili coast 640 km away. Without doubt,
Mapungubwe was the first state to emerge in the region.
Great Zimbabwe: A Territorial State?
Following the decline of Mapungubwe in the late thirteenth century, political
centralization shifted more than 300 km northward to south-central Zimbabwe
centered on Great Zimbabwe (Caton-Thompson 1970; Garlake 1983; Huffman
1996; Manyanga 2006; Ndoro 2001; Pikirayi 2001:123). Major centers controlled by
2
The evidence for this is inferred from several widespread burning episodes at the site before and after
1250 AD (Mitchell 2002: 302).
Afr Archaeol Rev
a powerful elite arose at Chivowa, Gumanye, and Great Zimbabwe Hill (Pikirayi
2001:123; Sinclair 1987). Usually known as Zimbabwe Tradition sites, after the
type-site at Great Zimbabwe Hill, they were generally located on fortified flat
hilltops in close proximity to permanent springs. As the leadership consolidated and
accumulated more wealth, they fortified these settlements creating hill-top palaces
that afforded leaders commanding views of the surrounding landscape (Manyanga
2006:82) while simultaneously camouflaging them from those below. In contrast to
K2/Bambandyanalo and Mapungubwe sites which were located on terraces, the
Zimbabwe settlements shifted to more elevated hilltops with fortifications and water
(Manyanga 2006:82). The decline of Mapungubwe partially caused by drought
related stress, severe water shortage, famine, pestilence, and warfare ushered in the
rise of Great Zimbabwe as the most powerful urban complex. In time, Great
Zimbabwe was to become the center of a powerful hegemony of allied lineages and
by far the largest precolonial state in southern Zambezia, and the most commanding
presence in southern Africa and, in modern times, the pride of Africa (Pikirayi
2001:124).
The city of Great Zimbabwe covered 720 ha and was one of the largest in subSaharan Africa (Sinclair et al. 1993). A perimeter wall enclosed two main
complexes, the Hill Complex and the Great Enclosure. Outside these elite residences
lived up to 20,000 people, doubtless commoners or vassals of the elites. Significant
evidence of ritual is found in elite residences of the Hill Complex, including six
soapstone birds distinctive to this site. Spatial segregation, stone architecture, and the
prominence of the Zimbabwe Hill were all meant to establish and separate elite
spaces and elite decision-making. The extent of the Great Zimbabwe state has been
estimated at 50,000 km2, including much of the Save-Runde catchment in southcentral Zimbabwe, which was a major conduit to the coastal trading ports such as
Sofala. This hinterland is dotted with stone ruins, home to vassals that controlled
distinct territories and exchanged gold for cattle, cloth and beads with the kings of
Zimbabwe.
Archaeologists place the founding of Great Zimbabwe, by Leopard’s Kopje
farmers (the ancestors of modern Shona), to the tenth century AD (Beach 1980,
1998; Pikirayi 2001:124). Researchers believe the settlement was constructed over
several centuries starting from approximately 900 AD (Huffman 1996; Ndoro
2001:22). Elite monopolization of resource rich areas including well-watered and
fertile pasture land, coupled with investment in the crafts and control of local and
regional market systems, would have created opportunities for accumulating much
wealth. The accumulated wealth was in the form of cattle, food, women, and
children. Like Mapungubwe, earlier Great Zimbabwe elites financed the local craft
industries including iron and gold mining, smelting and smithing, elephant hunting
and ivory carving and stone cutting and carving. These crafts were labor intensive
and were carried out on a scale that required astute leadership with managerial skills
to compel farmers and herders to work on public works. By the 12th century, Great
Zimbabwe elites had extended their networks east and added the coast as a major
trading partner (Kusimba 2007; Pikirayi 2001:125; Pwiti 2005; Sinclair 1982;
Sinclair and Hakansson 2000). Wealth drawn from regional and coastal trade was
large enough to encourage elite investment in dry masonry stone architecture.
Transformations in building technology from mud and daub structures to dry stone
Afr Archaeol Rev
masonry residences surrounded by perimeter walls during the 13th century
institutionalized inequality, whose foundation had been laid several centuries
earlier beginning with Bambandyanalo. By 1270 an elaborate urban complex
covering many villages and smaller towns and centered on Great Zimbabwe had
emerged in southern Africa (Pikirayi 2001:125). This complex was engaged with its
hinterland that extended in all directions into modern day Botswana, South Africa,
Zambia, and Mozambique.
The city of Great Zimbabwe was a metropolis with many neighborhoods,
including elite residences, ritual centers, public ceremonial courts, public forums,
markets, as well as houses for commoners and artisans (Pikirayi 2001:129). The site
covered a broad area, housing a large population within a complex of massive
walled structures (Pikirayi 2001:129). The city can be divided into three main
architectural zones, these being the Hill Complex, the Great Enclosure and the
Valley ruins (Ndoro 2001:24). In between the Hill Complex, the Great Enclosure,
and the Valley was a large, open, and apparently unoccupied space (Pikirayi
2001:131). Perimeter walls were constructed to enclose these sites, with peripheral
settlements outside of the walls accommodating population growth and other urban
functions. “The analyses of spatial organization delineate a complex social
organization in which authority and power were delegated to an elite sector, that
demonstrated its status not only by being in the forefront of political and ceremonial
displays but through the medium of its residences” (Pikirayi 2001:134). Functional
as well as social divisions occurred in the main precincts of the urban center—the
Hill Complex, the Great Enclosure, the Valley complexes, and the peripheral
residences were delineated (Pikirayi 2001:134).
From an archaeological standpoint Great Zimbabwe carried the features of a state.
The society was clearly socially stratified, with considerable wealth and power
concentrated in the hands of an upper strata of community members. As noted by
Flannery (1998), material markers for states include both temples and royal residences.
Both are visible within the urban complex. While there is debate about exactly where
some of these structures were located within the urban complex, it is clear is that royalty
and spiritual mediums resided within enclosures built of stone, protected from public
view and access. Furthermore, it is clear that the city and its surrounding vicinity
featured a significant population. Estimated population sizes for the settlement and its
immediate area have ranged from a few thousand up to 30,000 (Hall 1990:116; Ndoro
2001:22). Pikirayi (2001:130) estimates a population of between 11,000 and 18,000,
with the majority living in houses outside of the stone enclosures.
According to Trigger (2003:99), urban centers often contained 20% to 80% or
more of the total population of a city-state. These urban centers were often
surrounded by a number of smaller and dispersed villages or hamlets. Several
hundred villages, for example, surrounded Benin. We can thus extrapolate that the
territory of the Great Zimbabwe polity was extensive, including lands between the
eastern Kalahari and the Indian Ocean, and the heartland was a high plateau between
the Zambezi and Limpopo rivers (Hall 1990:91). Settlement stratification is
exhibited in the area for the state of Zimbabwe, with a number of towns that had
stonewalls of distinctive design and patterns of decorations. “These constructions are
known as madzimbahwe (singular, dzimbahwe), the Shona term for the residence of
a chief” (Hall 1990:92; Hannan 1974).
Afr Archaeol Rev
More than 50 madzimbahwe are known, mostly on the edges of the Zimbabwe
plateau overlooking the lowlands of the Limpopo and Sabi rivers to the south and east
and the Zambezi valley to the north (Hall 1990:92). Hall (1990:92) maintains that
Great Zimbabwe may have operated as a capital for regional madzimbahwe
settlements throughout the area. The madzimbahwe were usually located on elevated
hilltops and often enclosed by walling connecting boulders and other natural features.
“The carefully built stone walls clearly served to set those who lived in the
dzimbahwe apart from the majority of the population and occasionally there are
other indications of high status” (Hall 1990:93). These indications include items of
copper, beads, and ivory. At Chumnungwa, a dzimbahwe located near the southern
edge of the Zimbabwe plateau, seven burials have been unearthed containing gold
grave goods (Garlake 1973). It is probable that the wood-and-plaster houses of the
ordinary populace once surrounded all the walled hills, and this interpretation has
been supported by archaeological work at several sites (Hall 1990:93). Much, and
perhaps the majority, of the population of early Zimbabwe lived away from the
madzimbahwe in small villages that fell within the political and economic domain of
the regional centers of the capital itself, and there may have been hundreds of
villages fitting this pattern (Hall 1990:93).
Given the archaeological record for the Zimbabwe state, it is apparent that the
state operated as a set of regional centers from which members of the nobility
signified their authority over the mass of the population by lavish public
architecture, symbols of status, and ideological control. Pikirayi (2008, personal
communication) maintains that Great Zimbabwe architecture played an important
ideological role throughout history, and has always been a potent symbol of wealth,
status, and power. According to Hall (1990:95), the basis and the object of the
political control was control over the economy—“the network of transactions that
linked peasant villages, madzimbahwe and the capital and, beyond this, the state
itself with the wider commercial world.” In such a system, the peasant villages were
the main source of surplus production and tribute. “Thus the internal economy of the
Zimbabwe state must have involved agropastoral production beyond the needs of the
ordinary village community, generating an economic surplus which formed the basis
of the transactions that constituted the political economy” (Hall 1990:96).
As stated earlier, Trigger (2003:92) writes that territorial states possess a ruler
who governed a larger region through a multileveled hierarchy of provincial and
local administrators in a corresponding hierarchy of administrative centers. We argue
that the material record suggests that at its height Great Zimbabwe may have
operated as more of a territorial state. Other researchers seem to concur. For instance,
Pikirayi (2006, personal communication) maintains that Great Zimbabwe cannot be
perceived simply as a city-state for the obvious reasons connected with settlement
patterns and hierarchies in south-central Zimbabwe between 1300 and 1450 AD.
“These hierarchies assert to the primacy or dominance of Great Zimbabwe over
other zimbabwe (royal) settlements in the region” (Pikirayi 2006, personal
communication). Furthermore, according to Ndoro (2001:22), Great Zimbabwe’s
power was based mainly on cattle husbandry, crop cultivation, and the domination of
trade routes between the gold fields on the Zimbabwe plateau and the Indian Ocean
in the east, with trade contacts between Zimbabwe and the Swahili coast having
been established well before 900 AD (Kusimba 1999).
Afr Archaeol Rev
Accordingly, given the large territorial spatial distribution of Zimbabwe Culture
sites and the control of trade routes to the coast, we would categorize Great
Zimbabwe as a territorial state. More regional archaeological surveys followed by
intensive and extensive horizontal and vertical excavations are necessary to
completely map site distribution, chronology, intrasite and intersite relationships
(Sinclair et al 1993). Material remains recovered from such a long-term study will
serve to place Great Zimbabwe’s real influence and authority over its surrounding
and adjacent hinterlands and move us beyond the current posturing that has been
part of the debate of Zimbabwe’s greatness (e.g., Garlake 1973, 1978; Huffman
1982, 1996; Ndoro 2001; Pikirayi 2001). Our favorite comparative example would
be the Inca territorial state in Andean South whose leaders resided at Cuzco and held
power and authority over a number of provincial administrative centers scattered
throughout the polity’s hinterland (Arkush and Stanish 2005; D’Altroy 2002;
Hemming 1970; Moseley 2001). It seems very likely as Manyanga (2006) points
out, that Great Zimbabwe operated in a similar capacity, albeit on a different scale
than its South American counterpart.
Discussion: Pathways to Complexity
In this article we have made a strong case in support of the hypothesis that the
inhabitants of the Shashi-Limpopo Basin had attained some level of socio-political
sophistication by the 11th century AD which had readied their societies to depart
from largely acephalous and egalitarian formations to heterarchical and hierarchical
societies (Huffman 2005, 2007; Manyanga 2006:142). The cultural sequences of
southern Zambezia, from Bambandyanalo to Mapungubwe and finally to Great
Zimbabwe, make it apparent that a combination of factors propelled the region on a
path to social complexity: sedentism, livestock production, agriculture, region and
interregional trade, and crafts specialization (Garlake 1982:13). Favorable climatic
conditions and increased interaction amongst communities pursuing different but
complementary subsistence strategies made possible the sharing and exchange of ideas
and systems of knowledge that once held sway within specific ethnic and subsistence
groups. Migration and settlement of pastoralists in what was previously a huntergatherer domain and the later settlement of agrarian communities, along with the
incorporation of knowledge from all the groups, created a vibrant community that would
elevate individuals and personalities from the groups to leadership positions.
Complex society archaeologists have emphasized the generation of surplus,
accumulation of wealth and investment in craft specialization as among the most
important indices for determining levels of complexity. Generation of surplus is
indicative of societal efficiency in subsistence production and higher levels of
investment. Food surplus means food security, which frees people to engage in other
forms of labor on a part-time and full-time basis, for example, elephant hunting, iron
and gold working, ivory and stone carving, masonry, basketry, pottery making, and
trade. Food security and increasing sedentism especially amongst previously forager
and pastoral communities would have naturally encouraged demographic changes.
Households that had more children were thus more likely to take advantage of their
number to gain access to more resources.
Afr Archaeol Rev
The archaeological record clearly shows that Bambandyanalo, Mapungubwe,
Great Zimbabwe and many other contemporaneous sites were located in resourcerich areas with good water as well as arable and grazing land. These settlements
offered more opportunities and attracted more people from the surrounding regions.
Settlement aggregation in these settlements required management of resources and
people, thus creating opportunities for investment in more highly specialized crafts,
local and interregional trade, accumulation of wealth, power, and status. What were
the specific variables that led to statehood, both for Mapungubwe and Great
Zimbabwe?
The rulers of Mapungubwe controlled substantial herds of livestock, which they
traded widely (Hall 1990:89). But would cattle alone lead to more centralization?
After all, anyone could own cattle and there is no way of telling whether ownership
of cattle was restricted to only the dominant households. Thus, Hall (1990:89)
believes that neither trade nor cattle wealth explanations are sufficient to account for
emergent complexity, since they stress the role of the forces of production without
taking into account the relations of production. Could the incipient leaders in the
early settlements in the Limpopo basin seek forms of wealth that were qualitatively
different from livestock, that provided them with a means of breaking out of the
chiefly cycle of fusion and fission? Mapungubwe leaders would have monopolized
trade in beads and cloth and converted the revenue into real wealth and power. The
other means through which real power could have been exercised was through
coercion or threat of military power to compel the smaller-scale communities to
enter into a tributary relationship with more powerful societies. Hall (1990:89)
hypothesizes that “the rulers of Mapungubwe and other states that were to follow in
southern Africa commanded military power, even though the evidence for this is, at
present, unknown in the archaeological record.” What resulted was a Mapungubwe
state that created a pattern in which cattle, military service and other forms of tribute
would have flowed inwards to the major centers of power, while beads, cloth and
other valued signifiers of high status would have moved outwards to regional centers
and to local chiefs who acknowledged the suzerainty of the Mapungubwe kings
(Hall 1990:90). Hence the Mapungubwe case demonstrates that the means to compel
and coerce a sizable population were important for rulers of an emerging state.
In the case of Great Zimbabwe, pioneer researchers emphasized the prominence
of long-distance trade with the Swahili coast as the primary impetus in Great
Zimbabwe’s transformation to the state (Garlake 1982:10; Huffman 1972, 1982,
1986a). Great Zimbabwe exported gold and imported a variety of items, including
glass beads, cloth, porcelain, stoneware, and earthenware (Pikirayi 2001:20).
Advocates of the long-distance trade model maintain that the managerial elite at
Great Zimbabwe monopolized trade with rural and frontier zones with Great
Zimbabwe playing a central role as collection, processing, and distribution center for
gold, ivory, copper, and iron to regional and coastal entrepots on the Mozambican
coast (Huffman 1972, 1996).
It is understandable that modern perspectives are critical of the view privileges
trade as the primary transformative factor. Pikirayi (2001:21) sees the emergence and
organization of agriculture, management of cattle, propagation of culture, and
control of trade as the key factors contributing to the rise of states in southern Africa.
He maintains that imports played a minimal role in the Shona economy since gold
Afr Archaeol Rev
mining was not a full-time specialty. He favors the internal dynamism argument,
proposing that long distance trade interlocked a regional economy that was already
thriving, thus giving rise to a hierarchical society. This view is echoed by Manyanga
(2006:114) who sees long-distance trade as a major builder of wealth and status
among Great Zimbabwe’s leadership. However, Manyanga holds that trade was not
the primary cause for the social cleavages. In his words: “it is unlikely that external
trade and the products thereof made a sudden appearance on local systems that were
in a state of fragmentation without any form of centralized organization” (Manyanga
2006:144). This new perspective is supported by archaeological evidence indicative
of elite control of strategic resources such as cattle, ivory, iron ore, copper, and gold.
Further evidence points to elite investment and monopolization of extractive crafts
technologies such as iron working and gold panning, mining and processing, and
possibly specialized elephant hunting (e.g., Denbow 1984; Miller 1996; Wilmsen
1989). The wealth generated from these ventures presumably facilitated trade with
the coast (Mitchell 2002:327, 328; Pikirayi 2001:35). Thus the emergence of
Mapungubwe and Great Zimbabwe as capitals for powerful polities in southern
Zambezia in the second millennium AD was a consequence of many variables.
The ability to control access to resources or trade routes is an important element
of emergent complexity and political centralization. Embedded in this proposition is
power of a physical and military nature, a power that can restrict access to wealthcreating resources and production. For instance, control over iron production was
quite important for rulers, and the physical force needed to monopolize access to
iron resources that were beyond the political control of the state capital was probably
a necessity. Archaeologists have hinted that coercion, of one kind or another,
certainly played a role in maintaining elite power in Southern Zambezia (Mitchell
2002:329).
Once asymmetries in power had been established, what strategies do leaders use
to stay in power and accumulate more wealth, status, and power? The pathways and
mechanisms through which power is centralized will differ from case to case. In this
case, elites initially accumulated wealth in cattle and later invested in gold and ivory
trade with the coast. To do so, they also invested in local and regional infrastructures
that made it viable for communities living and exploiting different but complementary resources to be willing to comply and be incorporated into the regional political
economy. Both peaceful and coercive means were used to extend elite power to the
frontier chiefdoms and minor states. The point about peaceful means has been
strongly made by Thomas Huffman (1972:365): “As the paramount chief’s wealth
increased, the population of the royal settlement would swell, partly because of the
prestige of living in the settlement and the chance that some of the wealth might find
its way through the normal redistributive channels.” Ironically while Huffman
acknowledges the need for the military and public works to control this increased
population, he neglects to discuss the possibility that involuntary and coercive means
of incorporating peripheral societies were options open to a determined and
increasingly powerful leadership. Ideology and coercion played just as important a
role as did agriculture, long-distance trade, livestock production, and metalworking.
We need to recognize the possibility of internal conflicts, coercion, and exploitation
between classes and social segments within emergent complex societies. For
instance, while some researchers see the use of elevated hilltops by elites for their
Afr Archaeol Rev
residences as a means for delineating social status, we believe this kind of settlement
pattern may also reflect concerns over security and threat (Kim and Kusimba,
forthcoming). Similarly, while many of the walls for Mapungubwe and Great
Zimbabwe may have served a social demarcation function, we believe that they may
also have been constructed for defensive functions as well. We are convinced that
sufficient evidence exists in the available archaeological evidence to address the
possibility of warfare and coercion (Kim and Kusimba, forthcoming).
Concluding remarks
In sum, the evidence discussed in this article points to the emergence of social
stratification and complexity during the Iron Age in southern Africa. Mapungubwe
and Great Zimbabwe exhibit clear signatures of social inequality symbolized in
monumentality. Mapungubwe was a nascent state on its way to statehood but this
trajectory was interrupted, and the settlement abandoned for reasons that archaeologists have not fully explained. In the case of Great Zimbabwe, it is evident that
the polity was a territorial state at its height of power. Great Zimbabwe exercised
political and economic influence beyond its urban core to the hinterland and frontier
societies. How extensive was Great Zimbabwe’s hegemony? What forms of power
did it cultivate, and through what means?
Archaeologists have emphasized peaceful and non-violent means as primary
factors contributing to the rise of social complexity in southern Africa. However, a
significant amount of physical force is necessary for any society to control resources,
enforce labor, exact taxation and tribute, enforce cleavages in social rank,
accumulate and hoard wealth, create specialized goods, and engage in regional and
international trade which required protection of traders and trade routes. Although
the use or threat to use military force to compel commoners to do one’s bidding is
sometimes necessary, leaders understand that ideological persuasion to complement
brute force is preferable. In our next paper we will explore the importance of warfare
and ideology in the emergence and maintenance of state like polities in southern
Africa.
Acknowledgements We are grateful to a number of colleagues who have shared their time and insights
with us in completing this paper, including Robert Carneiro, Lawrence Keeley, and Innocent Pikirayi.
References
Arkush, E., & Stanish, C. (2005). Interpreting conflict in the ancient Andes. Current Anthropology, 46(1),
3–28.
Beach, D. (1980). The Shona and Zimbabwe, 900–1850: an outline of shona history. Gweru: Mambo.
Beach, D. (1998). Cognitive archaeology and imaginary history at Great Zimbabwe. Current
Anthropology, 39(1), 47–72.
Bent, T. (1892). The ruined cities of Mashonaland. Bulawayo: Books of Rhodesia.
Bilman, B. (2002). Irrigation and the origins of the Southern Moche State on the north coast of peru. Latin
American Antiquity, 13(4), 371–400.
Blanton, R. (1998). Beyond centralization: steps toward a theory of egalitarian behavior in archaic states.
In G. Feinman, & J. Marcus (Eds.), Archaic States (pp. 135–172). Santa Fe: School of American
Research Press.
Afr Archaeol Rev
Blanton, R., et al. (1996). A dual-processual theory for the evolution of Mesoamerican civilization.
Current Anthropology, 37(10), 1–14.
Carneiro, R. (1970). A Theory of the Origin of the State. Science, 169, 733–738.
Carneiro, R. (1981). The chiefdom: precursor of the state. In G. Jones, & R. Kautz (Eds.), The Transition
to Statehood in the New World (pp. 37–79). New York: Cambridge University Press.
Carneiro, R. (1990). Chiefdom-level warfare as exemplified in Fiji and the Cauca Valley. In J. Haas (Ed.),
The Anthropology of War (pp. 190–211). New York: Cambridge University Press.
Caton-Thompson, G. (1970). The Zimbabwe culture: ruins and reactions. New York: Negro Universities
Press [reprint of 1931 edition].
D’Altroy, T. (2002). The Incas. Malden: Blackwell.
Davison, C. C. (1972). Glass beads in African archaeology: Results from neutron activation analyses.
Unpublished PhD Dissertation, University of California, Berkeley.
Denbow, J. (1984). Cows and kings: a spatial and economic analysis of a hierarchical Early Iron Age
settlement system in eastern Botswana. In M. Hall, et al. (Ed.), Frontiers: southern African
archaeology today (pp. 24–39). Oxford: British Archaeological Reports.
Earle, T. (1997). How Chiefs Come to Power. Stanford: Stanford University Press.
Ehrenreich, R. N., et al. (1995). Heterarchy and the analyses of complex societies. Archaeological Paper
Number 5. Arlington: American Anthropological Association.
Fagan, B. (1964). The Greefswald sequence: Bambandyanalo and Mapungubwe. The Journal of African
History, 5(3), 337–361.
Feinman, G. (2001). Mesoamerican political complexity: the corporate–network dimension. In J. Haas
(Ed.), From Leaders to Rulers (pp. 151–175). New York: Kluwer Academic/Plenum.
Flannery, K. (1998). The Ground Plans of Archaic States. In G. Feinman, & J. Marcus (Eds.), Archaic
States (pp. 15–57). Santa Fe: School of American Research Press.
Fried, M. (1967). The evolution of political society: an essay in political anthropology. New York:
Random House.
Galloway, A. (1959). Skeletal remains of Bambandyanalo. Johannesburg: University of Witwatersrand.
Garlake, P. (1973). Great Zimbabwe. London: Thames and Hudson.
Garlake, P. (1978). Pastoralism and Zimbabwe. The Journal of African History, 19(4), 479–493.
Garlake, P. (1982). Prehistory and Ideology in Zimbabwe. Africa: Journal of the International African
Institute, 52(3), 1–19 Past and Present in Zimbabwe (1982).
Garlake, P. (1983). Early Zimbabwe: From the Matopos to Inyanga. Harare: Mambo.
Haas, J. (2001). Warfare and the evolution of culture. In G. Feinman, & T. D. Price (Eds.), Archaeology at
the millennium: a sourcebook (pp. 329–350). New York: Kluwer/Plenum.
Hall, M. (1990). Farmers, kings and traders: the peoples of southern Africa, 200–1860. Chicago:
University of Chicago Press.
Hannan, J. (1974). Standard Shona Dictionary. Salisbury: Rhodesia Literature Bureau.
Hemming, J. (1970). The Conquest of the Incas. San Diego: Harcourt.
Herlehy, T. J. (1984). Ties that bind. International Journal of African Historical Studies, 17(2), 285–308.
Holl, A. F. C. (1985). Background to the Ghana Empire: archaeological investigations on the transition to
statehood in the Dhar Tichitt region (Mauritania). Journal of Anthropological Archaeology, 4, 73–115.
Holl, A. F. C. (1997). Western Africa: the prehistoric sequence. In J. O. Vogel (Ed.), Encyclopedia of
Precolonial Africa (pp. 305–312). Walnut Creek: AltaMira.
Holl, A. F. C. (2000). The Diwan Revisited: Literacy, State Formation, and the Rise of Kanuri Domination
(AD 1200–1600). London: Kegan Paul.
Holl, A. F. C. (2003). The land of Houlouf: Genesis of a Chadic chiefdom (1900BC–1800AD). Ann Arbor:
University of Michigan Museum of Anthropology.
Huffman, T. (1972). The rise and fall of Zimbabwe. The Journal of African History, 13(3), 353–366.
Huffman, T. (1982). Archaeology and ethnohistory of the African Iron Age. Annual Review of
Archaeology, 11, 133–150.
Huffman, T. (1986a). Archaeological evidence and conventional explanations of Southern Bantu
settlement. Africa: Journal of the International African Institute, 56(3), 280–298.
Huffman, T. (1986b). Cognitive studies of the iron age in southern Africa. World Archaeology, 18(1), 84–
95 Perspectives in World Archaeology (June 1986).
Huffman, T. (1986c). Iron Age settlement patterns and the origins of class distinction in Southern Africa.
Advances in World Archaeology, 5, 291–338.
Huffman, T. (1996). Snakes and crocodiles: power and symbolism in ancient Zimbabwe. Johannesburg:
Witwatersrand University Press.
Afr Archaeol Rev
Huffman, T. (2005). Mapungubwe: ancient African civilization on the Limpopo. Johannesburg: Wits
University Press.
Huffman, T. (2007). Handbook to the Iron Age: The archaeology of pre-colonial farming societies in
southern Africa. Scottsville: University of KwaZulu-Natal Press.
Johnson, A., & Earle, T. (2000). The evolution of human societies: from foraging group to agrarian state
(2nd ed.). Stanford: Stanford University Press.
Keeley, L. (1996). War Before Civilization. Oxford: Oxford University Press.
Kusimba, C. (1999). The rise and fall of Swahili states. Walnut Creek: AltaMira.
Kusimba, C. (2007). The collapse of coastal city-states of East Africa. In A. Ogundiran, & T. Falola
(Eds.), Archaeology of Atlantic Africa and the African Diaspora (pp. 160–184). Bloomington: Indiana
University Press.
Kusimba, C., & Oka, R. (2008). Trade and polity in East Africa: re-examining elite strategies for acquiring
power. In J. Rawley (Ed.), Africa in India, Indian in Africa. Bloomington: Indiana University Press.
Kusimba, C., & Kusimba, S. (2005). Mosaics and Interactions: East Africa 2000 BP to the Present. In A.
B. Stahl (Ed.), African Archaeology: a critical introduction (pp. 394–419). Oxford: Blackwell.
Mabogunje, A. (1962). Yoruba Cities. Ibadan: University of Ibadan Press.
Mann, M. (1986). The Sources of Social Power. Volume 1. A History of power from the Beginning to A. D.
1760. Cambridge: Cambridge University Press.
Manyanga, M. (2006). Resilient Landscapes: socio-environmental dynamics in the Shashi-Limpopo Basin,
southern Zimbabwe c. AD 800 to the present, Studies in Global Archaeology 11, Department of
Archaeology and Ancient History. Uppsala: Uppsala University.
Marcus, J. (1998). The Peaks and Valleys of Ancient States: An Extension of the Dynamic Model. In
G. Feinman, & J. Marcus (Eds.), Archaic States (pp. 59–94). Santa Fe: School of American Research
Press.
McIntosh, S. K. (1999). Pathways to Complexity: An African Perspective. In McIntosh, S. K. (ed.) Beyond
Chiefdoms: pathways to complexity in Africa pp. 1–30. Cambridge: Cambridge University Press.
Meyer, A. (1998). The archaeological sites of Greefswald. Pretoria: University of Pretoria Press.
Miller, D. (1996). The Tsodillo jewellery: Metalwork from Northern Botswana. Cape Town: University of
Cape Town Press.
Mitchell, P. (2002). The Archaeology of southern Africa. Cambridge: Cambridge University Press.
Mitchell, P. (2005). African Connections: Archaeological perspectives on Africa and the Wider World.
Walnut Creek: Altamira.
Morris, C. (1998). Inka Strategies of Incorporation and Governance. In G. Feinman, & J. Marcus (Eds.),
Archaic States (pp. 293–309). Santa Fe: School of American Research Press.
Moseley, M. (2001). The Incas and their Ancestors. New York: Thames and Hudson.
Ndoro, W. (2001). Your Monument our Shrine: The preservation of Great Zimbabwe. Uppsala: Uppsala
University Studies in African Archaeology 19, Department of Archaeology and Ancient History.
Pikirayi, I. (1993). The archaeological identity of the Mutapa state: Towards an historical archaeology of
northern Zimbabwe, Studies in African Archaeology 6, Uppsala, Societas Archaeologica Upsaliensis.
Pikirayi, I. (2001). The Zimbabwe Culture: Origins and Decline of Southern Zambezian States. Walnut
Creek: AltaMira.
Pikirayi, I. (2006a). The demise of Great Zimbabwe, AD 1420–1550: an environmental re-appraisal. In A.
Green, & R. Leech (Eds.), Cities in the World, 1500–2000. The Society for Post-Medieval
Archaeology Monograph 3 (pp. 31–47). Leeds: Maney.
Popelka, R. S., et al. (2005). Laser ablation ICP-MS of African glass trade beads. In R. J. Speakman,
& H. Neff (Eds.), Laser ablation ICP-MS in archaeological research (pp. 84–93). Albuquerque:
University of New Mexico Press.
Pwiti, G. (2005). Southern Africa and the East Africa Coast. In A. B. Stahl (Ed.), African Archaeology: A
Critical Introduction (pp. 378–391). Oxford: Blackwell.
Rightmire, G. P. (1970). Iron Age Skulls from Southern Africa Reassessed by Multiple Discriminant
Analysis. American Journal of Physical Anthropology, 33, 147–167.
Robb, J. (1999). Material Symbols in Prehistory. Carbondale: Southern Illinois University Press.
Robinson, K. (1966). The Iron Age in Kapula Vlei, near Msuma Dam, Wankie Gane Reserve, Rhodesia.
Arnoldia, 2, 1–7.
Service, E. (1975). Origins of the State and Civilization: The Process of Cultural Evolution. New York:
Norton.
Sinclair, P. (1982). Chibuene—an early trading site in Southern Mozambique. Festschrift for J. Kirkman.
Paideuma, 28, 149–164.
Afr Archaeol Rev
Sinclair, P. (1987). Space, Time and Social Formation: a territorial approach to the archaeology and
anthropology of Zimbabwe and Mozambique, c. 0–1700 AD. Uppsala: Societas Archaeologica
Upsaliensis.
Sinclair, P. J. J., et al. (1993). Urban Trajectories on the Zimbabwean Plateau. In T. Shaw, et al. (Ed.), The
Archaeology of Africa: food, metals and towns (pp. 705–731). London: Routledge.
Sinclair, P., & Hakansson, T. (2000). The Swahili city-state culture. In M. Hansen (Ed.), Comparative
study of thirty city-state cultures (pp. 461–482). Copenhagen: Reizels Forlog.
Smith, J., Lee-Thorp, J., & Hall, S. (2007). Climate change and agropastoralist settlement in the ShasheLimpopo River Basin, southern Africa. South African Archaeological Bulletin, 62(186), 115–125.
Soper, R. (2006). The Terrace Builders of Nyanga. Avondale: Weaver.
Spencer, C., & Redmond, E. (2004). Primary State Formation in Mesoamerica. Annual Review of
Anthropology, 33, 173–199.
Stahl, A. B. (2001). Making History in Banda. Anthropological Visions of Africa’s Past. Cambridge:
Cambridge University Press.
Stahl, A. B. (2005). African Archaeology: A Critical Introduction. Oxford: Blackwell.
Steyn, M. (1997). A reassessment of the human skeletons from K2 and Mapungubwe (South Africa).
South African Archaeological Bulletin, 51, 14–20.
Steyn, M., & Henneberg, M. (1995a). The health status of the people of the Iron Age sites at K2 and
Mapungubwe (South Africa). Revista di Antropologia, 73, 133–143.
Steyn, M., & Henneberg, M. (1995b). Pre-Columbian presence of treponemal disease: a possible case
from Iron Age southern Africa. Current Anthropology, 36(5), 869–873.
Steyn, M., & Henneberg, M. (1996). Skeletal growth of children from the Iron Age site at K2 (South
Africa). American Journal of Physical Anthropology, 100(3), 389–396.
Steyn, M., & Henneberg, M. (1997). Cranial growth in the prehistoric sample from K2 and Mapungubwe
(South Africa) is population specific. Homo, 48(1), 62–71.
Summers, R. (1967). Archaeological Distributions and a Tentative History of Tsetse Fly Infestation in
Rhodesia and Adjacent Territories. Arnoldia, 3, 1–18.
Trigger, B. (2003). Understanding Early Civilizations. Cambridge: Cambridge University Press.
Tyson, P. D., & Lindesay, J. A. (1992). The climate of the last 2000 years in southern Africa. The
Holocene, 2, 271–278.
Tyson, P. D., Karlen, W., Holmgren, K., & Heiss, G. (2000). The Little Ice Age and medieval warming in
South Africa. South African Journal of Science, 96, 121–126.
Tyson, P. D., Lee-Thorp, J., Holmgren, K., & Thackeray, J. F. (2002). Changing gradients of climate
change in southern Africa during the past millennium: implications for population movements.
Climate Change, 52, 29–135.
Usman, A. A. (2001). A View From the Periphery: Northern Yoruba Villages During the Old Oyo Empire,
Nigeria. Journal of Field Archaeology, 27, 43–61.
Van der Merwe, N. J. (1969). The Carbon-14 Dating of Iron. Chicago: The University of Chicago Press.
Walker, N. J. (1995). Late Pliestocene and Holocene Hunter-gatherers of he Matopos. (Studies in African
Archaeology 10). Uppsala: Societas Archaeologica Upsaliensis.
Webster, D. L. (1998). Warfare and Status Rivalry. In G. Feinman, & J. Marcus (Eds.), Archaic States
(pp. 311–351). Santa Fe: School of American Research Press.
Wilmsen, E. (1989). Land Filled with Flies: A Political Economy of the Kalahari. Chicago: University of
Chicago Press.
Wood, M. (2000). Making connections: relationships between international trade and glass beads from the
Shashe-Limpopo area. In M. Leslie, and T. Maggs (eds.) Africa Naissance: The Limpopo Valley
1000 years ago. The South African Archaeological Society Goodwin Series, 8, 78–90.
Yoffee, N. (2005). Myths of the Archaic State: Evolution of the Earliest Cities, States, and Civilizations.
Cambridge: Cambridge University Press.