Metals in Past Societies

Full text

SpringerBriefs in Archaeology
Contributions from Africa
Series Editor
Ann B. Stahl
Department of Anthropology
University of Victoria
Victoria, British Columbia
Canada
shadreck.chirikure@uct.ac.za

More information about this series at http://www.springer.com/series/13523
shadreck.chirikure@uct.ac.za

Shadreck Chirikure
Metals in Past Societies
A Global Perspective on Indigenous African
Metallurgy
1 3
shadreck.chirikure@uct.ac.za

ISSN 1861-6623 ISSN 2192-4910 (electronic)
SpringerBriefs in Archaeology
ISBN 978-3-319-11640-2 ISBN 978-3-319-11641-9 (eBook)
DOI 10.1007/978-3-319-11641-9
Library of Congress Control Number: 2014959261
Springer Cham Heidelberg New York Dordrecht London
© The Author 2015
This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part
of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations,
recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or
information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar
methodology now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication
does not imply, even in the absence of a specific statement, that such names are exempt from the relevant
protective laws and regulations and therefore free for general use.
The publisher, the authors and the editors are safe to assume that the advice and information in this book
are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the
editors give a warranty, express or implied, with respect to the material contained herein or for any errors
or omissions that may have been made.
Printed on acid-free paper
Springer is part of Springer Science+Business Media (www.springer.com)
Shadreck Chirikure
Department of Archaeology
University of Cape Town
Cape Town
South Africa
shadreck.chirikure@uct.ac.za

For our sons Tawana and Tadana
shadreck.chirikure@uct.ac.za

vii
Acknowledgements
When asked to be part of this important project, I was very surprised as much as I
was excited. Unreservedly, I owe a big debt of gratitude to Professor Ann Stahl, the
Series Editor, for the invitation and for very stimulating, inspirational, intelligent
and resolute guidance throughout the development of this work and the publication
process. Ndinotenda (thank you so much) Ann. Teresa Krause at Springer laboured
tirelessly together with Ann and the Series Advisory body to ensure that this much
needed series materialized.
This work benefited from a thorough review by two anonymous readers whose
incisive comments strengthened the initial ideas and resulted in the production of a
more effective and geographically balanced outcome. I would be remiss if I did not
acknowledge the immortal contribution of the late Professor P. J. Ucko who always
and in his own ways challenged me to articulate aspects of African archaeology to
a global audience.
It is almost 15 years since I first knocked on Professor Thilo Rehren’s door on
the third Floor of the Institute of Archaeology at UCL. That day marked my first in-
duction into the study of metals in archaeology. Since then, Thilo has been a consis-
tent, very patient and untiring mentor in all technical aspects of archaeometallurgy,
from the use of analytical equipment to interpreting the results. Even in the draft-
ing of this book, Thilo as always, had many words of wisdom to share amidst his
many other innumerable commitments. Andrew Reid made sure that the technical
details that I learnt from Thilo were placed in their social context. Professors David
Killick and Simon Hall took me under their wing when I had just graduated from
the Institute of Archaeology in 2005. Since then, David gave me countless tutorials
in determinative mineralogy and assisted me in assembling a reference collection of
mineralogy books and mineral specimens. Simon allowed me to participate in his
many fieldwork projects in and outside South Africa. David and Simon generously
shared their unpublished work which helped me to navigate profitable research ave-
nues. Dr. Duncan Miller, another longtime supporter, is always available to attend to
my frequent requests and also gives me unlimited access to his vast and staggering
experience. I therefore feel very lucky to have Simon Hall, David Killick, Duncan
Miller, Andrew Reid and Thilo Rehren as faithful mentors, colleagues, critics and
above all friends. To them I say, thank you immensely; without your generosity and
shadreck.chirikure@uct.ac.za

viii Acknowledgements
goodwill, I would be very far away from where I am today. I must also acknowledge
the many discussions that I had with Dr. Foreman Bandama and Ms. Abigail (Moff)
Moffet in the Archaeological Materials Laboratory at the University of Cape Town
which also gave this work its present character. Dr. Tim Maggs also contributed in
many ways and generously gave me access to his vast archive and experience which
made life a lot easier.
Along the way many colleagues supported me in various ways but I would like to
single out Judith Sealy, Gilbert Pwiti, Bill Dewey, Robert Heimann, Chap Kusimba,
David Reid, Mark Pollard, Munyaradzi Manyanga, Innocent Pikirayi, Jane Hubert,
Andrew Reid, Phil De Barros, Per Ditlef Fredriksen, Godfrey Mahachi, Darlington
Munyikwa, Thomas Thondhlana, K T Chipunza, Marcos Martinon-Torres, Dana
Drake Rosenstein, Tom Fenn, Webber Ndoro, Randi Haaland, Susan McIntosh,
Seke Katsamudanga, Joseph Chikumbirike, Mukundi Chifamba, Abigail Moffet,
Bertram Mapunda, Mzee Edwinus Lyaya, Jane Humpris, Paul Sinclair, Martin
Carver, Chris Scarre, Paul Lane, Rob Morrell, Pamela Eze-Uzomaka, Augustine
Holl, Akin Ige, Ian Freestone, Adria La Violette, Mark Horton and others too nu-
merous to mention. Although space precludes me from naming everybody by name,
I sincerely appreciate and value the support provided by colleagues and critics alike.
The National Research Foundation of South Africa funded many of my research
projects and provided grants which made the writing of part of this book possi-
ble. The NRF, through an Indigenous Knowledge Systems Grant, partly funded
my sabbatical at Harvard and as such, I am extremely grateful for their support.
The University of Cape Town Research Committee and the University of Cape
Town Research Office through the Africa Knowledge Project and the Programme
for Enhancement of Research Capacity has also been a consistent funder. I thank
Professor Danie Visser, the UCT Deputy Vice Chancellor for Research, Dr. Marilet
Sienaert the Executive Director for Research at UCT together with Professor Rob
Morrell, the coordinator of PERC. Special thanks also go to Professor Thandabantu
Nhlapo for resolute support and encouragement. Additional funding from the UCT
Faculty of Science Awards is acknowledged with sincere gratitude. I therefore thank
Professor Anton Le Roex, the Dean of the Faculty of Science for his goodwill. The
Institute for Archaeometallurgical Studies (IAMS), Institute of Archaeology, UCL
is an unwavering supporter. I therefore sincerely thank the trustees of IAMS. The
Wenner Gren Foundation played an important role in my early career development
and continues to do so in various direct and indirect ways. At the Wenner Gren
Foundation, I extend my appreciation to President Leslie Aiello, Judy Kreid, Mike
Muse and their colleagues for the sterling effort in promoting all types of anthropo-
logical research.
I also acknowledge the trustees of the Mandela-Harvard Fellowship for making
my stay at Harvard possible. Some of the draft chapters for this book were writ-
ten during my time as a Mandela-Mellon Fellow at Harvard University’s Hutchins
Centre for African and African American Studies in 2012. At the Hutchins Centre,
I would like to thank Professor Louis ‘Skip’ Gates, Krishna Lewis, Abi Wolf and
Donald Yacovone. Emmanuel Akeyampong, Jean Comaroff, Suzanne Blier, Rowan
Flad and Ben Lewis also shared with me their knowledge of various issues that
shadreck.chirikure@uct.ac.za

ixAcknowledgements
made my time at Harvard productive. While at Harvard, Charles van Onselen and
my fellow fellows Celia Cussen and David Bindman, were fantastic sparring part-
ners.
Dr. Pamela Smith kindly granted permission to publish the roped bronze object
from Igbo Ukwu from the late pioneer of African archaeology, Professor Thurstan
Shaw’s estate. Thank you immensely, Dr. Smith, for your support to Thurstan’s vi-
sion and interest in a fully-fledged African archaeology. The Barbier-Mueller Gold
of Africa Museum in Cape Town granted permission to publish outstanding images
of masterpieces in African gold working. Professor Phil de Barros another longtime
colleague generously availed photographs from West Africa. Dr. Jane Humpris also
granted me permission to publish her Meroe images while Dr. Tim Maggs provided
the image of the twin furnaces from kwaZulu-Natal. Dr. Foreman Bandama drew
all the maps and other line drawings (except where acknowledged) in the book at a
time when he had to juggle many other important things. Ndinotenda babamudiki—
thank for your support.
On so many occasions, the writing of this book took me away from my family
and friends. To them I say thank you immensely for your unwavering support. May
God bless all of you!
I would be remiss if I do not mention the immense contribution which my best
friend and companion Geraldine always makes in my life. To you my wife, thank
you so much for taking care of everything when I was busy with this project.
While these individuals and institutions assisted, errors and omissions that re-
main should solely be attributed to me.
shadreck.chirikure@uct.ac.za

xi
Contents
1 Metals and the Production and Reproduction
of Society ..................................................................................................... 1
Introduction ................................................................................................. 1
Data Sources ................................................................................................ 3
Towards an Integrated View of Global Metallurgy ..................................... 10
Organization of Work .................................................................................. 12
References .................................................................................................... 13
2 Origins and Development of Africa’s Preindustrial
Mining and Metallurgy.............................................................................. 17
Introduction ................................................................................................. 17
Origins of Metallurgy in Egypt and Adjacent Areas ................................... 19
Metal from Somewhere: On the Origins of Metallurgy in West,
Central and East Africa ............................................................................... 20
Why are Iron and Copper Earlier than Gold and Bronze
in Sub-Saharan Africa? Some Provocative Thoughts ................................. 28
Conclusion ................................................................................................... 30
References .................................................................................................... 31
3 Mother Earth Provides: Mining and Crossing the Boundary
Between Nature and Culture .................................................................... 35
Introduction ................................................................................................. 35
History of Mining: A Global Perspective .................................................... 38
The First Step: Ore Exploration and Prospecting in Precolonial Africa ..... 39
Methods of Crossing the Nature–Culture Boundary:
Mining of Ores in Preindustrial Africa ........................................................ 40
Surface Collection ....................................................................................... 40
Alluvial Mining ........................................................................................... 41
Open Mining ............................................................................................... 44
Underground Mining ................................................................................... 45
Preindustrial Hoisting and Beneficiation .................................................... 51
Mining Equipment and Other Paraphernalia ............................................... 53
shadreck.chirikure@uct.ac.za

xii Contents
The Anthropology of Mining: A Global Outlook ...................................... 53
Conclusion ................................................................................................. 56
References .................................................................................................. 57
4 Domesticating Nature .............................................................................. 61
Introduction: Transforming Ore into Metal ............................................... 61
Raw Materials ........................................................................................... 63
Brief Overviews: Metal Smelting in Preindustrial Africa—
Egypt, Nubia, North Africa and the Horn of Africa .................................. 67
West Africa ................................................................................................ 74
Central Africa ............................................................................................ 78
East Africa ................................................................................................. 82
Southern Africa ......................................................................................... 84
Anthropology of Smelting ......................................................................... 87
Conclusion ................................................................................................. 92
References .................................................................................................. 93
5 Socializing Metals ..................................................................................... 99
Introduction: Fabricating Metal into Cultural Products ............................ 99
Metal Fabrication in Egypt and Nubia ...................................................... 101
Forging, Smithing and Casting in Sub-Saharan Africa: West and
Central Africa ............................................................................................ 105
East and Southern Africa with Occasional References
to Sudan and Ethiopia ............................................................................... 111
Metals in Society: The Anthropology of Smithing and Metal Objects ..... 119
Conclusion ................................................................................................. 121
References .................................................................................................. 122
6 The Social Role of Metals ........................................................................ 125
Introduction ............................................................................................... 125
General Impact of Metals in Society ......................................................... 126
Metals, Sociopolitical Complexity and Urbanism .................................... 130
Metallurgy, Culture Contact (Interaction), Proto-Globalization
and Technology Transfer ........................................................................... 138
Conclusion ................................................................................................. 145
References .................................................................................................. 146
7 Bridging Conceptual Boundaries, A Global Perspective ...................... 151
Introduction ............................................................................................... 151
African Metallurgy and the Bridging of Conceptual Boundaries
Between Technology, Society and Culture ................................................ 153
Bridging Analytical Boundaries: From Sources of Ethnographies
to Domains of Integrated Studies .............................................................. 154
Local Responses to Technology Transfer and Knowledge Dispersal ....... 157
shadreck.chirikure@uct.ac.za

xiiiContents
African Metals: Land and Sea Links and Protoforms of Globalization .... 158
Changing Contexts of Knowledge Production and the Future of
African Preindustrial Metallurgy .............................................................. 159
Conclusion ................................................................................................. 161
References .................................................................................................. 162
Index ................................................................................................................ 165
shadreck.chirikure@uct.ac.za

xv
Shadreck Chirikure graduated with a Master of Arts in artefact studies degree
and a PhD in archaeology from the Institute of Archaeology, University College
London. Chirikure was born into one of the most senior houses of the Gutu-Rufura
people in rural Zimbabwe. During his childhood years, Chirikure’s grandmother
was a potter and a number of men forged scrap iron in his village. Therefore, nego-
tiating through rituals and taboos embedded in pottery making and other categories
of practice were part of Chirikure’s experience growing up. Unknown to him, these
would be part of his academic routine when he later became an archaeologist. Be-
cause of this village experience, Chirikure always attempts to gestate archaeologi-
cal reconstructions that are tempered with local realities where nothing was fixed in
space and time. His main research marries techniques from earth and engineering
sciences with those from more humanistic disciplines to study high temperature
technologies such as metallurgy and pottery making to enlighten their contribution
to societal development. Currently, Chirikure’s work on the metallurgy of the world
heritage sites of Great Zimbabwe and Khami in Southern Africa is throwing new
light on the contribution of metals to culture contact, interaction and social dif-
ferentiation. The work shows that metals, like cattle were a pivot on which society
achieved growth and renewal. Shadreck has published extensively on the subject
including a book, multiple journal articles, book chapters and contributions to pres-
tigious encyclopaedias. In the process, he won several national and international
awards for his contributions to African Iron Age research.
About the Author
shadreck.chirikure@uct.ac.za

xvii
List of Figures
Fig. 1.1 Links between metal production and use with various
dimensions of society ...................................................................... 2
Fig. 1.2 Selected well-known sites and metal working groups by
region. Sudan, Ethiopia and Eritrea are clustered under
North Africa because they share the same metallurgical
transition from copper through bronze to iron ................................ 4
Fig. 1.3 Photomicrograph of iron-smelting slag from Mapungubwe
Hill, Southern Africa showing a magnetite skin near top
right corner, egg-shaped wüstite, skeletal fayalite on a
glass matrix and some voids in dark black. Magnetite skins
form when slag from furnaces is exposed to cool air and is
indicative of smelting and in some cases slag tapping. ................... 8
Fig. 1.4 Integrated view of preindustrial metal production and use
outlining the inputs and outputs at various stages and the
accompanying anthropological factors ........................................... 11
Fig. 2.1 Map of Africa showing metalworking sites with some of the
most important highlighted by number ............................................. 18
Fig. 2.2 Ellingham diagram shows the ease with which metals and
sulphides can be reduced. The position of the line for a
given reaction on the Ellingham diagram shows the stability
of the oxide as a function of temperature. Reactions closer
to the top of the diagram are the most ‘noble’ metals (for
example, gold and platinum), and their oxides are unstable
and easily reduced. Moving towards the bottom of the
diagram, metals become progressively more reactive and
their oxides harder to reduce ............................................................. 24
Fig. 3.1 Location of mines worked in African antiquity. Note that
because of its abundance, iron was worked in more areas
than illustrated here ........................................................................... 36
Fig. 3.2 Akan gold miners diving into the Ankobra River to
extract diamond-rich sand which was panned on the river bank. ..... 42
shadreck.chirikure@uct.ac.za

xviii List of Figures
Fig. 3.3 Cross section of the Aboyne gold mine, Central Zimbabwe. ............ 47
Fig. 3.4 Umkondo copper mine in Zimbabwe where miners
strategically backfilled and opened up new shafts to create
pillars (unmined blocks) for structural stability. ............................... 48
Fig. 3.5 Wooden bucket found inside ancient gold mine in
Southwestern Zimbabwe and donated to Natural History
Museum, Zimbabwe. Exact provenance unknown. .......................... 51
Fig. 3.6 Porcupine quills used for storing gold. ............................................. 51
Fig. 3.7 Undated grinding stone with dolly holes used to crush
copper ores at Phalaborwa. ............................................................... 52
Fig. 4.1 Known metal smelting groups in Africa. Numbers indicate
key sites and landscapes .................................................................... 62
Fig. 4.2 Approximate distribution of bowl, shaft and natural draught
furnace types in Africa. .................................................................... 66
Fig. 4.3 Distribution of bellows across sub-Saharan Africa. .......................... 68
Fig. 4.4 Location of early Egyptian and Nubian smelting sites. .................... 69
Fig. 4.5 Location of sites with smelting evidence in North Africa,
Egypt, Nubia and Ethiopia ................................................................ 70
Fig. 4.6 An Old Kingdom pictorial showing six smelters blowing
into two crucibles (Mastaba of Mereruka, fifth Dynasty). ................ 71
Fig. 4.7 Bowl and shaft furnaces used in dynastic Egypt. .............................. 72
Fig. 4.8 Furnace F5 excavated by Shinnie (1985) at Meroe. ......................... 73
Fig. 4.9 Large iron slag mound at Meroe. ...................................................... 73
Fig. 4.10 Smelting sites in West Africa ............................................................ 74
Fig. 4.11 Second-millennium AD furnaces used in Bassar, Togo. ................... 75
Fig. 4.12 Second-millennium AD slag mounds at Bassar, Togo. ..................... 76
Fig. 4.13 Remnants of a circular shaft furnace from an iron smelting
furnace in Senegal, dated between the twelveth and
fourteenth centuries AD. The black material is slag that
solidified within the furnace. ............................................................ 77
Fig. 4.14 Location of Central African sites ...................................................... 79
Fig. 4.15 Cross section of Mafa down draught furnace. .................................. 80
Fig. 4.16 Location of East African sites and groups ........................................ 83
Fig. 4.17 Location of Southern African sites and groups ................................. 85
Fig. 4.18 Twin-bowl furnaces used in the Later Iron Age (c. AD 1700
onward) of KwaZulu-Natal, South Africa. ........................................ 86
Fig. 4.19 Anthropomorphic low shaft iron smelting furnace from
Nyanga, Eastern Zimbabwe. Note the molded breasts, navel
and waist belt for enhancing fertility. ................................................ 89
Fig. 4.20 Decorated furnaces, one depicting a woman giving birth.
Redrawn from furnaces on display at Natural History
Museum, Bulawayo .......................................................................... 90
Fig. 4.21 Anthropomorphic drum, granary and iron smelting furnace
from Bent (1896) ............................................................................... 91
shadreck.chirikure@uct.ac.za

xixList of Figures
Fig. 5.1 Metal working groups and important sites discussed in the text .... 100
Fig. 5.2 Dish bellows connected to a clay nozzle with a reed ...................... 103
Fig. 5.3 Location of West and Central sites and groups discussed in
the text ............................................................................................. 106
Fig. 5.4 Traditional iron forging workshop at Bitchabe, Togo, using
concertina bellows. .......................................................................... 107
Fig. 5.5 Leaded tin bronze Igbo Isiah roped vessel produced using
the lost wax method ........................................................................ 109
Fig. 5.6 Pectoral from the tumulus of Rao Senegal dating between
seventeenth and eighteenth centuries. ............................................. 110
Fig. 5.7 Photograph showing earliest gold earing from Jenne Jeno,
Inland Niger Delta, Mali. ................................................................ 110
Fig. 5.8 Akan crocodile and lizard sword ornaments ................................... 111
Fig. 5.9 Location of smithing sites in East and Southern Africa .................. 112
Fig. 5.10 Illustration of a Mambari smith by Holub (1881) showing
bellows leading to tuyeres and a very small forge. Note also
the tongs illustrated ......................................................................... 113
Fig. 5.11 Illustration of pot bellows used for smithing by Mambari
smiths (Holub 1881) ........................................................................ 113
Fig. 5.12 Goat skin bellows similar to the ones used by Njanja.
Approximate size 80 cm long. Natural History Museum,
Bulawayo, Zimbabwe ..................................................................... 114
Fig. 5.13 X-shaped copper ingots in use in much of Southern
Zambezia in the Iron Age, particularly after AD1000 .................... 115
Fig. 5.14 Musuku ingot housed at Iziko Museums, Cape Town .................... 115
Fig. 5.15 Half of a cross-shaped copper ingot mold (carved from
steatite?) used to produce ingots in 5.15 recovered from
Zimbabwe, on display at the Natural History Museum,
Bulawayo, Zimbabwe ..................................................................... 117
Fig. 5.16 Second-millennium AD soapstone ingot mold with a gold
pellet insert, Natural History Museum, Bulawayo .......................... 118
Fig. 5.17 Femur and bangles from a high-status individual excavated
from the dry-stone-walled Zimbabwe culture site of
Danangombe (AD1680–1850), Central Zimbabwe. The
excavators retrieved this bone with those high numbers of
copper or copper alloy bangles, which are quite numerous.
It is on display in the Zimbabwe Museum of Human Sciences. ..... 118
Fig. 6.1 Location of metal working groups and places
discussed in the text ........................................................................ 128
Fig. 6.2 Iron gong recovered from Great Zimbabwe. Natural History
Museum, Bulawayo ........................................................................ 129
Fig. 6.3 Gold leaf bowl from Mapungubwe, Southern Africa. It is
believed that the leaf was attached to a wooden core which
has since decayed (Miller 2001). .................................................... 129
shadreck.chirikure@uct.ac.za

xx List of Figures
Fig. 6.4 Mapungubwe golden rhino made of gold leaf (size, c.
10 cm). It symbolized the “majesty” of kingship. ........................... 129
Fig. 6.5 Location of prominent urban centers in Africa ............................... 132
Fig. 6.6 Indian Ocean cowrie and sea shells made their appearance
in Southern Africa from AD 700 onwards ...................................... 133
Fig. 6.7 Ingots from burials excavated from the Upemba depression
in the modern-day DRC. ................................................................ 135
Fig. 6.8 Iron projectiles used by the Buluba of the Katanga region
of the Democratic Republic of Congo. Natural History
Museum of Zimbabwe, Bulawayo .................................................. 137
Fig. 6.9 Diamond-shaped hoes produced by Phalaborwa smiths in
Northern South Africa between c. AD 1600 to 1900. The
hoes were used as currency ............................................................. 139
Fig. 6.10 Location of Rooiberg in relation to capitals .................................... 140
Fig. 6.11 Ming Dynasty porcelain from Zimbabwe housed at Iziko
Museum, Cape Town. ..................................................................... 142
Fig. 6.12 Regional and trans-continental connections between
Southern and Western Africa and the trans-Saharan and
Indian Ocean worlds ....................................................................... 143
shadreck.chirikure@uct.ac.za

xxi
List of Tables
Table 2.1 shows some of the earliest dates for the appearance of
metallurgy in Africa. Calibrated using OxCal version 4.2.3
Bronk-Ramsey (2013) and IntCal13 (Reimer et al. 2013) ............. 21
Table 3.1 Techniques and tools used in precolonial mining across
Africa. Adapted and modified after Hammel et al. (2002, p. 51) .. 54
Table 4.1 Chronology of ancient Egypt and Nubia ....................................... 69
shadreck.chirikure@uct.ac.za

1
Chapter 1
Metals and the Production and Reproduction
of Society
© The Author 2015
S. Chirikure, Metals in Past Societies, SpringerBriefs in Archaeology,
DOI 10.1007/978-3-319-11641-9_1
Introduction
As with archaeology, the development of archaeometallurgy—the study of pre-
industrial metal production, distribution and consumption—is richly varied, from
place to place and time to time (Childs and Killick 1993; Chirikure 2014; Craddock
1995; Herbert 1993; Killick and Fenn 2012; Kiriama 1987; Kusimba 1993; Morton
and Wingrove 1969; Okafor 1993; Rehren and Pernicka 2008; Rothenberg 1970;
Summers 1969). Traditionally, archaeometallurgy was preoccupied with exploring
the materiality of ore to metal and metal to artefact transformations from combined
technological, economic and environmental perspectives. It was less concerned
with the sociocultural dimensions of those materialities and processes (Hauptmann
2007; Joosten et al. 1998; Morton and Wingrove 1969; Rehren et al. 2007). How-
ever, it is self-evident that archaeometallurgy can no longer afford to ignore the an-
thropological dimensions which constituted a resilient component of preindustrial
metal production technological repertoires and styles (Childs 1991; Cline 1937;
Lechtman 1997; Rehren et al. 2007; Rickard 1939). Therefore, it is universally ac-
knowledged that metal production and use in the past was simultaneously techni-
cal and sociocultural, with the result that it produced and reproduced society. This
makes it highly problematic to study materials without considering the active role
of ‘materialities’ in social life (Childs and Herbert 2005; Childs and Killick 1993).
Modern archaeometallurgical work continually demonstrates that as an integral
component of human society and culture, the advent and subsequent entrenchment
of metallurgy enabled humanity to respond to various necessities and fuelled ap-
petites for luxuries (Fig. 1.1). In the process, the production and consumption of
metals gestated sociopolitical inequalities which strategically positioned individu-
als and collectives on different socioeconomic and political gradients (Chirikure
2007; Chirikure and Rehren 2006; de Barros 1988; Hauptmann 2007; Miller 2002;
Pollard and Bray 2007; Rothenberg 1999; Schmidt 1997; Srinivasan 1994). The
production and use of metals also connected localities, regions and continents
for millennia, initiating ‘proto-globalization’. For example, trade in African gold
played a pivotal role in the rise and fall of centres of power, trade and empires on
shadreck.chirikure@uct.ac.za

21 Metals and the Production and Reproduction of Society
the continent throughout ancient times (Klemm and Klemm 2012). By 2400BC,
Egyptians were trading gold from the Eastern Desert with Mesopotamia, generat-
ing immense wealth which was invested in monumental architecture and objects
of conspicuous consumption (Garrard 2011). Similarly, West and Southern African
gold financed developments in the Islamic world and adjacent territories from the
mid-first millennium AD onwards (Levtzion 1973; Phimister 1974). This initiated
cultural, technological and values exchange which, at different temporal scales, af-
fected directly and indirectly the societies participating in this trade. The mining,
production, consumption and distribution of different metals created new dynamics
of power and interaction, which in turn shaped the evolution of society at various
levels.
Due to a very high number of researchers per capita in some regions compared
to others, our knowledge of the quartet of mining, smelting, smithing and distribu-
tion is uneven (Chirikure 2014). Rather than being a constraint, this differential rate
of disciplinary emphasis and growth is a massive opportunity for reflection and
inspiration because of the varied research trajectories in different world areas. For
example, a brief navigation of the landscape of studies of preindustrial metallurgy
illustrates vividly that we know considerably more about the origins of metallurgy
in some parts of Eurasia than elsewhere in the world (Pringle 2009; Radivojević
et al. 2010; Roberts 2009; Zangato and Holl 2010). While our grasp of the be-
ginnings of metallurgy in Africa is mostly shrouded in mists of uncertainty, the
anthropology of metal production and consumption on the continent leads the way
(Childs and Herbert 2005; Schmidt 2009). Comparatively, the study of the technol-
ogy of metal production, distribution and consumption in Europe is reasonably well
Fig. 1.1 Links between metal production and use with various dimensions of society
shadreck.chirikure@uct.ac.za

3Data Sources
known, particularly because of interest shown among chemical, earth and engineer-
ing sciences (Gale and Stos-Gale 1982; Morton and Wingrove 1969; Pernicka et al.
1997; Pollard et al. 2011).
Despite this uneven emphasis, Childs and Herbert (2005) and Killick (2014a)
contend that studies of sub-Saharan Africa’s preindustrial metallurgy are well ahead
of those of other world areas including Egypt, Nubia, Ethiopia and North Africa.
Therefore, as varied as the contours of knowledge between the world’s geographi-
cal regions might be, Africa’s long but richly varied history of encounters with
both cultural and technological aspects of metallurgy will help global archaeol-
ogy to catch up in this respect. This provides the motivation for this work: with
well-established cadastral points, such as the stages in the production cycle and the
synergy between technological and anthropological factors, this volume aims to
articulate and contour salient features of preindustrial African metallurgy onto the
global map (Fig. 1.2).
Data Sources
Given Africa’s vast extent and the very long and remarkably variegated history of
metal production and use, no source can, on its own, provide a credible story of the
role of metals in past societies. However, a number of sources are available which,
when considered in combination, provide an intellectually nourishing picture of
Africa’s preindustrial metallurgy and its multiple lessons for global archaeology.
These sources range from archaeology and archaeometallurgy to art history, history
(including historical linguistics) and ethnography. Traditionally, specialists from
these individual disciplines work within closed boundaries, resulting in disciplinary
isolation and compartmentalized knowledge of the same phenomena. With a view
to developing synergy between various sources and disciplines, this volume draws
on diverse perspectives, as outlined below.
Historical Sources: Documentary and Oral History Written records are an
important archive of information on preindustrial African metallurgy. Their util-
ity and availability vary from time to time and area to area. For example, while
literacy has a 5000-year history in Egypt and a 4000 year one in Nubia (Edwards
2004), written records for regions such as Southern Africa only started in the last
1000 years (Summers 1969). In Egypt and Nubia, there exist some documents and
wall paintings that describe metalworking deep in antiquity (Davies 1943). Piet-
erse (1998) describes Egyptian wall paintings that depict black Africans involved
in copper working around 2500 BC. Metallurgical activities including the gold
trade between Egypt and Mesopotamia are also captured in the Bible (Emery 1963;
Scheel 1989). For other regions, written records appear sporadically from the
time of Islamic contact around 700 AD when Islamic scholars wrote about West
Africa (Levtzion and Hopkins 2000). For East Africa, there are occasional docu-
ments such as the Periplus of the Erythrean Sea (written in the first century AD)
shadreck.chirikure@uct.ac.za

4 1 Metals and the Production and Reproduction of Society
which discusses interaction between Greco-Romans and residents of towns such
as Rhapta located on the Indian Ocean coast (Horton and Middleton 2000). By the
early second millennium AD, Islamic chroniclers such as Ibn Battuta and Al Masudi
provided detailed descriptions of events and places in West and East Africa. From
the fifteenth century AD, written records appeared with increasing frequency as the
Portuguese sailed the Atlantic and Indian Ocean littoral and set up forts and trad-
ing stations (Beach 1980; Mudenge 1974). Because they settled in areas such as
Elmina in modern Ghana, and Sofala in Mozambique, the Portuguese left accounts
containing various types of information about trade, politics and other aspects of
Fig. 1.2 Selected well-known sites and metal working groups by region. Sudan, Ethiopia and
Eritrea are clustered under North Africa because they share the same metallurgical transition from
copper through bronze to iron
shadreck.chirikure@uct.ac.za

5Data Sources
life (DeCorse 1992), though these reports must be treated with caution because of
inherent biases on the part of observers (Beach 1980).
Throughout sub-Saharan Africa, documentary records became more abundant
during the last 300 years leading up to colonization (seventeenth, eighteenth and
nineteenth centuries AD), some explicitly commenting on metallurgy. Mungo Park,
David Livingstone and many other explorers and missionaries left reports which
now form a critical archive of data on the various stages in the metal production
cycle and their relations with other aspects of society (Cline 1937). The advent of
industrialization as a form of organizing production in Europe promoted an early
disappearance of preindustrial metal working processes there when compared to
places such as Africa and Asia (Craddock 1995). However, despite the richness
of the European historical record, which sometimes vividly captured metalwork-
ing events (see, e.g., Georgius Agricola’s De Re Metallica, Hoover and Hoover
1950), few studies have drawn insight from history, myths and legends to develop
a broad-based and informative understanding of Europe’s precapitalist metallurgy
(for exceptions, see Haaland 2004a; Martinon-Torres and Rehren 2009a). In India,
attempts are being made to engage with written sources, but not on a scale compa-
rable to Africa (see Tripathi 2013).
For the more recent periods in sub-Saharan Africa, oral history too is an impor-
tant source of information for the study of preindustrial African metallurgy, particu-
larly in areas with no written records. Oral history encompasses traditions, myths
and legends and is transmitted by word of mouth from one generation to the next
(Vansina 1985). In some areas of West Africa, griots or court historians passed on
traditions for a very long time such that testimonies may go back to the first millen-
nium AD (Levtzion 1973). Schmidt (1978) used oral historical accounts to develop
a long-term perspective on Buhaya iron smelting, extending from the recent past
to the early first millennium AD. In Southern Africa, historians such as Phimister
(1974) used oral historical data to explore gold production in second millennium
AD Southern Zambezia and its role in early state formation.
Like any other source, oral and documentary sources must be used with caution
for they are often affected by a number of weaknesses, some source specific but
others universally applicable. For example, oral traditions are easily forgotten, can
be chronologically telescoped and may be edited to suit prevailing contexts (Beach
1980; Vansina 1985). Written records are also affected by observer bias, e.g. most
Portuguese documents on the Shona of Northern Zimbabwe which describe silver
mines that never existed (Mudenge 1988). Nowadays, there is a strong feedback
problem whereby oral sources are affected by written history. With increasing lit-
eracy, African people read about their past, knowledge of which may filter into
memory. Effectively, oral sources became documentary ones, and what was written
became part of oral histories in a maze that is difficult to disentangle (Beach 1980).
Fortunately, a large number of traditions recorded in the twentieth century are still
available and can be used to cross check details. Whatever their limitations, his-
torical sources often provide important details about events and historical processes
of varying duration, something that archaeology struggles to address because time
aggregates promoted by techniques such as radiocarbon dating compress events of
shadreck.chirikure@uct.ac.za

61 Metals and the Production and Reproduction of Society
different duration (Chirikure et al. 2012). Thus, when oral and documentary sources
are combined with archaeology, essential complementary information is generated
(Schmidt 1978; Stahl 1994).
Ethnographies Preindustrial methods of metal production and use in some areas
of Africa ended only after industrialization introduced at conquest and in others
persisted into the 1980s and 1990s (David et al. 1989; Huysecom and Augustoni
1997; Van der Merwe and Avery 1987). Thus, ethnographies are a rich source of
technological and cultural details of metals and their role in society (David 2012).
For example, ethnographic recording of the production chain of iron among the
Dogon of Mali in West Africa has yielded a holistic understanding of technology in
its social context (Huysecom and Augustoni 1997). In West-Central Africa, David
et al. (1989) re-enacted Mafa iron smelting in the Mandara Mountains of Cam-
eroon. The Mafa smelted magnetite sand in down-draught furnaces to produce a
mix of cast iron and soft iron. Other ethnographic studies illuminated the symbolic
and cultural attributes of preindustrial metal production and use among the Shona
of Zimbabwe and Haya of Tanzania (Dewey 1991; Schmidt 1997). The value of
these ethnographic studies lies in the fact that unlike the fragmentary archaeologi-
cal record, they provide robust insight into technical and symbolic aspects of metal
production and use. As such, ethnographies make it possible to explore how metal
production was socially embedded. For example, through metal production, smelt-
ers reproduced and articulated ideas about fertility, witchcraft, power of ancestors,
significance of medicines and pollution which fundamentally were rooted in societ-
ies that produced the metal (see Bent 1896 and Chap. 4).
There is a misperception that ethnographies are only useful in studying the prein-
dustrial metallurgy of Africa. In India, significant amounts of ethnographic material
were recorded over the centuries, but the application of ethnography to animate
the archaeological record is still in its infancy (Keen 2013; Tripathi 2013). Similar
observations can be extended to Latin America where, before and immediately af-
ter Columbus, indigenous methods of working silver were still widely present and
documented (Cohen et al. 2010; Schultze et al. 2009). Indeed, ethnographies can
also be applied to understanding Europe’s preindustrial metallurgy. For example,
Hansen (1986) makes it explicit that magic, myths and legends fundamentally lay
at the heart of Medieval Europe’s social and technological processes. In fact, Insoll
(2008) has demonstrated that insight into ritual practice in Africa holds potential
within limitations to inform on ritual practice in the European Bronze Age. There-
fore, the rest of the world can learn from Africa, with its ethnography providing a
source of information for understanding the metallurgical record.
Despite being useful, an uncritical application of ethnography is dangerous given
the potential to extrapolate a presentist view onto the past. In continents such as
Africa, the ‘curse of the ethnography’ is often criticized for creating a picture of a
static continent devoid of any history and dynamism (Lane 2005). However, be-
cause some cultural principles are resilient, a careful application of ethnography
supported by good contextual, spatial and temporal control has demonstrated the
utility of combining ethnographies with archaeological data to distil more nuanced
shadreck.chirikure@uct.ac.za

7Data Sources
interpretations, particularly of processes that left little in terms of tangible remains
(David and Kramer 2001; Stahl 1994, 2001). Therefore, the value of comparative
insight drawn from historical and ethnographic sources, within well-defined con-
texts, can be a pivot on which researchers can base their comparison of different
metallurgies and their impact on society across space and time.
Archaeology In spite of dealing with a fragmentary record, and lacking the finer-
grained temporal resolution provided by ethnographies and histories, archaeology
is essential for understanding metal production in past societies. Archaeology pro-
vides a long-term perspective critical for understanding the development of metal
working through space and time. Furthermore, it provides context-specific exam-
ples of how various peoples and regions engaged with the world around them to
produce and use metals. For example, Pleiner (2000) has on the basis of archaeol-
ogy presented a diachronic picture of the evolution of iron production in Central
Europe during the Iron Age. Pleiner observed changes in furnace types from slag
pit furnaces characteristic of the Early Iron Age to slag taping furnaces typical of
the Late Iron Age. In West Africa, Okafor (1993) noted significant changes in fur-
nace operation between the Early and Late Iron Ages of Nsukka in Nigeria. While
Nsukka Late Iron Age (AD1000–1600) furnaces were slag tapping, their prede-
cessors in the Early Iron Age (600BC to AD1000) were non-slag tapping, thereby
providing a picture of technological change over time. So too have differentials in
the quantity of slag found at Early Iron Age sites when compared to those of the
Late Iron Age in Southern and West Africa proffered insights into scale of produc-
tion and evolution of specialization in the regions (Chirikure 2007; de Barros 2013).
Archaeology also generates samples that can be studied using archaeometallurgical
techniques and methods.
Archaeometallurgy Archaeometallurgy is an inter-disciplinary method for study-
ing remains from past metal production and use (Rehren and Pernicka 2008).
Because they are products of high-temperature processes, remains from preindus-
trial metalworking such as slags contain within their microstructures and composi-
tion partial histories of the processes which they have undergone (Bachmann 1882).
Archaeometallurgists often deploy techniques from chemical, earth and engineering
sciences to read important technical details such as the temperatures achieved in fur-
naces, the quality of ores and skills in manipulating redox conditions. Furthermore,
by adopting more anthropologically oriented approaches that consider metallurgy
to be simultaneously social and technical, it is possible to situate the ‘human factor’
in metalworking by looking at the decisions that smelters and smiths made for the
process to be a success (Rehren et al. 2007). Even more pertinent is the observation
that different stages in the chaîne opératoire of metalworking (Childs and Herbert
2005) left signatures on the landscape. When the material and cultural properties
of process remains are considered in relation to the site’s geography and surround-
ing landscape, a holistic view of technology in society is achieved. Finished metal
objects too are significant for they also can be interrogated to provide information
on provenance, trade and exchange as well as metal consumption (Fenn et al. 2009).
shadreck.chirikure@uct.ac.za

81 Metals and the Production and Reproduction of Society
A number of complementary methods have been developed to gain informa-
tion from archaeometallurgical remains, but archaeological fieldwork forms an im-
portant pedestal for all interpretations. As such, archaeometallurgy requires high
standards of data recovery in the field. Data from laboratory techniques only make
sense when interpreted in relation to the context of recovery and associated milieu.
According to Chirikure (2014), the significance of fieldwork in archaeometallurgy
cannot be overemphasized and is underlined by the fact that no matter how sophisti-
cated laboratory methods are, the information from such techniques is meaningless
if the methods of field recovery are poor. Where possible, it is crucial for archae-
ologists to have an expert archaeometallurgist in the field (Killick 2014b). Projects
such as the Mandara Archaeological Project that included archaeometalurgists in
their interdisciplinary teams (David et al. 1989) are an example to follow.
Generally, laboratory archaeometallurgy thrives on a combination of microscop-
ic and compositional techniques (Bachmann 1982). Optical microscopic methods
are based on the principle that when exposed to light under a plane-polarizing mi-
croscope (in reflected or transmitted plane-polarized modes), different phases in
a polished sample of a material reflect light differently (Deer et al. 1992; Phillips
and Griffen 2004;). For instance, the various phase characteristics of metallurgical
slags provide information on efficiency of reduction (Morton and Wingrove 1972),
furnace-operating temperatures (Chirikure and Rehren 2006) and the skills of the
extinct smelters in balancing ore-to-fuel ratios and in reducing more oxide to metal
(Rehren et al. 2007) (Fig. 1.3).
Fig. 1.3 Photomicrograph of iron-smelting slag from Mapungubwe Hill, Southern Africa showing
a magnetite skin near top right corner, egg-shaped wüstite, skeletal fayalite on a glass matrix and
some voids in dark black. Magnetite skins form when slag from furnaces is exposed to cool air and
is indicative of smelting and in some cases slag tapping. (Photo credit: Author)
shadreck.chirikure@uct.ac.za

9Data Sources
While useful, the limitations of light microscopes in relation to magnification
(maximum of 1000 X) and an inability to determine the chemical composition of
various phases can be transcended by recourse to analytical techniques such as
scanning electron microscopy with attached energy dispersive spectrometer (SEM-
EDS). SEM-EDS combines high image resolution techniques with analysis of the
characteristic of X-rays produced when the sample is bombarded with primary elec-
trons. This marriage produces a very powerful analytical tool suitable for analyz-
ing small regions of solid materials and detecting spatial variation in composition.
For example, crucibles can be characterized by looking at the composition of the
clay matrix, the crucible slag and metal droplets to identify the metals worked and
to understand the various contributions of various inputs to the metallurgical pro-
cess (Bayley and Rehren 2007). The refractoriness of the clay as determined by the
quantities of heat-resistant minerals such as kaolin informs on the skill of metal-
lurgists in selecting suitable raw materials (Martion-Torres et al. 2006). Although
possessing these distinct advantages, compositional data from the SEM are often
not of a resolution low enough to detect some essential minor and trace elements in
archaeometallurgical materials.
This limitation is addressed by resorting to techniques such as X-ray florescence
(XRF) analysis. XRF exploits the properties of X-rays to identify the major, mi-
nor and trace element composition of archaeometallurgical materials (Pollard et al.
2011, pp. 93–118). Such compositional data are important for reconstructing raw
material sources, and estimating furnace operating temperatures and the contribu-
tion of different variables to slag formation. Another useful technique is X-ray dif-
fraction (XRD) analysis. XRD uses X-rays of known wavelengths to determine the
lattice spacing in crystalline structures, and in the process, it indirectly identifies
chemical compounds (Pollard et al. 2011, pp. 93–118). Raman spectrometry too is
becoming an important technique in archaeometallurgy. It exploits the phenomenon
that as radiation passes through a transparent medium, a small proportion of the
incident beam is scattered in all directions (Muralha et al. 2011). The difference in
wave length between the incident and scattered radiation is a characteristic of the
material responsible for scattering (Pollard et al. 2011, pp. 83–85). Neutron activa-
tion analysis (NAA) is a powerful multi-element nuclear technique that allows the
determination of the concentration of a large number of elements in archaeometal-
lurgical remains. In essence, NAA involves converting some atoms of the elements
within a sample into artificial radioactive isotopes by irradiation with neutrons (Pol-
lard et al. 2011, pp. 195–208). The radioactive isotopes formed then decay to form
stable isotopes at a rate which depends on their half-life. Measurement of the decay
allows the identification of the nature and concentration of the original elements in
the sample. Solid samples can be analysed and whole samples can be irradiated. In-
ductively coupled plasma-mass spectrometry (ICP-MS) too is best suited for rapid-
trace element-level elemental analysis in archaeometallurgical materials.
All these geochemistry techniques have been applied with varying degrees of
success in reconstructing the raw materials used in ancient metal production across
the world. Matching the trace elements in the metal or alloy and ore is a common
tool of tracing the movement of metal and ores on the landscape through trade and
shadreck.chirikure@uct.ac.za

10 1 Metals and the Production and Reproduction of Society
exchange relationships (Chikwendu et al. 1989; Pernicka et al. 1997). However, all
analytical techniques do not have the same capabilities and detection limits, mak-
ing it standard practice to combine one or more techniques in a single study. For
example, Chirikure et al. (2010a) combined bulk chemical analyses using WDXRF
with trace element analysis by ICP-MS, individual phase characterization using
electron microprobe analysis (EMPA) and SEM to investigate the technology of tin
smelting utilized at Rooiberg in the Southern Waterberg of South Africa. In other
studies, Fenn et al. (2009) used lead isotope analyses in conjunction with composi-
tional techniques to demonstrate the existence of contact between West Africa and
Roman North Africa.
In summary, although different sources of information are available for study-
ing preindustrial metal production and its role in society, they must be combined
to create balanced picture of the past. As such, researchers must cross disciplinary
boundaries to exploit interpretive leads offered by other disciplines. Fortunately, for
studies of preindustrial metallurgy, archaeometallurgy is one of the most inter-dis-
ciplinary methods, combining data and techniques from history, archaeology, earth
and engineering sciences, anthropology and sociology (Killick and Fenn 2012;
Rehren et al. 2007). Such a broad-based view allows synergy to be developed be-
tween scientific techniques and archaeological or historical approaches to explore
the socially embedded nature of metals in society.
Towards an Integrated View of Global Metallurgy
Although anthropologists such as Mauss (1954) have long realized that technology
is socially mediated, it took a very long time for Western science to accept that it is
impossible and in fact unwise to separate technical and cultural aspects of any given
technology (Appadurai 1986). The tendency in Western science in the twentieth
century was to bifurcate technology into the scientific and magical aspects. Ironi-
cally, before the onset of Europe’s scientific revolution, which began after the Mid-
dle Ages, magic formed a key component of Western worldview (Hansen 1986).
Because of its association with symbolism, indigenous African metallurgy was of-
ten derided as magical and therefore unworthy of proper scientific study (Schmidt
1996). From the 1960s onwards, Africanists realized that the so-called magic was
just as important as the control of air pumped into the furnace or the quality of the
ore. The rituals and taboos were a technology of practice that enabled smelters to
take control of the process through learned behavior (Herbert 1993). Not surpris-
ingly, when studies of African metalworking flourished, a conscious attempt was
made to combine both the scientific and symbolic aspects of the technology (Her-
bert 1993; Killick 1990; Schmidt 1978; Van der Merwe and Avery 1987) (Fig. 1.4).
While these developments were taking place in Africa, the tendency in Anglo-
American technological studies was to isolate science from anthropology. It was
only in the early 1990s with the emergence of conceptual approaches such as chaîne
opératoire that it was realized that technology is made up of cultural and scientific
shadreck.chirikure@uct.ac.za

11Towards an Integrated View of Global Metallurgy
attributes (Lechtman 1997; Lemonnier 1993). The chaîne opératoire framework
focuses attention on the sequence of steps in artefact production and the embedded
cultural choices. As such, it takes technological studies out of a black box that up
to that point had diverted attention from the cultural dimensions of technological
production in Western technological studies (Dobres 2000). More recently concepts
such as ‘materiality’ have elaborated different elements of the chaîne opératoire ap-
proach (Jones 2004). As an analytical concept, materiality simultaneously considers
the technological and anthropological factors of artefacts and technologies that pro-
duced them and falls within the broader social construction of technology paradigm
(Killick 2004; Martinón-Torres and Rehren 2009b).
Another useful analytical concept is that of technological style or technology of
practice which is specific to groups of people across space and time (Stahl 2009).
Technological style and technology of practice situate the repertoire of metalwork-
ing within communities, making it possible to explore diachronic and synchronic
variation. All this can be enveloped within the broader social constructivism move-
ment in anthropology which seeks to consider both scientific and related cultural
factors (Killick 2004). While global archaeology has now realized that dissecting
technology into the scientific and magical is wholly unnecessary, studies of African
metallurgy have long realized this. Indigenous African metallurgy is imbued with
metaphors of birth and reproduction (Childs 1991; Ndoro 1991; Schmidt 2006);
it is also a technical process involving thermochemically gaining elemental metal
from ore (Childs and Herbert 2005). All these processes are rooted in society and
social structure and entail both human–human relations as well as human material
relations.
Fig. 1.4 Integrated view of preindustrial metal production and use outlining the inputs and outputs
at various stages and the accompanying anthropological factors
shadreck.chirikure@uct.ac.za

12 1 Metals and the Production and Reproduction of Society
This book continues with this proud Africanist tradition of considering the quar-
tet of mining, smelting, smithing and distribution simultaneously as a scientific and
cultural process, but pushes a bit further. A central strand that runs through the
chapters is the need to open conceptual boundaries and close gaps between the
technological and sociological dimensions of metallurgy (cf. Schmidt 2009; also
David 2012). Radiating from this centre is a strong focus on regional variation, the
social engineering role of metals in society and the general societal implications
emanating from the adoption of metallurgy. Examples from other areas of the world
will, where possible, be brought into the discussion to forge a rich comparative
perspective.
Organization of Work
This book considers metals and alloys worked in North African and sub-Saharan
antiquity and explores variability in their working and uses from area to area within
the wider archaeology of the continent. It seeks to communicate to both a global
and local audience the key attributes of African metallurgy, including technologi-
cal variation across space and time, methods of mining and extractive metallurgy
and the fabrication of metals within and between Africa and the outside world. Al-
thouigh metallurgy was introduced at different points in time across the continent,
the metals and alloys shaped in a significant and permanent way the development
and organisation of metalworking societies and their destinies.
Chapter 2 discusses the origins and development of metal production and use in
various parts of the African continent. Chapter 3 focuses on the anthropology and
technology of mining ores worked in precolonial Africa. It shows that the tech-
niques of extracting ore from the ground were a cultural venture, mediated by an-
cestors who made it possible to cross the boundary into the earth, to haul out its
riches. Chapter 4 deals with the process of transforming through heat ore into metal
during reduction, combining scientific with ethnographic observations. Smelting
was socioculturally embedded and was an important transformative process that
turned an object from nature (ore) into a cultural product (metal). The technology
used across the continent varied according to local specifics such as geology, scale
of production and many others. Chapter 5 discusses the fabrication of metals into
objects and how metals acquired social and cultural roles. Chapter 6 discusses the
broader issues associated with metals in society, including how they were distrib-
uted and their consequences for producers, consumers and society at large. It shows
that overland links developed between communities on the continent, while sea
links enchained communities to Indian Ocean and later Atlantic trading systems.
This promoted urbanization at places such as the Swahili city states of East Africa,
Great Zimbabwe in Southern Africa and, among others, ancient Ghana in West Afri-
ca. Chapter 7 emphasizes lessons for global anthropology that emanate from Africa.
shadreck.chirikure@uct.ac.za

13References
References
Appadurai, A. (Ed). (1986). The social life of things: Commodities in cultural perspective. Cam-
bridge: Cambridge University Press.
Bachmann, H. (1982). The Identification of slags from archaeological sites. Occasional Publica-
tion no 6. London: Institute of Archaeology.
Bayley, J., & Rehren, T. (2007). Towards a functional and typological classification of crucibles. In
S. La Niece, D. Hook, & P. Craddock (Eds.), Metals and mines: Studies in archaeometallurgy
(pp. 46–55). London: Archetype.
Bent, J. T. (1896). The ruined cities of Mashonaland: Being a record of excavation and exploration
in 1891. London: Longmans, Green, and Co.
Beach, D. N. (1980). The Shona and Zimbabwe, 900–1850: an outline of Shona history. London:
Heinemann.
Chikwendu, V. E., Craddock, P. T., Farquhar, R. M., Shaw, T., & Umeji, A. C. (1989). Nigerian
sources of copper, lead and tin for the Igbo-Ukwu bronzes. Archaeometry, 31(1), 27–36.
Childs, S. T. (1991). Style, technology and iron smelting furnaces in Bantu-speaking Africa. Jour-
nal of Anthropological Archaeology, 10(4), 332–359.
Childs, S. T., & Herbert, E. W. (2005). Metallurgy and its consequences. In A. B. Stahl (Ed.), Afri-
can archaeology: Acritical introduction (pp. 276–300). Oxford: Blackwell.
Childs, S. T., & Killick, D. (1993). Indigenous African metallurgy: Nature and culture. Annual
Review of Anthropology, 22, 317–337.
Chirikure, S. (2007). Metals in society: Iron production and its position in Iron Age communities
of Southern Africa. Journal of Social Archaeology, 7(1), 72–100.
Chirikure S. (2014). Geochemistry of ancient metallurgy: Examples from Africa and elsewhere.
In H. D. Holland & K. K. Turekian (Eds.), Treatise on Geochemistry, Second Edition, vol. 14
(pp. 169–189). Oxford: Elsevier.
Chirikure, S., & Rehren, T. (2006). Iron smelting in precolonial Zimbabwe: Evidence for dia-
chronic change from Swart Village and Baranda, Northern Zimbabwe. Journal of African Ar-
chaeology, 4, 37–54.
Chirikure, S., Heimann, R. B., & Killick, D. (2010). The technology of tin smelting in the Rooi-
berg Valley, Limpopo Province, South Africa, ca. 1650–1850 CE. Journal of Archaeological
Science, 37(7), 1656–1669.
Chirikure, S., Manyanga, M., & Pollard, A. M. (2012). When science alone is not enough: Radio-
carbon timescales, history, ethnography and elite settlements in Southern Africa. Journal of
Social Archaeology, 12(3), 356–379.
Cline, W. B. (1937). Mining and metallurgy in Negro Africa (No. 5). Menasha: George Banta
Publishing Company.
Cohen, C., van Buren, M., Mills, B., & Rehren, T. (2010). Current silver smelting in the Bolivian
Andes: A review of the technology employed. Historical Metallurgy, 44, 153–162.
Craddock, P. T. (1995). Early metal mining and production. Washington, DC: Smithsonian Uni-
versity Press.
David, N. (2012). Metals in Mandara mountains society and culture. Trenton NJ: Africa World
Press.
David, N., & Kramer, C. (2001). Ethnoarchaeology in action. Cambridge: Cambridge University
Press.
David, N., Heimann, R., Killick, D., & Wayman, M. (1989). Between bloomery and blast furnace:
Mafa iron-smelting technology in North Cameroon. African Archaeological Review, 7(1),
183–208.
de Barros, P. (1988). Societal repercussions of the rise of large-scale traditional iron production: A
West African example. African Archaeological Review, 6(1), 91–113.
de Barros, P. (2013). A comparison of early and later Iron Age societies in the Bassar region of
Togo. In J. Humpris & Th. Rehren (Eds.), The World of Iron (pp. 10–21). London: Archetype.
de Garis Davies, N. (1943). The Tomb of Rekh-mi-Reâ at Thebes (Vol. 2). New York: Arno Press.
shadreck.chirikure@uct.ac.za

14 1 Metals and the Production and Reproduction of Society
DeCorse, C. R. (1992). Culture contact, continuity, and change on the Gold Coast, AD 1400–1900.
African Archaeological Review, 10(1), 163–196.
Deer, W. A., Howie, R. A., & Zussman, J. (1992). Rock-forming minerals: Vol 2 chain Silicates.
Longman.
Dewey, W. J. (1991). Weapons for the ancestors. Des Moines, University of Iowa, Department of
Art History.
Dobres, M. A. (2000). Technology and social agency: Outlining a practice framework for archae-
ology. Oxford: Blackwell.
Edwards, D. N. (2004). The Nubian past: An archaeology of the Sudan. London: Routledge.
Emery, W. B. (1963). Egypt Exploration Society preliminary report on the excavations at Buhen,
1962. Kush, 11, 116–120.
Fenn, T. R., Killick, D. J., Chesley, J., Magnavita, S., & Ruiz, J. (2009). Contacts between West Af-
rica and Roman North Africa: Archaeometallurgical results from Kissi, Northeastern Burkina
Faso. In S. Magnavita, L. Koté, P. Breunig, & O. A. Idé (Eds.), Crossroads/Carrefour Sahel:
Cultural and technological developments in first millennium BC/AD West Africa (pp. 119–
146). Frankfurt am Main: Africa Magna Verlag. (Journal of African Archaeology Monograph
Series 2).
Gale, N. H., & Stos-Gale, Z. A. (1982). Bronze Age copper sources in the Mediterranean: A new
approach. Science, 216(4541), 11–19.
Garrard, T. F. (2011). (1989). African gold: Jewellery and ornaments from Ghana, Côte d’’Ivoire,
Mali and Senegal in the collection of the Barbier-Mueller Museum. Munich: Prestel.
Haaland, R. (2004a). Technology, transformation and symbolism: Ethnographic perspectives on
European iron working. Norwegian Archaeological Review, 37(1), 1–19.
Hansen, B. (1986). The complementarity of science and magic before the scientific revolution.
American Scientist, 74(2), 128–136.
Hauptmann, A. (2007). The archaeometallurgy ofcCopper: Evidence from Faynan, Jordan. New
York: Springer.
Herbert, E. W. (1993). Iron, gender, and power: Rituals of transformation in African societies.
Bloomington: Indiana.
Hoover, H. C., & Hoover, L. R. (trans.) 1950. Georgius Agricola’s De Re Metallica. New York:
Dover Publications.
Horton, M. C., & Middleton, J. (2000). The Swahili: The social landscape of a mercantile society.
Oxford: Blackwell.
Huysecom, E., & Agustoni, B. (1997). Inagina: l’ultime maison du fer/the last house of iron. Ge-
neva: Telev. Suisse Romande. Videocassette, 54 min.
Insoll, T. (2008). Negotiating the archaeology of destiny. An exploration of interpretive possibili-
ties through Tallensi Shrines. Journal of Social Archaeology, 8(3), 380–403.
Jones, A. (2004). Archaeometry and materiality: Materials-based analysis in theory and practice.
Archaeometry, 46(3), 327–338.
Joosten, I., Jansen, J. B. H., & Kars, H. (1998). Geochemistry and the past: Estimation of the
output of a Germanic iron production site in the Netherlands. Journal of Geochemical Explora-
tion, 62(1), 129–137.
Keen, J. (2013). Smelting: A sacred process. Observations of iron smelting in Madhya Pradesh,
India. In J. Humpris & T. Rehren (Eds.), The world of Iiron (pp. 97–103). London: Archetype.
Killick, D. (2004a). Review essay: What do we know about African iron working? Journal of
African Archaeology, 2(1), 97–112.
Killick, D. (2014a). Cairo to Cape: The spread of Metallurgy through Eastern and Southern Africa.
In B. W. Roberts & C. Thornton (Eds.), Archaeometallurgy in global perspective (pp. 507–
527). New York: Springer.
Killick, D. (2014b). From ores to metals. In B. W. Roberts & C. Thornton (Eds.), Archaeometal-
lurgy in global perspective (pp. 11–45). New York: Springer.
Killick, D. J. (1990). Technology in its social setting: Bloomery iron-smelting at Kasunga, Malawi,
1860–1940. University Microfilms.
Killick, D., & Fenn, T. (2012). Archaeometallurgy: The study of preindustrial mining and metal-
lurgy. Annual Review of Anthropology, 41, 559–575.
shadreck.chirikure@uct.ac.za

15References
Kiriama, H. O. (1987). Archaeo-metallurgy of iron smelting slags from a Mwitu Tradition site in
Kenya. The South African Archaeological Bulletin, 125–130.
Klemm, R., & Klemm, D. D. (2012). Gold and gold mining in ancient Egypt and Nubia: Geoar-
chaeology of the ancient gold mining sites in the Egyptian and Sudanese eastern deserts. New
York: Springer.
Kusimba, C. M. (1993). The archaeology and ethnography of iron metallurgy on the Kenya coast
Unpublished Doctoral dissertation, Bryn Mawr College.
Lane, P. J. (2005). Barbarous tribes and unrewarding gyrations? The changing role of ethnographic
imagination in African archaeology. In A. B. Stahl (Ed.), African archaeology: A critical Intro-
duction (pp. 24–54). Oxford: Blackwell.
Lechtman, H. (1977). Style in technology—some early thoughts. In H. Lechtman & R. Merrill
(Eds.), Material culture: Styles, organization and dynamics of technology (pp. 3–20). New
York: West.
Lemonnier, P. (1993). Introduction. In P. Lemonnier (Ed.), Technical choices (pp. 1–35). London:
Methuen.
Levtzion, N. (1973). Ancient Ghana and Mali (Vol. 7). London: Methuen.
Levtzion, N., & J.F.P. Hopkins, J. F. P. (Eds.). (2000). Corpus of early Arabic sources for West
African history. Princeton: Markus Wiener Publishers.
Martinón-Torres, M., Rehren, T., & Freestone, I. C. (2006). Mullite and the mystery of Hessian
wares. Nature, 444(7118), 437–438.
Martinón-Torres, M., & Rehren, T. (Eds.). (2009a). Archaeology, history and science: Integrating
approaches to ancient materials. Walnut Creek: Left Coast Press.
Martinón-Torres, M., & Rehren, T. (2009b). Post-medieval crucible production and distribution: A
study of materials and materialities. Archaeometry, 51(1), 49–74.
Mauss, M. (1954). The gift: Forms and functions of exchange in archaic societies (No. 378). New
York: WW Norton & Company.
Miller, D. (2002). Smelter and smith: Iron Age metal fabrication technology in Southern Africa.
Journal of Archaeological Science, 29(10), 1083–1131.
Morton, G. R., & Wingrove, J. (1969). Constitution of bloomery slags: Part 1: Roman. Journal of
the Iron and Steel Institute, 207, 1556–1564.
Morton, G. R., & Wingrove, J. (1972). Constitution of bloomery slags: Part II: Medieval. Journal
of the Iron and Steel Institute, 210, 478–487.
Mudenge, S. I. (1974). The role of foreign trade in the Rozvi empire: A reappraisal. Journal of
African History, 15(3), 373–391.
Mudenge, S. I. (1988). A political history of Munhumutapa c 1400–1902. Harare: Zimbabwe Pub-
lishing House.
Muralha, V. S., Rehren, T., & Clark, R. J. (2011). Characterization of an iron smelting slag from
Zimbabwe by Raman microscopy and electron beam analysis. Journal of Raman Spectroscopy,
42(12), 2077–2084.
Ndoro, W. (1991). Why decorate her. Zimbabwea, 3, 5–13.
Okafor, E. (1993). New evidence on early iron-smelting in Southeastern Nigeria. In T. Shaw, P.
Sinclair, B. Andah, & A. Okpoko (Eds.), The archaeology of Africa: Food, metals, and towns
(pp. 432–448). London: Routledge.
Pernicka, E., Begemann, F., Schmitt-Strecker, S., Todorova, H., & Kuleff, I., (1997). Prehistoric
copper in Bulgaria. Its composition and provenance. Eurasia Antiqua, 3, 41–180.
Phillips, W. R., & Griffen, D. (2004). Optical mineralogy: Non-opaque minerals. New York: W.
H. Freeman and Company.
Phimister, I. R. (1974). Alluvial gold mining and trade in nineteenth-century South Central Africa.
The Journal of African History, 15(3), 445–456.
Pieterse, J. N. (1998). White on black: images of white and blacks in western popular culture. New
Haven: Yale University Press.
Pleiner, R. (2000). Iron in archaeology: the European bloomery smelters. Archeologický ústav
AVČR.
shadreck.chirikure@uct.ac.za

16 1 Metals and the Production and Reproduction of Society
Pollard, A. M., & Bray, P. (2007). A bicycle made for two? The integration of scientific techniques
into archaeological interpretation. Annual Review of Anthropology, 36, 245–259.
Pollard, A. M., Batt, C. M., Stern, B., & Young, S. M. M. (2011). Analytical chemistry in archaeol-
ogy. Cambridge: Cambridge University Press.
Pringle, H. (2009). Archaeology. Seeking Africa’s first iron men. Science, 323(5911), 200–202.
Radivojević, M., Rehren, T., Pernicka, E., Šljivar, D., Brauns, M., & Borić, D. (2010). On the ori-
gins of extractive metallurgy: New evidence from Europe. Journal of Archaeological Science,
37(11), 2775–2787.
Rehren, T., & Pernicka, E. (2008). Coins, artefacts and isotopes–archaeometallurgy and archaeom-
etry. Archaeometry, 50(2), 232–248.
Rehren, T., Charlton, M., Chirikure, S., Humphris, J., Ige, A., & Veldhuijzen, H. A. (2007). Deci-
sions set in slag: The human factor in African iron smelting. In S. La Niece, D. Hook, & P.
Craddock (Eds.), Metals and mines: Studies in archaeometallurgy (pp. 211–218). London:
Archetype.
Rickard, T. A. (1939). The primitive smelting of iron. American Journal of Archaeology, 43(1),
85–101.
Rothenberg, B. (1970). An archaeological survey of South Sinai: First Season 1967/1968, Prelimi-
nary Report. Palestine Exploration Quarterly, 102(1), 4–29.
Rothenberg, B. (1999). Archaeo-metallurgical researches in the Southern Arabah 1959–1990. Part
2: Egyptian new kingdom (Ramesside) to early Islam. Palestine Exploration Quarterly, 131,
149–175.
Scheel, B. (1989). Egyptian metalworking and tools. Oxford: Shire Publications.
Schmidt, P. R. (1978). Historical Archaeology: A structural approach in an African culture. West-
port: Greenwood Press.
Schmidt, P. R. (1996). The culture and technology of African iron production. Gainesville: Uni-
versity Press of Florida.
Schmidt, P. R. (1997). Iron technology in East Africa: Symbolism, science, and archaeology.
Bloomington: Indiana University Press.
Schmidt, P. R. (2009). Tropes, materiality, and ritual embodiment of African iron smelting fur-
naces as human figures. Journal of Archaeological Method and Theory, 16(3), 262–282.
Schultze, C., Stanish, C., Scott, D., Rehren, T., Kuehner, S., & Feathers, J. (2009). Direct evidence
of 1,900 years of indigenous silver production in the Lake Titicaca Basin of Southern Peru.
Proceedings of the National Academy of Sciences, 106, 17280–17283.
Srinivasan, S. (1994). Wootz crucible steel: A newly discovered production site in South India.
Institute of Archaeology, University College London, 5, 49–61.
Stahl, A. B. (1994). Change and continuity in the Banda area, Ghana: The direct historical ap-
proach. Journal of Field Archaeology, 21(2), 181–203.
Stahl, A. B. (2001). Making history in Banda: Anthropological visions of Africa’s past. Cambridge:
Cambridge University Press.
Stahl, A. B. (2009). The archaeology of African history. The International Journal of African His-
torical Studies, 42(2), 241–255.
Summers, R. (1969). Ancient mining in Rhodesia and adjacent areas. Salisbury: Trustees of the
National Museum of Rhodesia.
Tripathi, V. (2013). An ethno-archaeological survey of iron working in India. In J. Humpris & T.
Rehren (Eds.), The world of Iiron (pp. 1094–1115). London: Archetype.
Van der Merwe, N. J., & Avery, D. H. (1987). Science and magic in African technology: Tradi-
tional iron smelting in Malawi. Africa, 57(2), 143–172.
Vansina, J. M. (1985). Oral tradition as history. Madison: University of Wisconsin Press, New
York: Springer.
Zangato, E., & Holl, A. F. (2010). On the iron front: New evidence from North-Central Africa.
Journal of African Archaeology, 8(1), 7–23.
shadreck.chirikure@uct.ac.za

17
Chapter 2
Origins and Development of Africa’s
Preindustrial Mining and Metallurgy
© The Author 2015
S. Chirikure, Metals in Past Societies, SpringerBriefs in Archaeology,
DOI 10.1007/978-3-319-11641-9_2
Introduction
From antiquity, Africa has been simultaneously a continent of similarities and dif-
ferences. Geographically and to some extent culturally, Africa can be divided into
discrete regions: West, Central, East, and Southern Africa, as well as the Horn and
North Africa including Egypt. Egypt, North Africa and the Horn have a long history
of participation in the metallurgical, ceramic, glass and other high temperature tech-
nological traditions of the Near East and the Mediterranean worlds. Other regions,
primarily in sub-Saharan Africa, form a distinctly different area, which, although
interacting with North Sahara, particularly after 500 BC (see Stahl 2014a and refer-
ences therein), forms a distinct cultural and technological block.
Egypt and adjacent regions closely mimic the metallurgical trajectories of the
nearby Middle East. Egyptian metallurgy started with the working of copper around
4000 BC. By 3000 BC, the Bronze Age was fully established with iron appear-
ing much later in the last millennium BC (Scheel 1989). Because Egypt had cul-
tural interactions with regions to the south of the Nile, metallurgy was established
in Nubia by 2600 BC (Emery 1963). Iron smelting appeared much later in Egypt
(c. 600 BC; Scheel 1989) when compared to the rest of the Middle East and was
established even later (c. 500 BC) at places such as Meroe in the Sudan (Rehren
2001). In North Africa, the Phoenician settlements at Carthage were established by
600 BC (Fig. 2.1). The development of metallurgy in Carthage is not clearly un-
derstood, but it is clear that by 600 BC or shortly after, Carthaginians worked iron,
copper and bronze (Alpern 2005).
Sub-Saharan Africa differs from this picture in that metallurgy in this part of the
continent began with the working of iron and in some cases iron and copper (Holl
2009). This is especially true in West Africa, Central Africa, East Africa and South-
ern Africa. The advent of metallurgy in sub-Saharan Africa is a highly contentious
topic because for every possibility, there are two or more contradictions (Craddock
2010). Metallurgy in West, East and Central Africa began sometime between 800
and 400 BC in the radiocarbon black hole created by fluctuations in atmospher-
ic concentration of radiocarbon (Clist 2013; Killick 2004a). In Southern Africa,
shadreck.chirikure@uct.ac.za

18 2 Origins and Development of Africa’s Preindustrial Mining and Metallurgy
metallurgy only appeared with the advent of agriculturalists early in the first millen-
nium AD (Phillipson 2005). After almost a thousand years, bronze and gold made
their appearance in sub-Saharan Africa when the region was directly integrated into
the Islamic trade via the Sahara and the Indian Ocean littoral. This difference with
the picture north of the Sahara precipitated the development of a raging and largely
unresolved debate regarding the origins of sub-Saharan African metallurgy, par-
ticularly whether it is local or external in origin (Alpern 2005). Whatever the case
Fig. 2.1 Map of Africa showing metalworking sites with some of the most important highlighted
by number
shadreck.chirikure@uct.ac.za

19Origins of Metallurgy in Egypt and Adjacent Areas
maybe, the differences in the adoption of metallurgy in Africa’s different regions
provide important lessons for innovation, technology transfer and cultural interac-
tion. Once established, metallurgy was neither static nor homogenous throughout
antiquity. It developed locally and regionally, creating a very richly varied history
of local innovation and cross-cultural borrowing.
Origins of Metallurgy in Egypt and Adjacent Areas
The earliest evidence for metallurgy in Africa comes from the Nile Delta in Egypt
and is associated with the Maadi culture dating between 4000 and 3200 BC (Killick
2014a; Scheel 1989). Evidence suggests that copper substituted for flint as the raw
material for making heavy-duty tools during this period. The paucity of copper de-
posits in this area coupled to its proximity to the Sinai Desert and Southern Jordan
presents a persuasive but untested hypothesis that the copper of the Arabah Desert
was used by Maadi people. Elsewhere in Egypt, rare ornaments and implements
of copper metal were recovered in the middle Nile during the Badarian period (ca.
4400–4000 cal BC). However, no archaeometallurgical studies were carried out to
determine whether they were made of smelted or native copper (Killick 2014a).
Still in the middle Nile, although copper oxides were used in the Naqada I period
(4000–3500 cal BC), heavy-duty copper tools such as axes and blades were more
common during the Naqada II phase (3500–3200 cal BC) (Scheel 1989). The ore
used to make these objects probably came from the lower Nile near Nubia. Gold
and silver also appear at low frequency in Naqada II graves (Midant-Renes 2000).
It is possible that some if not all of this gold came from the Eastern Desert and later
from Nubia (Klemm et al. 2003).
Copper and gold artefacts initially appeared in lower Nubia (the region be-
tween the First and Second Cataracts) in graves of the Middle A Group, which
are dated from ca. 3600–3300 cal BC (Killick 2014a). These are associated with
Naqada pottery and other items of Egyptian provenance, suggesting that they too
were imported. By 3000 cal BC, copper beads, awls and pins were found as far
south as the Third Cataract. Interestingly, it appears as if all cutting implements
were still made of stone. The earliest evidence of the production of metals in Nubia
is from Old Kingdom contexts (ca. 2600 BCE) at Buhen (Emery 1963) and within
the temple precinct further upstream at Kerma, in contexts dated by radiocarbon to
2200–2000 cal BC. Ancient Egyptians forged meteoric iron (iron in its native state)
from c. 3000 BC onwards to produce beads and other decorative items (Rehren
et al. 2013). Indeed, sporadic iron objects were found at Egyptian sites, but it is gen-
erally accepted that iron smelting began in Egypt after its invasion by the Assyrians
in 691 BC. Iron smelting then gradually filtered down the Nile to Kerma, Meroe
and other places and was well established by c. 500 BC. Craddock (2010) argues
that given the antiquity of its metallurgy, Kerma is a possible source of sub-Saharan
metallurgy but more research is required to substantiate this thinking.
shadreck.chirikure@uct.ac.za

20 2 Origins and Development of Africa’s Preindustrial Mining and Metallurgy
The Phoenicians are credited with introducing knowledge of metallurgy to North
Africa, particularly to modern-day Tunisia and Libya. Around 1101 BC, the Phoeni-
cians established the trading port of Utica in Tunisia and by 814 BC had established
Carthage nearby (Alpern 2005). There is a great deal of debate regarding the met-
allurgical history of Carthage, but it is clear that iron was worked together with
copper and bronze by 300 BC. Alpern (2005) cites unsubstantiated reports of iron
smelting at Carthage dating to 800 BC. Unless corroborated by written texts, this
dating too may be affected by the radiocarbon ‘black hole’ where the calibration
curve flattens between cal 800 and 400 BC resulting in uncertain dates (Killick
2014a) and, like similar dates elsewhere in Africa, must be treated with caution.
Seemingly, Phoenician ventures into the western Mediterranean were motivated by
a desire to identify sources of gold, silver, copper and tin for trade purposes. This
was crucial because the Egyptians had virtual monopoly over the gold from Nubia
and the Eastern desert. Although copper is available at Akjoujt in Mauretania and
tin in Niger’s Aïr Mountains, it seems that Phoenicians never knew of these sources,
preferring the tin of Cornwall that is believed to have featured in Carthaginian trade
(Alpern 2005). Carthage features strongly in debates over origins of sub-Saharan
metallurgy, with proponents of external origins speculating that it may have been a
conduit in knowledge transfer. I return to this point after presenting the evidence for
early metallurgy in sub-Saharan Africa.
Ethiopia and Eritrea are poorly understood as far as the development of metal-
lurgy is concerned (Mapunda 1997; Phillipson 2005). It has been suggested that
the Horn of Africa follows the progression witnessed in Egypt, Nubia and Arabia.
As such, gold, copper, and silver and bronze were known in Ethiopia by the last
centuries BC. Aksum witnessed the height of its power from the early first millen-
nium AD and minted its own coinage in gold and silver (Phillipson 2005, p. 230).
The beginning of iron working in Ethiopia was also late relative to adjacent regions,
starting around cal 300 BC (Mapunda 1997). Indeed, the available evidence sug-
gests close interaction between the Kingdom of Kush in its various stages and the
Horn of Africa on the one hand and Egypt and the Mediterranean world on the other
via the Red Sea trade.
Metal from Somewhere: On the Origins of Metallurgy
in West, Central and East Africa
In the studies of Africa’s later prehistory, no topic evokes as much debate and emo-
tion as the origins of sub-Saharan metallurgy (Alpern 2005; Zangato and Holl 2010
and responses therein). When compared to the Middle East and the adjacent Balkans,
which are widely believed to be independent centres of metallurgy (Radivojević
et al. 2010), sub-Saharan metallurgy started simultaneously with the working of
copper and iron (Van der Merwe and Avery 1982) (Table 2.1 & Fig. 2.1). The path-
way to metallurgy in Middle Eastern and Balkan centres of metallurgical origins, as
well as in Egypt, began with the intentional heating of oxide and carbonate copper
shadreck.chirikure@uct.ac.za

21Metal from Somewhere: On the Origins of Metallurgy in West …
Site name Lab nos. Uncalibrated
dates
Calibrated dates
at 95 % confi-
dence interval
Sources
Termit Massif,
Niger
Do Dimmi 16
a M
Do Dimmi 15
a F
UPS
IFAN
2590 ± 120
2630 ± 120
978–404 BC
1031–410 BC
Person and
Quenchon 2004,
p. 122
Gara Tchia Bo
48 E
Pa 810 3260 ± 100 1770–1290 BC Person and
Quenchon 2004,
p. 122
Gara Tchia B
48 W
Pa 811 3265 ± 100 1775–1294 BC Person and
Quenchon 2004,
p. 122
Tchire Ouma 147 Pa 320 3300 ± 120 1895–1370 BC Person and
Quenchon 2004,
p. 122
Termit Ouest 96
b M
Pa 481 3100 ± 100 1611–1107 BC Person and
Quenchon 2004,
p. 122
Termit Ouest 8-b Pa 688 2880 ± 120 1322–819 BC Person and
Quenchon 2004,
p. 122
Nsukka Region,
Nigeria
Opi OxA-3201 2305 ± 90 596–166 BC Okafor 1993,
p. 347
OxA2691 2170 ± 80 396–40 BC Okafor 1993,
p. 347
Oxa3200 2080 ± 90 361 BC–70 AD Okafor 1993,
p. 347
Lejja Ua 34416 1715 ± 35 244–398 AD Eze-Uzomaka
2013
Ua 34417 2370 ± 40 545–380 BC Eze-Uzomaka
2013
Ua 34415 4005 ± 40 2631–2458 BC Eze-Uzomaka
2013
Taruga BM938 2541 ± 104 846–403 BC Calvacoressi and
David 1979
Taruga BM942 2291 ± 123 596–98 BC
Togo
Dekpassanware Beta 252674 2970 ± 40 1297–1051 BC De Barros 2013
Cameroon
Olinga Beta 31414 2820 ± 70 1131–827 BC Essomba 2004,
p. 140
Table 2.1 shows some of the earliest dates for the appearance of metallurgy in Africa. Calibrated
using OxCal version 4.2.3 Bronk-Ramsey (2013) and IntCal13 (Reimer et al. 2013)
shadreck.chirikure@uct.ac.za

22
Site name Lab nos. Uncalibrated
dates
Calibrated dates
at 95 % confi-
dence interval
Sources
Ly4978 2380 ± 110 792–347 Essomba 2004,
p. 140
Ly4979 1954 ± 250 544 BC–590 AD Essomba 2004,
p. 140
Beta 31412 1860 ± 70 2–345 AD Essomba 2004,
p. 140
Central African
Republic
Obui Pa 2223 3645 ± 35 2136–1921 BC Zangato and
Holl 2010
Obui Pa 2130 3635 ± 35 2058–1903 BC Zangato and
Holl 2010
Gbabiri Pa 1446 2670 ± 40 898–797 BC Zangato and
Holl 2010
Rwanda
Rwiyange HV 1296 2250 ± 125 593–20 BC Van Grunder-
beek et al. 2001
Mozambique
Matola R1327 1880 ± 50 19–246 AD Huffman 2007,
p. 163
St8546 1720 ± 110 70–550 AD Huffman 2007,
p. 163
South Africa
Silver leaves Pta 2360 1760 ± 50 137–386 AD Huffman 2007,
p. 163
Pta 2459 1700 ± 40 246–416 AD Huffman 2007,
p. 163
Broederstroom KN 2643 1600 ± 50 344–569 AD Huffman 2007,
p. 163
1350 ± 80 547–880 AD Huffman 2007,
p. 163
Zimbabwe
Mabveni SR79 1380 ± 110 425–886 AD Huffman 2007,
p. 163
Gokomere SR26 1420 ± 120 386–886 AD Huffman 2007,
p. 163
Table 2.1 (continued)
2 Origins and Development of Africa’s Preindustrial Mining and Metallurgy
ores in temperature- and environment-regulated apparatuses to gain a usable prod-
uct (Craddock 1995; Pernicka et al. 1997; Radivojević et al. 2010; Scheel 1989).
The Bronze Age started with the working of arsenical copper followed by the alloy-
ing of tin with copper to produce bronze, with the more complicated iron appearing
shadreck.chirikure@uct.ac.za

23Metal from Somewhere: On the Origins of Metallurgy in West …
around 1500 BC (Tylecote 1976; Craddock 2000). Egypt and areas under its influ-
ence along the Nile broadly followed this trajectory of copper, bronze and iron tran-
sition. Gold and other metals such as lead were also worked during this time, such
that a long-distance trade had evolved by 2000 BC. Despite its advantages, iron was
not universally accepted in the Middle East because Egypt only fully embraced it
around 700 BC, more than six centuries after its adversaries, neighbours and trading
partners adopted it (Craddock 2000; Holl 2000). The Cushite Egyptian Pharaohs
were defeated by iron-armed Assyrians in 691 BC. This supports the argument that
the adoption of metallurgy, such as technology in general, is culturally mediated;
no matter how many perceived advantages there are, society determines what is and
what is not acceptable.
The path to metallurgy in the Middle East and adjacent regions indicates that
discovery and innovation followed the easiest methods through which the very first
metals could be worked (Craddock 2010). Such a picture partly intersects with the
laws of physics and chemistry as summarized by the Ellingham diagram (Fig. 2.2)
which presents the temperature at which oxide ores are reduced to metal in rela-
tion to the levels of carbon monoxide sufficient for reduction. According to Killick
(2014b), pioneer metals such as copper and tin could be easily reduced at low tem-
peratures, while latecomers such as iron required much higher temperatures and
delicate control of furnace atmosphere to reduce their ores. Following a technical
logic, this seems to account for why copper and tin were smelted earlier than iron.
However, it is not just a temperature issue, but also one of redox–carbon monoxide
is not strong enough to reduce ‘modern’ metals (Th. Rehren pers comm 2014).
The Ellingham diagram does not fully explain the sequence of metallurgical in-
novation in antiquity (Killick 2014b, p. 35). For example, metals such as cobalt
and nickel (Fig. 2.2) have a lower melting point when compared to iron and are
reducible at even lower temperatures. Yet, they were only smelted in the nineteenth
century. In fact, nickel is more abundant than copper in the earth’s crust, while co-
balt is more abundant than lead (Killick 2014b). There are many possibilities that
account for why nickel and cobalt were not smelted in the known centres of metal-
lurgical origins. The most important one is that nickel and cobalt oxides are quite
soluble in water and thus are almost never found in gossans (intensely oxidized,
weathered and exposed/upper part of an ore deposit or mineral vein), making the
fact that nickel and cobalt oxides are relatively easy to reduce irrelevant and there
were no oxide or carbonate ores of these elements available (Killick 2014b). This
also demonstrates that laws of physics and chemistry do not always fully explain
the evolution of cultural phenomena. In fact, the laws themselves are cultural phe-
nomena which were discovered at various points, explaining why most metals were
discovered much later, and most of them not in any order that respects the known
affinities between them.
In sub-Saharan Africa, tin, bonze and gold were worked more than a millennium
after iron and copper were introduced. This period coincided with the integration
of the subcontinent into the fledging long-distance trading network rooted in the
Persian Gulf and the Indian subcontinent (Miller and Van der Merwe 1994). The big
question, therefore, is where did knowledge of metalworking in sub-Saharan Africa
shadreck.chirikure@uct.ac.za

24
Fig. 2.2 Ellingham diagram shows the ease with which metals and sulphides can be reduced. The
position of the line for a given reaction on the Ellingham diagram shows the stability of the oxide
as a function of temperature. Reactions closer to the top of the diagram are the most ‘noble’ met-
als (for example, gold and platinum), and their oxides are unstable and easily reduced. Moving
towards the bottom of the diagram, metals become progressively more reactive and their oxides
harder to reduce
2 Origins and Development of Africa’s Preindustrial Mining and Metallurgy
originate? Is it local or external in origin? As a follow on, what are the mecha-
nisms for the dispersal of knowledge of metallurgy across the sub-Saharan lati-
tudes? Based on thermodynamic theory and the lateness with which the continent
shadreck.chirikure@uct.ac.za

25Metal from Somewhere: On the Origins of Metallurgy in West …
embraced metallurgy in comparison with regions such as the Middle East (Fig. 2.2),
one school of thought argues that knowledge of West and East African metalwork-
ing was external in origin (Killick 2004a; McIntosh and McIntosh 1988; Phillipson
2005).
The fulcrum of this position is that it is very difficult to start smelting a techni-
cally complicated metal such as iron without exposure to easier metals or com-
parable pyrotechnology as illustrated by the Middle Eastern trajectory (Craddock
2000; Killick 2014a). Furthermore, there are technical problems with early dates for
West and East African metallurgy which may have been affected by the old wood
problem (Alpern 2005) and the problems with calibration for dates falling between
800 and 400 BC. The old wood problem emanates from the fact that some of the
trees in sub-Saharan Africa lived for long periods due to the gradual desertifica-
tion of the Sahara between 4500 and 2000 BC (Childs and Herbert 2005; Killick
1987). If charcoal from old wood was used for smelting and subsequently dated by
archaeologists, the dates produced reflect when the trees died and not necessarily
the metalworking episodes. This has led to a rejection of most of the early dates for
African metallurgy, some of which were additionally compromised by uncertainty
of contexts (Clist 2013). Then there is the fact that the calibration curve flattens
between 2300 and 2600BP which gives a very long tail between 800 and 400 BC
and with that a great deal of uncertainty (Alpern 2005). It has long been advocated
that researchers must use alternative dating techniques such as luminescence dating
(Killick 2004a) though few have heeded the call (Darling 2013).
If West, Central and East African metallurgy emerged from outside, as posited
by the external origins theory, what transmission routes did it follow? The site of
Akjoujt in Mauritania has yielded copper working objects dating to 800 BC, sug-
gesting a possible introduction from Morocco and a copper to iron transition (Miller
and Van der Merwe 1994). However, earlier thoughts suggested that Egypt and Car-
thaginian settlements in North Africa were possible conduits for a north-to-south
transmission of knowledge (Childs and Herbert 2005). Alpern (2005) believes that
iron working was well established at Carthage by c. 800 BC, making it possible
that West African metallurgy diffused from there. The only problem on the basis of
current knowledge is that the dating is not universally agreed on and that iron be-
came established in Egypt after the invasion by Assyrians in 691 BC (Scheel 1989).
Furthermore, the evidence for the appearance of iron in Carthage postdates that of
the supposedly receiving areas of West Africa (Darling 2013: 158; Eze-Uzomaka
2013, p. 4). Another conundrum is the many outward differences between the fur-
nace types used in Egypt, Carthage and other possible source areas when compared
to those utilized at places such as Taruga in Nigeria (Tylecote 1975). If the sub-
Saharans obtained knowledge of metallurgy from Carthage, Egypt or somewhere,
then the rapidity with which they adapted the technologies to the local situations,
without evident experimentation, thereby distinguishing their technology from its
sources at the same time they were adopting it, is remarkable.
These anomalies became fodder for viewpoints that consider African metallurgy
to be local in origin. The local origins hypothesis contends that because Africa has
always been a centre of technological development throughout human history, there
is no reason why metallurgy could not have been independently developed here.
shadreck.chirikure@uct.ac.za

26 2 Origins and Development of Africa’s Preindustrial Mining and Metallurgy
More importantly, various communities on the African continent were aware of
the transformative potential of fire since the middle Pleistocene times. The local
origins viewpoint seemed to gain momentum in the late 1970s and early 1980s
when Danilo Grébénart and his team excavated significant sites with evidence of
early metallurgy in the Agadez region of Niger (Grébénart 1987). The dating evi-
dence, when combined with archaeometallurgical analyses of remnant furnaces and
slags, seemed to indicate that there was an earlier Copper Age, named Copper 1
(2000–1000 BC), followed by a later Copper 2 (1000 BC) phase. According to
Grébénart (1987), iron working started in the Copper 2 period, suggesting that in
similar fashion to other areas in the Old World, African metallurgy started with an
apprenticeship phase of copper working, followed by smelting of the more techni-
cally complicated iron. This seemed to refute the argument that Africans could not
have developed metallurgy independently because they lacked experience with an
easier metallurgy. However, a meticulous re-investigation of the Agadez material
by Killick et al. (1988) demonstrated that what were thought to be remnant fur-
naces associated with Grébénart’s Copper 1 period were vitrified tree stumps. The
re-examination further highlighted that all reliable evidence of metallurgy dated to
the Copper 2 period, later than 1000 BC. For a while, the critique of this evidence
seemed to tilt the pendulum into the direction of external origins.
More recently, indications from the work carried out by Zangato and others in
Central Africa seem to challenge again the external origins thinking. Excavations at
places such as Ôboui in the Central African Republic revealed artefacts and forges
which were dated to a much earlier time period, between 2300 and 1900 cal B.C.—
long before the Anatolians were working iron (Zangato and Holl 2010). Archaeo-
metallurgical studies of the microstructure of iron objects revealed that they were
made of bloomery iron. This Central African evidence generated intense debate,
with critics arguing that although the dates formed a nice cluster, they were prob-
ably from old wood. Furthermore, it was argued that given the acidic nature of soils
in tropical Africa, the iron objects seem remarkably well preserved for their age
(Clist 2013). Other authorities such as Craddock (2010) dismissed the possibility
that Africa started its own metallurgy with iron, suggesting that this is about as
likely as a baby walking without first crawling.
While scholars continue to debate the relevance of these dates, a set of dates
from the Leija sites in Nsukka Nigeria has not yet been considered in full. These
dates are Ua-34415, 4005 ± 40, and Ua-37422, 3445 ± 40 (Eze-Uzomaka 2013). The
Leija date Ua-34415 calibrates to the second millennium BC and was obtained from
charcoal embedded in slag in a stratified context over a meter deep. Clist (2013)
notes that the Nsukka dates are some of the best in terms of association between
the dated material and the events of metalworking. Also in Nigeria, Darling (2013)
dated samples of fired slag pit furnaces at the Durham Thermoluminescence Labo-
ratory, producing very early dates of 2400 BC ± 1100 for Fitola (Dur TL57–2AS)
and 1400 BC ± 850 (Dur TL57–3AS) for the site of Matanfada in the Hausaland area
of Northern Nigeria. The Durham TL dates require some comment because they
have unusually very high error terms. According to Darling (2013), the laboratory
could not find any sources of error and control samples are currently being dated.
Until this dating is properly resolved, the Fitola dates must be viewed with caution.
shadreck.chirikure@uct.ac.za

27Metal from Somewhere: On the Origins of Metallurgy in West …
A point that begs deeper consideration is that the Leija sites seem to be far away
from Phoenician influence and considerably pre-date Carthage and other centres.
This makes a strong statement that dates alone will probably never solve the origins
issue. What will happen if luminescence dates from Leija and other sites also pro-
duce second millennium BC dates as in the case of Darling (2013)? Are we prepared
as scholars to accept that there is nothing wrong with iron smelting starting indepen-
dently in West Africa? Yes, iron smelting may be complicated, but there is so much
diversity in furnace types—including some very rudimentary bowls simply made
of banana stems (see Celis and Nzikobanyanka 1976)—that suggest that iron can be
reduced anywhere provided temperature and air were sufficient. As such, perhaps
the argument that iron is complex to smelt is a presentist assumption. In any case,
the laws of physics and chemistry, which are so lucid today, were mostly hap-hazard
before the nineteenth century such that today people tend to think that discovery and
innovation in the past followed the nice and neat picture depicted in the periodic
table of elements. Indeed, smelting started with less complicated metals but skipped
nickel and cobalt and jumped to iron in Eurasia (Killick 2014b). Perhaps if we shed
off our presentist arguments, it is possible that iron was independently developed
in Africa. The contexts favoring the development of certain technologies are very
different. That being said, further work is required to produce more dates, interro-
gate contexts of recovery and explore possible interconnections between the various
parts of Africa to increase the confidence in the research community.
In East Africa, van Grunderbeek et al. (2001) and Schmidt (1997) have produced
important evidence relating to the beginning of metallurgy in this region. Despite
being close to the Sudan, which was heavily influenced by Egypt, it seems that
the development of metallurgy in the Great Lakes Region emerged as something
remarkably different (Humphris and Iles 2013; Killick 2014a). The Great Lakes re-
gion has traditionally been seen as one of the independent centres for the evolution
of metallurgy in sub-Saharan Africa. Interestingly, van Grunderbeek et al. (2001)
have backed away from the dates of > 2800 BP from Rwanda and Burundi because
one cannot prove that these dates are not contemporary with the furnaces (Killick
2014a). On his part Schmidt (1997, p. 14) has disavowed the three oldest dates (
> 3000 BP) from Buhaya. There remain, however, at least four radiocarbon dates
between ca. 2350 BP and ca. 2650 BP, each in good association with furnaces and
on charcoal of small diameter, which should eliminate the possibility of an ‘old
wood’ error. However, these all fall within a well-known ‘black hole’ in the radio-
carbon calibration curve, and thus, all calibrate at two sigma to a calendar range of
approximately 800–400 cal BCE. This is exactly the same range of calibrated age
as for the earliest radiocarbon dates for iron slag at Meroe (Shinnie 1985; Rehren
2001) and appears to be contemporary with some West African and Carthaginian
dates (Alpern 2005; Holl 2009).
As things stand, there is no consensus regarding the origins of metalworking
in Africa—the knowledge came from somewhere, but that somewhere could be
within or outside the continent. This is because the two competing paradigms have
potent gaps which cannot be overlooked, regardless of how persuasive their propo-
nents are. What Africa has had are very good reviews with a lot of common sense
that unfortunately will not decisively solve the problem at hand. As such, and as
shadreck.chirikure@uct.ac.za

28 2 Origins and Development of Africa’s Preindustrial Mining and Metallurgy
Killick (1987, 2014a) has long advocated, Africa requires more research, backed
up by careful and robust dating programs to understand the origins and dispersal of
metallurgy. One such fruitful area is Nigeria where good association has been estab-
lished between the dated material and iron smelting. So far, evidence from Carthage
presented by Alpern (2005) suggests an early Phoenician presence and metallurgy
around 800 BC. However, as traders, it is not clear why the Phoenicians could not
take advantage of mineral wealth in West Africa and why they would introduce only
iron and copper and not bronze when local sources of tin were available in Niger.
The old wood problem, while making much sense, must be further investigated to
establish the time frames by which it affects the usability and acceptability of early
dates within the context of limitations and error limits of chronometric dating tech-
niques. Clist (2013) argues that it is prudent to identify the tree species responsible
for the charcoal being dated, for it may assist with interpretation. The only chal-
lenge is that charcoal is chemically inert and can thus persist for millennia beyond
the life of trees that produced it.
While the origins debate largely concerns West, East and Central Africa, South-
ern Africa appears to be relatively controversy free. Here, it is acknowledged that
metallurgy consisting of iron and copper working was introduced by the southward
migration of Bantu-speaking people early in the first millennium AD (Kiriama 1987;
Pwiti 1996). As with the pattern in West and Central Africa, gold, tin and bronze
appeared after the integration of the region into the long-distance trade network.
Why are Iron and Copper Earlier than Gold and Bronze
in Sub-Saharan Africa? Some Provocative Thoughts
Whatever the source of sub-Saharan Africa’s preindustrial metallurgy, one telling
observation is that gold, tin, and the alloys bronze and brass were only worked af-
ter 700 AD when the region was integrated into the trading system centred on the
Indian Ocean (including the Persian Gulf) (Miller and Van der Merwe 1994). What
prompts deeper thought is the observation that the so-called source areas north of
the Sahara already knew how to work bronze, tin and gold, making the question
why it took a millennium before Africans adopted these metals and alloys a relevant
one. Egyptian art, dating to 2500 BC and much later, shows black Africans working
and smelting copper. These individuals probably came from the southern neighbors
of Egypt, and indeed, ebony and other commodities from black Africa also ended
up in Egypt and the broader Middle Eastern region (Pieterse 1998). Therefore, con-
nections existed between sub-Saharan Africa and the Middle East via North Africa
(Wilson 2012) and down the Nile (Edwards 2004). It is difficult not to accept that
the parts of sub-Saharan Africa interacting with Mediterranean Africa at least knew
about metalworking. During Roman times, rock art in Mauretania when coupled
with compositional analyses of copper objects indicate the existence of indirect
interaction between parts of West and North Africa (Fenn et al. 2009). The only
complication is that the earliest dates for metallurgy (c. 800–400 BC), on the basis
shadreck.chirikure@uct.ac.za

29Why are Iron and Copper Earlier than Gold and Bronze …
of current evidence, seem to be from Niger, Nigeria and the Central African Repub-
lic, and not from any of the supposed conduit areas. Indeed, the early dates from
Akjoujt, which also fall into the radiocarbon black hole, and objects stylistically
related to those from Morocco may demonstrate contact and confirms that there was
a flow of ideas between regions adjacent to North Africa. If that is the case, why
did this not happen in Niger, or Great Lakes East Africa which is closer to areas that
had undisputed evidence of metallurgy? Again, this underscores the point that Af-
rica is poorly known as far as metallurgical innovation is concerned, placing most
of the existing knowledge on the edge of speculation. Other questions that arise are
if knowledge of metalworking originated from Middle Eastern communities which
were traversing distant territories in search of precious metals as early as 2000 BC,
why did they not take advantage of the vast mineral wealth in sub-Saharan Africa
as their successors did a thousand years later? And, why did Africa not keep on
borrowing technologies after the establishment of increased contact between the
different areas? Several reasons may account for this but it is doubtful whether the
late development of trade in precious metals was a question of delayed consumption
on the part of Middle Eastern communities. It seems most likely that Middle East-
erners either had rich alternative sources of precious metals or they were ignorant
of the metallurgical resources of sub-Saharan Africa which strongly questions the
depth and strength of the diffusionist argument for the origins of African metallurgy.
This paradox prompted Cline (1937, p. 11) to comment that ‘it is curious that
the metal most active in drawing the attention of Arab and European worlds to
Negro Africa was the one which the Negroes themselves generally value the least’.
Certainly, this situation has something to do with the dynamics of values and tech-
nology transfer. In much later periods, gold played an important role in the value
system of the Asante in present-day Ghana while copper was more valuable in many
Central African societies to the extent that it was the ‘red gold of Africa’ (Her-
bert 1984). It seems, in most cases, that technology transfer and innovation was a
complex endeavor based on serendipity and neither followed ‘rule books’ nor was
it premised on ‘common sense’ because some ‘advanced’ technologies appeared,
only to disappear and be rediscovered thousands of years later. Thornton and Reh-
ren (2009) discuss the engineering and geochemical characteristics of a fourth mil-
lennium BC highly refractory copper working crucible recovered at Tepe Hissar,
North-Eastern Iran. Although demonstrating advanced refractoriness and contain-
ing material properties sought after thousands of years later, Tepe Hissar crucibles
were neither widely adopted nor did their manufacture last for very long (Thornton
and Rehren 2009). What is interesting is that such a technology only surfaced in Eu-
rope, in a completely different context, in the postmedieval period (Martinón-Torres
et al. 2006; Martinón-Torres and Rehren 2009). In sub-Saharan Africa, the Igbo
Ukwu bronzes—arguably one of the finest achievements in metal casting the world
over and dating back to the late first millennium AD (Shaw 1970)—form another
example of advanced metallurgical techniques that surfaced and disappeared, only
to emerge again centuries later (Chikwendu et al. 1989). The Yoruba of Southeast-
ern Nigeria independently re-invented the technology of glass making between the
eleventh and thirteenth centuries AD (Lankton et al. 2006). These examples seem to
shadreck.chirikure@uct.ac.za

30 2 Origins and Development of Africa’s Preindustrial Mining and Metallurgy
indicate that the appearance of seemingly advanced technologies in various places
is a random variable, and possibly one with varying combinations and permuta-
tions. Presumably, the reason why, within culture-specific contexts, technologies
can appear and disappear, is that their acceptance passes through cultural filters that
determine whether or not they are acceptable (Thomas 1991). Demography is also
a factor—if more people take on an innovation, it becomes dominant. Otherwise, it
disappears. However, continuities and discontinuities in knowledge transmission/
apprenticeship could presumably be a factor as well.
Conclusion
Although metallurgy runs deep in the pulse of African empires and kingdoms from
Dynastic Egypt to Kush, ancient Ghana, Mali, the Swahili towns of East Africa
and Great Zimbabwe to mention a few examples (Fenn et al. 2009; Horton 1996;
Levtzion 1973; Pikirayi 2001; Summers 1969), the adoption of metallurgy was very
different in these areas. In Egypt, Nubia, Ethiopia, Eritrea and North Africa, the
metals known include copper, gold, silver, iron, lead, mercury and tin. This con-
trasts with sub-Saharan Africa where only iron and copper were known for 1000
years before the adoption of tin, gold and possibly silver. Interestingly, in South-
east Asia, the picture neatly follows that in the Middle East, raising the question
why is sub-Saharan Africa different if it also received its knowledge from the same
source—the Middle East?
Perhaps it is now time to move away from the presentist argument that iron
smelting is complex. It may be today, but may not have been to ancient sub-Saha-
rans. The argument that smelting must progress from less to more complicated met-
als contradicts even the way chemistry developed, which neither followed rules of
simplicity nor easiness given that most elements with known chemical relationships
and affinities were discovered at different times and often by accident.
The main take-home message from this chapter is that there are many unknowns
as far as the origins of African metallurgy are concerned. It is impossible on the
basis of current evidence to argue definitively for or against local and external ori-
gins. More research is required to produce new insights, backed up by well resolved
dating and good association between the dated events and metal production. Frag-
mentary as our knowledge of it is, the origins of African metallurgy is of global
importance because it highlights that the adoption of the technology was neither
homogenous nor straightforward. Instead, it followed different contours mediated
by cultural, economic, political and other factors. This book, however, is about the
nature of the technology, its evolution over time and the associated sociocultural
permutations and impacts within a global picture.
shadreck.chirikure@uct.ac.za

31References
References
Alpern, S. B. (2005). Did they or didn’t they invent it? Iron in sub-Saharan Africa. History in
Africa, 32, 41–94.
Bronk Ramsey, C. (2013). OxCal Program, version 4.2 (http://c14.arch.ox.ac.uk/oxcal.html), Ra-
diocarbon Accelerator Unit. University of Oxford.
Calvocoressi, D., & David, N. (1979). A new survey of radiocarbon and thermoluminescence dates
for West Africa. Journal of African History, 20(1), 1–29.
Celis, G., & Nzikobanyanka, E. (1976). La métallurgie traditionnelle au Burundi: Techniques et
croyances (Archives d’anthropologie, no. 25). Tervuren: Musée royal de l’Afrique centrale.
Chikwendu, V. E., Craddock, P. T., Farquhar, R. M., Shaw, T., & Umeji, A. C. (1989). Nigerian
sources of copper, lead and tin for the Igbo-Ukwu bronzes. Archaeometry, 31(1), 27–36.
Childs, S. T., & Herbert, E. W. (2005). Metallurgy and its consequences. In A. B. Stahl (Ed.), Afri-
can archaeology: Acritical introduction (pp. 276–300). Oxford: Blackwell.
Cline, W. B. (1937). Mining and metallurgy in Negro Africa (No. 5). Menasha: George Banta
Publishing Company.
Clist, B. (2013). Our iron smelting 14C dates from central Africa: From plain appointment to a
full blown relationship. In J. Humpris & Th. Rehren (Eds.), The world of iron (pp. 22–28).
London: Archetype.
Craddock, P. T. (1995). Early metal mining and production. Washington, DC: Smithsonian Uni-
versity Press.
Craddock, P. T. (2000). From hearth to furnace: Evidences for the earliest metal smelting technolo-
gies in the Eastern Mediterranean. Paléorient, 26(2), 151–165.
Craddock P. T. (2010). New paradigms for old iron: Thoughts on É. Zangato & A.F.C. Holl’s “On
the Iron Front”. Journal of African Archaeology, 8, 29–36.
Darling, P. (2013). The world’s earliest iron smelting? Its inception, evolution and impact in North-
ern Nigeria. In J. Humpris & Th. Rehren (Eds.), The world of iron (pp. 156–167). London:
Archetype.
de Barros, P. (2013) A comparison of early and later Iron Age societies in the Bassar region of
Togo. In J. Humpris & Th. Rehren (Eds.), The World of Iron (pp. 10-21). London: Archetype.
Edwards, D. N. (2004). The Nubian past: An archaeology of the Sudan. London: Routledge.
Emery, W. B. (1963). Egypt exploration society preliminary report on the excavations at Buhen,
1962. Kush, 11, 116–120.
Essomba, J.–M. (2004). Status of iron age archaeology in Cameroon. In H. Bocoum (Ed.), The
origins of iron metallurgy in Africa (pp. 135–148). Paris: ENESCO Publishing.
Eze-Uzomaka, P. (2013). Iron and its influence on the prehistoric site of Leija. In J. Humpris & Th.
Rehren (Eds.), The world of iron (pp. 3–9). London: Archetype.
Fenn, T. R., Killick, D. J., Chesley, J., Magnavita, S., & Ruiz, J. (2009). Contacts between West Af-
rica and Roman North Africa: Archaeometallurgical results from Kissi, Northeastern Burkina
Faso. In S. Magnavita, L. Koté, P. Breunig, & O. A. Idé (Eds.), Crossroads/Carrefour Sahel:
Cultural and technological developments in first millennium BC/AD West Africa (pp. 119–
146). (Journal of African Archaeology Monograph Series 2). Frankfurt a. M.: Africa Magna.
Grébénart, D. N. H. (1987). Characteristics of the final neolithic and metal ages in the region of
Agadez (Niger). In A. Close (Ed.), Prehistory of Arid North Africa: Essays in honour of Fred
Wendorf (pp. 287–3I6). Dallas: Southern Methodist University Press.
Herbert, E. W. (1984). Red gold of Africa: Copper in precolonial history and culture. Madison:
University of Wisconsin Press.
Holl, A. (2000). Metals and precolonial African society. In M. Bisson, S. T. Childs, P. de Barros,
& A. Holl (Eds.), Ancient African metallurgy: The sociocultural context (pp. 1–81). Walnut
Creek: AltaMira Press.
Holl, A. F. (2009). Early West African metallurgies: New data and old orthodoxy. Journal of World
Prehistory, 22(4), 415–438.
shadreck.chirikure@uct.ac.za

32 2 Origins and Development of Africa’s Preindustrial Mining and Metallurgy
Horton, M. (1996). Shanga: The archaeology of a muslim trading community on the coast of East
Africa. London: British Institute in Eastern Africa Memoir 14.
Huffman, T. N. (2007). Handbook to the iron age: The archaeology of precolonial farming societ-
ies in Southern Africa. Durban: University of Kwazulu-Natal Press.
Humpris, J. and Iles, L. (2013). Pre-colonial iron production in Great Lakes Africa: recent re-
search at UCL Institute of Archaeology. In J. Humpris & Th. Rehren (Eds.). The World of Iron
(pp. 56–64). London: Archetype.
Killick, D. J. (1987). On the dating of African metallurgical sites. Nyame Akuma, 28, 29–30.
Killick, D. (2004a). Review essay: What do we know about African iron working? Journal of
African African Archaeology, 2(1), 97–112.
Killick, D. (2004b). Social constructionist approaches to the study of technology. World Archaeol-
ogy, 36(4), 571–578.
Killick, D., Van der Merwe, N. J., Gordon, R. B., & Grébénart, D. (1988). Reassessment of the
evidence for early metallurgy in Niger, West Africa. Journal of Archaeological Science, 15(4),
367–394.
Kiriama, H. O. (1987). Archaeo-metallurgy of iron smelting slags from a Mwitu Tradition site in
Kenya. The South African Archaeological Bulletin, 42, 125–130.
Klemm, D., Klemm, R., & Murr, A. (2001). Gold of the Pharaohs–6000 years of gold miningin
Egypt and Nubia. Journal of African Earth Sciences, 33(3), 643–659.
Lankton, J., Ige, A., & Rehren, T. (2006). Early primary glass production in southern Nigeria.
Journal of African Archaeology, 4, 111–138.
Levtzion, N. (1973). Ancient Ghana and Mali (Vol. 7). London: Methuen.
Mapunda, B. B. (1997). Patching up evidence for ironworking in the Horn. African Archaeological
Review, 14(2), 107–124.
Martinón-Torres, M., & Rehren, Th. (2009): Post-medieval crucible production and distribution: a
study of materials and materialities. Archaeometry, 51, 49–74.
Martinón-Torres, M., Rehren, T., & Freestone, I. C. (2006). Mullite and the mystery of Hessian
wares. Nature, 444(7118), 437–438.
McIntosh, S. K., & McIntosh, R. J. (1988). From stone to metal: New perspectives on the later
prehistory of West Africa. Journal of World Prehistory, 2(1), 89–133.
Midant-Renes, B. (2000). The Naqada period (c. 4000–3200 BC). In I. Shaw (Ed.), The Oxford
history of ancient Egypt (pp. 44–60). Oxford: Oxford University Press.
Miller, D. E., & Van der Merwe, N. J. (1994). Early metal working in sub-Saharan Africa: A review
of recent research. Journal of African History, 35(1), 1–36.
Okafor, E. (1993). New evidence on early iron-smelting in Southeastern Nigeria. In T. Shaw, P.
Sinclair, B. Andah, & A. Okpoko (Eds.), The archaeology of Africa: Food, metals, and towns
(pp. 432–448). London: Routledge.
Pernicka, E., Begemann, F., Schmitt-Strecker, S., Todorova, H., Kuleff, I. (1997). Prehistoric cop-
per in Bulgaria. Its composition and provenance. Eurasia Antiqua, 3, 41–180.
Person, A & Quechon, G. (2004). Chronometric and chronological data on metallurgy at Termit:
Graphs for the study of the ancient iron ages. In H. Bocoum (Ed.), The origins of iron metal-
lurgy in Africa. (pp. 119–126). Paris: UNESCO Publishing.
Phillipson, D. W. (2005). African archaeology. Cambridge: Cambridge University Press.
Pikirayi, I. (2001). The Zimbabwe culture: Origins and decline in Southern Zambezian states
(Vol. 3). Walnut Creek: AltaMira.
Pieterse, J. N. (1998). White on black: images of white and blacks in western popular culture. New
Haven: Yale University Press.
Pwiti, G. (1996). Continuity and change: An archaeological study of farming communities in
Northern Zimbabwe AD 500–1700. Uppsala: Societa Archaeologica Uppsaliensis.
Radivojević, M., Rehren, T., Pernicka, E., Šljivar, D., Brauns, M., & Borić, D. (2010). On the ori-
gins of extractive metallurgy: New evidence from Europe. Journal of Archaeological Science,
37(11), 2775–2787.
Rehren, Th. (2001). Meroe, iron and Africa. Der Antike Sudan, 12, 102–109.
shadreck.chirikure@uct.ac.za

33References
Rehren, T., Belgya, T., Jambon, A., Káli, G., Kasztovszky, Z., Kis, Z., & Szőkefalvi-Nagy, Z.
(2013). 5000 years old Egyptian iron beads made from hammered meteoritic iron. Journal of
Archaeological Science, 40(12), 4785–4792.
Reimer, P. J., Bard, E., Bayliss, A., Beck, J. W., Blackwell, P. G., Ramsey, C. B., ... & van der
Plicht, J. (2013). IntCal13 and Marine13 radiocarbon age calibration curves 0–50,000 years cal
BP. Radiocarbon, 55(4), 1869–1887.
Scheel, B. (1989). Egyptian metalworking and tools. Oxford: Shire Publications.
Schmidt, P. R. (1997). Iron technology in East Africa: Symbolism, science, and archaeology.
Bloomington: Indiana University Press.
Shaw, T. (1970). Igbo-Ukwu: An account of archaeological discoveries in Eastern Nigeria (Vol. 2).
Evanston: Northwestern University Press.
Shinnie, P. L. (1985). Iron working at Meroe. In R. Haaland & P. L. Shinnie (Eds.), African iron
working-ancient and traditional (pp. 28–35). Oslo: Norwegian University Press.
Stahl, A. B. (2014a). Africa in the World: (Re)centering African history through archaeology. Jour-
nal of Anthropological Research, 70(1), 5–33.
Summers, R. (1969). Ancient mining in Rhodesia and adjacent areas. Salisbury: Trustees of the
National Museums of Rhodesia.
Thomas, N. (1991). Entangled objects: exchange, material culture and colonialism in the Pacific.
Cambridge: Harvard University Press.
Thornton, C. P., & Rehren, T. (2009). A truly refractory crucible from fourth millennium Tepe His-
sar, Northeast Iran. Journal of Archaeological Science, 36(12), 2700–2712.
Tylecote, R. F. (1975). The origin of iron smelting in Africa. West African Journal of Archaeology,
5, 1–3.
Tylecote, R. F. (1976). A history of metallurgy. London: Metals Society.
van Grunderbeek, M.-C., Roche, E., & Doutrelepont, H. (2001) Un type de fourneau de fonte de
fer associé à la culture Urewe (Aˆ ge du Fer Ancien) au Rwanda et au Burundi. Mediterranean
Archaeology, 14, 291–297.
Van Der Merwe, N. J., & Avery, D. H. (1982). Pathways to Steel: Three different methods of mak-
ing steel from iron were developed by ancient peoples of the Mediterranean, China, and Africa.
American Scientist, 146-155.
Wilson, A. (2012). Saharan trade in the Roman period: Short-, medium-and long-distance trade
networks. Azania: Archaeological Research in Africa, 47(4), 409–449.
Zangato, E., & Holl, A. F. (2010). On the iron front: New evidence from North-Central Africa.
Journal of African Archaeology, 8(1), 7–23.
shadreck.chirikure@uct.ac.za

35
Chapter 3
Mother Earth Provides: Mining and Crossing
the Boundary Between Nature and Culture
© The Author 2015
S. Chirikure, Metals in Past Societies, SpringerBriefs in Archaeology,
DOI 10.1007/978-3-319-11641-9_3
Introduction
Ores are rocks whose metal content is economically exploitable using available
technology (Chirikure 2010a). Ore deposits are often formed where subsurface
geological processes have removed metals from common rock or from masses of
molten magma, and have redeposited them in other locations at much higher con-
centrations (Killick 2014b) (Fig. 3.1). Alternatively, ore deposits are formed where
surface processes have eroded minerals from rocks and concentrated them else-
where. These processes created two major ore forms: epigenetic and syngenetic.
Epigenetic ores were introduced into their surrounding rocks after the host rock had
already formed. Usually, such ores mineralized in the form of lodes and veins. In
contrast, syngenetic deposits were formed at the same time as the host rock. The
erosion of the two types of ore bodies through fluvial processes often transports
small particles of ore which when found in sufficient concentration may be mined
as alluvial or eluvial deposits.
Because ores abound in nature, the process of ore procurement in Africa involved
negotiations between the living, the dead and the deities through the mediation of
intermediaries such as spirits of the land. For the living to cross the nature–culture
boundary to extract ores from the earth’s belly, a number of rituals and taboos were
conducted to propitiate ancestors. There are few archaeological traces of rituals
associated with mining. However, ethnographically, in most parts of sub-Saharan
Africa from the Dogon of Mali in West Africa to the Njanja of Zimbabwe, miners
used medicines to neutralize malevolent influences during the process of mining
(Chirikure 2006; Huysecom and Augustoni 1997). The concept of pollution often
expressed through forbidding menstruating women from mines was important in
sub-Saharan mining (Cline 1937; Haaland 2004a; Herbert 1993; Schmidt 2009).
However, as with everything else, the diversity of African practice makes it danger-
ous to generalize ethnographically and archaeologically because women worked
open pit copper mines in Katanga, Democratic Republic of Congo, in the historical
and ethnographic periods, and archaeologically, female skeletons have been found
shadreck.chirikure@uct.ac.za

3 Mother Earth Provides: Mining and Crossing the Boundary Between …36
in collapsed gold mines dating to the mid-second millennium AD in Zimbabwe
(Summers 1969).
In cases where miners failed to obtain suitable ores, sacrifices were also of-
fered, resulting in the deities unlocking the earth by providing ore (Haaland 2004b;
Huysecom and Augustoni 1997). It was the ancestors and deities who had the pow-
er to mediate between nature and culture and between the known and unknown
(Mbiti 1990). The intervention of deities in mining as well as the miners’ belief in
Fig. 3.1 Location of mines worked in African antiquity. Note that because of its abundance, iron
was worked in more areas than illustrated here
shadreck.chirikure@uct.ac.za

Introduction 37
supernatural powers is not unique to Africa. In current-day Bolivia, El Tio, a devil
figure has power over the mines and miners and is believed to have simultaneous-
ly offered protection and destruction of miners since pre-Columbian times (Nash
1993). At Timna in Israel, copper miners invoked the power of gods and the super-
natural as evidenced by the presence of an Egyptian temple which serviced miners
at the mining site (Rothenberg 1999; Rothenberg and Bachmann 1988).
Armed with intent to produce metal, and the power of supernatural forces, pre-
historic miners proceeded to extract the ore from the ground (Childs 1998). Most of
the time, miners were smelters who knew the desired quality of the ore. In Africa,
as elsewhere, there exist many metallogenic provinces and geological bodies rich
in exploitable iron, copper, tin and gold (Robb 2009). Examples of these include
the Lubumbashi (formerly Katanga) region in the Democratic Republic of Congo,
rich in copper and iron; the South African Phalaborwa Carbonatite Complex, rich in
iron and copper ores; and the Bambuk goldfields of West Africa situated on the up-
per Senegal River area (Bisson 2000; Curtin 1973; Phimister 1974; Robb 2009; see
Fig. 3.1). In Africa, archaeological signatures dating from as early as the Early Iron
Age indicate that these types of ore bodies were exploited for almost 2000 years
before the onset of colonialism in the late nineteenth century (Summers 1969). The
mining techniques were not static, and neither was the scale and organization of
production throughout antiquity. In some cases, free labour was used, but slavery
was also a source of labour in ancient Egypt and the Roman times. As such, the be-
liefs and values that were prevalent in society were also produced and reproduced
during mining.
Different methods of ore extraction were adopted, depending on the nature of the
mineralization. In regions where natural processes of erosion resulted in the deposi-
tion in downstream areas of rich but fine grains of ores, alluvial mining was prac-
tised, as in the Egypt Eastern Desert around 2000 BC. The Mafa of Northern Cam-
eroon (David et al. 1989) and the Kikuyu (Brown 1995) of Kenya panned magnetite
sand for bloomery iron smelting in the more recent past. Similarly, the Soninke peo-
ple of ancient Ghana panned gold from the rich goldfields of Bambuk between cal
AD800 and 1000 (Levtzion 1973). In areas where rich outcrops of ore existed such
as at Thabazimbi, the iron mountain of South Africa, no excavation was required,
resulting in mere surface collection of ores (Friede and Steel 1976). As surficial
resources were depleted, miners would follow the lodes horizontally or vertically,
resulting in either open or underground mining. Geological constraints such as the
nature of the mineralization motivated for the use of similar mining techniques in
both the Old and New Worlds, be it at Potosi, Rio Tinto, Timna, Phalaborwa or
Zawar (Chirikure 2010a; Hammel et al. 2000; Hauptmann 2007; Summers 1969).
Importantly, not all ‘ore bodies’ were worked due to limitations posed by the avail-
able extractive metallurgy. For example, sulphidic copper ores were avoided by
smelters in much of sub-Saharan Africa because the widely available single-stage
smelting process was not optimized to deal with them (Miller and Killick 2004).
In sum, the process of finding ores and extracting them from the ground as doc-
umented in the historical period in Africa and Latin America and in antiquity at
Timna in the Arabah Desert required the intervention of ancestors and deities as
shadreck.chirikure@uct.ac.za

3 Mother Earth Provides: Mining and Crossing the Boundary Between …38
well as an understanding of the technical knowledge of the local geology and the
behaviour of the ore body. This intersection between technology and culture made
preindustrial mining across the globe a socially embedded process.
History of Mining: A Global Perspective
The point that preindustrial mining was not static throughout antiquity needs no
emphasis. However, where and when did mining begin in the world? This broad
question cannot be answered with certitude but it is now a well-established fact
that during the Middle Stone Age between 200,000 and 40,000 years ago, early
humans residing in Africa extracted and modified iron oxides for use as pigments
(Watts 2002). The Ngwenya mines on the Bomvu Ridge in Swaziland which were
dated by radiocarbon to 40,000 BP are further testimony to the antiquity of mining
in Southern Africa (Dart and Beaumont 1967). However, the earliest intentional
mining of oxide and carbonate ores to gain metal through a controlled application
of heat did not occur in Africa (Craddock 2000). The available evidence suggests
that around 5000 BC, communities living in the Middle East and the neighbouring
Balkans were smelting copper ores to produce metal for tool making. Mines such as
Rudna Glava and Ai Bunar in Bulgaria and Serbia were worked during this period
(Radivojević et al. 2010). Initially, while ores were possibly surface-collected, as
outcropping deposits got depleted and populations increased, more complex meth-
ods of mining developed, resulting in open and underground mining at places such
as Faynan in Jordan, Timna in Israel, Laurion in Greece and Rio Tinto in Spain
(Hauptmann 2007).
In Africa, the available evidence indicates that mining with intent to obtain ores
for metallurgical reasons was established by the fourth millennium BC in Egypt
(Klemm and Klemm 2012) and by first millennium BC in regions such as West
Africa (Holl 2009; Killick et al. 1988). In Southern Africa, the southward migration
of the Bantu introduced metallurgically motivated mining to this region early in
the first millennium AD. In these regions, the nature of the geology too resulted in
humanity responding to the need to extract ore in identical and varying ways.
As elsewhere in the Old World, ancient mining was a key component of pre-
colonial Africa’s social, technical and economic systems. It became the pivot on
which long-distance trade, local trade and the rise and fall of empires were anchored
(Chirikure 2007; Phimister 1974). For example, the depletion of gold fields un-
der its control precipitated the decline of ancient Ghana at the onset of the second
millennium AD, while at the same time promoting the fortunes of Mali (Levtzion
1973). Similarly, the rise of lucrative gold mining in Northern Zimbabwe has been
strongly implicated in the decline of Great Zimbabwe (Pikirayi 2001). The most per-
tinent observation, however, is that unlike some twentieth-century industrial mines
across Africa, which became centres of urbanization, in precolonial Africa as else-
where, early mining landscapes were not always associated with urban settlements
(Mackenzie 1975; Summers 1969; Van der Merwe and Scully 1971). It seems that
shadreck.chirikure@uct.ac.za

The First Step: Ore Exploration and Prospecting in Precolonial Africa 39
at places such as Timna and Faynan, metal workers were associated with village-
scale settlement but some settlements developed over time as the scale of produc-
tion increased (Hauptmann 2007). The absence of rich ore sources near the Early
Iron Age site of Swart Village (cal 700 to 1200 AD) near Mt Darwin in Northern
Zimbabwe suggests that smelters at the site brought ore from elsewhere (Chirikure
and Rehren 2006). With time, large-scale smelting took place at sources, resulting
in large mounds of slag around Phalaborwa in South Africa, Meroe in the Sudan,
the Dogon of Mali (> 300,000 m3 of slag) (Robion Brunner et al 2013), Bassar in
Togo (> 80,000 m3 of slag) (de Barros 2013), at Laurion in Greece, in Cyprus and
the Alps region in Austria. According to Summers (1969), most modern mines in
Africa as elsewhere in the world are situated on sites of preindustrial mining, show-
ing continuity in humanity’s dependence on minerals since metallurgy began. This
overlay of later evidence on earlier traces has destroyed valuable early evidence.
The First Step: Ore Exploration
and Prospecting in Precolonial Africa
According to Cline (1937), most well-known ore deposits across much of sub-Sa-
haran Africa were worked for successive millennia. Iron, gold, tin and copper are
some of the metals whose ores were continuously worked preindustrially. Mining
landscapes such as Katanga in the Democratic Republic of Congo, the Copper belt
in Zambia, Bambuk on the upper Senegal River, the historic Gold Coast, the Zim-
babwe plateau and Phalaborwa are but few examples of metallogenic provinces
that best fit this description. For example, the copper province of Central Africa
(Katanga and Copperbelt of Zambia) was worked from c.AD200–1900 (Bisson
2000) while the gold deposits on the Zimbabwe plateau were worked between
c.AD1000 and 1900. To identify mineralization, preindustrial Africans would have
prospected and explored the land. In geological terms, mineral prospecting is the
process of exploring the landscape in search of exploitable mineral deposits (Robb
2009). Knowledge of the earth’s physical landscape was important to the prospec-
tors because different metals are geologically specific and are associated with dif-
ferent types of host rocks and vegetation regimes (Summers 1969). In some cases,
outcropping ores may have provided some additional leverage. Without doubt, it is
clear that these extinct prospectors were advanced in their reading of the landscape
to distinguish ore-bearing from non-ore-bearing rocks. The colour of the ore-bear-
ing rocks would certainly have helped (Hauptmann 2007). Successful exploration
would also have demanded complex methods of organization and mobilization of
labour. In ethnographically known examples, sizeable groups of people scoured
the landscape for viable ore sources and the head smelters certified the quality of
the ore through visual inspection (Childs 1998; Mackenzie 1975). Once identified,
certain ore bodies were worked successively over millennia.
Prospecting too was guided by the intervention of ancestors in the ethnographic
record (Childs 1998) such that amongst the Tsara of Ethiopia, the discovery of a
shadreck.chirikure@uct.ac.za

3 Mother Earth Provides: Mining and Crossing the Boundary Between …40
mine was followed by a series of rituals including pouring beer on the ground to
acknowledge the role of ancestors in guiding the prospectors (Haaland 2004b).
Conversely, failure to find a rich ore deposit prompted the Dogon prospectors of
Mali to slaughter animals for ritual cleansing and to look for medicines to chase
away bad luck and evil spirits. It is possible that preindustrial prospecting was asso-
ciated with a range of rituals and beliefs pertinent in contexts of success just as they
were in failure. Whereas an absence of written documents and a lack of tangible
evidence makes it difficult to reconstruct the nature of prospecting and exploration
in much of preindustrial Africa, it was nonetheless a key component of the chaîne
opératoire of metal working.
Methods of Crossing the Nature–Culture Boundary:
Mining of Ores in Preindustrial Africa
Once ore bodies were located, numerous techniques were used to dig up the ore,
and these were largely motivated by the nature of the mineralization and the op-
portunities and or constraints that it imposed. For example, alluvial and lode depos-
its demanded not just varying techniques of mining, but also different tools used
in separating the ore from the host rock or material. Within variation, four major
types of ore mining were used across the world before industrialization (Chirikure
2010a; Craddock 1995). These are surface collection, alluvial mining, open mining
and underground mining. It would seem that with very rich deposits, miners often
started working outcropping ore and followed the mineralization into the ground
once surficial deposit was exhausted (Hammel et al. 2000). The nature of the lode
determined whether open or underground methods were to be used. Alluvial meth-
ods were also used to mine placer deposits, usually after the rainy seasons (Curtin
1973; Phimister 1974). On a global level, the methods for mining in the preindus-
trial world were context dependent but almost universal, constrained by the avail-
able techniques, the underlying geology and the prevailing sociocultural context.
Surface Collection
In areas with rich ore mineralization, there was no need to dig deep into the ground
to extract suitable ore. Instead, the miners simply surface collected high grade ores
for smelting. Craddock (2000) argues that the earliest smelted copper ores in Egypt
c. 4000 BC may have been very pure because they left little slag and may have
been surface-collected. There is no tangible evidence for surface collection in the
archaeological record beyond this inference. Even in Africa’s recent past, there are
numerous examples where miners collected rich nodules of iron ore and smelted
them to gain elemental metal. Across the continent, the surface collection of iron
seems to have been a very popular activity (Cline 1937). For example, nineteenth-
shadreck.chirikure@uct.ac.za

Alluvial Mining 41
century miners in Nyanga, Northeastern Zimbabwe surface collected lateritic ores
and hematite for smelting (Chirikure and Rehren 2004). At Musina in modern-day
South Africa, the Venda and Lemba people also surface-collected rich iron ores and
smelted them to produce metal (Mamadi 1940). Surface collection was practised
in numerous metallogenic provinces of sub-Saharan Africa such as Katanga and
Bassar where, respectively, copper and iron exist in abundance (Cline 1937; de
Barros 1988). Surface collection did not leave much in the form of traces in an-
tiquity but it is tempting to speculate that it was one of the earliest methods of
accessing ore and that with increased scale of production other methods came into
play. Continual surface collection, however, exhausted outcropping deposits, which
motivated finding methods that required excavation into the earth’s crust and the
accompanying but varying intensities of earth moving.
Alluvial Mining
Alluvial mining was one of the most common methods of ore extraction in prein-
dustrial Africa, and indeed in the whole world (Summers 1969). Alluvial deposits
form along waterways and rivers, when small quantities of ore are washed from
source and deposited on river beds and gravels (Robb 2009). Normally, this took
place in cases where rivers flow near or across auriferous and other ore-rich out-
crops. There are a few metallogenic provinces that since time immemorial were the
center of alluvial mining in South Central and West Africa (Curtin 1973; Phimister
1974) (Fig. 3.1). Examples of these include Bambuk on the upper Senegal River,
Mandara Mountains of Cameroon, Northern Zimbabwe and tons of magnetite scree
(eluvial) around the former Lolwe Hill at Phalaborwa which may have been ex-
ploited since AD1000 (Killick and Miller 2014).
Ethnographic and Historical Descriptions of Alluvial Mining The ethnographic
examples collated by Cline (1937) indicated that gold, tin and iron were the only
metal ores that were extracted from alluvial ore in precolonial Africa. Alluvial
deposits were worked using a process known as panning which involved digging
ore-rich sand on the river beds and scooping it into hemispherical buckets or bowls
for density separation through shaking or winnowing (Chirikure 2010a). During
this process, the lighter sand matrix stayed on top and was thrown away while the
heavy metal settled at the bottom. Across many areas of Africa, runoff from heavy
rains often carried with it small magnetite particles that were deposited along river
banks and other erosion channels. In regions such as Northern Cameroon, Kenya,
and Southwestern Nigeria, these magnetite sands were collected in exploitable
quantities. Not surprisingly, the Mafa of Northern Cameroon, the Yoruba of South-
western Nigeria and the Kikuyu of Kenya panned rich magnetite sands which they
smelted to produce a mix of cast and soft iron in the recent past (Brown 1995; Cline
1937; David et al. 1989; Ige and Rehren 2003). Cline (1937) vividly described
how Kikuyu men and women diverted water into small artificial lakes to wash the
magnetite-rich sands, thereby separating the ore from the surrounding sandy matrix.
shadreck.chirikure@uct.ac.za

3 Mother Earth Provides: Mining and Crossing the Boundary Between …42
Alluvial gold mining was recorded in detail by the Portuguese who operated in
Northern Zimbabwe from the sixteenth century AD onwards. In this region, the
Shona people worked the alluvial deposits along rivers such as Mukaradzi and
Mazowe in Northern Zimbabwe (Phimister 1974; Swan 1994). Ellert (1992) de-
scribes a very complicated method of working alluvial gold deposits used by the
Shona in Northern Zimbabwe in the nineteenth and twentieth centuries. This is
the process of underwater dredging whereby experienced divers draped weights on
their backs and dived into the waters of the Mazowe and Rwenya rivers. Here, they
scooped auriferous sand into receptacles and took it to the surface where women
and children separated the gold from the sand (Chirikure 2010a). Interestingly,
the Dutch painter Olfert Dapper painted a seventeenth-century Akan community
dredging auriferous sand from the Ankobra River in Southern Ghana, suggesting
that the technique may have some wide expression throughout Africa (Fig. 3.2)
(Garrard 2011, p. 116).
It seems that alluvial mining was practised by both men and women. In particu-
lar, men dug the earth while women winnowed, but this division of labour varied
from context to context. It should be pointed out that there are rare instances where
alluvial miners dug wide and deep holes into the sand, resembling open mining
(Cline 1937). The ore–debris separation was similarly carried out by women using
density separation techniques such as winnowing. It has been argued that women
Fig. 3.2 Akan gold miners diving into the Ankobra River to extract diamond-rich sand which
was panned on the river bank.. (Redrawn from Garrard 2011, p. 116, original from Dapper 1668)
shadreck.chirikure@uct.ac.za

Alluvial Mining 43
featured prominently in panning because the principle of the process was rooted
in winnowing, an occupation gendered as female across many African societies
(Summers 1969). This is one remarkable example of crossovers in gender roles
which, as we have seen in the extraction of magnetite sands by Kikuyu women,
deviates from generalizations which see no space and place for women in African
mining.
Archaeological Insights into Alluvial Mining Alluvial mining is difficult to
explore archaeologically because traces of mining were often overwritten after
the following year’s rains. In an archaeologically rare case, Wagner and Gordon
(1929) recorded undated late second millennium AD finds of alluvial tin ore at a
few tin-smelting archaeological locations in the Blaauwbank Donga near Rooiberg
in the Southern Waterberg of South Africa (Fig. 3.1). In most cases, what we
know about alluvial mining in African archaeology is derived either from writ-
ten sources or from secondary sources such as the geochemistry of objects and
slags. In Southern Africa, the geochemistry of tin slags dating from AD1600
revealed that some slags—particularly those from the Blaauwbank Donga—were
rich in zircon, ilmenite and other heavy detrital minerals (Miller and Hall 2008).
This contrasted with the composition and mineralogy of slags from Smelterskop
which lacked detrital minerals. Chirikure et al. (2010) concluded on the basis of
these differences that Rooiberg tin miners exploited both alluvial and hydrothermal
tin.
Arabic writers such as Ibn Battuta indicate that the fabled Bambuk gold fields
in West Africa were worked alluvially during the time of the Soninke Empire of
ancient Ghana in the late first to early second millennium AD. Later its successor
empires Mali and Songhai tapped into the gravels of the upper Senegal River and
adjacent tributaries, extracting tons of ore which ended up in the Islamic world via
the trans-Saharan trade (Curtin 1973). In southern Africa, Al Masudi noted that the
inhabitants of Sofala widely believed to be the Zimbabwe plateau exploited alluvial
gold in the early second millennium AD. As indicated in textual evidence, alluvial
mining was one of the methods practised by ancient Egyptians and Nubians from
4000 BC onwards (Klemm and Klemm 2012).
Summary Hammel et al. (2000) contend that alluvial mining was a fairly simple
method of ore extraction, despite the fact that it required complex methods of deci-
sion-making and organization. For instance, in the ethnographic record, panners
had to understand rainfall cycles as well as to separate rich deposits from poor ones
so as to make their labour input worthwhile. Furthermore, the miners had to care-
fully schedule their activities depending on seasons to free labour for critical pur-
suits such as agriculture in the rainy season (Curtin 1973; Phimister 1974; Summers
1969). Consequently, most panning activities—whether associated with magnetite
sands or gold extraction—were confined to the winter months when the water table
was low and communities had just finished harvesting. Annually, the end of alluvial
mining cycle coincided with the beginning of the rainfall season when labour was
reallocated to pursuits such as farming and cattle herding.
shadreck.chirikure@uct.ac.za

3 Mother Earth Provides: Mining and Crossing the Boundary Between …44
Open Mining
Across the African continent, the nature of the ore mineralization mandated that
trenches had to be excavated into the ground to extract ores. According to Summers
(1969), once surface outcropping ore was depleted, miners followed the deposit into
the ground. If the mineralization was horizontal, long trenches were dug, but if the
mineralization was vertical, miners would dig deeper creating open pits of varying
sizes. In contrast to underground mines explained below, open mines are more hori-
zontal and could reach a distance of more than 380 m (Bisson 2000). Copper, tin,
gold and iron were all extracted using the technique of stope mining. The process of
stope mining involved excavating ore mineralization from the ground using basic
tools such as hoes, shovels and picks. The ore was separated from the host rock,
resulting in the creation of mining dumps of varying sizes on the sides of the mines.
Most of these preindustrial mines were not backfilled; because of this, significant
amounts of tangible evidence littered different parts of Africa before the onset of
modern mining (Cline 1937; Summers 1969). In cases where miners encountered
very hard rocks, difficult to break with the limited tools in their arsenal, they fired
those rocks to very high temperatures and poured water on them while they were
still hot leading to a cracking of the rock which can then be mined more easily. This
technique known as ‘fire setting’ worked remarkably well and conferred the added
advantage to archaeologists in that it left charcoal which can be radiometrically
dated to estimate the age of the mine or the period when the mine was worked.
Ethnographic and Historical Descriptions One of the most detailed cases of
open mining relates to copper mining by the Yeke of the Democratic Republic of
Congo and the Kaonde of Zambia (Fig. 3.1). According to Cline (1937), in the nine-
teenth and early twentieth centuries, the renowned copper workers such as the Yeke
and Kaonde worked open copper mines, digging into the ground to extract the ore
from the lodes. These copper miners of Central Africa also practised fire setting to
break hard host rocks at mines such as Etoile and Dikuluwe (Bisson 2000). The
bigger lumps were broken by hammerstones into smaller pieces which were eas-
ily transportable to the surface. Interestingly, one of the expeditions was led by a
woman (Bisson 2000).
Mackenzie (1975) discusses open mining for banded iron stone on the histori-
cally famous Hwedza range of Mountains in the eighteenth and nineteenth centuries
which rarely exceeded 4 m in diameter. Baskets were used to hoist the ore and as-
sociated rocks. Where possible, the ore was separated from the host rock inside the
mine. In and around Oyo in Nigeria, a number of long trenches representing open
mines have also been reported (Bellamy and Habord 1904) (Fig. 3.1).
Archaeological Descriptions One of the most detailed and vivid descriptions of
open mining in precolonial Africa is provided by Summers (1969) who studied
in detail the ancient copper and gold workings found in Zimbabwe and adjacent
territories dating between the late first and mid-second millennium AD. Summers
explains that the epigenetic quartz vein ore mineralization presented notable con-
straints that shaped open gold mining on the Zimbabwe plateau from the late first
shadreck.chirikure@uct.ac.za

Underground Mining 45
millennium AD onwards. For instance, the small width of the lodes and veins dic-
tated that most gold mines were only a few metres in diameter. Such a width was
also determined by the need to create a comfortable working space for the miners.
Occasionally, when two gold reefs were adjacent to each other, and were detached
by soft rock, they were mined simultaneously using the technique of side stoping.
At Thakadu in North-Eastern Botswana, the technique of open mining was widely
used to extract copper from AD1450 onwards (Huffman et al. 1995).
The available evidence indicates that women worked some of the open mines.
For example, amongst the Shona people of Zimbabwe, women worked inside gold
mines. Similarly, women also worked the Katanga copper mines in central Africa.
At Kansanshi in Zambia, copper was mined using open mining methods from AD
300 onwards (Bisson 2000). The earliest example of open mining in Africa comes
from Egypt and Nubia and is associated with gold extraction (Klemm and Klemm
2012). There are a number of gold mines in the Eastern Desert which represent
abandoned open but not very deep shafts worked from dynastic periods to later
times. Hammerstones and ore milling infrastructure was also documented here. In
sub-Saharan Africa, the fact that most mines were worked over generations result-
ed in the obliteration of earlier evidence by most recent mining. Summers (1969)
thinks that some open gold mines on the Zimbabwe plateau date to the late first
millennium AD. Superficially, this may appear to contradict the earliest finds of
gold which date to the early second millennium AD. On closer observation, this
may be true given that glass beads from the East African coast start appearing in
Southern Africa from AD700 onwards. Arabic writers mention that gold was one
of the commodities sourced in Southern Africa. The only controversial and wildly
speculative conclusion from Summers (1969) is the innuendo that Indians worked
the gold mines on the Zimbabwe plateau. This is invalidated by skeletal evidence
which shows that local people worked the mines. In any case, the nature of the
mineralization dictated that miners had to dig into the ground resulting in similar
techniques across the world throughout antiquity. There is simply no other way to
extract subsurface ore other than digging.
Underground Mining
The technique of open mining was not well suited to work horizontal lodes which
lay at some depth underground. To exploit these, precolonial peoples first made
vertical shafts into the ground and mined out those underground reefs, giving rise to
underground mining. During preindustrial times, it appears that different smelting
communities across the world—be it at Timna in Israel (Rothenberg 1962), Zawar in
India (Willies et al. 1984), Bambuk in West Africa or the Late Bronze Age mines on
the Austrian Alps—developed underground mining when the open mines reached
great depths. The underground deposits were mined out using procedures identical
to those employed during open mining. The only difference is that underground, ore
mineralization often branched into various directions which, when mined, created
shadreck.chirikure@uct.ac.za

3 Mother Earth Provides: Mining and Crossing the Boundary Between …46
a network of tunnels, galleries, adits and mined out pockets beneath the ground. As
with open mining, copper, tin, iron and gold were extracted using the technique of
underground mining. However, there are very few instances in precolonial Africa
where miners created underground tunnels and galleries during iron mining. This
probably stems out of the metal’s relative abundance on the earth’s crust.
Ethnographic and Historical Descriptions Cline’s (1937) compilation of eth-
nographic cases of mining in sub-Saharan Africa exposed that underground min-
ing was practised by many groups to work gold, copper and possibly iron and tin.
The Venda and Lemba of Musina extracted carbonate copper ores from the copper
mines around Musina in Northern South Africa in the nineteenth century (Stayt
1931). These mines have since been destroyed by modern mining. Similarly, the
Ba-Phalaborwa of South Africa people also extracted copper ores from the under-
ground mines on Lolwe Hill. One of the largest copper mines in precolonial Africa,
Etoile, near modern-day Lubumbashi in the Democratic Republic of Congo, was
also mined using underground methods in the late nineteenth and early twentieth
centuries (Bisson 2000).
Archaeological Descriptions Predevelopment impact assessments carried out
before the onset of modern mining at Harmony Block in Northeastern South Africa
and at Lolwe in Phalaborwa resulted in some of the most detailed archaeological
descriptions of precolonial underground mining (Fig. 2.1). Amongst the most dra-
matic examples of preindustrial underground mines in Southern Africa is the cal
thirteenth-century AD copper mine previously located on the Harmony block in the
North-Eastern Limpopo Province of South Africa (Evers and van den Berg 1974).
In the 1970s, an interdisciplinary team of archaeologists and geologists led by Evers
carried out salvage work in advance of modern copper mining. The rescue team
documented in remarkable detail the evidence from the mine. Research indicated
that the Harmony copper mine consisted of one open stope which branched into at
least 25 shafts, pockets and galleries underground. These galleries provided deep
insights into indigenous people’s understanding of structural geology, mining and
the need for ventilation. In Unit 22, Evers and van den Berg (1974) discovered size-
able quantities of hardwoods, probably used as ladders for accessing the mine and
or for hoisting ore to the surface. Amazingly, three wooden props used to provide
a structural support to forestall collapse were still in place. In another section of
the mine, the miners strategically left blocks of unmined rock as pillars to further
provide structural support. These techniques are akin to those used in contemporary
mining and demonstrate how advanced these preindustrial miners were.
This practice of leaving unmined host rock as pillars underground to provide
support was also noted at the Aboyne Gold Mine in Zimbabwe showing that it was
widely used in Southern Africa (Summers 1969) (Fig. 3.3). However, such interven-
tions often fell short. The Aboyne Mine collapsed and in the process killed at least
four individuals, some of them women. Presumably, the dangers associated with
this activity required the intervention of supernatural powers. This is why ancestors
and the supernatural seem to have played a significant role in mining in preindus-
trial Africa and elsewhere (see Nash 1993 for Bolivia).
shadreck.chirikure@uct.ac.za

Underground Mining 47
The Rooiberg tin mines, situated in the Thabazimbi district of Limpopo Prov-
ince, South Africa, provides another graphic example of underground mining. The
mines were worked between the fifteenth and nineteenth centuries (Chirikure et al.
2007; Hall 1981), but new research has extended the beginning to the thirteenth
century AD (Bandama 2013). According to Baumann (1919), the Rooiberg tin
mines were established by digging vertical and inclined shafts into the earth’s crust.
While underground, the tin lodes were followed by drives to pockets (Trevor 1906).
It has been estimated that about 18,000 t of ore were mined by indigenous peoples
before the onset of modern mining. Interestingly, some of the underground adits
at Rooiberg were very narrow, implicating the use of child labour and or very un-
comfortable working conditions for adults (Friede and Steel 1976). As with other
African underground mines, backfilling was done primarily to reduce the volume
to be hoisted, provide stability to the underground building and manage air flow in
order to close off ‘dead’ parts of the mine (Fig. 3.4).
Ladders may have been used for access (Hall 1981). The ore-rich rock was
sometimes processed underground using grinding stones similar to those used for
processing grain. Examples of these were recovered inside the Rooiberg mines. In
terms of division of labour, this may suggest the involvement of women since grind-
ing stones were mainly associated with them just as winnowing and alluvial mining.
The ore was likely transported to the top using baskets, while the gangue material
was used as backfill. Because underground miners sometimes encountered poison-
ous gases, some of which were life-threatening, backfilling was one of the interven-
tions made secondarily to neutralize their harmful effects (Hammel et al. 2000). To
improve ventilation, the miners at Rooiberg dug narrow and vertical shafts from the
ground into the mining chambers. It is possible that a fire was lit underground to
Fig. 3.3 Cross section of the Aboyne gold mine, Central Zimbabwe. (Redrawn from Summers
1969, p. 24, Fig. 5)
shadreck.chirikure@uct.ac.za

3 Mother Earth Provides: Mining and Crossing the Boundary Between …48
facilitate the use of one shaft as an up-draught chimney which drew air down the
other one using the principle of convection (Miller 2000). These techniques were
also used by New Kingdom Egyptians used for mining copper at Timna in Israel
(Rothenberg 1962), showing that geological constraints homogenized approaches
to ore extraction in many areas.
Fig. 3.4 Umkondo copper mine in Zimbabwe where miners strategically backfilled and opened
up new shafts to create pillars (unmined blocks) for structural stability. (Redrawn from Summers
1969, p. 29, Fig. 6)
shadreck.chirikure@uct.ac.za

Underground Mining 49
The Rooiberg mines also provide additional insights into the methods used to
transport personnel and the ore to the surface (Recknagel 1906; Baumann 1919). In
one of the ancient stopes, early mining geologists discovered an expertly cut tree
whose branches had been strategically removed for ease of stepping during use as a
ladder (Chirikure et al. 2007). Furthermore, some of the vertical shafts connecting
to underground galleries and chambers had steps incised from the top to bottom for
easy access to the work place.
The now destroyed precolonial copper mine on Lolwe Hill, Phalaborwa, repre-
sents one of the earliest dated underground mining complexes in sub-Saharan Afri-
ca. Van der Merwe and Scully (1971) dated charcoal from one of the galleries to the
eighth century AD. Other absolute dates, material culture and historical evidence
from in and around the mine suggest that the carbonate copper deposits on Lolwe
Hill were exploited by local people up to the late nineteenth century (Van der Mer-
we and Scully (1971). As a site of modern copper mining operated by the Palabora
Mining Company, the site is now one of the largest open cast mines ever developed
in the world. The archaeology around Lolwe Hill suggests that at different points in
time, beginning in the late first millennium AD until the nineteenth century, various
communities dipped into the mines, but occupation was concentrated away from the
source at places such as Kgopolwe (Van der Merwe and Scully 1971) and Shankare
(Thondhlana 2012). It is possible that Early Iron Age copper mining on Lolwe Hill
initially began with surface collection developed through open mining and ended up
being a sophisticated underground operation.
During the archaeological research in the area, it was clear that the underground
mines had started as open mines that only intersected rich veins and pockets of
mineralization beneath the ground at depths of between 15 and 20 m (Miller et al.
2001). As the mineralization was mined out, a maze of galleries and adits was left
underground. In one case, a wide vertical shaft ended abruptly in a round chamber,
while in others inclined shafts branched off into horizontal galleries and adits (Van
der Merwe pers comm 2013). Remarkably, Van der Merwe and his team observed
that the miners at Lolwe followed the ore veins with great accuracy, resulting in
the creation of narrow shafts underground which could only be worked by gracile
people and or children. As demonstrated at Rooiberg and other places, child labour
was probably widely used in African mining.
The mining engineering at Lolwe was very advanced, for the site was littered
with very deep vertical shafts of not more than 45 cm in diameter that connected
to underground galleries. These ventilation shafts provided air underground which
was crucial for breathing and for driving away poisonous gases from fire setting and
other mining activities. Inside the mine shafts and galleries, archaeologists found
characteristic mining tools such as iron gads, chisels and dolerite hammerstones.
A dense charcoal concentration found on the floor of yet another gallery on Lolwe
Hill was presumably the result of fire setting which was often utilized to break hard
rock from the gallery walls. The details of these precolonial workings on Lolwe Hill
are not just evidence of the mining techniques used by the indigenous miners in the
first and second millennium AD, but are also an indication of their geologically con-
strained and technologically inspired choices. Again, the underground mines in the
shadreck.chirikure@uct.ac.za

3 Mother Earth Provides: Mining and Crossing the Boundary Between …50
Egyptian desert (c. 4000 onwards) represent Africa’s earliest underground mines
which may have been worked by slaves (Klemm and Klemm 2012). As in later
cases of mining, fire setting was an important technique for breaking the host rock.
The descriptions of underground mining in Africa have deep resonance and cor-
respondence with the practices documented at other famous mining landscapes such
as the copper mines at Timna in the Arabah desert of Israel (Hauptman 2007). Here,
underground mining followed shafts and adits with good accuracy, while narrow
ventilation shafts supplied air to the mining chambers underground and were im-
portant for sucking from underground the abundant poisonous gases. This picture
was also observed at other early mining landscapes in places such as Cornwall in
England during the Middle Ages. This similarity does not and should not necessar-
ily imply the diffusion of ideas from multiple areas into Africa. Rather, it shows
that when confronted with similar geological opportunities and constraints, ancient
miners at different points in time, whether in Katanga in Democratic Republic of
Congo, Zawar in India, Potosi in Bolivia, Cornwall in England and Sinai Desert in
the Middle East selected technological choices that left identical results. Although
there are many technological solutions (Sillar and Tite 2000), it seems that as far as
mining is concerned and as far as geology dictates, there is one way in which below
the ground ore can be mined regardless of time and place. Humanity had to sink
shafts into the earth, and to make provisions for necessities such as air, and safety.
The problems affecting underground mining were mostly related to structural
collapse (Fig. 3.3), flooding and lighting. The burning of logs and grasses may
have provided lighting underground. Timbering inside the mines (e.g., at the Har-
mony Block), digging ventilation shafts (Lolwe and Timna) and creating pillars of
unmined earth (Fig. 3.4) were technological and practical choices aimed at alleviat-
ing these problems. Arguably, the level of skill, technology and decision-making
involved in underground mining is in excess of that invested in either alluvial or
simple open mining.
With mining—whether surface, open or underground—the main challenge for
archaeologists lie in establishing tight sequences of deposit exploitation. This is
because later activities are always overwritten on earlier activities, such that in most
cases the available dates are for more recent activities and not the earliest ones.
Lolwe Hill is one of the few places where an adit was left with charcoal and other
material characteristic of a single time horizon. It is also difficult to speculate what
the earliest technique of mining was on the basis of perceived simplicity or com-
plexity. This is because often all methods were simultaneously used by one group
at the same locality (Hammel et al. 2000; Rothenberg 1999). As such, depending
on needs, and other technological or cultural considerations, mining was a vast
enterprise aimed at obtaining ores from the ground through surface collection, and
digging underground.
shadreck.chirikure@uct.ac.za

Preindustrial Hoisting and Beneficiation 51
Preindustrial Hoisting and Beneficiation
In terms of haulage and hoisting, small wooden bowls and buckets (Fig. 3.5) were
used to carry the ore and earth. Sometimes, buckets made of wood were attached
to a rope and hoisted into deeper mines to transport the ore and gangue materials.
In the case of Shona open gold mining, miners made up a human conveyor belt,
passing the baskets from person to person (Ellert 1992). Sometimes, animal draught
was used by the Njanja of Zimbabwe to carry sacks of ore from mines to smelt-
ing areas (Chirikure 2006). After processing, gold dust was placed in porcupine
quills (Fig. 3.6) for later consolidation in crucibles. However, in times of increased
Fig. 3.6 Porcupine quills used for storing gold. (Photo Credit Author)
Fig. 3.5 Wooden bucket found inside ancient gold mine in Southwestern Zimbabwe and donated
to Natural History Museum, Zimbabwe. Exact provenance unknown. (Photo credit: Aluthor)
shadreck.chirikure@uct.ac.za

3 Mother Earth Provides: Mining and Crossing the Boundary Between …52
demand, the Islamic traders operating at places such as Tewdaghoust often bought
gold dust in quills (Curtin 1973), just as the Portuguese did in Southern Zambezia
from AD1500 onwards.
One of the most important stages in the chaîne opératoire of mining was the pro-
cess of beneficiating the ore, which involved cleaning it to remove occluded waste
materials known as gangue. Once recovered, the miners dressed the ore to remove
waste materials. This process is known as beneficiation, and it varied from metal to
metal. Often, high-grade and low-grade ores were blended together to create more
balanced ore in terms of exploitable metal and impurities essential for promoting
slag formation in self-fluxing smelting technologies (Chirikure and Rehren 2004).
The evidence from the hard rock gold mining in Southern Zambezia suggests
that fairly complex grinding and crushing methods were used to recover gold from
its ore. The Shona people crushed gold-rich ore on hard rock surfaces where they
ground it into fine sand for panning. The rock surfaces known as ruware in Shona
acted as the female part while the upper grinding stone was the male (Summers
1969). This process left smoothly abraded but shallow depressions known as dolly
holes (Fig. 3.7), some of which contained smears and traces of gold (Summers
1969; Swan 1994). In other cases, the ore was ground on grinding stones similar to
the ones used for processing grain. The fine sand was then panned to recover gold
dust which was then packed in porcupine quills (Fig. 3.6). According to Curtin
(1973), such a technological solution was also practised at the Bambuk goldfields
on the upper Senegal River in the first and second millennium AD.
At Rooiberg, in South Africa, the tin-rich host rock was crushed into small pieces
which were ground on conventional grinding stones similar to those used for grain.
The product from this process was then washed and panned to recover the cas-
siterite through density separation. Examples of such grinding stones were found
in fairly large quantities at the large-scale tin-smelting site known as Smelterskop
(Chirikure et al. 2010). Much larger rock-grinding stones were used in Zimbabwe
gold mines and were illustrated by Summers (1969).
Fig. 3.7 Undated grind-
ing stone with dolly holes
used to crush copper ores at
Phalaborwa. (Photo credit:
Author)
shadreck.chirikure@uct.ac.za

The Anthropology of Mining: A Global Outlook 53
The literature on copper is not very detailed, but at Phalaborawa and Musina,
where malachite is associated with magnetite, a very meticulous process of sepa-
rating the iron from the copper carbonate was practised. Failure to achieve this re-
sulted in co-smelting of iron and copper resulting in a very weak and useless copper
iron alloy called musina by the Venda (Stayt 1931; see also Craddock and Meeks
1987 for a more technical description of iron in copper). The recovered material was
ready for metallurgical processes such as smelting in clay-walled furnaces.
Klemm and Klemm (2012) published a number of mortars used for beneficiating
gold ores in the Eastern Desert and Nubia. These mortars resemble the ones that
were used in other parts of Africa, at much later periods suggesting that identical
technological solutions were often applied in different contexts as dictated by the
task at hand.
Mining Equipment and Other Paraphernalia
Generally, the tools and equipment used in preindustrial African mining served
multiple purposes: excavating the earth, hoisting and haulage. Digging was accom-
plished using iron hoes attached to wooden handles, iron gads, chisels and wedges.
Hammerstones were used to drive the chisels into the ore body and to break the re-
sultant rocks into small-sized lumps. Shovels were important in scooping earth into
carriers. From the Eastern Desert in Egypt to Nubia and Bambuk, through to Katan-
ga and Phalaborwa in different parts of Africa, mining tools have been recovered by
archaeologists, although preservation favoured inorganic equipment (Table. 3.1).
The Anthropology of Mining: A Global Outlook
In the late nineteenth and early twentieth centuries, there was an established belief
in the Western academy that African technologies such as mining were shrouded in
magic (Beuster 1881 cited in Rickard 1939; see Collett 1993 for critique). This con-
trasted significantly with perceptions of science-based Western technologies (e.g.,
Austen and Headrick1983). In this section, efforts are made to recap the cultural
or, rather broadly speaking, the anthropological aspects of African mining and to
compare them to practices elsewhere. The result demonstrates that, regardless of
location, mining was a heavily ritualized process, requiring the intervention of dei-
ties, spirits of the land and the supernatural (Nash 1993).
When discussing cultural attributes associated with African mining, it is impor-
tant to separate iron and copper from gold, and it is crucial to keep in mind that
most graphic examples are from ethnographic cases, and the detailed recordings in
Ancient Egypt and Nubia. The archaeological evidence is simply not that well pre-
served to allow a detailed reconstruction. The mining of copper and iron was vari-
ably associated with rituals and taboos. For example, amongst the Dogon of Mali,
shadreck.chirikure@uct.ac.za

3 Mother Earth Provides: Mining and Crossing the Boundary Between …54
Technique/
Tool Description
Description Purpose
Stone hammers and
pounders
Dolerite, diorite or—more rarely—quartz-porphyry Rock-breaker. Also used to ‘greenstone’ or
granite drive gads
Iron gads/chisels Smaller ones were pointed at both ends, one end being held in a wooden handle. Larger
ones were driven with stone hammers
Used to split ore
Stone wedges Very few examples from ancient mines. Wooden wedges, although unknown in the
archaeological record due to their impermanence, would seem a more likely tool
Used to enlarge or deepen crack in rock
Fire setting The process of heating rock and then cooling it rapidly (with water) Used to crack rock
Digging Tools
Hoe A heavy diamond-shaped iron tool hafted in a wooden handle. Hoes worn down in the
fields could be adapted for mine use
Used to break, dig or draw soil. Gathered
together broken rock to draw into a basket,
or other carrier
Shovel Iron, with cranked shanks, Smaller iron scoops, known from Zambian, copper mines To gather ore
Haulage and Hoisting
Small wooden bowls Not many examples. It seems that in shallow diggings, mined ore was passed in bowls
from hand to hand until it reached the surface
Haulage in shallow mines
Buckets (wood/bark/
leather/hide)
Fes examples, a small bucket (approx. 3-litre capacity) with three holes for suspension
has been found (Hammel et al. 2000)
Haulage and hoisting in deeper mines
Baskets Musina: evidence of baskets made of tightly woven palm ribs and reinforced with a
leather cover
Carrying ore
Transport
Baskets, sacks Baskets made of vegetal material or sacks of leather Carrying ore from mines to smelting places
Oxen Sacks were placed on the back of oxen Carrying ore from mines to smelting places
Sledges Sacks and baskets loaded on a sledge for transportation Carrying ore
Storage
Porcupine quills Quills used to store gold dust Storage of gold dust for melting and trade
Pots Clay pots Carry ore, food drink for miners
Lighting
Logs Carbonized logs were found in adits suggesting their possible use as torches Providing light
Table 3.1 Techniques and tools used in precolonial mining across Africa. Adapted and modified after Hammel et al. (2002, p. 51)
shadreck.chirikure@uct.ac.za

The Anthropology of Mining: A Global Outlook 55
menstruating women were forbidden from the mines (Huysecom and Augustoni
1997). The presence of women in general too was limited or encouraged depend-
ing on context. Similarly, amongst the Phoka of Malawi, women were not allowed
near mines and neither were they permitted to touch the ore (Van der Merwe and
Avery 1987). This practice was also noted amongst various other groups such as the
Toro in Uganda (Childs 1998). However, at Katanga in DRC, women of all ages
participated in copper mining (Bisson 2000). Therefore, generalizations should not
subsume context-specific variations.
Because mining involved crossing the boundary between the surface and the
belly of the earth, it was seen as a dangerous activity, characterized by many un-
knowns. Miners across Africa and elsewhere often required the intervention of the
deities via intermediaries such as spirit mediums. Medicines were an important part
of the mining ritual, for good medicines kept malevolent influences at bay while
making it easier to find ore. In the African ethnographic record, miners did not al-
ways manage to find good ores; there were instances of failure. Under such circum-
stances, offerings and sacrifices were made to the ancestors, e.g., in the form of beer
or chicken. Once ancestors were propitiated, the earth would release the ore. There-
fore, mining required the intervention of both the dead and the living to help cope
with uncertainty. The involvement of ancestors was important for another reason;
mining took place in nature, conceptually away from the realm of culture. In this
liminal space, it was only the ancestors who could guide the living. Furthermore,
amongst many societies, the land on which mines were located was also under the
realm of ancestors, who made it rain, and made land fertile and productive.
Gold mining seems to have differed from copper and iron as far as rituals and ta-
boos were concerned. It relied heavily on the labour of women and children and yet
involved substantial underground excavation (Summers 1969). Why did different
communities treat this metal differently? Presumably, this stemmed from the fact
that gold exists in pure form in nature, thereby contrasting with copper and iron that
underwent dangerous heat-mediated chemo-thermal transformations to produce us-
able metal. Often, as mentioned above, there were also some contexts where women
worked inside iron and copper mines. Men and women were therefore different ac-
tors, who networked with each other within context given norms to achieve success.
These rich contrasts warn researchers against making broad generalizations at the
expense of culture-specific variation and milieus.
Arguably, precolonial African mining was simultaneously technological in as
much as it was transformative and sociocultural. A brief comparison with well-
documented cases in other parts of the world indicates an overwhelming presence
of gods and supernatural forces in mining. Rothenberg and Bachmann (1988) dis-
cuss the Egyptian temple at Timna which catered for the needs of the miners, pro-
tecting them from harm. As hinted above, El Tio still plays and played an impor-
tant role in ancient silver mining in South America (Nash 1993). The same applies
to Nepal where religion and beliefs were central in technologies such as mining
(Haaland 2004a). Therefore, magic was a strong feature of mining be it in Africa,
Latin America and other parts of the world. It is only that most of the evidence has
been destroyed while not enough recourse was made to historical sources to reclaim
shadreck.chirikure@uct.ac.za

3 Mother Earth Provides: Mining and Crossing the Boundary Between …56
these important details (Schmidt 1997). Mining was embedded in the broader so-
ciocultural fabric, and as such, it exhibited beliefs that were common in society: the
belief in ancestors and deities; the power of evil forces; and concepts of pollution.
Taking away this ritual from technological studies removes the oil that lubricated
the system by empowering people to go beyond the limitations of the everyday so
as to extract material from the other world. Today, we see ourselves as guided by
technological concepts, but in the past, rituals and taboos played a fundamental role
in routinizing learned behaviour in various processes, technical or otherwise.
Another important variable in preindustrial mining relates to the organization
of labour. In preliterate contexts, it is difficult to fully understand how mining was
organized without speculation. However, the Romans, and possibly the Egyptians,
used slave labour in the gold mines of the Eastern Desert (Klemm and Klemm
2012). De Barros (1986) has argued that Bassar iron smelting was highly special-
ized with villages specializing in getting charcoal and ore, in smelting and in forg-
ing.
Conclusion
In conclusion, as the first stage in the chaîne opératoire of metal production and
use, mining occupies an important place in technological and anthropological stud-
ies. The many rich metallogenic provinces of Africa contain substantive footprints
of ore extraction spanning close to 5000 years old in Egypt and Nubia and 2000
years old in sub-Saharan Africa. By comparison, Timna, Faynan and other Eurasian
landscapes have more layers but the similarity in techniques from surface collec-
tion through open to underground mining suggest that the geology to a large extent
determined the techniques of mining and the technocultural solutions selected to
cross the boundary between culture and nature. Far from promoting an unneces-
sary geological determinism, this view is supported by the observation that similar
mineralization was worked using the same methods independent of spatial and tem-
poral frameworks. For instance, if underground veins were excavated, there were
high chances that the ensuing structural instability would result in collapse. This
motivated various techniques of provisioning structural support in the mines, from
the use of pillars, strategic backfilling to using wooden props. This can be seen
from Phalaborwa, Katanga, in Africa to Potosi in the New World and Timna and
Rio Tinto in Eurasia.
Finally, wherever humanity was mining ore in the world, rituals and belief sys-
tems formed part of the mining enterprise, and it is therefore myopic to think that
this was a quintessentially African experience. The gods and spirits empowered the
living to cross into a different conceptual space (underground). Therefore, pollu-
tion through menstruation or malevolent influences was unwelcome, just as these
forces were feared in day-to-day life. There were risks associated with crossing
this nature–culture boundary: death being one of the misfortunes that could be-
fall miners. Because risk is both danger and opportunity, when miners managed to
shadreck.chirikure@uct.ac.za

57References
obtain the ore, how did they transform this commodity from the ‘other world’ into
a cultural product for everyday use? Because the ore existed as a compound of dif-
ferent oxides, it was important to win elemental metal from the ore. This motivated
for smelting to break the bond between the metal and the oxygen. This geochemi-
cal engineering was achieved through smelting in a heat and atmosphere regulated
environment and is an important phase in the chaîne opératoire of metalworking,
which is the focus of the next chapter.
References
Austen, R. A., & Headrick, D. (1983). The role of technology in the African past. African Studies
Review, 26, 155–161.
Bandama, F. (2013). The archaeology and technology of metal production in the Late Iron Age of
the Southern Waterberg, Limpopo Province, South Africa. Unpublished Doctoral dissertation,
University of Cape Town.
Baumann, M. (1919). Ancient tin mines of the Transvaal. Journal of the Chemical, Metallurgical
and Mining Society of South Africa, 19, 120–132.
Bellamy, C. V., & Harbord, F. W. (1904). A West African smelting house. Journal of the Iron and
Steel Institute, 66, 99–126.
Bisson, M. (2000). Precolonial copper metallurgy: sociopolitical context. In M. Bisson, P. de Bar-
ros, T. C. Childs, & A. F. C. Holl (Eds.), Ancient African metallurgy: The sociocultural context
(pp. 83–146). Walnut Creek: AltaMira Press.
Brown, J. (1995). Traditional metalworking in Kenya. Oxford: Oxbow.
Childs, S. T. (1998). ‘Find the ekijunjumira’: Iron mine discovery, ownership and power among
the Toro of Uganda. In A. B. Knapp, V. C. Piggott, & E. W. Herbert (Eds.), Social approaches
to an industrial past: The archaeology and anthropology of mining (pp. 123–137). Abingdon:
Routledge.
Chirikure, S. (2006). New light on Njanja iron working: Towards a systematic encounter between
ethnohistory and archaeometallurgy. South African Archaeological Bulletin, 61, 142–161.
Chirikure, S. (2007). Metals in society: Iron production and its position in Iron age communities
of Southern Africa. Journal of Social Archaeology, 7(1), 72–100.
Chirikure, S. (2010a). Indigenous mining and metallurgy in Africa. Cambridge: Cambridge Uni-
versity Press.
Chirikure, S., & Rehren, T. (2004). Ores, furnaces, slags, and prehistoric societies: Aspects of iron
working in the Nyanga agricultural complex, AD 1300–1900. African Archaeological Review,
21(3), 135–152.
Chirikure, S., & Rehren, T. (2006). Iron smelting in pre-colonial Zimbabwe: Evidence for dia-
chronic change from Swart village and Baranda, Northern Zimbabwe. Journal of African
Archaeology, 4, 37–54.
Chirikure, S., Hall, S., & Miller, D. (2007). One hundred years on: what do we know about tin and
bronze production in Southern Africa? In: S. La Niece, D. Hook, P. Craddock (Eds.), Metals
and mines, studies in archaeometallurgy (pp. 112–119). London: British Museum Press.
Chirikure, S., Heimann, R. B., & Killick, D. (2010). The technology of tin smelting in the Rooi-
berg Valley, Limpopo Province, South Africa, ca. 1650–1850 CE. Journal of Archaeological
Science, 37(7), 1656–1669.
Cline, W. B. (1937). Mining and metallurgy in Negro Africa (No. 5). Menasha: George Banta.
Collett, D. P. (1993). Metaphors and representations associated with precolonial iron—smelting
in Eastern and Southern Africa. In T. Shaw, P. Sinclair, B. Andah, & A. Okpoko (Eds.), The
archaeology of Africa: Food, metals and towns (pp. 499–511). London: Routledge.
shadreck.chirikure@uct.ac.za

3 Mother Earth Provides: Mining and Crossing the Boundary Between …58
Craddock, P. T. (1995). Early metal mining and production. Washington, DC: Smithsonian Uni-
versity Press.
Craddock, P. T. (2000). From hearth to furnace: Evidences for the earliest metal smelting technolo-
gies in the Eastern Mediterranean. Paléorient, 26(2), 151–165.
Craddock, P. T., & Meeks, N. D. (1987). Iron in ancient copper. Archaeometry, 29(2), 187–204.
Curtin, P. D. (1973). The lure of Bambuk gold. Journal of African History, 14(4), 623–631.
Dart, R. A., & Beaumont, P. (1967). Amazing antiquity of mining in Southern Africa. Nature, 216,
407–408.
David, N., Heimann, R., Killick, D., & Wayman, M. (1989). Between bloomery and blast furnace:
Mafa iron-smelting technology in North Cameroon. African Archaeological Review, 7(1),
183–208.
de Barros, P. (1986). Bassar: A quantified, chronologically controlled, regional approach to a tra-
ditional iron production centre in West Africa. Africa, 56(02), 152–174.
de Barros, P. (1988). Societal repercussions of the rise of large-scale traditional iron production: A
West African example. African Archaeological Review, 6(1), 91–113.
de Barros, P. (2013) A comparison of early and later Iron Age societies in the Bassar region of
Togo. In J. Humpris & Th. Rehren (Eds.), The world of iron (pp. 10–21). London: Archetype.
Ellert, H. (1992). Rivers of gold. Gweru: Mambo Press.
Evers, T.M., & Van den Berg, R.P. (1974). Ancient mining in Southern Africa, with reference to
a copper mine in the Harmony Block, North-Eastern Transvaal. Journal of the South African
Institute of Mining and Metallurgy, 74, 217–226.
Friede, H. M., & Steel, R. H. (1976). Tin mining and smelting in the Transvaal during the Iron Age.
Journal of the South African Institute of Mining and Metallurgy, 76(12), 461–470.
Garrard, T. F. (2011) (1989). African Gold: Jewellery and ornaments from Ghana, Côte d’Ivoire,
Mali and Senegal in the collection of the Barbier-Mueller Museum. Munich: Prestel.
Haaland, R. (2004a). Technology, transformation and symbolism: Ethnographic perspectives on
European iron working. Norwegian Archaeological Review, 37(1), 1–19.
Haaland, R. (2004b). Iron smelting—a vanishing tradition: Ethnographic study of this craft in
South-West Ethiopia. Journal of African Archaeology, 2(1), 65–79.
Hall, S. L. (1981). Iron Age sequence and settlement in the Rooiberg, Thabazimbi area. Masters
thesis, University of the Witwatersrand, Johannesburg.
Hammel, A., White, C., Pfeiffer, S., & Miller, D. (2000). Pre-colonial mining in Southern Africa.
Journal of South African Institute of Mining and Metallurgy, 100(1), 49–56.
Hauptmann, A., Heitkemper, E., Pernicka, E., Schmitt-Strecker, S., & Begemann, F. (1992). Early
copper produced at Feinan, Wadi Araba, Jordan: The composition of ores and copper. Archeo-
materials, 6(1), 1–33.
Herbert, E. W. (1993). Iron, gender, and power: Rituals of transformation in African societies.
Bloomington: Indiana University Press.
Holl, A. F. (2009). Early West African metallurgies: New data and old orthodoxy. Journal of World
Prehistory, 22(4), 415–438.
Huffman, T. N., Van der Merwe, H. D., Grant, M. R., & Kruger, G. S. (1995). Early copper min-
ing at Thakadu, Botswana. Journal of the South African Institute of Mining and Metallurgy,
95, 53–61.
Huysecom E., & Agustoni B. (1997). Inagina: l’ultime maison du fer/the last house of iron. Ge-
neva: Telev. Suisse Romande. (Videocassette, 54 min).
Ige, A., & Rehren, Th. (2003). Black sand and iron stone: Iron smelting in Modakeke, Ife, South-
Western Nigeria. Institute for Archaeo-Metallurgical Studies, 23, 15–20.
Killick, D. (2004b). Social constructionist approaches to the study of technology. World Archaeol-
ogy, 36(4), 571–578.
Killick, D., & Miller, D. (2014). Smelting of magnetite and magnetite-ilmenite iron ores in the
Northern Lowveld, South Africa, ca. 1000 CE to ca. 1880 CE. Journal of Archaeological Sci-
ence, 43, 239–255.
shadreck.chirikure@uct.ac.za

59References
Killick, D., Van der Merwe, N. J., Gordon, R. B., & Grébénart, D. (1988). Reassessment of
the evidence for early metallurgy in Niger, West Africa. Journal of Archaeological Science,
15(4),367–394.
Klemm, R., & Klemm, D. D. (2012). Gold and gold mining in ancient Egypt and Nubia: Geoar-
chaeology of the ancient gold mining sites in the Egyptian and Sudanese eastern deserts. New
York: Springer.
Levtzion, N. (1973). Ancient Ghana and Mali (Vol. 7). London: Methuen.
Mackenzie, J. M. (1975). A pre-colonial industry: The Njanja and the iron trade. Nada, 11(2),
200–220.
Mamadi, M. F. (1940). The copper miners of Musina. In N. J. van Wamelo (Ed.), The copper
miners of Musina and the early history of the Zoutpansberg. Ethnological Publications VIII
(pp. 81–87). Pretoria: Department of Native Affairs, Union of South Africa.
Mbiti, J. S. (1990). African religions & philosophy. Portsmout: Heinemann.
Miller, D. (2000). 2000 years of indigenous mining and metallurgy in Southern Africa—a review.
South African Journal of Geology, 98, 232–238.
Miller, D. E., & Hall, S. L. (2008). Rooiberg revisited–the analysis of tin and copper smelting
debris. Historical Metallurgy, 42(1), 23–38.
Miller, D., & Killick, D. (2004). Slag identification at Southern African archaeological sites. Jour-
nal of African Archaeology, 2(1), 23–47.
Miller, D., Killick, D., & Van der Merwe, N. J. (2001). Metal working in the Northern Lowveld,
South Africa AD 1000–1890. Journal of Field Archaeology, 28(3–4), 3–4.
Nash, J. C. (1993). We eat the mines and the mines eat us: Dependency and exploitation in Bolivi-
antin mines. New York: Columbia University Press.
Phimister, I. R. (1974). Alluvial gold mining and trade in nineteenth-century south central Africa.
The Journal of African History, 15 (3), 445–456.
Pikirayi, I. (2001). The Zimbabwe culture: Origins and decline in Southern Zambezian states. Vol.
Walnut Creek: Altamira.
Radivojević, M., Rehren, T., Pernicka, E., Šljivar, D., Brauns, M., & Borić, D. (2010). On the ori-
gins of extractive metallurgy: New evidence from Europe. Journal of Archaeological Science,
37(11), 2775–2787.
Recknagel, R. (1906). On some mineral deposits in the Rooiberg district. Transactions of the Geo-
logical Society of South Africa, 11, 83–106.
Rickard, T. A. (1939). The primitive smelting of iron. American Journal of Archaeology, 43(1),
85–101.
Robb, L. (2009). Introduction to ore-forming processes. London: Wiley.
Robion-Brunner, C, Serneels, V, & Perret, S. (2013). Variability in iron smelting practices: as-
sessment of technical, cultural and economic criteria to explain the metallurgical diversity in
the Dogon area (Mali). In J. Humpris & Th. Rehren (Eds.), The world of iron (pp. 257–265).
London: Archetype.
Rothenberg, B. (1962). Ancient copper industries in the Western Arabah: An archaeological survey
of the Arabah, Part 1. Palestine Exploration Quarterly, 94(1), 5–71.
Rothenberg, B. (1999). Archaeo-metallurgical researches in the Southern Arabah 1959-1990.
Part 2: Egyptian new kingdom (Ramesside) to early Islam. Palestine exploration quarterly,
131(JULDEC), 149–175.
Rothenberg, B., & Bachmann, H. G. (1988). The Egyptian mining temple at Timna, (Vol. 1). Lon-
don: Institute for Archaeometallurgical Studies.
Schmidt, P. R. (1997). Iron technology in East Africa: Symbolism, science, and archaeology.
Bloomington: Indiana University Press.
Schmidt, P. R. (2009). Tropes, materiality, and ritual embodiment of African iron smelting fur-
naces as human figures. Journal of Archaeological Method and Theory, 16(3), 262–282.
Sillar, B., & Tite, M S. 2000. The challenge of ‘technological choices’ for materials science ap-
proaches in archaeology. Archaeometry, 42, 2–20.
Stayt, H. A. (1931). The Bavenda (No. 58). International Institute of African languages & cultures.
Oxford: Oxford University Press.
shadreck.chirikure@uct.ac.za

3 Mother Earth Provides: Mining and Crossing the Boundary Between …60
Summers, R. (1969). Ancient mining in Rhodesia and adjacent areas. Salisbury: Trustees of the
National Museums of Rhodesia.
Swan, L. (1994). Early gold mining on the Zimbabwean plateau. Studies in African Archaeology,
9. Uppsala: Societas Archaeologica Upsaliensis.
Thondhlana, T. (2012). Metalworkers and smelting precincts: Technological reconstructions of
second millennium copper production around Phalaborwa, Northern Lowveld of South Africa.
Unpublished Doctoral dissertation, University College of London.
Van der Merwe, N. J., & Avery, D. H. (1987). Science and magic in African technology: Tradi-
tional iron smelting in Malawi. Africa, 57(2), 143–172.
Van der Merwe, N. J., & Scully, R. T. (1971). The Phalaborwa story: Archaeological and ethno-
graphic investigation of a South African Iron Age group. World Archaeology, 3(2), 178–196.
Wagner, A., & Gordon, S. (1929). Further notes on ancient bronze smelters in the Waterberg dis-
trict, Transvaal. South African Journal of Science, 26, 563–574.
Watts, I. (2002). Ochre in the Middle Stone Age of Southern Africa: Ritualised display or hide
preservative? The South African Archaeological Bulletin, 57, 1–14.
Willies, L., Craddock, P. T., Gurjar, L. J., & Hegde, K. T. M. (1984). Ancient lead and zinc mining
in Rajasthan, India. World Archaeology, 16 (2), 222-233.
shadreck.chirikure@uct.ac.za

61
Chapter 4
Domesticating Nature
© The Author 2015
S. Chirikure, Metals in Past Societies, SpringerBriefs in Archaeology,
DOI 10.1007/978-3-319-11641-9_4
Introduction: Transforming Ore into Metal
According to Herbert (1993), the role of ancestors, ritual, and the supernatural in the
pre-industrial African metal production chain did not end with the process of min-
ing. Instead, the power of ritual, at least in the ethnography of sub-Saharan Africa,
was important in the reduction of oxide or carbonate ores to metal. The longevity
of this practice is suggested by excavation of furnaces dating to the first and second
millennium AD associated with holes where medicines were strategically planted
at the bottom of the iron smelting furnaces in Central, Eastern and Southern Africa
(Mapunda 1995; Rowlands and Warnier 1993; Schmidt 1997) (Fig. 4.1). In Egypt
and Nubia, Gods such as Maat were important throughout the history of metalwork-
ing in the ancient past (Scheel 1989) and smelting was often done within temple
precincts (El Rahman et al. 2013). Within the context of glass making, Robson
(2001, p. 54) argued that the boundaries between science and religion, medicine
and magic were always blurred in the ancient Near East: The spiritual was insepa-
rable from the rational. This demonstrates continuity in various practices across
Africa but does not in any way suggest that African extractive metallurgy was static
through time.
Smelting is a fundamental step in the chaîne opératoire of metal production. For
the ore to be successfully tamed into a cultural product—metal—great skill was
needed in selecting and assembling suitable raw materials such as ore, air (sup-
plied by blowing, bellows or naturally using the principle of convection), clay for
making combustion vessels and receptacles used during smelting and charcoal fuel.
These raw materials are generic and cross-cut metals worked pre-industrially. Food,
labour and in some cases music were important but archaeologically invisible com-
ponents of the metal production process (Dewey 1991).
Preindustrial extractive metallurgy techniques differed significantly from the in-
dustrial processes now currently in use (Rostoker and Bronson 1990). The produc-
tion of iron in modern blast furnaces passes through two stages: (1) reduction of
high-grade iron ores using coke and limestone flux to produce liquid cast iron and
shadreck.chirikure@uct.ac.za

62 4 Domesticating Nature
very glassy slag and (2) further processing of cast iron in reveberatory furnaces to
moderate carbon levels to create usable metal ranging from wrought iron to medi-
um-and high-carbon steels (Rostoker and Bronson 1990). This indirect technique
differs significantly from the bloomery process where ores were reduced to pro-
duce solid metal and liquid slag in a single operation (Tylecote 1980). This ‘direct
method’, as it is known, was the principal method of metal reduction for millennia
in much of the Old World, with the exception of China which had the blast furnace
method from early on Wagner (2008).
Fig. 4.1 Known metal smelting groups in Africa. Numbers indicate key sites and landscapes
shadreck.chirikure@uct.ac.za

63Raw Materials
The inventory of metals smelted in preindustrial Africa varied from region to re-
gion. In Egypt, Nubia and North Africa, the seven metals of antiquity were worked:
gold, copper, silver, tin, lead, iron and mercury. Because of its different history, the
metals worked in sub-Saharan Africa were limited to four: iron and copper (first
millennium BC to recent times i.e. AD 1900) and, from the second millennium AD,
tin and gold (Cline 1937; Killick 2014). Some sources suggest that lead was worked
in the Benue region of Nigeria in the late first and early second millennium AD
and in the Democratic Republic of Congo between the seventeenth and nineteenth
centuries (Bisson 2000; Chikwendu et al. 1989). These metals were worked in di-
verse furnace types that differed from group to group, time to time and area to area.
Often, copper and tin were mixed to produce an alloy known as bronze in Egypt
since the Old Kingdom, as on the Jos plateau of Nigeria (c. AD 900–1200) and in
Rooiberg in South Africa (c. AD 1450–1850). Because gold is a noble metal that
exists in metallic state in nature, it was treated differently from iron, copper and tin.
Instead, the gold dust or nuggets from the mines were melted in ceramic crucibles
to consolidate them into usable pellets (Ellert 1993; Nixon et al. 2011; Ogden 2000;
Oddy 1984; Summers 1969).
The waste products from smelting are remains of high-temperature processes,
which contain within their chemical composition and microstructures partial histories
of the processes that they have undergone (Bachmann 1982; Craddock 1995; Rehren
et al. 2007). Not surprisingly, they form the staple of archaeometallurgy—a sub-
discipline of archaeology that studies metal production and consumption in past soci-
eties (Rehren and Pernicka 2008; Roberts et al. 2009). Archaeometallurgical studies
have primarily been concerned with reconstructing the technology of the process as
revealed by slags, tuyeres, collapsed furnaces and remnants of ore (Bachmann 1982;
Bayley and Rehren 2007; Hauptmann 2000; Heimann et al. 2010; Morton and Wing-
rove 1969). Such studies shed light on the quality of the ore used (Killick 1990), the
efficiency of reduction (metal recovery) (Chirikure 2005, 2006) and the refractory
nature of the clays used to make technical ceramics (tuyeres and furnaces) (Miller
and Killick 2004). Because archaeometallurgical methods of investigation are rooted
in earth and engineering sciences, until recently, the results of this technique were
highly technical, largely ignoring the cultural attributes of the technology and its role
in society (Rehren et al. 2007). Today, however, it is recognized that any study of
preindustrial metalworking must focus on the materials and their materiality in order
to understand the technology in its social context. The rest of the chapter is organized
as (1) a brief outline of the raw materials needed for smelting, (2) a discussion of
the chemistry of reduction, (3) a brief overview of metal smelting in West, Central,
Southern, East and North Africa is provided within a diachronic framework, after
which (4) the chapter concludes with a discussion of the anthropology of smelting.
Raw Materials
For reduction to succeed, and for smelters to transform nature into culture, they had
to gather all the fundamental raw materials. These include the clay for fashioning in-
frastructure to hold the charge, bellows to pump air, and charcoal whose combustion
shadreck.chirikure@uct.ac.za

64 4 Domesticating Nature
was essential for initiating and sustaining endothermic reactions and reducing the
ore to metal. Suitable ore was indispensable, for its reduction produced metal. The
available evidence shows changing preferences in raw material selection as ancient
metallurgists became more and more skilled in metal production (Craddock 2000).
Successful reductive smelting demanded ores of sufficient quality which, as we
saw in Chap. 3, were sourced through various forms of mining. With respect to iron,
a wide variety of ores were worked across Africa, from hematite at Dekpassan-
ware in Togo during the Early Iron Age (c. 400 BC to AD 1000) (de Barros 2013),
through ilmenite-rich magnetite sands in Yorubaland (c. AD 1700–1850) (Ige and
Rehren 2003), to laterites in Nyanga Northeastern Zimbabwe (AD 1700–1850)
(Chirikure and Rehren 2004) (Fig. 4.1). In terms of copper, malachite, cuprite and
azurite were the most frequently used types across all ages from Wadi Dara in Egypt
c. 4000 BC (Craddock 2000) to Musina in Venda, South Africa, in the nineteenth
century. Sulfide ores were generally avoided in most of sub-Saharan Africa (Bisson
2000; Miller and Killick 2004), but were worked in Egypt and Nubia. Cassiterite
predominantly appears to be the only tin ore that was smelted in sub-Saharan Af-
rica and possibly Egypt (El Rahman et al. 2013; Heimann et al. 2010). Overall, the
different ores that were smelted and the constraints and opportunities which they
conferred to smelters often resulted in the development throughout the history of
metallurgy in Africa of localized and context-specific smelting techniques and reci-
pes (Chirikure and Bandama 2014).
Wood charcoal harvested and processed in a variety of ways acted as fuel whose
combustion supplied heat to spur chemothermal reactions in the furnaces. Charcoal
was produced through the incomplete incineration of wood (Horne 1982). Most of
what we know about charcoal comes from the ethnographic and historical records
because very little work has been done on the anthracology of smelting sites (Lyaya
2013; Mapunda 1995). In the twentieth century, the Dogon of Mali cut down dry
trees and burned them before covering the inferno with sand to create a reducing
environment (Huysecom and Augustoni 1997). This dry distillation of wood pro-
duced charcoal of good quality, and the method was also utilized by the Shona of
Zimbabwe and the Kaonde of Zambia in the recent past (Cline 1937), but methods
of producing charcoal in antiquity have not been explored in full. Wood sources had
to be managed because overharvesting resulted in environmental and ecological
degradation. The ratio of wood to charcoal in some West African communities was
10 to 1 (Goucher 1981), such that in the Mema region of ancient Ghana (modern
day Mali), intensive iron smelting in the late first millennium AD severely depleted
trees resulting in massive soil erosion (Haaland 1980), and there appears to have
been a reduction in forest cover around Begho in modern Ghana (Goucher 1981).
However, Njanja metal workers of Zimbabwe managed the woodlands through al-
ternating tree species, thereby avoiding upsetting the ecological balance as was the
case with Mema (Chirikure 2006). In Europe, intensive metal production was as-
sociated with negative ecological consequences, such as the depletion of forests
that stimulated woodland management techniques such as coppicing (Joosten et al.
1998).
shadreck.chirikure@uct.ac.za

65Raw Materials
Smelting required access to good clays for making charge receptacles and tuy-
eres (blow pipes) for supplying air into the furnaces. The types of clays used varied
from area to area, but in general, they had to be sufficiently refractory to maintain
mechanical integrity of the furnace during smelting and at the same time gradually
melt to contribute to slag formation (David et al. 1989; Martinón-Torres and Reh-
ren 2014). Furthermore, selected clays were supposed to keep furnaces insulating
enough to keep heat loss during reduction to a minimum (Crew 1991). It is possible
that crucible furnaces were used in the early stages of copper smelting in Egypt
and the Middle East (Craddock 2000; Hauptmann 2007). Smelting furnaces used
in precolonial Africa have been grouped into four broad classes: crucible, bowl,
low shaft, and high shaft types (Chirikure et al. 2009; Kense 1985; Miller and Van
der Merwe 1994; Van der Merwe 1980; see Fig. 4.2 for distribution). Crucible fur-
naces comprised ceramic vessels that were fired from inside using blowpipes and
were mostly used in ancient Egypt (Ogden 2000). In sub-Saharan Africa, crucible
furnaces were used for smelting copper at Kansanshi in Zambia between AD 1000
and 1200 (Bisson 2000). Bowl furnaces consisted of a pit without any protruding
superstructure, and low shaft and tall shaft furnaces stood to a height not exceeding
1, 5 and 7 m respectively (Chirikure et al. 2009). These are merely broad categories
and contain variations within each group.
The archaeology of sub-Saharan Africa suggests a broad developmental trajec-
tory from the non-slag tapping shaft and bowl furnaces used in the first millennium
BC to the introduction of natural draught furnaces by the late first millennium AD
(de Barros 2013; Robion-Brunner et al. 2013). The temperatures for reducing var-
ied ores in these equally diverse furnaces ranged between 1100 and 1200°C. The
furnaces that were worked when sub-Saharan African metallurgy began somewhere
in the first millennium BC are easily identifiable by slag blocks and were non-slag
tapping (Alpern 2005). Bowl furnaces consisted of a semicircular depression in
the ground lined with refractory materials. A variant of this type had superimposed
short shafts aimed at providing high volumes and better draught when compared
to the ordinary bowl type (Miller and Van der Merwe 1994). Bowl furnaces were
both slag tapping and non-slag tapping (Ackermann et al. 1999; Cline 1937). Slag
tapping is the process of draining the molten slag out of the furnace as the smelting
process unfolded and had the advantage that furnaces could be used continuously
and that more output could be obtained without the furnace getting full (Craddock
1995). In contrast, non-slag tapping furnaces had no such provision. In some de-
signs, the slag drifted down into a specially designed pit (slag pit furnaces) or was
raked out after the smelting was complete. The low shaft furnace type stood to
between one and one and half meters above the ground and the diameter at the base
too varied (Kense 1985; see David et al. 1989 for a 2.7-m-high exception). The
shaft acted as the combustion vessel and was insulating enough to promote heat
retention during smelting. Further distinctions have been made on these low shaft
furnaces between those that had a provision for slag tapping and those without this
feature (Van der Merwe 1980). The third group mostly consisted of high shaft fur-
naces that stood between 1.5 and 4 m above the ground. In contrast to the bowl and
low shaft varieties that were operated by forced draught, these huge furnaces were
shadreck.chirikure@uct.ac.za

66 4 Domesticating Nature
universally powered by natural draught (Chirikure et al. 2009; Kense 1985; Van der
Merwe 1980). Some natural draught furnaces used in West and Central Africa had
a provision for slag tapping (e.g., Killick 1990, 1991), while others were non-slag
tapping (Huysecom and Augustoni 1997). Archaeologically, it is difficult to distin-
guish between different types of shaft furnaces, but often natural draught furnaces
have tuyeres fused in multiples (Prendergast 1975).
Fig. 4.2 Approximate distribution of bowl, shaft and natural draught furnace types in Africa.
(From Chirikure et al. 2009)
shadreck.chirikure@uct.ac.za

67Brief Overviews: Metal Smelting in Preindustrial Africa
Slag tapping is seen as an advanced feature of smelting in Egypt from Rames-
side times. In Europe, it was mostly popular from Roman times onwards (Pleiner
2000). Within an evolutionary perspective, it appears that slag removal in Central
European furnaces passed through various stages. The first involved slag solidifi-
cation inside the furnaces, while the second involved slag pits that collected slag
underneath the furnaces. The third process of tapping appeared much later (Joosten
et al. 1998; Pleiner 2000).
Copper and perhaps tin were smelted in bowl and low shaft furnaces, employing
tapping and non-slag tapping technologies. These metals were rarely smelted in
natural draught furnaces because they are highly reducing, which may have pro-
duced unwanted alloys of copper and iron and tin and iron (Craddock and Meeks
1987; Killick 1991).
The air for initiating and sustaining combustion was generated either artificially
by blowing (in ancient Egypt) and pumping bellows (forced draft furnaces) or natu-
rally by letting the furnace act as a chimney, which drew in air using the principle of
convection (natural draught furnaces). In general, two types of bellows were used
in precolonial Africa: bag and pot/drum varieties (see Fig. 4.3). A variant of the pot
type known as concertina bellows was restricted to West Africa and was very effi-
cient in generating air (Cline 1937). The bellows were connected to clay tuyeres or
blow pipes, which directed air to the furnace. In Uganda, the Toro people gendered
the tuyeres as male (Childs 1998). Schmidt and Avery (1983) believe that preheat-
ing the air in the tuyeres raised temperatures resulting in the production of high-
carbon steels (see Rehder 1986 for alternative view). One of the earliest evidence of
bellows used in smelting comes from Egypt and Egyptian paintings that show drum
bellows from the third millennium BC onwards. At Meroe in the Sudan, Shinnie
(1985) excavated an iron smelting furnace (Furnace 5) dating to AD 300 with pots
used as bowls still in place. Elsewhere in Africa, bellows have rarely survived, par-
ticularly in cases where they were made using perishable materials.
In summary, the work carried out in different parts of Africa showed that sources
of raw materials may have changed from time to time just as the furnace types and
their method of operation were not static over time. Different scales of production
motivated for the development of new innovations such that the story of African
metallurgy is one of dynamism and change.
Brief Overviews: Metal Smelting in Preindustrial Africa—
Egypt, Nubia, North Africa and the Horn of Africa
Chronology of Egypt and Nubia The earliest metal smelting evidence in Africa
comes from Egypt. The very first Egyptian metal artifacts were recovered from
Neolithic settlements such as Badari situated south of modern Asyut in Middle
Egypt and were probably forged from native copper (Ogden 2000; Scheel 1989).
Most of the objects recovered from Badarian sites were isolated objects from burial
contexts. The evidence from Naqada, situated about 27 km (17 miles) north of
shadreck.chirikure@uct.ac.za

68 4 Domesticating Nature
modern Luxor, shows that during the predynastic (Table 4.1), between c. 4000 and
3000 BC, metal processing became more common during the chalcolithic Naqada
1–111 phases (Emery 1970; Scheel 1989). In the Naqada 11 and 111 phases (about
3500 to 3050 BC), the evidence for copper smelting became more common. Naqada
I1 culture spread over the entire Nile valley from north of Hierakonpolis into the
Delta, and Egypt was networked with the surrounding regions via long-distance
trade (Scheel 1989) (Figs. 4.4 and 4.5).
Fig. 4.3 Distribution of bellows across sub-Saharan Africa. (From Chirikure et al. 2009)
shadreck.chirikure@uct.ac.za

69Brief Overviews: Metal Smelting in Preindustrial Africa
The earliest stages in Egyptian extractive metallurgy are represented by a non-
slagging process where reduction took place in crucibles. This was followed by a
slagging process evident in the early third millennium ВС at smelting sites such as
Wadi Dara in the eastern desert of Egypt as well as somewhat further afield at Timna
and Faynan (Craddock 2000). From the beginning of the dynastic period in Egypt
at about 3050 BC, metal smelting techniques were continuously developed and re-
fined. With the centralization of the Egyptian administration and the formation of a
cultural centre at the royal capital, various professions and trades were established.
Fig. 4.4 Location of early
Egyptian and Nubian smelt-
ing sites. (After Scheel 1989)
Table 4.1 Chronology of ancient Egypt and Nubia. (After Scheel 1989)
Period Dates
Predynastic 5050–3050 BC
Old kingdom 2613–2181 BC
New kingdom 1570–1070 BC
Late period 713–332 BC
Ptolemaic period 332 BC–AD 395
shadreck.chirikure@uct.ac.za

70 4 Domesticating Nature
The first pictorial and inscriptional sources relating to the metalworker’s craft come
from Egyptian mastaba tombs at the beginning of the Old Kingdom (c. 2600 BC)
(Scheel 1989). The first scenes of metalworking were found in Giza, but more pic-
torials appear on the tombs of officials in all periods of Egyptian history. Pictorial,
inscriptional and archaeological sources, including the metal artifacts themselves,
together serve as the basis for our investigation into Egyptian extractive metallurgy.
The furnaces for smelting copper evolved over time. During the Old and Middle
Kingdom, smelting took place in crucibles appearing in the form of ceramic bowls
(Fig. 4.6). Bowl furnaces developed over time. In the course of the smelting pro-
cess, a mixture of crushed malachite and charcoal—the charge—was roasted and
reduced to small prills of rich copper ore embedded in a copper slag. The prills
were extracted by crushing the slag and then were melted together to form copper
ingots. In Ramesside times, sophisticated pot bellow-driven shaft furnaces made of
bricks were used, which enabled higher temperatures of about 1200°C to be reached
(Fig. 4.7). These brick-built furnaces were fitted with a tap hole for draining slag
(Scheel 1989). In the course of the smelting process, the copper droplets sank to the
furnace bottom, forming an ingot. At the end of the process, the lighter slag above
the smelted metal could be tapped into a slag pit through the tap hole in the furnace
wall. Afterwards, the copper ingot was removed from the bottom of the furnace.
The methods of air supply also changed with time, from the use of foliage in the pre-
dynastic period to the use of simple mouth blow piece made of reed and tipped with
clay (Scheel 1989). During the Middle Kingdom, skin bellows of a goat or gazelle
actuated by feet and cords were used with drum, pot and dish bellows.
Fig. 4.5 Location of sites with smelting evidence in North Africa, Egypt, Nubia and Ethiopia
shadreck.chirikure@uct.ac.za

71Brief Overviews: Metal Smelting in Preindustrial Africa
Although meteoric iron is known in Egypt from c. 3000 BC, the actual reductive
smelting of ores to produce usable iron was very late. Unlike other Middle Eastern
areas which adopted iron after 1500 BC, the earliest indications of iron smelting in
Egypt were found in the delta region, particularly at the sites of Naucratis (Figs. 4.4,
4.5) and Defenna where archaeologists recovered a large quantity of iron slag and
some ore. The site of Naucratis can be dated to about 580 BC. The iron was worked
by Greek and Carian mercenaries after the expulsion of the Kushites (Arkell 1966;
El Rahman et al. 2013). Iron was also smelted in the central eastern desert at Wadi
Abu Gerida as shown by the presence of furnaces and olivine-rich slags dating to
Ptolemaic period (El Rahman et al. 2013; Ogden 2000).
In lower Nubia, the region between the First and Second Cataracts, copper and
gold items first appeared in graves of the Middle A-Group, which are dated from ca.
3600–3300 cal BCE and by 3000 cal BC copper beads, awls, and pins were reach-
ing as far south as the Third Cataract (Edwards 2004). The earliest evidence of the
production of metals in Nubia is from Old Kingdom context (ca. 2600 BC) at Buhen
(Emery 1963) (Fig. 4.4) and within the temple precinct further upstream at Kerma,
in contexts dated by radiocarbon to 2200–2000 cal BC (Bonnet 1986). At Buhen,
the copper carbonate malachite was smelted to metallic copper using charcoal from
acacia trees. The later Kerma furnace is a rectangular platform, originally covered
by a vault, and heated from below. This was used for melting bronze in crucibles.
Fig. 4.6 An Old Kingdom pictorial showing six smelters blowing into two crucibles (Mastaba of
Mereruka, fifth Dynasty). (Redrawn from Duel 1938, plate 30)
shadreck.chirikure@uct.ac.za

72 4 Domesticating Nature
No evidence of smelting has yet been found at Kerma, and the sources of the copper
and tin used are unknown (Killick 2014).
Like in Egypt, iron appeared comparatively late in Nubia with the earliest date
being a radiocarbon date of 514 + /− 73 BC (Shinnie 1985, p. 30). Meroe, the capital
of the kingdom of Kush, yielded huge quantities of smelting slag, slag, possible
smithing hearths, and tuyeres (Rehren 2001). Iron production continued until the
early first millennium AD when Meroe was abandoned. Meroitic furnaces were
constructed of fired brick, and Shinnie (1985) found good evidence for the nature of
the bellows when he exposed pot cylinders associated with furnace F5 (Fig. 4.8). It
appears that Meroitic furnaces were derived from Egypt and were introduced by the
Romans. Meroitic iron smelting was large scale (Fig. 4.9) and may have produced
more than 5000 t of iron, which is clearly beyond local needs (Rehren 2001).
Fig. 4.7 Bowl and shaft furnaces used in dynastic Egypt. (Redrawn from Scheel 1989, p. 16,
Figs. 8 and 9)
shadreck.chirikure@uct.ac.za

73Brief Overviews: Metal Smelting in Preindustrial Africa
Fig. 4.9 Large iron slag mound at Meroe. (Source: Jane Humpris)
Fig. 4.8 Furnace F5 excavated by Shinnie (1985) at Meroe. (Redrawn from Shinnie 1985, p. 33)
shadreck.chirikure@uct.ac.za

74 4 Domesticating Nature
In Ethiopia and the Horn of Africa, gold and copper appeared in the last century
BC. The available evidence suggests that iron and copper metallurgy was estab-
lished by c. 500 BC, possibly as a result of influence from either via Egypt and
Nubia or via Arabia. The production of copper and iron was typically through the
bloomery process, and Severin et al. (2011) studied the slags from Aksum dating to
AD 300 and suggested that some of them were from iron smelting. The extractive
metallurgy of North Africa is closely related to that of Egypt and the Mediterranean.
In Morocco, the metallurgy closely mirrors that of the adjacent Spain with the ob-
ject forms resembling those used in the European Bronze Age (Alpern 2005). This
shows that different areas were networked and were thus not closed to each other.
West Africa
West Africa (Fig. 4.10) possesses one of the longest records of metal production in
Africa south of the Sahara, excluding Nubia. Starting with the ethnographic period,
a lot is known about the production of metal in West Africa (Cline 1937). Huy-
secom and Augustoni (1997) reconstructed iron smelting in natural draft furnaces
by the Dogon of Mali encompassing the entire chaîne opératoire from raw material
selection through smelting to actual forging. Similar reconstructions were also per-
formed at Banjeli in Togo (Goucher and Herbert 1996). Archaeologically, detailed
work was carried out in a number of areas in West Africa, which enhanced our
understanding of the technology of the process within a diachronic perspective (see
for example, de Barros 2013; Eze-Uzomaka 2013).
Fig. 4.10 Smelting sites in West Africa
shadreck.chirikure@uct.ac.za

75West Africa
De Barros (1986, 2003, 2013) carried out regional surveys in the Bassar region
of Togo. In the process, he explored the development of iron production from the
Early Iron Age (c. 500 BC to AD 1000) until the recent periods. The most interesting
evidence comes from the multicomponent site of Dekpassanware in Bassar, Togo.
Dekpassanware yielded an impressive record of smelting; often punctuated by dis-
continuities in the stratigraphy between c. 500 BC until recent times (Fig. 4.10). Ac-
cording to de Barros (2013), from c. 500 BC up to sometime in the mid-to late first
millennium AD, iron smelting furnaces used at Dekpassanware were very small and
were possibly powered by bellows. But from the late first millennium AD, smelting
took place in natural draught furnaces (Fig. 4.11), which consumed large quantities
of ore, clay and labour. As the second millennium AD unfolded, the formation of
large-scale polities was associated with large-scale iron production and was charac-
terized by division of labour with some villages specializing in ore preparation, oth-
ers in smelting and some in smithing (de Barros 1986). Iron production in regions
such as Banjeli and Bassar peaked up to reach up to 80,000 cubic m (Fig. 4.12),
which surpasses production at places such as Meroe (de Barros 1986). No detailed
studies of Early and Late Iron Age slags from Bassar have been carried to date, but
analyses by Goucher indicated that fifteenth-and sixteenth-century AD slags from
this area were typical bloomery waste products, while preliminary metallographic
analyses of objects showed them to be made of low-carbon steels also consistent
with bloomery process (de Barros 2013).
Fig. 4.11 Second-millennium AD furnaces used in Bassar, Togo. (Source: Philip de Barros)
shadreck.chirikure@uct.ac.za

76 4 Domesticating Nature
From the late 1990s, a team led by Huysecom initiated a long-term study of
iron production in the Dogon region of Mali (Robion-Brunner et al. 2013). The
team was able to establish five iron smelting traditions characterized by different
furnace types and scales of production within this limited area. Natural draught
furnaces in this area possibly developed in the second half of the first millennium
AD. The scale of production was also very high, particularly from the second mil-
lennium AD, and is comparable with that of places such as Banjeli in Togo, which
were associated with increasing political centralisation. Robion-Brunner et al.
(2013) speculate that the caste system may have developed in the late second mil-
lennium AD in this region.
Research in various parts of Nigeria unearthed important data which documents
the development of iron working in this part of sub-Saharan Africa. In Nsukka,
Okafor (1993) chronicled the evolution of iron smelting at places such as Opi dat-
ing from c. 600 BC to the late first millennium AD (Fig. 4.1). During this Early Iron
Age, the furnaces were non-slag tapping. In the Late Iron Age (AD 1000–1800), tap-
ping furnaces were used with the result that they increased output from the furnaces
through reduction efficiency. This observation is supported by microstructural stud-
ies of slag from the two periods, which indicated that Late Iron Age slags had less
residual iron oxide when compared to those produced in Early Iron Age furnaces
(Okafor 1993). The tapped slag from Opi had macroscopically visible tap lines,
while microscopically a series of magnetite skins, which develop when hot slag is
exposed to cool air typically after a smelt, were detected. Elsewhere in Nigeria, Ige
Fig. 4.12 Second-millennium AD slag mounds at Bassar, Togo. (Source: Philip de Barros)
shadreck.chirikure@uct.ac.za

77West Africa
and Rehren (2003) discuss the smelting of ilmenite-rich magnetite sand by the Yo-
ruba from the eighteenth and nineteenth centuries. According to Killick and Miller
(2014), such a capability showcases the versatility of the bloomery process because
modern blast furnaces cannot smelt iron ores with more than 2 wt % titanium.
As these examples make clear, iron metallurgy was not static from its introduc-
tion in the first millennium BC up to the present. Sites such as Waldadé in Senegal
have produced iron tools and hint of iron production on two mounds producing
calibrated dates of 800–200 BC which fall in the radiocarbon black hole (Deme and
McIntosh 2006) (Fig. 4.1). Earliest evidence from Waldadé indicates transitional
use of iron between cal 800 and 550 BC, but the evidence shows that smelting was
fully established by 200 cal BC. Studies by Killick and others in the Middle Senegal
River identified many furnaces dating to the late first and early second millennium
AD (Fig. 4.13) In Mauretania, there is evidence that iron was smelted between
760 and 400 cal BC (MacDonald et al. 2009). These early Mauritanian furnaces
were non-slag tapping, supporting a trajectory from non-slag-tapping to tapping as
recorded in areas such as Bassar. One important observation based on examples of
iron production in West Africa mentioned here and elsewhere is the phenomenal
boom in scale of production from the late first and early second millennium AD
onwards. The volume of iron slags recorded by Robion-Brunner (et al. 2013) in the
Dogon region of Mali was estimated to be in excess of 50,000 m3 of slag, relatively
comparable with the equally staggering 80,000 m3 for Bassar and Banjeli areas of
Fig. 4.13 Remnants of a circular shaft furnace from an iron smelting furnace in Senegal, dated
between the twelveth and fourteenth centuries AD. The black material is slag that solidified within
the furnace. (Photo credit: David Killick)
shadreck.chirikure@uct.ac.za

78 4 Domesticating Nature
Togo. Several reasons may account for this, but there appears to be a correlation
between large-scale production and a boom in demography, mainly visible after AD
1400 (de Barros 1986); furthermore, expanding networks of connection in relation
to initially trans-Saharan trade and later the Atlantic-based trade (Stahl 2014a). It is
equally possible that some of the iron production may have served the demands of
the slave trade (MacEachern 1993).
In contrast to the relatively robust evidence for iron working, only a few cop-
per working localities in West Africa have been described in detail. Perhaps the
most well-known example of copper working in this region seems to be at Akjoujt
(850–300 BC) in Mauretania, which also represents the earliest copper south of the
Sahara outside Nubia (Lambert 1983; Wood house 1998). At Azelik in Niger, cop-
per objects were found in Late Stone Age and Iron Age contexts. The dating of the
earliest appearance of copper in this region is controversial, but the main indication
is that copper smelting may have started somewhere after c. 1000 BC (Killick et al.
1988). The earliest furnaces used for smelting copper were low shaft and non-slag
tapping (Bisson 2000, p. 88). Copper working in Niger continued well into the sec-
ond millennium AD, with interruptions in between, such that the Arab chronicler
Ibn Battuta recorded copper working in this area c. AD 1300 (Herbert 1984). In Ni-
geria, copper was also worked in the Benue valley, which was the source of copper
used in the castings at Igbo Ukwu in the late first and early second millennium AD
(Chikwendu et al. 1989). Because most West African regions lacked copper, trade
in this metal was the focus of interaction between communities on the two sides of
the Sahara (Fenn et al. 2009).
Central Africa
Central Africa (Fig. 4.14) provides a range of insightful examples of iron and cop-
per smelting in Africa. Warnier and Fowler’s (1979) study of the Cameroonian
Grassfields on the Bamenda Plateau provides one of the most staggering examples
of ninettenth-century large-scale iron production. Here the scale of iron production
implies mobilization of large labour inputs and quantities of ore and other raw ma-
terials resulting in a production that unmistakably was geared beyond local needs
just as Bassar and Dogon discussed above. This iron production resulted in the
accumulation of debris in a very short space of time estimated at 163,000 m3 and
therefore remains one of the largest examples of iron production documented in
Africa (see de Barros 1986).
Research in the Mandara region of Cameroon has yielded rich insights through
the work of first Renee Gardi and later Nic David and his Mandara archaeology
project. Gardi recorded ethnographic cases of iron production, illustrating furnace
types, elucidating embedded sociocultural metaphors and describing how the pro-
cess unfolded. While Gardi’s observations formed a useful archive of cultural ma-
terial, David et al. (1989) persuaded one of the last Mafa smelters Dokwaza to
re-enact iron smelting in a forced down draught furnace which stood to a height of
shadreck.chirikure@uct.ac.za

79Central Africa
2.7 m and was uniquely located to exploit the nature of the topography (Fig. 4.15).
Two pot cylinders with an attached sheep skin diaphragm were used to generate air,
which was directed vertically by a single tuyere into the furnace. Mafa smelting
exploited the much iron-rich magnetite sand panned from water courses by women
(Cline 1937). Over the course of smelting, the tuyere had to be trimmed of viscous
slag that intermittently blocked the air hole; slag was also removed from a series
of vent holes cut into the furnace wall at increasing heights. During the smelting,
the tuyere lost about a third of its length (0.5 m), contributing to slag formation.
After ten and half hours of nonstop smelting that consumed 82.3 kg of charcoal
and 18.0 kg of ore, 15.7 kg of iron bloom was recovered (David et al. 1989). The
carbon content of the bloom varied from 0.05 % carbon to cast iron of more than
4 % carbon. The cast iron was decarburized together with unsorted fragments of
other metallic products in an open crucible. The carbon content of the final product
ranged from 0.2 to 0.8 %, making it a low-to high-carbon steel. It is because of this
versatility and ability to produce cast iron that this Mafa iron smelting has been
described as a hybrid between the bloomery and blast furnace (David et al. 1989).
Other regions of Central Africa such as Gabon have also produced important
information regarding metal production (Clist 2013), but in the interest of regional
balance, it is more prudent to shift focus to the Democratic Republic of Congo and
adjacent areas. Most work in this region was carried out by de Maret and his stu-
dents (de Maret 1988). It is clear that iron metallurgy in this region appeared much
later than in Cameroon, Gabon and the Central African Republic (Clist 2013). The
Fig. 4.14 Location of Central African sites
shadreck.chirikure@uct.ac.za

80 4 Domesticating Nature
archaeological sites in the Upemba Depression yielded iron and copper objects in
the early first millennium AD. As with other parts of the continent, once established,
metallurgy developed over time, resulting in very rich diversity of furnace types
from slag tapping bowl furnaces to low shaft and natural draught furnaces.
Although the Democratic Republic of Congo, Angola and Congo-Brazzaville
have important iron production signatures, they also host significant evidence of
copper production dating from the early first millennium AD to the early twentieth
century. Furthermore, copper production in this area was studied within a diachron-
ic perspective that allows us to explore in detail the development of copper smelting
better than any other region in Africa (see Bisson 2000). Regions such as Lubum-
bashi (formerly Katanga) in the Democratic Republic of Congo (DRC), Bembe in
Angola, and the copper belt in Zambia host significant ethnographic and archaeo-
logical evidence of copper smelting (Herbert 1984). This broad region in the heart
Fig. 4.15 Cross section of Mafa down draught furnace. (From David et al. 1989, p. 188, Fig. 3)
shadreck.chirikure@uct.ac.za

81Central Africa
of Africa possesses a rich data set of ethnographic examples of copper smelting
and its associated sociological factors. According to Bisson (2000), from the nine-
teenth century onwards, European travelers documented copper smelting practices
by groups such as the Sanga, Yeke, and Luba communities who worked in one of
the largest copper deposits in the world around Lubumbashi. Ethnographically, the
Sanga copper smelters (Fig. 4.1) reduced malachite in shaft furnaces that were one
and three quarter meters tall and about a meter wide at the base. These bellows-
driven furnaces were molded out of clay from termite mounds and had a shallow
depression in the ground where molten metal collected. Upon the completion of the
smelt, the furnace rake channel was opened and the solidified metal was skilfully
removed without destroying the superstructure. The metal from the furnace was
refined in a secondary furnace powered by four pairs of bellows to ensure that high
level of heat was generated. Copper smelting was the preserve of certain kin groups,
although mining could be done by both men and women of all strata (Bisson 2000).
The Luba, another group in today’s DRC, had an extraordinary copper smelt-
ing technology that directly tapped molten metal into ingot molds as soon as the
reduced metal reached the furnace bottom (Bisson 2000; Herbert 1984). The shaft
furnace was connected to numerous but shallow X-shaped molds lined with ashes.
After each mold was filled, the smelting assistants opened other vents and re-orient-
ed the channel until all the molds were full and the furnace was empty, after which
the process was started all over. This semicontinuous process resulted in the pro-
duction of a significant number of crosses which were traded over wide distances.
Although the archaeological record of Central Africa is rich in evidence of cop-
per smelting from the first millennium AD, there has been little technical study of
remains from the production process. To date, the earliest evidence of copper smelt-
ing south the equator comes in the form of pieces of malachite, slag, broken pots,
crucibles, heavily vitrified and slag-encrusted tuyeres, and remnants of furnaces
from Naviundu Springs (Fig. 4.14) near Lubumbashi in DRC (de Maret 1982; Her-
bert 1984). The site dates to the third millennium AD. More evidence came from
archaeological excavations at Sanga near Lake Kisale in the Upemba Depression.
According to de Maret (1985), the Sanga burials represented three groups (1) An-
cient Kisalian (c. eighth to tenth centuries AD), (2) Classic Kisalian (c.eleventh to
fourteenth centuries AD) and (3) the Kabambian (c. fifteenth to eighteenth centuries
AD), which had evidence of copper working, mostly objects. The Sanga copper
funerary evidence revealed increasing social differentiation across the three periods
and validated Herbert (1984)’s observation that copper was ‘the red gold of Africa’
and thus the prestige metal of choice.
One of the most detailed archaeological examples of preindustrial copper min-
ing comes from Zambia at Kansanshi mine, which was worked from the early first
millennium AD up to fairly recent times (Bisson 2000). Copper smelting took place
away from the mine on a 150,000-square-meter-large site characterized by three
major phases of occupation. Archaeological excavations by Bisson (1976) revealed
that the earliest evidence for copper working designated Kansanshi Phase 1 dated
to AD440 + /-90. This phase yielded a crucible and a big block of slag. This was
followed by Phase 2 dating from the eighth to the tenth century AD. According
shadreck.chirikure@uct.ac.za

82 4 Domesticating Nature
to Bisson (2000, p. 140), one of the furnaces used during this phase had multiple
tuyeres, suggesting that it was likely powered by natural draught. This furnace dif-
fered from earlier and later ones, which had single-or two-tuyere pots indicating
that they were forced draught driven. The last Phase 3 was dated between AD 1200
and 1600. The smelting debris from this area has not yet been studied in detail,
but holds potential to yield insight into the operations of natural draught furnaces.
These furnaces are generally believed to be heavily reducing, and in copper, smelt-
ing would have reduced more iron, which was undesirable (Craddock and Meeks
1987). It is therefore tempting to speculate that the practice of smelting copper in
natural draught furnaces was abandoned for these technical reasons; however, only
empirical research can validate or refute this proposition. Nonetheless, the scale of
production at Kansanshi was considerable and easily reached 130 000 kg of slag for
the last phase showing production beyond localized needs (Bisson 2000).
East Africa
Mapunda (1995; 2003) combined oral traditions, ethnohistories, and archaeological
evidence to investigate iron production from AD 1600 up to the 1950s in Ufipa near
Lake Tanganyika, South-western Tanzania. His work identified three traditions: (1)
the pre-Bantu Katukutu or dwarf technology (c. sixteenth to eighteenth centuries
AD) found on the southeastern shore of Lake Tanganyika (Mapunda 2003, p. 72);
(2) the Malungu (tall furnaces), dating to the nineteenth and twentieth centuries AD,
and located mainly on the escarpment and the Fipa plateau, and (3) the Barongo-type
technology, dating to the nineteenth century and located along the lakeshore. These
technologies or traditions also extended into parts of neighboring regions in coun-
tries such as DRC, Malawi and Zambia. The Malungu tradition was associated with
the vintengwe furnaces used for refining blooms from the tall furnaces. Mapunda
(2003, p. 82) argues that the authors of the katukutu tradition were of hunter–gather
extraction known locally as mbonelakuti on the basis of linguistic, genetic and his-
torical evidence. This, argues Mapunda, indicates that iron working technologies
were not only limited to groups conventionally labeled Bantu(Fig. 4.16).
Lyaya et al. (2012) and Lyaya (2013) carried out archaeometallurgical studies
combining optical microscopy with compositional techniques to understand iron
production in Southern Tanzania, an area encompassing Fipa and Unyiha areas.
More interestingly, Lyaya studied the slags and tuyeres from the Malungu and vin-
tengwe furnaces to argue powerfully for a three-stage iron production process in
this part of Africa. The Malungu process produced a sintered matrix that was further
processed in the refining or vintengwe furnaces. The blooms from this second stage
were smithed in the third and final phase of processing. According to Lyaya et al.
(2012), this three-stage process differs from the two-stage (smelting and smithing)
process, which generally excluded refining furnaces and was recorded in numer-
ous African parts of sub-Saharan Africa. This indicates great variability in African
bloomery processes. Lyaya also found evidence that Fipa smelters could produce
shadreck.chirikure@uct.ac.za

83East Africa
cast iron and various grades of steel, further exposing the diversity of iron produc-
tion techniques across Africa.
Schmidt (1997) and collaborators carried out a detailed ethnoarchaeological
study of iron production in Buhaya, Northwestern Tanzania. These studies were
supplemented by nine experiments that were recorded with the aim of generating
analogues for interpreting the archaeological record. The Haya people roasted their
ore before smelting, used charcoal from the hard wood Mucwezi tree and inserted
their tuyeres deep into furnace bowl with the implication that air was pre-heated
in tuyeres before introduction to the furnace. These practices were combined to
produce steels of variable carbon content (Schmidt 1997, p. 106–110). Schmidt
and Avery (1978) argued that this preheating technology had a deep antiquity in the
Iron Age, extending to between c. 500 BC and 1000 AD. In particular, the sites of
KM1, KM2 and KM3 linked to the Kaiija shrine at Rugomora Mahe dating to c. 500
BC possess this evidence, though claims for preheating have been challenged (see
Rehder 1986). These Early Iron Age sites were made of clay bricks, while those in
the Later Iron Age were made of termite earth and slag blocks. The earliest smelting
technology in Buhaya was related to that recorded in the adjacent areas of Rwanda
and Burundi where Urewe pottery was recovered (Humphris and Iles 2013).
In the Interlacustrine region of East and Central Africa, van Noten (1985) re-
enacted iron smelting at Madi in Northeastern Democratic Republic of Congo and
the adjacent Gasara region of Rwanda. While the Madi experiment produced iron,
the Gasara one failed demonstrating how colonial processes interrupted knowledge
transmission. Archaeologically, Rwanda and Burundi have generated important sites
crucial for our understanding of the evolution of metallurgy. The Early Iron Age in
Fig. 4.16 Location of East African sites and groups
shadreck.chirikure@uct.ac.za

84 4 Domesticating Nature
this area dates to c. 500 BC and perhaps earlier (Humphris and Iles 2013). The typi-
cal Early Iron Age furnaces were made of either bricks or coils of refractory clays.
Interestingly, as observed in Buhaya Tanzania and other areas of Africa, van Noten
found a small pot buried beneath the furnace pit (Van Noten 1985, p. 104), suggest-
ing the use of medicines to neutralize malevolent forces. Some of the furnace bricks
were decorated with incisions and/or dimples. The evidence suggests that the super-
structure was broken to extract iron while slag accumulated in the pits. In the Late
Iron Age, slag was tapped into pits outside the furnaces. Van Noten (1985) specu-
lated that iron production in the Early Iron Age of Rwanda had strong ecological
consequences for the area of Kabuye was deforested. In the mid-second millennium
AD, increased demography and political centralization precipitated large-scale iron
smelting in Buganda and surrounding areas. In fact, concepts of power, fertility and
iron were integrated with ideas of smith-kings as was practiced in neighboring areas
of DRC, Rwanda and Tanzania (Humphris and Iles 2013, p. 59).
Iles and Martinón-Torres (2009) studied iron production by pastoralist groups
in Kenya and noted that they used bowl furnaces. The pastoral groups placed less
value on copper (Bisson 2000). Overall, the evidence for copper smelting is very
rare, but copper objects were recovered around Kalambo Falls in Tanzania and other
areas (Mapunda 2003).
Southern Africa
Southern Africa (Fig. 4.17) presents yet another story of diversity as far as the work-
ing of metals is concerned. A number of studies aimed at understanding iron produc-
tion were carried out in different parts of the subcontinent. Dewey (1991) organized
re-enactments of Njanja iron smelting at Ranga in Central Zimbabwe. He persuaded
the descendants of Zinwamhanga, the head smelter who features prominently in
MacKenzie’s (1975) study to re-enact traditional iron smelting. Zinwamhanga had
also carried out a series of re-enactments in low shaft furnaces decorated with fe-
male anatomical features at the Museum of Human Sciences (formerly Queen Vic-
toria Memorial Museum). Dewey’s film of the re-enactment shows Njanja women
singing and dancing together with men to entertain and perhaps help bellows opera-
tors to maintain their rhythm. While Njanja is historically depicted as a center for
specialized iron production, not many large mounds of slag comparable with those
recorded in West and Central Africa were recorded.
Chirikure (2006) carried out interviews with Ranga people and carried out ar-
chaeological work around Hwedza Mountains, but failed to locate large mounds
comparable with those recorded in West and Central Africa. His study of Njanja
smelting remains documented a very efficient reduction technology which left little
residual iron oxide in the slag.
Maggs (1992) discusses a very interesting story of large-scale iron production us-
ing bowl furnaces within the context of the nineteenth-century historical Zulu state.
According to Maggs (1992), Shaka Zulu reduced the once-independent specialist
shadreck.chirikure@uct.ac.za

85Southern Africa
metal producers into dependent specialists who made iron to meet the requirements
of his ever swelling army. These specialists produced iron in small bowl furnaces
that fully equipped Shaka’s marauding armies numbering in excess of 50,000. In-
terestingly, although field research may alter this supposition, no large-scale slag
mounds have been noted in this area, consistent with the oral evidence that substan-
tive production was achieved through many small-scale installations. The implica-
tion for global archaeology is that substantive outputs (large-scale production) do
not necessarily correlate with concentrated production centers. Production can still
be dispersed but still specialized in orientation.
There is evidence that, once established in Southern Africa, iron production de-
veloped in both versatility and complexity. It seems that bowl furnaces and shaft
furnaces were used north and south of the Limpopo. Bowl furnaces were recorded
in the Early Iron Age of KwaZulu-Natal (between AD 300 and 1000) and after-
wards (Fig. 4.18). In other areas, such as Phalaborwa, different types of shaft fur-
naces were used in the Early and Late Iron Ages, showing the development of iron
and copper smelting. While it is notoriously difficult to use furnaces as a proxy
for group identity, Miller et al. (2001) present interesting results which associated
various groups in the Northern Lowveld with specific types of furnaces particularly
after AD 1400. The Venda were linked with cylindrical furnaces, while the Phaba-
lorwa are associated with the triangular furnace. However, there are many dangers
associated with conflating recent observations into the deep past.
Fig. 4.17 Location of Southern African sites and groups
shadreck.chirikure@uct.ac.za

86 4 Domesticating Nature
Ndoro (1994) studied iron smelting remains recovered at Chigaramboni near
Great Zimbabwe and concluded that, because tuyeres fused in multiples, the furnac-
es were powered by natural draught. No scientific studies have been performed on
this material, but it is possible that this site may have supplied iron to the inhabitants
of Great Zimbabwe (Ndoro 1994). Prendergast (1975) excavated a furnace with
tuyeres fused in multiples dating to the fourteenth and fifteenth centuries near Dar-
wendale in Zimbabwe. This is one of the earliest known natural draught furnaces
south of the Zambezi. In the Tswapong Hills of Botswana, there exist tuyeres fused
in multiples, which motivated Kiyaga-Mlindwa (1993) to argue that these late first
millennium AD sites were naturally draught driven, but more work may be required
to substantiate this. Interestingly, ethnographic evidence of natural draught furnaces
is mostly restricted to areas north of the Zambezi, suggesting complex patterns of
continuity and change in need of more substantive investigation.
Owing to the work of Miller and others before and after him, Southern Africa has
witnessed a number of archaeometallurgical studies that enhanced our understand-
ing of iron, copper and bronze technology as reconstructed from the production re-
mains. The work of Miller and Killick (2004) around Phalaborwa identified a very
versatile technology, which thrived on exploiting very high-grade but titanium-rich
magnetite. It is possible that sand was added as a flux to reduce these high-grade
ores. Chirikure and Rehren (2006) analyzed the archaeometallurgical remains from
Fig. 4.18 Twin-bowl furnaces used in the Later Iron Age (c. AD 1700 onward) of KwaZulu-Natal,
South Africa. (Photo credit: Tim Maggs)
shadreck.chirikure@uct.ac.za

87Anthropology of Smelting
Swart Village near Mt Darwin in Northern Zimbabwe. They too documented a very
efficient iron reduction practice in the region.
Copper smelting is less well investigated in part because it is difficult to dis-
tinguish copper from iron smelting slags visually and in some cases chemically
(Miller and Killick 2004). One of the most detailed works on archaeological copper
smelting slags from Southern African sites was carried out by Thondhlana (2012)
at Phalaborwa as part of a project directed by the present author. Thondhlana exca-
vated a midden and a slag mound at the site of Shankare and identified both cop-
per and iron smelting dating to Kgopolwe (c. 1100 to 1300) and Letaba (c. 1700
to 1900). Archaeometallurgical analyses revealed that while iron smelting slags
from the site were enriched in titanium, copper slags lacked this element. Further-
more, in contrast to the wustite-dominated iron smelting slags, the copper slags
had magnetite spinels and small copper prills. The copper from Shankare furnaces
was melted in ceramic crucibles visually indistinguishable from normal pottery.
Petrographic work on the crucibles revealed that some of them were tempered with
slag inclusions identifiable through the presence of wustite and the olivine fayalite
(Thondhlana 2012). Based on these interesting results, ongoing work as part of
Abigail Moffet’s PhD thesis at the University of Cape Town is exploring similari-
ties and differences between Kgopolwe and Letaba iron and copper smelting at the
site to determine whether the two metals were associated with similar or different
technological styles. Without imposing any relationships, cultural or otherwise, slag
tempering of pottery has been noted at archaeological sites of various periods in
Banda, modern day Ghana (Stahl pers comm 2014).
Anthropology of Smelting
One of the most important topics in African metal smelting relates to the symbolic
and cultural aspects of the process, which clearly were not only restricted to this
process but extended to broader society (Chirikure 2007). A number of important
works, for example, Herbert (1993), have articulated the inextricable nature of pro-
cesses such as iron smelting and rituals of power, gender and transformation across
sub-Saharan Africa. In fact, as a heat-mediated routine, smelting transforms objects
of nature into culture. This act of transformation is often conceived to be a dan-
gerous process, with implications for the social position of smelters which varied
geographically and through time. In some societies, particularly as documented in
the ethnography of Central Africa, smelters were held in awe, while in others such
as Ethiopia, they were despised. Smelters in Mali and Ethiopia formed an endoga-
mous caste that only married potters and were of a lowly social position (Haaland
2004; Tamari 1991). At the same time, smelters and smiths could occupy privileged
positions through which they accumulated wealth in Central Africa (de Maret 1985).
Regardless of variation, smelters were associated with beliefs in magic and deities
across the continent but in different contexts. In light of the power of evil spirits
to influence smelts, smelters were motivated to place medicines beneath furnaces.
shadreck.chirikure@uct.ac.za

88 4 Domesticating Nature
As such, some furnaces in many parts of Africa, from the Grasslands of Camer-
oon, through Central Africa to Eastern and Southern Africa, both ethnographic and
archaeological, had medicine holes strategically placed underneath (Killick 2014;
Mapunda 1995; Rowlands and Warnier 1993; Schmidt and Mapunda 1995).
One of the most important themes in the anthropology of iron smelting, par-
ticularly in Bantu Africa, is the metaphoric association between reduction and hu-
man reproduction and the implications for spatial location of smelting in relation
to settlements (Ndoro 1991). Iron smelting furnaces in East and Southern Africa
are often decorated with female anatomical features such as breasts (Figs. 4.19 and
4.20). Some ethnographies argue that, because of its link with reproduction, smelt-
ing took place in secluded areas away from settlements (Van der Merwe and Avery
1987). This was partly meant to enforce sexual abstinence, for intercourse with
their real wives would be tantamount to adultery which gestated failed smelts (Her-
bert 1993). Indeed, there are some furnaces such as those in the Matopos region of
Southwestern Zimbabwe and Nyanga which were located away from settlements
(Soper 2002). However, most of the studies on this topic have lacked some nuance
and variation because Hatton (1967) has reported a case in nineteenth-century Zim-
babwe where furnaces were often located in homesteads with women and children
watching. Hatton (1967, p. 39) argues that ‘often, a wife would help her husband
by pumping the bellows’. This variation contradicts some ethnography that sees no
role for women in smelting. The Kalanga case described by Hatton (1967) exem-
plifies improvisation in preindustrial African metalworking (cf. Stahl 2014b). This
improvisation enabled communities to transcend the usual and is often missed by
an adherence to sweeping generalisations. This tempers the view that in Southern
and Eastern Africa from the beginning of metallurgy until recent times, smelting
was always carried outside settlements, regardless of time and place because of its
association with metaphors of reproduction and concepts of pollution.
In fact, there are numerous areas in the recent past where smelting was carried
out either in relatively secluded areas within the centre of villages or in full view
of everybody (MacKenzie 1975; Chirikure unpublished field notes). My own field
research carried out in the Njanja area reached concordance with Hatton’s obser-
vations in the Matopos. My informant Headman Ranga categorically stated that
smelting was carried out in the villages because the labour of women and children
was important in times of high demand. Furthermore, Njanja women also sang and
danced to help bellow workers maintain rhythm (Dewey 1991). Archaeologically,
there are well-documented examples where smelting took place near or within vil-
lages as well as outside villages from the distant to the near past. Generally, the
indicators of smelting are partially reduced ore, flow slag, remnants of furnaces, and
vitrified tuyeres (Miller and Killick 2004). These cannot be mistaken for smithing
for the two processes are outwardly different (Bachmann 1982; Serneels and Perret
2003).
An intact furnace surrounded by a heap of slag mixed with broken tuyeres, vitri-
fied furnace wall and partially reduced ore was found at a nineteenth-century site
in upland Nyanga, Eastern Zimbabwe, in association with a decorated furnace with
breasts and a waist belt (Fig. 4.19) (Chirikure and Rehren 2004; Soper 2002). This
shadreck.chirikure@uct.ac.za

89Anthropology of Smelting
furnace was in a low enclosure within the homestead precinct as indicated by pit
structures and house floors (Chirikure and Rehren 2004). The ceramics found on
the metal smelting area of the settlement were the same as those from the house
floors indicating some contemporaneity. This example demonstrates that metaphors
of reproduction were important in this smelting as shown by the elaborately deco-
rated furnace (Fig. 4.19). Indeed, the low enclosure may have had a practical func-
tion of demarcating space, but equally it may have been part of a structure which
shielded smelters from general view and was therefore an expression of symbolic
Fig. 4.19 Anthropomorphic low shaft iron smelting furnace from Nyanga, Eastern Zimbabwe.
Note the molded breasts, navel and waist belt for enhancing fertility. (Photo credit: Author)
shadreck.chirikure@uct.ac.za

90 4 Domesticating Nature
seclusion. Many other nineteenth-century furnaces in lowland Nyanga were located
away from homesteads (Bernhard 1962; Chirikure and Rehren 2004). This varia-
tion in the spatial location of smelting in relation to settlements shows that it is not
the physicality of the location that matters, but an adherence to cultural principles
and values. As such, it may not have mattered whether a furnace was contiguous
to, or far from a settlement, at the time as smelters as actors were always producing
and reproducing ideas prevalent in society. Archaeological evidence therefore has
great potential to illuminate not only various trajectories of the spatial and symbolic
association between smelting and settlements but also temporal improvisation and
change in those ideas and values (see Stahl 2014b).
This thinking is further substantiated by the observations made by Bent (1896)
who recorded items of material culture such as houses, granaries, drums and head
rests decorated with anatomical features such as breasts, navels and other fertility
iconography (Fig. 4.21). All these were within the context of the homestead, and
the decorations resembled those on furnaces situated away from settlements. If fer-
tility symbolism was so important in determining the remote location of furnaces
together with the associated taboos, why then were houses and other household
material culture similarly decorated? (see also Collett 1993) Instead, we should con-
sider metal smelting as an integral element of society which also produced and re-
produced ideas that pervaded society such as fertility. Therefore, whether smelting
was within a village or outside may be immaterial; what is material is that the pro-
cess produced and reproduced ideas associated with reproduction, witchcraft and so
on. Equally, we should recognize that insights from ethnographically documented
Fig. 4.20 Decorated furnaces, one depicting a woman giving birth. Redrawn from furnaces on
display at Natural History Museum, Bulawayo
shadreck.chirikure@uct.ac.za

91Anthropology of Smelting
practices of the recent past may not provide a reliable guide to past variation. At the
same time, it should be noted that it was not only cultural factors that determined
the location of furnaces but also more practical ones such as availability of ore and
fuel and also labour (Maggs 1982).
Archaeological evidence provides a key means for probing continuities and
change in practice, including the relationship between smelting and residential
space. Schmidt (1997) excavated Early Iron Age sites in Buhaya, Tanzania, where
house floors contiguous to iron smelting furnaces were found at both Rugomora
Mahe and KM sites in Northwestern Tanzania. Schmidt (1997, p. 17) categorically
stated that ‘I want to be cautious that a false dichotomy is not drawn between settled
life and iron production. Such a dichotomy appears to be based partly on the hidden
inference that iron smelting would have been conducted outside village precincts,
an idea that is drawn from inappropriate and incomplete ethnographic analogy:
that iron smelting is always conducted in secret outside of villages. In fact, iron
smelting was sometimes conducted within village precincts (cf. Killick 1990) and in
Fig. 4.21 Anthropomorphic drum, granary and iron smelting furnace from Bent (1896)
shadreck.chirikure@uct.ac.za

92 4 Domesticating Nature
the case of recent Haya smelting, in a zone contiguous to village precincts…. Simple
logic also informs us that people must have lived nearby their places of work’.
Indeed, Haaland (1994) also excavated iron smelting debris within the center of
a settlement at Dakawa in Tanzania. In Southern Africa, sites such as Swart Village
have yielded smelting evidence from the center and precincts of the village together
with many more other examples. It is now widely accepted that the Early Iron Age
in Eastern and Southern Africa is closely related (Huffman 2007; Phillipson 2005),
such that it is reasonable to infer that we are dealing with ancestral Bantu peoples.
That smelting within villages has been documented at related sites in the north (East
Africa) and in some places such as Swart Village in Northern Zimbabwe should be
enough evidence to warn archaeologists against the ‘tyranny’ of selective use of the
ethnography (Wobst 1978). The inference does not stand that if smelting took place
in villages, then it was not associated with metaphors of reproduction. Equally, it
is inappropriate to deny any involvement of women in smelting. They cooked and
brought food to the smelters; they often participated in mining, charcoal preparation
and in some cases even took part in smelting (Dewey 1991). In Lubumbashi, one
woman was very famous for leading delegations that included men who extracted
copper ore (Bisson 2000). This variation indicates that there is no reason to select
only one aspect of the ethnography and extrapolate it to the deeper past to the ex-
clusion of other patterns that may—or may not—be documented in ethnographic
sources.
In ancient Egypt, it has been argued that women were generally spared smelt-
ing and metalworking because they are heavy tasks (Scheel 1989). However, most
smelters were linked to temples, and deities performed an important role in ensuring
success.
Conclusion
In conclusion, not only is Africa characterized by a great diversity in metals worked
across regions, but evidence suggests that methods developed in different ways
between the various regions. Egypt, Nubia, North Africa and the Horn of Africa
seem to mirror a similar trajectory. Egypt’s copper working history demonstrates
diverse approaches, from use of crucibles to shaft furnaces, and introducing air
through simple blowing by mouth to use of varieties of pot bellows. In contrast,
West, Central, East and Southern Africa show greater diversity and innovation in
iron working with furnaces appearing in different forms. The scale of production
clearly increased through time with varying output in West, East and Central Africa.
The output in areas of Central Africa even surpasses that of Meroe by rough calcula-
tion (Warnier and Fowler 1979). It seems that demography played an important role
in increased scale because Southern Africa appears to have relatively low popula-
tions when compared to East and West Africa. As such, no large-scale mounds of
slag have been reported even in landscapes historically associated with specializa-
tion such as Njanja. The variation in Africa’s iron smelting furnaces is matched by
shadreck.chirikure@uct.ac.za

93References
the products which range from soft iron, to cast iron and steel, which can only be
documented through detailed metallurgical analyses.
Besides this technological diversity, metal smelting was associated with rituals,
deities and beliefs. It is absolutely critical to avoid ‘selective use of the ethnogra-
phy’ (Lane 2005) by searching for macro-and micro-variation. For example, women
may not have played a direct role in the actual process of smelting in some com-
munities (Hatton 1967), but they prepared food for smelters and in some cases even
sang during smelting (Dewey 1991). The spatial location of smelting varied and
was determined by diverse variables including location of ore, fuel, clay and so on.
Wherever smelting took place, it produced and reproduced ideas in society; thus,
we should avoid assuming that smelting that took place outside villages was more
ritually charged than smelting practiced within villages.
References
Ackerman, K. J., Killick, D. J., Herbert, E. W., & Kriger, C. (1999). A study of iron smelting at
Lopanzo, Equateur Province, Zaıre. Journal of Archaeological Science, 26(8), 1135–1143.
Alpern, S. B. (2005) Did they or didn’t they invent it? Iron in sub-Saharan Africa. History in
Africa, 32,41–94.
Arkell, A. J. (1966). The iron age in the Sudan. Current Anthropology, 7(4), 451–452.
Bachmann, H. (1982). The Identification of slags from archaeological sites. Occasional Publica-
tion no 6. London: Institute of Archaeology.
Bayley, J., & Rehren, T. (2007). Towards a functional and typological classification of crucibles. In
S. La Niece, D. Hook, & P. Craddock (Eds.), Metals and mines: Studies in archaeometallurgy
(pp. 46–55). London: Archetype.
Bent, J. T. (1896). The ruined cities of Mashonaland: Being a record of excavation and exploration
in 1891. London: Longmans, Green, and Co.
Bernhard, F. O. (1962). Two types of iron smelting furnaces on Ziwa Farm (Inyanga). South
African Archaeological Bulletin, 17, 235–236.
Bisson, M. S. (1976). The prehistoric copper mines of Zambia. Ann Arbor, Michigan: University
Microfilms International.
Bisson, M. (2000). Precolonial copper metallurgy: sociopolitical context. In M. Bisson, P. de Bar-
ros, T. C. Childs, & A. F. C. Holl (Eds.), Ancient African metallurgy: The sociocultural context
(pp. 83–146). Walnut Creek: AltaMira.
Bonnet, C. (1986). Un atelier de bronziers à Kerma. In M. Krause (Ed.), Nubische Studien: Ta-
gungsakten der 5. Internationalen Konferenz der International Society for Nubian Studies, Hei-
delberg, 22.-25. September 1982 (pp. 19-23). Mainz am Rhein: von Zabern.
Chikwendu, V. E., Craddock, P. T., Farquhar, R. M., Shaw, T., & Umeji, A. C. (1989). Nigerian
sources of copper, lead and tin for the Igbo-Ukwu bronzes. Archaeometry, 31(1), 27–36.
Childs, S. T. (1998). ‘Find the ekijunjumira’: Iron mine discovery, ownership and power among
the Toro of Uganda. In A. B. Knapp, V. C. Piggott, & E. W. Herbert (Eds.), Social approaches
to an industrial past: the archaeology and anthropology of mining, (pp. 123–137). Abingdon:
Routledge.
Chirikure, S. (2005). Iron production in Iron Age Zimbabwe: stagnation or innovation. Unpub-
lished Doctoral dissertation, University College London.
Chirikure, S. (2006). New light on njanja iron working: Towards a systematic encounter between
ethnohistory and archaeometallurgy. South African Archaeological Bulletin, 61, 142–151.
Chirikure, S. (2007). Metals in society: Iron production and its position in Iron Age communities
of Southern Africa. Journal of Social Archaeology, 7(1), 72–100.
shadreck.chirikure@uct.ac.za

94 4 Domesticating Nature
Chirikure, S., & Bandama, F. (2014). Indigenous African furnace types and slag composition—Is
there a correlation? Archaeometry, 56(2), 296–312.
Chirikure, S., & Rehren, T. (2004). Ores, furnaces, slags, and prehistoric societies: aspects of iron
working in the nyanga agricultural complex, AD 1300–1900. African Archaeological Review,
21(3), 135–152.
Chirikure, S., & Rehren, T. (2006). Iron smelting in pre-colonial Zimbabwe: Evidence for dia-
chronic change from Swart village and Baranda, Northern Zimbabwe. Journal of African Ar-
chaeology, 4, 37–54.
Chirikure, S., Burrett, R., & Heimann, R. B. (2009). Beyond furnaces and slags: a review study of
bellows and their role in indigenous African metallurgical processes. Azania: Archaeological
Research in Africa, 44(2), 195–215.
Cline, W. B. (1937). Mining and metallurgy in Negro Africa (No. 5). Wisconsin: George Banta.
Clist, B. (2013). Our iron smelting 14C dates from Central Africa: From plain appointment to a
full blown relationship. In J. Humpris and Th. Rehren (Eds.), The world of iron (pp. 22–28).
London: Archetype.
Collett, D. P. (1993). Metaphors and representations associated with precolonial iron- smelting
in eastern and southern Africa. In T. Shaw, P. Sinclair, B. Andah, & A. Okpoko (Eds.), The
archaeology of Africa: food, metals and towns (pp. 499–511). London: Routledge.
Craddock, P. T. (1995). Early metal mining and production. Washington, DC: Smithsonian Uni-
versity Press.
Craddock, P. T. (2000). From hearth to furnace: Evidences for the earliest metal smelting technolo-
gies in the Eastern Mediterranean. Paléorient, 26(2), 151–165.
Craddock, P. T., & Meeks, N. D. (1987). Iron in ancient copper. Archaeometry, 29(2), 187–204.
Crew, P. (1991). The experimental production of prehistoric bar iron. Historical Metallurgy, 25(1),
21–36.
David, N., Heimann, R., Killick, D., & Wayman, M. (1989). Between bloomery and blast furnace:
Mafa iron-smelting technology in North Cameroon. African Archaeological Review, 7(1),
183–208.
de Barros, P. (1986). Bassar: A quantified, chronologically controlled, regional approach to a tra-
ditional iron production centre in West Africa. Africa, 56(2), 152–174.
de Barros, P. (1988). Societal repercussions of the rise of large-scale traditional iron production: A
West African example. African Archaeological Review, 6(1), 91–113.
de Barros, P. (2013) A comparison of early and later iron age societies in the Bassar region of Togo.
In J. Humpris & Th. Rehren (Eds.), The world of iron (pp. 10–21). London: Archetype.
De Maret, P. (1982). New survey of archaeological research and dates for West-Central and North-
Central Africa. Journal of African History, 23, 1–15.
de Maret, P. (1985). The smith’s myth and the origin of leadership in Central Africa. In R. Haaland
& P. Shinnie (Eds.), African iron working—ancient and traditional (pp. 73–87). Oslo: Norwe-
gian University Press.
Deme, A., & McIntosh, S. K. (2006). Excavations at Walaldé: New light on the settlement of the
middle senegal valley by iron-using peoples. Journal of African Archaeology, 4(2), 317–347.
Dewey, W. J. (1991). Weapons for the ancestors. Des Moines, University of Iowa, Department of
Art History.
Duell, P. (1938). The Mastaba of Mereruka. University of Chicago Oriental Institute.
Edwards, D. N. (2004). The Nubian past: An archaeology of the Sudan. London: Routledge.
El-Rahman, Y. A., Surour, A. A., El Manawi, A. H. W., Rifai, M., Motelib, A. A., Ali, W. K., & El
Dougdoug, A. M. (2013). Ancient mining and smelting activities in the wadi abu gerida area,
central eastern desert, Egypt: Preliminary results. Archaeometry, 55(6), 1067–1087.
Ellert, H. (1993). Rivers of gold. Gweru: Mambo Press.
Emery, W. B. (1963). Egypt exploration society preliminary report on the excavations at Buhen,
1962. Kush, 11, 116–120.
Emery, W. B. (1971). Preliminary report on the excavations at North Saqqâra, 1969-70. The Jour-
nal of Egyptian Archaeology, 57, 3–13.
shadreck.chirikure@uct.ac.za

95References
Eze-Uzomaka, P. (2013). Iron and its influence on the prehistoric site of Leija. In J, Humpris & Th.
Rehren (Eds.), The world of iron (pp. 3–9). London: Archetype.
Fenn, T. R., Killick, D. J., Chesley, J., Magnavita, S., & Ruiz, J. (2009). Contacts between West
Africa and Roman North Africa: Archaeometallurgical results from Kissi, northeastern Burki-
na Faso. In S. Magnavita, L. Koté, P. Breunig, and O. A. Idé (Eds.), Crossroads/Carrefour
Sahel: Cultural and technological developments in first millennium BC/AD West Africa (pp.
119–146). Journal of African Archaeology Monograph Series 2. Frankfurt am Main: Africa
Magna Verlag.
Goucher, C. L. (1981). Iron is iron’til it is rust: trade and ecology in the decline of West African
iron-smelting. The Journal of African History, 22(02), 179–189.
Goucher, C., & Herbert, E. (1996). The blooms of Banjeli: Technology and gender in West African
iron making. In P. R. Schmidt (Ed.), The culture and technology of African iron production
(pp. 40–57). Gainesville: University of Florida Press.
Haaland, R. (1980). Man’s role in the changing habitat of mema during the old kingdom of Ghana.
Norwegian Archaeological Review, 13(1), 31–46.
Haaland, R. (1994). Dakawa: An early Iron Age site in the Tanzanian hinterland. Azania, 29(1),
238–247.
Haaland, R. (2004). Iron smelting—a vanishing tradition: ethnographic study of this craft in south-
west Ethiopia. Journal of African Archaeology, 2 (1), 65–79.
Hatton, J. S. (1967). Notes on Makalanga iron smelting. Nada, 9(4), 39–42.
Hauptmann A. (2007). The archaeometallurgy of copper: Evidence from Faynan, Jordan. New
York: Springer.
Heimann, R. B., Chirikure, S., & Killick, D. (2010). Mineralogical study of precolonial (1650–
1850 CE) tin smelting slags from Rooiberg, Limpopo Province, South Africa. European Jour-
nal of Mineralogy, 22(5), 751–761.
Herbert, E. W. (1984). Red gold of Africa: Copper in precolonial history and culture. Madison:
University of Wisconsin Press.
Herbert, E. W. (1993). Iron, gender, and power: Rituals of transformation in African societies.
Bloomington: Indiana University Press.
Horne, L. (1982). Fuel for the metal worker. The role of charcoal and charcoal production in an-
cient metallurgy. Expedition, 25, 6–13
Huffman, T. N. (2007). Handbook to the Iron Age: The archaeology of pre-colonial farming societ-
ies in Southern Africa. Durban: University of Kwazulu-Natal Press.
Humpris, J. and Iles, L. (2013). Pre-colonial iron production in great lakes Africa: recent research
at UCL institute of archaeology. In J. Humpris & Th. Rehren (Eds.), The world of iron (pp. 56–
64). London: Archetype.
Huysecom E., & Agustoni B. (1997). Inagina: l’ultime maison du fer/the last house of iron. Ge-
neva: Telev. Suisse Romande. Videocassette, 54 min.
Ige, A., & Rehren, Th. (2003). Black sand and iron stone: iron smelting in Modakeke, Ife, South-
western Nigeria. Institute for Archaeo-Metallurgical Studies, 23, 15–20.
Iles, L., & Martinón-Torres, M. (2009). Pastoralist iron production on the Laikipia Plateau, Ke-
nya: wider implications for archaeometallurgical studies. Journal of Archaeological Science,
36(10), 2314–2326.
Joosten, I., Jansen, J. B. H., & Kars, H. (1998). Geochemistry and the past: estimation of the output
of a Germanic iron production site in the Netherlands. Journal of Geochemical Exploration,
62(1), 129–137.
Kense, F. (1985). The initial diffusion of iron to Africa. In R. Haaland & P. L. Shinnie (Eds.),
African Ironworking–ancient andtraditional (pp. 11–35). Oslo: Norwegian University Press.
Killick, D. J. (1990). Technology in Its Social Setting: Bloomery Iron-smelting at Kasunga, Ma-
lawi, 1860-1940. University Microfilms.
Killick D. (1991). A little-known extractive process: Iron smelting in natural-draft furnaces. JOM,
43(4), 62–64.
Killick, D. (2004). Review essay: What do we know about African iron working? Journal of Afri-
can Archaeology, 2(1), 97–112.
shadreck.chirikure@uct.ac.za

96 4 Domesticating Nature
Killick, D. (2014). Cairo to Cape: The spread of metallurgy through Eastern and Southern Africa.
In B. W. Roberts & C. Thornton (Eds.), Archaeometallurgy in global perspective (pp. 507–
527). New York: Springer.
Killick, D., & Miller, D. (2014). Smelting of magnetite and magnetite-ilmenite iron ores in the
Northern Lowveld, South Africa, ca. 1000 CE to ca. 1880 CE. Journal of Archaeological Sci-
ence, 43, 239–255.
Killick, D., Van Der Merwe, N. J., Gordon, R. B., & Grébénart, D. (1988). Reassessment of the
evidence for early metallurgy in Niger, West Africa. Journal of Archaeological Science, 15(4),
367–394.
Kiyaga-Mulindwa, D. (1993). The iron age peoples of East–Central Botswana. In T. Shaw, P.
Sinclair, B. Andah, & A. Okpoko (Eds.), The archaeology of Africa: Food, metals, and towns
(pp. 386–390). New York: Routledge.
Lambert, N. (1983). Nouvelle contribution à l’étude du Chalcolithique de Mauritanie. In N.
Échard (Ed.). Métallurgies Africaines, Nouvelles contributions (pp. 63–87). Mémoires de la
Sociéte des Africanistes 9. Paris.
Lane, P. J. (2005). Barbarous tribes and unrewarding gyrations? The changing role of ethnographic
imagination in African archaeology. In A. B. Stahl (Ed.), African archaeology: A critical Intro-
duction (pp. 24–54). Oxford: Blackwell.
Lyaya, E. (2013). Use of charcoal species for iron working in Tanzania. In J. Humpris & Th. Reh-
ren (Eds.), The world of iron (pp. 444–453). London: Archetype.
Lyaya, E.C., Mapunda, B. B., & Rehren, Th. (2012). The bloom refining technology in Ufipa, Tan-
zania (1850-1950). In C. Robion-Brunner & B. Martinelli (Eds.), Metallurgie du fer et Societes
africaines: Bilans et nouveaux paradigmes dans la recherche anthropoloque et archeologique
(pp. 195–207). Oxford: Archaeopress.
MacDonald, K. C., Vernet, R., Martinón-Torres, M., & Fuller, D. Q. (2009). Dhar Néma: From
early agriculture to metallurgy in Southeastern Mauritania. Azania, 44(1), 3–48.
MacEachern, S. (1993). Selling the iron for their shackles: Wandala-Montagnard interactions in
Northern Cameroon. Journal of African History, 34, 247–270.
Mackenzie, J. M. (1975). A pre-colonial industry: The njanja and the iron trade. Nada, 11(2),
200–220.
Maggs, T. (1992). ‘My father’s hammer never ceased its song day and night’: The zulu ferrous
metalworking industry. Southern African Humanities, 4, 65–87.
Mapunda, B. B. (1995). An archaeological view of the history and variation of ironworking in
Southwestern Tanzania. Unpublished Doctoral dissertation, University of Florida, Gainesville.
Mapunda, B. (2003). Fipa iron technologies and their implied social history. In C. M. Kusimba
& S. B. Kusimba (Eds.), East African archaeology: Foragers, potters, smiths, and traders
(pp. 71–86). Philadelphia: University of Pennsylvania Museum Press.
Martinón-Torres, M., & Rehren, T. (2014). Technical ceramics. In B. W. Roberts & C. Thornton
(Eds.), Archaeometallurgy in global perspective: Methods and syntheses (pp. 107–131). New
York: Springer.
Miller, D., & Killick, D. (2004). Slag identification at Southern African archaeological sites.
Journal of African Archaeology, 2(1), 23–47.
Miller, D. E., & Van Der Merwe, N. J. (1994). Early metal working in sub-Saharan Africa: A re-
view of recent research. Journal of African History, 35(1), 1–36.
Miller, D., Killick, D., & Van der Merwe, N. J. (2001). Metal Working in the Northern Lowveld,
South Africa AD 1000–1890. Journal of Field Archaeology, 28(3–4), 3–4.
Morton, G. R., & Wingrove, J. (1969). Constitution of bloomery slags: Part 1: Roman. Journal of
the Iron and Steel Institute, 207, 1556–1564.
Ndoro, W. (1991). Why decorate her. Zimbabwea, 3, 5–13.
Ndoro, W. (1994). Natural draught furnaces south of the Zambezi River. Zimbabwean Prehistory,
14, 29–32.
Nixon, S., Rehren, T., & Guerra, M. F. (2011). New light on the early Islamic West African gold
trade: coin molds from Tadmekka, Mali. Antiquity, 85(330), 1353–1368.
Oddy, A. (1984). Gold in the Southern African Iron Age. Gold Bulletin, 17(2), 70–78.
shadreck.chirikure@uct.ac.za

97References
Ogden, J. (2000). Metals. In P. Nicholson & I. Shaw (Eds.), Ancient Egyptian materials and tech-
nology (pp. 148–176). Cambridge: Cambridge University Press.
Okafor, E. (1993). New evidence on early iron-smelting in Southeastern Nigeria. In T. Shaw, P.
Sinclair, B. Andah, & A. Okpoko (Eds.), The archaeology of Africa: Food, metals, and towns
(pp. 432–448). London: Routledge.
Phillipson, D. W. (2005). African archaeology. Cambridge: Cambridge University Press.
Pleiner, R. (2000). Iron in archaeology: the European bloomery smelters. Archeologický ústav.
Prendergast, M. D. (1975). A new furnace type from the darwendale dam basin. Rhodesian Prehis-
tory, 7(14), 16–20.
Rehder, J. E. (1986). Use of preheated air in primitive furnaces: Comment on views of Avery and
Schmidt. Journal of Field Archaeology, 13(3), 351–353.
Rehren, Th. (2001). Meroe, iron and Africa. Der Antike Sudan, 12, 102–109.
Rehren, T., & Pernicka, E. (2008). Coins, artefacts and isotopes–archaeometallurgy and archaeom-
etry. Archaeometry, 50(2), 232–248.
Rehren, T., Charlton, M., Chirikure, S., Humphris, J., Ige, A., & Veldhuijzen, H. A. (2007). Deci-
sions set in slag: the human factor in African iron smelting. In S. La Niece, D. Hook, & P.
Craddock (Eds.), Metals and mines: Studies in archaeometallurgy (pp. 211–218). London:
Archetype.
Roberts, B., Thornton, C., & Pigott, V. (2009). Development of metallurgy in Eurasia. Develop-
ment, 83(322), 1012–1022.
Robion-Brunner, C, Serneels, V and Perret, S. (2013). Variability in iron smelting practices: Con-
frontation of technical cultural and economic criteria to explain the metallurgical diversity in
the Dogon Area (Mali). World of iron conference, Natural History Museum, London.
Robson, E. (2001). Society and technology in the Late Bronze Age: A guided tour of the cuneiform
sources.
Rostoker, W., & Bronson, B. (1990). Pre-industrial iron: Its technology and ethnology. Philadel-
phia: Archeomaterials.
Rowlands, M., & Warnier, J. P. (1993). The magical production of iron in the Cameroon Grass-
fields. In T. Shaw, P. J. J. Sinclair, B. Andah, & A. Okpoko (Eds.), The Archaeology of Africa:
Food, metals and towns (pp. 512–550). London: Routledge.
Scheel, B. (1989). Egyptian metalworking and tools. Oxford: Shire Publications.
Schmidt, P. R. (1996). The culture and technology of African iron production. Gainesville: Uni-
versity Press of Florida.
Schmidt, P. R. (1997). Iron technology in East Africa: Symbolism, science, and archaeology.
Bloomington: Indiana University Press.
Schmidt, P. R., & Avery D. H. (1978). Complex iron smelting and pre-historic culture in Tanzania.
Science, 201, 1085–1089.
Schmidt, P. R., & Avery, D. H. (1983). More evidence for an advanced prehistoric iron technology
in Africa. Journal of Field Archaeology, 10 (4), 421–434.
Schmidt, P. R., & Mapunda, B. B. (1997). Ideology and the archaeological record in Africa: In-
terpreting symbolism in iron smelting technology. Journal of Anthropological Archaeology,
16(1), 73–102.
Serneels, V. and Perret, S. 2003. Quantification of smithing activities based on the inves-
tigation of slag and other material remains. Archaeometallurgy in Europe. Proceedings of the
International Conference (Milano, September 24-26, 2003). Milano: Associazione Italiana di
Metallurgia, Vol. 1, pp. 469–478.
Severin, T., Rehren, T., & Schleicher, H. (2011). Early metal smelting in Aksum, Ethiopia: Copper
or iron?. European Journal of Mineralogy, 23(6), 981–992.
Shinnie, P. L. (1985). Iron working at Meroe. In R. Haaland & P. L. Shinnie (Eds.), African iron
working: Ancient and traditional (pp. 28–35). Oslo: Norwegian University Press.
Soper, R. (2002). Nyanga: Ancient fields, settlements and agricultural history in Zimbabwe. Nai-
robi: British Institute in Eastern Africa.
Stahl, A. B. (2014a). Africa in the World: (Re)centering African history through archaeology. Jour-
nal of Anthropological Research, 70 (1), 5–33
shadreck.chirikure@uct.ac.za

98 4 Domesticating Nature
Stahl, A. B. (2014b). Metal working and ritualization: Negotiating change through improvisational
practice in Banda, Ghana, AD 1300-1650. In A. B. Stahl, The materiality of everyday life.
Archaeology papers of the American Anthropological Society. Arlington: American Anthro-
pological Society.
Summers, R. (1969). Ancient mining in Rhodesia and adjacent areas. Salisbury: Trustees of the
National Museums of Rhodesia.
Thondhlana, T. (2012). Metalworkers and Smelting Precincts: Technological Reconstructions of
Second Millennium Copper Production Around Phalaborwa, Northern Lowveld of South Af-
rica. Unpublished Doctoral dissertation, University of College London.
Tylecote, F. F. (1980). Furnaces, crucibles and slags. In T. E. Wertime & J. D. Muhly (Eds.), The
coming of the age of iron (pp. 183–228). New Haven: Yale Univ. Press.
Tamari, T. (1991). The development of caste systems in West Africa. Journal of African History,
32 (2), 221–250.
Van der Merwe, N. J. (1980). The advent of iron in Africa. In T. E. Wertime & J. D. Muhly (Eds.),
The Coming of the Age of Iron, (pp. 463–506). New Haven CT: Yale Univ. Press.
Van der Merwe, N. J., & Avery, D. H. (1987). Science and magic in African technology: traditional
iron traditional iron smelting in Malawi. Africa, 57(2), 143–172.
Van Noten, F. (1985). Ancient and modern iron smelting in Central Africa: Zaire, Rwanda and Bu-
rundi. In R. Haaland & P. L. Shinnie, African Iron working—ancient and traditional (pp. 102–
120). Oslo: Norwegian University Press.
Wagner, D. 2008. Science and civilisation in China. Vol. 5: Chemistry and chemical technology.
Part 11: Ferrous metallurgy. Cambridge University Press. http://donwagner.dk/
Warnier, J. P., & Fowler, I. (1979). A nineteenth-century Ruhr in Central Africa. Africa, 49(4),
329–351.
Wobst, H. M. (1978). The archaeo-ethnology of hunter-gatherers or the tyranny of the ethno-
graphic record in archaeology. American Antiquity, 43(2), 303–309.
Woodhouse, J. (1998). Iron in Africa: Metal from nowhere. In G. Connah (Ed.) Transformations in
Africa: Essays on Africa’s later past (pp. 160–185). London: Leicester University Pres.
shadreck.chirikure@uct.ac.za

99
Chapter 5
Socializing Metals
© The Author 2015
S. Chirikure, Metals in Past Societies, SpringerBriefs in Archaeology,
DOI 10.1007/978-3-319-11641-9_5
Introduction: Fabricating Metal into Cultural Products
The diversity in smelting practices observed for different peoples and periods in
the history of different parts of Africa (Fig. 5.1), with their variably entangled and
distinct technological traditions, characterizes processes of smithing—the next and
culturally conditioned stage in the chaîne opératorie of metal working, to which
this chapter is devoted. Sub-Saharan Africa differs remarkably from some of the
Egyptian, Nubian and North African practices. In this region, as we have seen in
the previous chapter, the process of smelting was a cultural intervention that trans-
formed products of nature into culture. This act of transformation was symbolically
associated with giving birth. The concept of ‘bringing into life’ was a fundamental
one because, analogically speaking, the newly smelted metal (iron) passed through
several additional stages of transformation (Herbert 1993). For example, metals
were consolidated into ingots, or transformed into objects used in utilitarian, deco-
rative and ceremonial domains (Cline 1937; Miller 2002; Reid and McLean 1995;
Stayt 1931). In other contexts, ingots themselves were the final products used in
religious and expressive spheres. Overall, the metal from furnaces and crucibles
was traded sometimes as ingots but in other cases was smithed into objects. Egypt
participated in the long-distance trade around the Mediterranean and also extend-
ing to Nubia, particularly after 3000 BC. Similarly, various sub-Saharan regions
were networked locally and trans-regionally. Taken together, metals were situated
within the nexus of society in that their use extended to utilitarian and nonutilitarian
domains and were critical elements in the operations of economic, social, political,
cultural and religious domains.
But how were the metals from Africa’s diverse furnaces and cultures fabricat-
ed into disparate objects that addressed society’s needs? This chapter is primar-
ily aimed at addressing this question—it discusses the processes and techniques
involved in African preindustrial metal smithing and fabrication. Metallographic
and compositional work performed on a large corpus of objects and recovered from
different parts of Africa illuminate the techniques of hot and cold working metal
shadreck.chirikure@uct.ac.za

100 5 Socializing Metals
that were in use across the ages (Miller 1996, 2000; Scheel 1989). While hot work-
ing was practiced across the entire spectrum of metals and alloys that were manipu-
lated preindustrially, techniques such as casting were specific to those that could be
melted using available technology (Scott 1991).
In Egypt, pictorial paintings show the fabrication and casting of copper, gold,
bronze and silver from Dynastic to Ptolemaic times (Scheel 1989). According to
Miller (2002), techniques employed to fabricate metals in preindustrial sub-Saharan
Fig. 5.1 Metal working groups and important sites discussed in the text
shadreck.chirikure@uct.ac.za

101Metal Fabrication in Egypt and Nubia
Africa were fairly stable through time, although there were major instances of re-
gional and temporal variation. Steel, the alloy of iron and carbon, was the hardest
metal known to preindustrial sub-Saharan Africa (David et al. 1989). Iron and low-
carbon steels were smithed into utilitarian and nonutilitarian objects. The use of
finished objects was, however, not monolithic because so-called utilitarian items
such as hoes were often used as currency in trade as well as in ritual and ceremo-
nial settings (Dewey 1991). Other metals and alloys—copper, gold, bronze, brass
and tin—were either smithed or casted into symbolic and decorative objects. Their
colours appealed to various senses and valuations, resulting in a strong association
with different genders, ancestors and power (Herbert 1993).
As with elsewhere in the Old World, preindustrial African metal smithing in-
volved heating billets of metal in oxidizing environments (Crew 1991). This was
followed by cycles of hot and cold hammering, which brought the metal to shape.
The fabrication of copper, copper-based alloys and gold was designed to exploit
not only their relatively low melting temperatures but also their physical properties
such as ductility and malleability (Scott 1991). Routinely, these nonferrous metals
were cast to produce a wide range of spectacular objects such as ingots, sheets,
bangle blanks, beads and earrings (Miller 2002; Oddy 1984). Often, metals and al-
loys were drawn into wire or were hammered into thin sheets. In Southern Africa,
these were wound around a vegetal core to produce spectacular decorative products
(Miller 1996).
Over 7000 years of metal fabrication in Egypt and Nubia left residues and was
often painted on tomb walls, which has remarkably preserved the evidence. In sub-
Saharan Africa, over 2000 years of indigenous metal smithing has left ubiquitous
fingerprints in the form of finished objects and tools (Garrard 1980; Kusimba and
Killick 2003; Shaw 1970; Thondhlana and Martinón-Torres 2009). The contexts
from which such traces of metallurgy were recovered powerfully demonstrate that
as a technological solution, metal fabrication was precipitated by the need to achieve
social, economic and political ends through the use of metals. This is testament to
the important role played by metals in the preindustrial world. The integrated nature
of metallurgy and society not only promoted or demoted the fortunes of smiths, but
also enhanced the opulence enjoyed within varying gradients, by elites and common-
ers alike. In many sub-Saharan contexts, smiths were either feared or held in awe for
their supernatural powers (McNaughton 1993), but in Egypt and Nubia, they formed
a lowly class below that of learned officials and royalty (Scheel 1989), further under-
scoring the need to be attuned to variation in practice through time and across space.
Metal Fabrication in Egypt and Nubia
Because of their comparatively deep history of metallurgy—longer than anywhere
else on the African continent–Egypt and Nubia offer interesting perspectives on
the evolution of metal manipulation and fabrication. Scientific work suggests that
the earliest objects of gold, native copper and meteoric iron were hammered and
shadreck.chirikure@uct.ac.za

102 5 Socializing Metals
annealed (Craddock 2000; Rehren et al. 2013). With the advent of reductive smelt-
ing, metallic copper was sometimes hammered hot and cold but in other cases cast
to produce objects, for temples, royalty, commoners and the military to mention a
few (Ogden 2000). From the Old Kingdom (c. 3050BC) up to the end of Ptolemaic
times (AD 385), a significant amount of pictorial depictions and associated texts
in tombs demonstrates the techniques by which copper, tin bronze, gold, leaded tin
bronze and silver were worked (see for example de Graris Davis 1943). During this
period, metalworking was closely regulated by the state such that no independent
specialists existed. In retrieving and returning stock, metalworkers and officials had
to weigh opening and closing amounts and were always symbolically watched by
Maat, the god of justice, who crowned the balance masts (Scheel 1989). Gold, silver
and electrum served throughout Egyptian history as the basic materials for objects
of royal use or for funerary and temple equipment (Aldred 1971), while copper, ar-
senic copper and bronze were used for the production of tools for daily use. As part
of a chaîne opératoire, Egyptian and Nubian metalworkers melted large ingots or
pieces of metal from furnaces or acquired through trade to refine, alloy, cast or split
them up into smaller portions for further treatment by the smiths (Scheel 1989).
Exploiting the Behaviour and Properties of Metals: Melting,
Casting and Plate Production
Scientific analysis of Egyptian artifacts has revealed that copper casting was prac-
ticed as early as the late Naqada II and Naqada III, between 3300 to 3000 BC
(Scheel 1989). Copper tools and weapons were manufactured by hammering or
open-mold casting. From the beginning of the Dynastic Period in Egypt around
3050 BC, metal fabrication techniques continued to be developed and improved,
resulting in notable continuity and change. The metals (copper, tin, gold, silver, lead
and their alloys) were melted in one or more crucibles, depending on the amount
required (Scheel 1989). The pictorial depiction on the Sixth Dynasty tomb of the
Vizier Mereruka at Saqqâra shows six metalworkers fanning through blowpipes
into crucibles placed side by side. Although unverified, it is possible that fans of
foliage were employed to provide a draught. In the Old and Middle Kingdoms,
before the advent of artificial bellows, Egyptian metalworkers blew air into blow
pipes consisting of reeds tipped with clay. Subsequently, skin bellows were used to
provide the draught from the Middle Kingdom onwards. These were followed by
dish bellows whose earliest depiction appears inside the Eighteenth Dynasty tomb
of the priest Puyemre, the Second Prophet of the god Amun in Thebes (Fig. 5.2).
The introduction of dish bellows enabled large quantities of metal to be melted for
the casting of large metal objects as shown on a wall painting in the tomb of the
Vizier Rekhmire at Thebes (Scheel 1989).
Although richly illustrated on tomb walls, finds of metal working sites are very
rare. One ancient Egyptian foundry dating to Ptolemaic times was discovered in
the Theban necropolis. The most important finds include numerous mud-brick
hearths, which acted as receptacles for the burning charcoal onto which crucibles
shadreck.chirikure@uct.ac.za

103Metal Fabrication in Egypt and Nubia
were placed (Scheel 1989). The foundry was designed for small-scale and mass pro-
duction. Other finds include charcoal from acacia species, crucible sherds, broken
tuyeres, limestone mold and the nozzles of dish bellows.
After being melted, refined and subsequently divided into portions, the cooled
metal was passed to the smiths or blacksmiths for plate or sheet production. The
metal was beaten on basalt, diorite or granite anvils, which were placed on a wood-
en block to absorb the hammering. Ancient Egyptian metalworkers used flat (for
smoothing) and rounded hammer stones (for chasing). Evidence suggests that they
practiced the technique of annealing as early as the Predynastic Period. Work-hard-
ened metal was heated or annealed in a blowing or bellows-fanned fire to soften it
to restore ductility. One pictorial in the tomb of Rekhmire depicts gold beaters plac-
ing thin gold plates on stone blocks and using hammer stones to repeatedly beat the
gold leaf until the desired thinness was achieved (Scheel 1989). Silver and electrum
also were worked to foil, while objects of less rare material were often gilded or
silver-plated. Gold, silver and electrum foil or leaves were used to cover wooden
furniture, statues, coffins and models of daily life manufactured for funerary equip-
ment (de Graris Davis 1943). During the Roman Period, metalworkers practiced
fire gilding with gold amalgam. Gold amalgam was applied to the base metal object
to be gilded, and in the process, the mercury content of the amalgam vaporized,
leaving gold attached to the surface of the metal objects (Scheel 1989).
From the Fourth Dynasty, ancient Egyptians joined together various components
of objects using the techniques of soldering and riveting. It appears that Egyptian
smiths applied different mixtures of gold, silver and copper to produce solders of
different colours and melting points. The objects were polished on anvils using
stones such as agate to smooth uneven patches on metal objects. Often, finished
vessels were engraved with hieroglyphic text and other decorations. The engraver
worked out the outline drawing using a hammer stone and chisels of different sizes.
Casting is yet another important metalworking process practiced by ancient
Egyptians and Nubians. A diachronic study of casting shows that founders in Early
Dynastic times poured molten copper or arsenic copper into preformed stone or clay
molds to make simple tools and weapons. In the Old Kingdom, a more sophisticated
form of casting using two-part molds of clay and steatite or serpentine allowed both
Fig. 5.2 Dish bellows connected to a clay nozzle with a reed (redrawn from Scheel 1989)
shadreck.chirikure@uct.ac.za

104 5 Socializing Metals
faces of the object to be fashioned. Perhaps the most complicated method of cast-
ing is known as lost-wax or cire perdue casting, which enabled production of very
detailed and complex shapes (Scheel 1989). This technique produced complicated
jewelry, practical objects and statuettes for religious purposes. Initially, a beeswax
model of the object to be cast was produced, and then coated with clay. The com-
posite structure was heated in a charcoal fireplace to harden the clay and to melt
the wax. In the process, the clay mold remained and retained all the details of the
molten model. Egyptian founders then poured the molten metal into the clay mold.
Upon cooling, the clay mold was broken and the cast object was cleaned and pol-
ished. Jewelry, amulets or other valuable items made of gold, silver and electrum
were cast in this way. The technique of core casting was applied in casting large
objects (Scheel 1989). A core of clay or sand was covered by a layer of malleable
wax, shaped to form the mold of the object to be cast. The model, consisting of
the core and the formed wax layer, was coated with clay. In order to secure it in
position throughout the later casting process, the core had to be stabilized by pins
or wire fixed to the outer cover of clay. The wax was melted away, leaving in the
kiln the hardened clay mold with the fixed core. In casting, only the gap between
the outer clay mold and the inner core had to be filled with bronze or other molten
metal. Core casting was very common in the manufacture of larger objects during
the New Kingdom and reached its peak in the Late Period (713–332BC). All these
techniques were also exported to the Nubia (Emery 1963,1971 )
Wirework and Jewelry
Gold, silver, and bronze wire was an important product of Egyptian metalwork.
According to Scheel (1989), wire made its appearance during the First Dynasty
and was used for mundane activities in the household, temples, palaces and in con-
struction. Wire was produced through several methods. Initially, a metal piece was
hammered out to sheet metal, which was then cut into thin strips. These strips were
hammered out and cut again. Hammered wires show variations in diameter along
the length, a faceted surface and a solid but noncircular cross section (Scheel 1989).
This process of hammering was used to manufacture relatively thick wires. Another
technique utilized in ancient Egypt is that of block-twisting, which appeared in New
Kingdom times. The procedure involved hammering an ingot to create a square rod
which was twisted about its major axis to form a solid wire with a screw thread of
variable pitch. In the course of this twisting, the wire had to be annealed repeatedly
to preserve its ductility. Spiraling could be eliminated by rolling the wire between
two flat pieces of hard wood (Scheel 1989). In addition to hammering and block-
twisting, strip-drawing and strip-twisting were probably mastered by Egyptian wire
makers. A strip of metal foil was drawn through a number of holes of different
diameters so that the strip curled in upon itself, forming a hollow tube. The strip
could have been drawn through holes drilled through precious stones. In the final
stage of wire production by this method, the wire could be drawn through holes of
different diameters. Goldsmiths, silversmiths, and jewelers used wires of precious
shadreck.chirikure@uct.ac.za

105Forging, Smithing and Casting in Sub-Saharan Africa: West and Central Africa
metal to decorate valuable objects or to manufacture pieces of jewelry such as an-
klets, aprons, armlets, belts, collars, chains, necklaces and pectorals.
Organization of Metalworking
Ancient Egyptian and Nubian metalworkers were not independent specialists; they
were supervised by the state through a number of officials. Workshops were at-
tached to temples, royal palaces or to the household of a high official. Because of
their attached status, craftsmen in ancient Egypt were all dependent on their em-
ployers, who kept and allocated raw materials. Not surprisingly, metalworkers oc-
cupied a low position in the occupational specialization when compared to learned
officials (Scheel 1989). This low social position contrasts with some areas of sub-
Saharan Africa such as Central Africa where smiths were associated with royalty
(de Maret 1985). Metalworkers were skilled in a range of metals and in relation to
the physical properties of individual metals and alloys; they transferred skills from
gold and copper to electrum, silver and bronze and eventually iron. Therefore, the
copper worker was the gold worker who was also a bronze worker and later an iron
worker (Scheel 1989).
Forging, Smithing and Casting in Sub-Saharan Africa:
West and Central Africa
From their initial establishment in West and Central Africa, sometime in the ra-
diocarbon black hole between 800 BC and 400BC, iron prevailed for utilitarian
purposes, while copper was used for ornamental activities. Tin, bronze, gold and
brass, introduced a millennium after iron and copper, were restricted in use to orna-
mental functions. As discussed above, this developmental trajectory differed from
that of Egypt, Nubia, North Sahara, Ethiopia and Eritrea, which started with copper
followed by bronze and iron. While metalworkers at Akjoujt in Mauretania started
with copper, this was not followed by bronze, illustrating variation within the sub-
continent. Unlike Egypt where a long record of literacy and tomb paintings provides
insight into the fashioning of objects from metals, in West and other parts of sub-
Saharan Africa, we have to rely on ethnography, travelers’ reports and archaeologi-
cal and technological studies for insight.
Two decades ago, Miller and Van der Merwe (1994) argued that traditional smith-
ing and forging had received less attention in sub-Saharan technical literature than
smelting, and the picture has not changed very much. While much is known from
ethnographic sources, researchers should resist the present argument that ethno-
graphically documented techniques represent unchanging continuities from earlier
periods. Nevertheless, the ethnographic accounts of iron, copper and gold smithing
are a good starting point for approaching variation in the archaeological past.
shadreck.chirikure@uct.ac.za

106 5 Socializing Metals
As in Egypt and Nubia and elsewhere in the world, West African smiths manipu-
lated smelted, and melted metal in a number of ways: hammering (iron, gold and
copper and its alloys), casting (copper, bronze, gold and brass) and wire drawing
(copper, iron, gold, bronze and brass). In the area around Togo and Burkina Faso,
the forges used from the sixteenth century AD onwards were very small and pow-
ered by concertina bellows (Cline 1937; Fig. 5.3). These forges contrasted with the
very high natural-draught-powered smelting furnaces used in this area. The process
of fabricating objects of iron started with heating blooms from furnaces on anvils to
expel unwanted and occluded slag. The refined billet of metal was repeatedly ham-
mered and annealed to gain usable metal. Today, smiths at places such as Bitchabe
in Togo forge scrap iron using concertina bellows (Fig. 5.4).
Bellamy and Harbord (1904) described indigenous bloom forging and the smith-
ing practice of the Yoruba of Oyo in early twentieth century Southwestern Nigeria.
The forge was blown with a pair of valveless, circular, wooden bowl bellows cov-
ered with goat skin and pumped by centrally attached bamboo rods, feeding air to
the open hearth through two wooden tuyeres. An iron hammer was used to beat the
heated bloom, first on a large stone anvil and then on a smaller metal one to fashion
the required tools. The indigenous metal was considered to be superior and more
durable than imported English iron which contained sulfur from coal. In the north-
Fig. 5.3 Location of West and Central sites and groups discussed in the text
shadreck.chirikure@uct.ac.za

107Forging, Smithing and Casting in Sub-Saharan Africa: West and Central Africa
eastern village of Sukur, Northeastern Nigeria, smith’s heated iron fragments in the
forge for about half an hour (Sassoon 1964). The hot metal was beaten on a rock an-
vil with large hammer stones. The consolidated lump was then reheated and forged
into the required tools. According to David and Sterner (1997), during the peak of
the forging season, Sukur exported 60,000 hoes annually. In the nearby Mandara
Mountains of Cameroon, the Mafa first sorted the products from the down draught
furnaces into soft and cast iron. The cast iron was decarburized in crucibles to pro-
duce low-to high-carbon steels with typical martensitic microstructures. These were
forged to produce iron bars which were traded over long distances. These smiths
also produced shackles that were used in the slave trade (MacEachern 1993). In-
terestingly, while smelting in these areas took place in very large furnaces, forges
were comparatively much smaller, yet produced a significant amount of finished
products that fed into local, regional and in some cases long-distance trade. There-
fore, this challenges assumptions that large-scale production of necessity equates
with large-scale installations. Unlike smelting which could operate using natural
draught, forging was invariably powered by bellows (Chirikure et al. 2009).
The available but sporadic evidence indicates that the technique of hammering
was practiced in both the Early and Late Iron Ages of West and Central Africa. The
smithing technology of the Early and Late Iron Age components at Dekpassanware
was identical to ethnographically documented practice as revealed by finds of coni-
cal tuyeres, stone anvils, stone hammers and bloom crushing mortars (De Barros
Fig. 5.4 Traditional iron forging workshop at Bitchabe, Togo, using concertina bellows. (Photo-
graph Credit: Philip de Barros)
shadreck.chirikure@uct.ac.za

108 5 Socializing Metals
2013, p. 17). Metallographic analyses of an Early Iron Age bracelet from a burial at
Dekpassanware indicated that it was made of good-quality iron of variable carbon
content. The microstructural evidence showed that the object cooled slowly and was
therefore not quenched. The array of iron objects manufactured in West and Central
Africa included hoes, iron bars, bangles, anklets, shackles, chains, spears, sheets
and stout wire. Iron gongs, both single and double, were also made in some regions
where they were associated with the power of kings (Vansina 1969).
The available evidence indicates that copper, bronze and gold were also ham-
mered to produce a wide variety of objects in West and Central Africa including
sheets, blanks, beads and wire (Cline 1937). More importantly, copper, gold and
bronze and later brass were cast in open ceramic and steatite molds to produce
simple shapes. The famous Katanga crosses of Central Africa dating from the mid-
to late first millennium AD were cast using this method. By the late first millennium
AD, the ‘lost wax’ process was used to cast copper and its alloys, with an early ex-
ample supplied by bronzes at Igbo Ukwu in Nigeria dated to the nineth century AD
(Shaw 1970). The famous Benin ‘bronzes’ and Akan gold weights were similarly
produced through lost wax casting from the second half of the second millennium
AD onwards.
During the lost wax process, the desired object was either molded using wax or
wax was used to cover a clay model mold. Subsequently, support rods of wax were
attached to the exterior of the sculpture to create casting sprues on the wall of the
mold. These enabled the cast metal to flow without trapping air pockets. In similar
fashion to the practice of ancient Egyptian metal workers described above, the en-
tire structure was dipped in liquid clay mixed with finely ground ash and, when dry,
coated in a thick clay jacket. The dried mold was then fired to remove the wax/latex,
after which molten metal was poured into the resultant void. In some cases, the cru-
cible used to melt the metal was physically attached to the mold (Bisson 2000). On
cooling, the clay exterior was shattered, thereby exposing the sculpture. The object
then had the casting sprues removed and was subsequently cleaned and polished.
This technique was finely executed to produce the famous Igbo Ukwu copper and
bronze objects in Nigeria dating to between the nineth and thirteenth centuries AD
(Shaw 1970; Fig. 5.5). The Igbo Ukwu bronzes represent a masterpiece in terms of
metalworking, are easily the most technically skilled castings in all of sub-Saharan
Africa, and are among the finest art castings ever done anywhere in the world at any
time (Killick pers comm 2014). Thurstan Shaw’s excavations unearthed more than
three hundred elaborately cast objects including the renowned roped pot (Fig. 5.5),
immaculate bronze altar stands, spectacular bronze bowls, elaborate bracelets, cop-
per snakes, a bronze snail, bronze elephant, bracelets and much more. Over 80
objects from Igbo Ukwu were analysed, and those produced using lost wax were
leaded bronzes, while the hammered and chased ones were made of pure copper
(Shaw 1970). Other objects include very fine copper and bronze wire work. Igbo
Ukwu also yielded significant amounts of hammered iron objects, pieces of ivory
and thousands of trade beads, indicating that it represents high-status burials. Ini-
tially, it was believed that the Igbo Ukwu bronzes were not local in origin. However,
compositional and provenance studies indicated that the copper was obtained from
shadreck.chirikure@uct.ac.za

109Forging, Smithing and Casting in Sub-Saharan Africa: West and Central Africa
Fig. 5.5 Leaded tin bronze
Igbo Isiah roped vessel
produced using the lost wax
method (height: 32.2 cm).
(Photograph Credit: Estate of
Thurstan Shaw with permis-
sion of Dr Pamela Smith)
the Benue River valley, while the tin came from the Jos Plateau. Lost wax technique
continued to be practiced in what is today Nigeria well into the nineteenth century
and resulted in the equally outstanding bronze and brass objects from Ife and Benin
(Chikwendu et al. 1989).
According to Garrard (2011), the stories of abundant gold in West Africa, partic-
ularly from Islamic writers, has not matched the recovery of gold from archaeologi-
cal sites. Amongst the few gold objects recovered include a very fine pectoral from
Rao Senegal (Fig 5.6) and a seventh to nineth century earing from Jenne Jeni in the
Inland Niger Delta of Mali (Fig 5.7). Reports by Al Bakri in the eleventh century
indicate that gold was exported from West Africa to North Africa via the town of
Tewdagoust (Nixon et al. 2011). The scale of West African production was revealed
when Mansa Musa the king of ancient Mali (AD1200–1400) made pilgrimage to
Mecca with significant quantities of gold, which led to a dramatic fall of gold prices
in Cairo (Levtzion 1973) and ultimately whetting the European appetite for gaining
direct access to African gold. While historical focus has tended to be on the role
shadreck.chirikure@uct.ac.za

110 5 Socializing Metals
of gold as an export commodity, West African gold smiths produced technically
sophisticated objects of considerable beauty. One of the most spectacular is a gold
pectoral excavated by archaeologists at the tumulus of Rao in contexts dating be-
tween the seventeenth and eighteenth centuries (Fig. 5.6, Garrard 2011). In the area
around modern Ghana, the Akan people produced spectacular gold objects, which
like brass objects worked by the same group, were hammered into thin sheet or cast
(Garrard 2011, p. 132).
Just as the skilled metal workers were attached to temples/political authorities in
Egypt, in some West African contexts, prestige metals such as gold and brass were
Fig. 5.7 Photograph showing
earliest gold earing from
Jenne Jeno, Inland Niger
Delta, Mali. (Source: Rod
McInstosh)
Fig. 5.6 Pectoral from the
tumulus of Rao Senegal dat-
ing between seventeenth and
eighteenth centuries. (Source:
Gold of Africa Museum,
Cape Town)
shadreck.chirikure@uct.ac.za

111East and Southern Africa with Occasional References to Sudan and Ethiopia
intimately bound up in the institutions of state (Garrard 1980, p. 38). For example,
Akan goldsmiths produced gold objects such as sword ornaments (Fig. 5.8), which
were used in stately functions together with the Golden Stool (Garrard 1980, p. 63).
Interestingly, the gold smiths and other skilled metallurgists practiced as attached
specialists (Garrard 1980, p. 47).
East and Southern Africa with Occasional References
to Sudan and Ethiopia
Turning to fabrication processes in East and Southern Africa (Fig. 5.9), a short ac-
count of smithing in Sudan recorded by Crawhill (1933) documented important
features of forging.
The fragments of iron blooms from the smelting furnace were heated in open
clay crucibles until they were soft enough to be hammered together between stone
anvils and then shaped further using a handle-less iron hammer. The smiths used
four valve-less hand-operated bellows whose cylinders were made of clay and cov-
ered with goat skin diaphragms. In Ethiopia, Dimi forges of the twentieth century
consisted of a walled hearth driven by two drum bellows operated by hand that blew
air into a single tuyere (Todd 1985). A blunt iron hammer and stones were used to
fashion the desired implement by hammering the heated bloomery iron on a spe-
cially selected stone anvil. The final product was not quenched or tempered. Broken
tools were also repaired at the forge. Brown (1995) studied ethnographic iron work-
ing among the Kikuyu in Kenya in the twentieth century. The Kikuyu forged iron in
small forges powered by bellows. They used stone anvils and hammer stones. They
also quenched the finished objects in water.
Fig. 5.8 Akan crocodile and
lizard sword ornaments (size
21.5 cm, see Garrard 2011,
p. 234) (Source: Gold of
Africa Museum, Cape Town)
shadreck.chirikure@uct.ac.za

112 5 Socializing Metals
One of the best ethnographic descriptions of preindustrial iron smithing in Af-
rica is provided by Emil Holub, a Czech medical practitioner who journeyed from
Cape Town to the Barotseland plain of Zambia in the late 1870s. While in Zambia,
he recorded the activities of a Mambari iron smith, paying meticulous attention to
the pumping of the bellows and the tools used in the process. Holub (1881) made
impressive illustrations not only of these bellows but also of the other tools used by
the smith. These include hammers, chisels, tongs, anvils and some finished objects
(Figs. 5.10 and 5.11).
South of the Zambezi, Njanja iron forging was discussed by a number of re-
searchers. After a smelting re-enactment carried out by Headman Ranga in the
1950s, a bloom from the furnaces was consolidated in a forge, powered by goat
skin bellows (Fig. 5.12). The bloom was repeatedly hammered on an anvil until the
metal was usable. Iron hoes were made by forge welding pieces of metal together.
Forge welding resulted in the incorporation of surface scale into the weld lines
Fig. 5.9 Location of smithing sites in East and Southern Africa
shadreck.chirikure@uct.ac.za

113East and Southern Africa with Occasional References to Sudan and Ethiopia
Fig. 5.10 Illustration of a Mambari smith by Holub (1881) showing bellows leading to tuyeres
and a very small forge. Note also the tongs illustrated
Fig. 5.11 Illustration of pot
bellows used for smithing
by Mambari smiths (Holub
1881)
shadreck.chirikure@uct.ac.za

114
which created internal voids in the objects. This enables archaeometallurgists to
identify objects manufactured using preindustrial techniques. Furthermore, bloom-
ery iron consisted of slag stringers that are microscopically detectable. Objects such
as axes were quenched in water, which improved the strength of the metal with
some carbon in it. This may be why objects from Njanja smiths were highly regard-
ed in areas with their own smiths (Chirikure 2006; Mackenzie 1975). Interestingly,
Njanja scale of production is described as semi-industrial and was often character-
ized by the presence of itinerant smiths who smithed iron in different areas and in
the process amassed wealth. Igbo smiths in Nigeria were similarly itinerating.
As with iron forging, copper and gold smithings also feature prominently in eth-
nographic and historical accounts of Southern Africa. In most of Southern Zam-
bezia, copper was melted and cast in ingot molds that produced X-shaped copper
ingots similar to those produced in Central Africa (Fig. 5.13).
The most vivid ethnographic descriptions of copper working are those of the
Venda and Ba-Phalaborwa people, historically associated with parts of Southern
5 Socializing Metals
Fig. 5.12 Goat skin bellows
similar to the ones used by
Njanja. Approximate size
80 cm long. Natural History
Museum, Bulawayo, Zim-
babwe (Photograph Credit:
Author)
shadreck.chirikure@uct.ac.za

115East and Southern Africa with Occasional References to Sudan and Ethiopia
Zimbabwe and parts of Northern South Africa. The copper from furnaces was melt-
ed in ceramic crucibles that resemble domestic pots (Stayt 1931; Mamadi 1940).
Once refined, the metal was either cast to produce small items such as bangle blanks
or hammered to produce thin sheet. In some cases, the molten metal was cast to
produce the iconic musuku and lerale ingots (Stayt 1931; Miller et al. 2001). Lerale
ingots were made by pressing a 1-or 2-cm-diameter stick lengthwise into the soil.
At one end of the mold, a small hollow was carved to create a small head, some-
times with short arms protruding from it. Molten metal was then poured, which on
solidification produced lerale ingots (Miller et al. 2001). In some cases, the end of a
larger stick was thrust into the ground, producing an oblong hole with straight sides
and a flat bottom. The ends of smaller sticks were pressed into the bottom of this
hole to form a pattern of smaller holes, usually parallel lines (Stayt 1931). When
filled to overflowing with copper, this mold produced a musuku ingot, a short cylin-
der with studs on one end and an irregular flange where the copper had spilled out
of the top of the mold (Mamadi 1940 Fig. 5.14). These objects were used in ritual
and ceremonial settings associated with political authority and healing.
Fig. 5.14 Musuku ingot
housed at Iziko Museums,
Cape Town (Photograph
Credit: Author)
Fig. 5.13 X-shaped copper
ingots in use in much of
Southern Zambezia in the
Iron Age, particularly after
AD1000 (Photograph Credit
Author)
shadreck.chirikure@uct.ac.za

116 5 Socializing Metals
Ellert (1993) studied traditional gold production on the Zimbabwe plateau and
observed that gold dust and nuggets were melted in ceramic crucibles resembling
domestic pottery. It seems that very few specialized crucibles were used for gold
working. By contrast to West African examples such as Asante where gold was a
prerogative of state (Garrard 1980, pp. 135–136), gold production was not associ-
ated with any centralized authority for villagers could freely scour for ore and melt
it for a variety of purposes (Mudenge 1974).
Archaeologically, a number of researchers have studied the microstructures
and chemistry of finished metal objects in Eastern and Southern Africa. In par-
ticular, Kusimba et al. (1994) studied 180 iron artifacts from Swahili sites on the
East African coast. While the techniques were generally similar to those practiced
by other African communities, the late first millennium AD Swahili inhabitants of
Galu and Ungwana worked cast iron which was decarburised to the forgeable low-
carbon steels presumably in crucibles or oxidising hearths (Kusimba and Killick
2003, p. 180). Cold forging and annealing were also practiced together with pres-
sure welding or joining two hot pieces of metal together. It has also been surmised
that the Swahili at Galu may have produced crucible steel, which was used for more
mundane tools such as nails (Kusimba and Killick 2003). While the Swahili’s abil-
ity to produce high-carbon steel has never been questioned, it is possible that the
crucible steel may have been imported (Killick 2009).
In Africa south of the Zambezi, Miller (1996; 2002) carried detailed metallo-
graphic and compositional analyses of iron and copper objects from Early Iron Age
(AD200–1000) and iron, copper, bronze and gold objects from Late Iron Age sites
(AD1000–1850). Most of the Early Iron Age objects came from Nqoma and Di-
vuyu in Botswana, while analysed Late Iron Age ones had a larger geographical
spread. The inventory of iron objects made in the 2000-year history of fabrication in
Southern Africa include needles, awls, bracelets, bangles, spears, axes, arrowheads
and hoes among others. Miller’s detailed work revealed that small ferrous items
were often fully heated or annealed as testified by the presence of spheroids in their
microstructures, which normally result from cumulative overheating (Miller 1996).
Archaeological and metallographic work has shown that the basic fabrication tech-
nology used for working indigenous bloomery iron, copper and gold did not change
sharply over time, despite meaningful changes in the types of objects made (Miller
2002). While small objects prevailed at first-millennium AD sites, larger ones be-
came ubiquitous in the second millennium, at sites such as Bosutswe, Mapungubwe,
Khami and others. Presumably, during the first millennium AD, iron was still a rare
metal, which was consistently recycled, and because it was widely available in the
second millennium AD objects of various sizes could be found (Miller 1996).
Sometimes, softened (annealed) iron was forged into thin sheets and cut into
long strips to make helical wound bangles, often using a flexible fiber core as well
as clips and wrapped beads. Stout iron wire was made by hammering and rotating a
thin rod (Miller 1996). Metallographic and compositional analyses of objects from
Caton-Thompson’s excavations at Great Zimbabwe by Stanley (1931) identified
bronze, brass and leaded gun metal, which were all fabricated using indigenous
techniques. Bronze and brass were certainly imports, and if the early twelfth century
shadreck.chirikure@uct.ac.za

117East and Southern Africa with Occasional References to Sudan and Ethiopia
date is to be believed, imported metal was coming into the interior from the coast
much earlier than some researchers are prepared to accept. This is hardly surpris-
ing, for bronze appeared on the East African coast at AD 700 (LaViolette 2008), and
glass beads and other imports were coming into the interior from the eighth century
AD, if not earlier.
Miller’s (2002) study demonstrated that the techniques invested in fabricating
iron were similar to those invested in copper (AD900–1850), bronze, and gold
working (AD1000–1850). Routine copper, bronze and gold fabrication involved
the alternate heating and hammering of ingots into shape. Often the finished objects
were left to cool slowly, but sometimes hammered sheets of copper, bronze and gold
were cut into thin strips, which were wound around vegetal cores to make bangles.
In the Early and Late Iron Ages, copper beads were made by cutting short lengths
of metal strip or thin rods with a chisel and bending them around with the cut levels
on the inside to create a relatively smooth join without welding or soldering (Miller
2002; Thondhlana and Martinón-Torres 2009). After the second millennium AD,
gold was also hammered into foil, which was attached to wooden poles and objects
as shown by the examples from the Valley Enclosures at Great Zimbabwe (Chi-
punza pers comm) and Mapungubwe (Oddy 1984; Miller 2002).
Copper and gold were cast to produce preformed shapes. Molds and crucibles
for working gold and copper have been found in Southern Africa. Crucibles used
in most of Southern and Central Africa basically resembled normal pottery (Bisson
2000). Thondhlana (2013) excavated such crucibles dating to the Kgopolwe (early
second millennium AD) levels at Shankare in Phalaborwa. Miller and Hall (2008)
also found such crucibles at Rooikrans dating from the sixteenth century AD on-
wards. However, often there were also specialized crucibles such as the one from
Phalaborwa, which has a pourer (Thondhlana 2013). There are also examples of
sandstone crucibles, which were used for melting brass in the sixteenth and seven-
tieth centuries. However, few studies of crucibles have been conducted in Southern
Africa. Archaeologists have also recovered ingot molds (Fig. 5.15) below which
is on display at the Natural History Museum in Zimbabwe. X-shaped ingots were
Fig. 5.15 Half of a cross-
shaped copper ingot mold
(carved from steatite?) used
to produce ingots in 5.15
recovered from Zimbabwe,
on display at the Natural
History Museum, Bulawayo,
Zimbabwe (Photograph:
Author)
shadreck.chirikure@uct.ac.za

118 5 Socializing Metals
found at Great Zimbabwe indicating that the mold may date to any period in the
second millennium AD.
Gold dust was melted in ceramic crucibles resembling domestic pottery and was
often cast in soapstone molds to produce spherical beads (Fig. 5.16).
Of all the techniques used for copper, bronze and gold working, wire drawing
seems to have been the most challenging to execute. It was carried out using plates
some of which were oblong in cross section. According to Bisson (2000), wire draw
plates had circular holes ranging between 3.5 cm and 1 mm in diameter. During the
process of wire drawing, the draw plate was fixed to a stand. A heated copper rod,
with its end specifically tapered for the process, was pulled through the draw plate
using a pair of tongs. The resultant wire was pulled consecutively through a graded
series of holes in the iron plate until the desired diameter was achieved (Brown
(1995). Usually, the last grade was a millimeter thick, and this extremely fine gauge
copper wire was an important constituent of spirally wound bangles, which are
ubiquitous at Southern African archaeological sites. These are often found in large
quantities in burials of high-status individuals (Fig. 5.17) from capitals such as Da-
nangombe (AD1680–1850) and Mapungubwe.
Ingombe Ilede, a mid-second-millennium AD ossuary on the banks of the Zam-
bezi River in Zambia (Fig. 5.1) has yielded some of the best evidence of copper
Fig. 5.17 Femur and bangles from a high-status individual excavated from the dry-stone-walled
Zimbabwe culture site of Danangombe (AD1680–1850), Central Zimbabwe. The excavators
retrieved this bone with those high numbers of copper or copper alloy bangles, which are quite
numerous. It is on display in the Zimbabwe Museum of Human Sciences. Photograph: Author
Fig. 5.16 Second-millen-
nium AD soapstone ingot
mold with a gold pellet insert,
Natural History Museum,
Bulawayo (Photograph
credit: Author)
shadreck.chirikure@uct.ac.za

119Metals in Society: The Anthropology of Smithing and Metal Objects
working in Africa. Discovered in the 1960s, the site consists of various burial
chambers. The accompanying grave goods speak of specialized copper workers
who were buried with a complete set of wire drawing equipment, unused X-shaped
copper ingots and an amazing copper bar that represented the first stages of the wire
drawing process (Phillipson and Fagan 1969). To crown it all, finished copper wire
was also part of the funerary goods.
As in West and Central Africa, iron was used to make utilitarian objects such
as hoes, axes, hammers and among arrow heads while initially copper, but later
brass, bronze and gold were used to make decorative and ornamental objects such
as bracelets, beads, earrings, symbolic hoes such as the bronze ones from Khami
and Great Zimbabwe, and thin sheets/foil. However, this was not a static dichotomy
for iron was also used to make bangles, anklets and ceremonial objects such as
gongs, axes and spears.
One striking difference between metalworking in Southern and West Africa is
that while endogamous occupational groups or castes have deep roots in areas such
as Mali (Robion-Brunner et al. 2013), such castes are not known in Southern Af-
rica. The status of smiths was variable such that in some cases they occupied a high
status and even founded chiefdoms as discussed by Mackenzie (1975) in the case of
the seventeenth-century Njanja in Central Zimbabwe. Maggs (1992) discusses the
example of iron making in the Zulu state where the Amalala were reduced to a class
of attached specialists that produced metal for Shaka’s army. However, even in this
case, the Amalala did not transform into a caste.
This points to variability in the social status of metal workers across Africa and
through time. Smiths and potters in Ethiopia and Sudan belong to the same caste re-
garded as lowly at the same time as feared (Haaland 2004b). Intermarriage between
castes and non-caste members was forbidden. This case is analogous to that docu-
mented in West Africa by Tamari (1991) and others such as McNaughton (1993).
In a different context, ancient Egyptian and Nubian metalworkers, as we have seen,
occupied a low class, but this was not inherited. Anybody who worked hard and
became learned could be an official, while those who failed became artisans (Scheel
1989). These differences in the organization of metalworking at different points of
African history warn against the dangers of extrapolating observations from one
area and time period into another. Generalizations must be finely calibrated in rela-
tion to context-specific situations; otherwise, they create reconstructions perfectly
sensible to our minds but tangential with what may have happened in the past.
Metals in Society: The Anthropology of Smithing
and Metal Objects
As Mauss (1954) has submitted, technology is a complete social and cultural phe-
nomenon. As an important element of the chaîne opératoire of metal production
and use, smithing, like smelting and mining, was also associated with cultural ideas
of its own but which drew from the general symbolic load as articulated in broader
shadreck.chirikure@uct.ac.za

120 5 Socializing Metals
society. Gods such as Maat played an important role in Egyptian and Nubian met-
alworking just as they did in life from Dynastic to Ptolemaic times. The metals
themselves also had values associated with them from the valued gold, silver and
electrum used in palaces, temples and tombs to iron which when established was
used for day to day objects. The centralized organization of metal fabrication en-
abled through effective record keeping ensured that these states controlled metal
working in ways that probably did not characterize empires such as ancient Ghana
and Mali where gold producers appear to have operated independently and only
remitted tribute to the capital (Levtzion 1973). This was probably also the case in
relation to Mapungubwe, Great Zimbabwe, and polities that emerged afterwards.
Here, despite centralized organization, the indication and absence of large-scale
debris from capitals is that production took place in the hinterland. Lack of an effec-
tive means of record keeping may have comprised central control of production. As
such, Mudenge (1974) argues within the context of the Rozvi State (AD1680–1850)
that commoners produced and traded metal, only submitting taxes to the courts. The
outcome of this situation was that some individuals became very wealthy with the
result that metalworking was highly respectable in this area.
In terms of the human procreational paradigm, the reduced iron metal was meta-
phorically seen as a child in sub-Saharan Africa (Herbert 1993). Although smelting
was sometimes practiced away from residents because of taboos and other fac-
tors, smithing took place in the village (Schmidt 2009). The smithing of objects
took place in full view and was not shrouded in secrecy or mystery (Cline 1937).
A variety of objects were fashioned from metal and these assumed a persona, for
they played a central role in solving societal problems as well as in negotiating and
structuring personal relations.
Metal objects gained value depending on the context in which they were used.
For example, the iron hoe occupied an important role in Shona fertility and associ-
ated beliefs. A hoe or badza made it possible for the Shona to cultivate the land
to harvest food, so too was it important for weeding the fields and digging the
earth, placing it at the center of societal renewal and growth (Bent 1896; Bourdil-
lon 1976). Not surprisingly, the iron hoe fundamentally structured human relations,
particularly those related with fertility and growth. For some, iron hoes were an
important medium in marriage transactions, as for example among the Shona for
whom roora or lobola (bride wealth) is often referred to as badza or the hoe (Childs
1991). During marriage negotiations, a father whose daughter was about to be mar-
ried demanded the symbolic iron hoe to ensure continued productivity of his house-
hold. The iron hoe too was an important a symbol of divorce. If a husband wanted to
divorce his wife, he gave her a worn out hoe to take to her parents, a concept known
as gupuro in Shona (Bourdillon 1976). In between these human relations, a hoe was
a basic tool for achieving a variety of tasks in society, but because of these utilitar-
ian and symbolic roles, it served a nodal role in a web of social relations. Indeed,
the iron bundles known as bikie in Southern Cameroon also played a similar role in
this part of the world (Guyer 2012).
A fundamental way in which metal is socialized is through giving names, often
ones associated with the tasks and not the metal. This socialization is hardly sur-
prising because metal symbolically represented nature that had been tamed through
shadreck.chirikure@uct.ac.za

121Conclusion
reduction. Once smithed into artifacts, such objects acquired names and meanings
which had little or nothing to do with metal. A Shona copper or iron needle was
known as tsono and it was used for sewing ( kusona). A Shona iron axe was not
referred to as an iron axe, rather it was known as an axe or demo. This process of
socializing material culture equally applies to other African communities and those
beyond (see Appandurai 1986). The forging of objects and their use socialized metal
artifacts, placing them alongside other material culture whose role and function was
more dependent on context than material. As such, the objects are fundamentally
human and it is human to make, use, personify and discard the objects.
Of course, the role of metal objects far extends beyond the ideas summarized
here. Metal became the pivot on which trade and exchange, interaction, culture
contact, and even warfare were anchored. This broader theme is explored in detail
in Chap. 6.
Conclusion
Building on the history of metalworking in Africa’s multiple regions, copper in
parts of West, Central, East and Southern Africa was largely reserved for ornamen-
tation, or in some cases sculpture, while agricultural tools and weapons were always
made of iron. In this respect, sub-Saharan Africa was very different from Eurasia
in the Bronze Age, and even from Egypt and North Africa, where bronze weapons
were widely used. Even after the introduction of gold and bronze, these too were
restricted to the ornamental and ceremonial domains. In fact, it appears as if gold,
with few exceptions (e.g., Asante; Garrard 1980), did not appeal to commoners in
sub-Saharan Africa who preferred the red colour and tonality of copper when com-
pared to gold (Herbert 1993). As a result, control of gold production was not one
of the ways in which elites maintained their power. Rather, they controlled the land
and its fertility through a link to their ancestors (Mudenge 1974). This way copper,
iron and gold working together with agricultural success were linked to ancestors
who bestowed this through their descendants in power. Such a situation also differs
with Egypt and North Africa where rulers had to account for every bit of metal and
reserving for their exclusive use the highly valued gold, silver and electrum.
Linked to the large-scale iron production in West and Central Africa as well as
at Meroe in the Sudan is the semi-industrial production of iron billets at (1) Sukur
where iron was traded north into the Lake Chad Basin (David and Sterner 1997);
(2) Bassar in Togo for sale to Hausa caravans (de Barros 2013); (3) Meroe for Ptol-
emaic and Roman Egypt (Shinnie 1985); (4) the Cameroon grasslands (Warnier
and Fowler (1979); and (5) among the Dogon in present-day Mali (Robion-Brunner
et al. 2013). It is possible that high levels of production in West and Central Africa
were a response to the slave trade (MacEachern 1993), but it was also linked with
demographic increases, something which was not of consequence in Southern Af-
rica where the scale of production was hardly comparable.
Finally, there are substantial intersections between techniques employed in vari-
ous regions, just as there were major differences. For example, the lost wax technique,
shadreck.chirikure@uct.ac.za

122 5 Socializing Metals
which represents the pinnacle of copper and copper alloy casting in Egypt and West
Africa, was unknown in Southern and Eastern Africa. The techniques invested in
hot or cold working were similar across the metals and alloys used in African civi-
lizations. Differentials in the physical properties of the individual metal motivated
for the casting of copper, tin and gold. These differences were stimulated by local
innovations as well as by broader regional interactions about which we continue to
learn through archaeological sources (Mitchell 2005).
References
Aldred, C. (1971). Jewels of the pharaohs: Egyptian jewelry of the dynastic period. New York:
Praeger.
Appadurai, A. (Ed.) (1986). The social life of things: Commodities in cultural perspective. Cam-
bridge: Cambridge University Press. (Archetype).
Bellamy, C. V., & Harbord, F. W. (1904). A West African smelting house. Journal of the Iron and
Steel Institute, 66, 99–126.
Bent, J. T. (1896). The ruined cities of Mashonaland: Being a record of excavation and exploration
in 1891. London: Longmans, Green, and Co.
Bisson, M. (2000). Precolonial copper metallurgy: Sociopolitical context. In M. Bisson, P. de Bar-
ros, T. C. Childs, & A. F. C. Holl (Eds.), Ancient African metallurgy: The sociocultural context.
(pp. 83–146). Walnut Creek: AltaMira.
Bourdillon, M. F. (1976). The Shona people: An ethnography of the contemporary Shona with
special reference to their religion. Gweru: Mambo Press.
Brown, J. (1995). Traditional metalworking in Kenya. Oxford: Oxbow.
Chikwendu, V. E., Craddock, P. T., Farquhar, R. M., Shaw, T., & Umeji, A. C. (1989). Nigerian
sources of copper, lead and tin for the Igbo-Ukwu bronzes. Archaeometry, 31(1), 27–36.
Childs S. T. (1991). Style, technology and iron smelting furnaces in Bantu-speaking Africa. Jour-
nal of Anthropological Archaeology, 10(4), 332–359.
Chirikure, S. (2006). New light on Njanja iron working: Towards a systematic encounter between
ethnohistory and archaeometallurgy. South African Archaeological Bulletin, 61, 142–151.
Chirikure, S., Burrett, R., & Heimann, R. B. (2009). Beyond furnaces and slags: A review study of
bellows and their role in indigenous African metallurgical processes. Azania: Archaeological
Research in Africa, 44(2), 195–215.
Cline, W. B. (1937). Mining and metallurgy in Negro Africa ( No. 5). Menasha: George Banta.
Craddock, P. T. (2000). From hearth to furnace: Evidences for the earliest metal smelting technolo-
gies in the Eastern Mediterranean. Paléorient, 26(2), 151–165.
Crawhall, T. C. (1933). Iron working in the Sudan. Man, 48, 41–43.
Crew, P. (1991). The experimental production of prehistoric bar iron. Historical Metallurgy, 25(1),
21–36.
David, N., & Sterner, J. (1997). Water and iron: Phases in the history of Sukur. In H. Jungraith-
mayr, D. Barreteau, & U. Seibert (Eds.), L’homme et l’eau dan le basin du lac Tchad. (pp. 255–
270). Actes du Colloque Mega-Tchad. Paris: Colloques et Seminaires, Editions ORSTOM.
David, N., Heimann, R., Killick, D., & Wayman, M. (1989). Between bloomery and blast furnace:
Mafa iron-smelting technology in North Cameroon. African Archaeological Review, 7(1),
183–208.
de Barros, P. (2013) A comparison of early and later Iron Age societies in the Bassar region of
Togo. In J. Humpris & Th. Rehren (Eds.), The world of iron (pp. 10–21). London: Archetype.
de Garis Davies, N. (1943). The Tomb of Rekh-mi-Re⊂ at Thebes (Vol. 2). New York: Metropolitan
Museum of Art.
de Maret, P. (1985). The smith’s myth and the origin of leadership in Central Africa. In R. Haaland
& P. Shinnie (Eds.), African iron working — ancient and traditional (pp. 73-87). Oslo: Norwe-
gian University Press
shadreck.chirikure@uct.ac.za

123References
Dewey, W. J. (1991). Weapons for the ancestors. Des Moines: University of Iowa, Department of
Art History.
Ellert, H. (1993). Rivers of gold. Gweru: Mambo Press.
Emery, W. B. (1963). Egypt exploration society preliminary report on the excavations at Buhen,
1962. Kush, 11, 116–120.
Emery, W. B. (1971). Preliminary report on the excavations at North Saqqâra, 1969–70. The Jour-
nal of Egyptian Archaeology, 57, 3–13.
Garrard, T. (1980). Akan weights and the gold trade. London: Longman.
Garrard, T. F. (2011) (1989). African gold: Jewellery and ornaments from Ghana, Côte dʼIvoire,
Mali and Senegal in the collection of the Barbier-Mueller Museum. Munich: Prestel.
Guyer, J. I. (2012). Soft currencies, cash economies, new monies: Past and present. Proceedings of
the National Academy of Sciences, 109(7), 2214–2221.
Haaland, R. (2004b). Iron smelting—a vanishing tradition: Ethnographic study of this craft in
South-West Ethiopia. Journal of African Archaeology, 2(1), 65–79.
Herbert, E. W. (1993). Iron, gender, and power: Rituals of transformation in African societies.
History, 32(2), 221–50.
Holub, E. (1881). Seven years in South Africa: Travels, researches, and hunting adventures, be-
tween the diamond-fields and the Zambesi ( 1872–79) (Vol. 2). London: S. Low, Marston,
Searle & Rivington.
Killick, D. J. (2009). Agency, dependency and long-distance trade: East Africa and the Islamic
World, ca. 700–1500 C.E. In S. Falconer & C. Redman (Eds.), Polities and power: Archaeo-
logical perspectives on the landscapes of early states (pp. 179–207). Tucson: University of
Arizona Press.
Kusimba, C. M., & Killick, D. (2003). Iron working on the Swahili Coast of Kenya. In C. M.
Kusimba & S. B. Kusimba (Eds.) East African archaeology: Foragers, potters, smiths, and
traders (pp. 99–115). Philadelphia: University of Pennsylvania Museum of Archaeology and
Anthropology.
Kusimba, C. M., Killick, D. J., & Cresswell, R. G. (1994). Indigenous and imported metals at
Swahili sites on the coast of Kenya. In S. T. Childs (Ed.) Society, culture and technology in
Africa. Philadelphia: MASCA, University of Pennsylvania Museum of Archaeology and An-
thropology.
LaViolette, A. (2008). Swahili cosmopolitanism in Africa and the Indian Ocean world, AD 600–
1500. Archaeologies, 4 (1), 24–49.
LaViolette A., & Fleisher J. (2005) The archaeology of sub-Saharan urbanism: Cities and their
countrysides. In A. B. Stahl (Ed.), African Archaeology: A critical introduction (pp. 327–352).
Oxford: Blackwell.
Levtzion, N. (1973). Ancient Ghana and Mali (Vol. 7). London: Methuen.
MacEachern, S. (1993). Selling the iron for their shackles: Wandala-Montagnard interactions in
Northern Cameroon. Journal of African History, 34, 247–247.
Mackenzie, J. M. (1975). A pre-colonial industry: The Njanja and the iron trade. Nada, 11(2),200–
220.
Maggs, T. (1992). My father’s hammer never ceased its song day and night’: The Zulu ferrous
metalworking industry. Southern African Humanities, 4, 65–87.
Mamadi, M. F. (1940). The copper miners of Musina. In N. J. van Wamelo (Ed.), The copper min-
ers of Musina and the early history of the Zoutpansberg, (pp. 81–87). Pretoria (Department of
Native Affairs, Union of South Africa): Ethnological Publications VIII.
Mauss, M. (1954). The gift: Forms and functions of exchange in archaic societies (No. 378). New
York: Norton.
McNaughton, P. R. (1993). The Mande Blacksmiths: Knowledge, power, and art in West Africa.
Bloomington: Indiana University Press.
Miller, D. (1996). Tsodilo jewellery. Cape Town: University of Cape Town Press.
Miller, D. (2000). 2000 years of indigenous mining and metallurgy in Southern Africa—a review.
South African Journal of Geology, 98, 232–238.
Miller, D. (2001). Metal assemblages from Greefswald areas K2, Mapungubwe Hill and Mapun-
gubwe Southern terrace. The South African Archaeological Bulletin, 56, 83–103.
shadreck.chirikure@uct.ac.za

124 5 Socializing Metals
Miller, D. (2002). Smelter and smith: Iron Age metal fabrication technology in Southern Africa.
Journal of Archaeological Science, 29(10), 1083–1131.
Miller, D. E., & Hall, S. L. (2008). Rooiberg revisited–the analysis of tin and copper smelting
debris. Historical Metallurgy, 42(1), 23–38.
Miller, D. E., & Van der Merwe, N. J. (1994). Early metal working in sub-Saharan Africa: A review
of recent research. Journal of African History, 35(1), 1–36.
Miller, D., Killick, D., & Van der Merwe, N. J. (2001). Metal working in the Northern Lowveld,
South Africa AD 1000–1890. Journal of Field Archaeology, 28(3–4), 3–4.
Mitchell, P. (2005). African connections: Archaeological perspectives on Africa and the wider
world. Walnut Creek: AltaMira.
Mudenge, S. I. (1974). The role of foreign trade in the Rozvi empire: A reappraisal. Journal of
Oddy, A. (1984). Gold in the Southern African Iron Age. Gold Bulletin, 17(2), 70–78.
Nixon, S., Rehren, T., & Guerra, M. F. (2011). New light on the early Islamic West African gold
trade: coin molds from Tadmekka, Mali. Antiquity, 85 (330), 1353–1368.
Ogden, J. (2000). Metals. In P. Nicholson & I. Shaw (Eds.), Ancient Egyptian materials and tech-
nology (pp. 148-176). Cambridge: Cambridge University Press.
Phillipson, D. W., & Fagan, B. M. (1969). The date of the Ingombe Ilede burials. The Journal of
African History, 10(2), 199–204.
Rehren, T., Belgya, T., Jambon, A., Káli, G., Kasztovszky, Z., Kis, Z., & Szőkefalvi-Nagy, Z.
(2013). 5,000 years old Egyptian iron beads made from hammered meteoritic iron. Journal of
Archaeological Science, 40(12), 4785–4792.
Reid, A., & MacLean, R. (1995). Symbolism and the social contexts of iron production in Karag-
we. World Archaeology, 27(1), 144–161.
Robion-Brunner, C, Serneels, V, & Perret, S. (2013). Variability in iron smelting practices: As-
sessment of technical, cultural and economic criteria to explain the metallurgical diversity in
the Dogon area (Mali). In J. Humpris & Th. Rehren (Ed.), The world of iron (pp. 257–265).
London: Archetype.
Sasson, H. (1964). Iron-smelting in the hill village of Sukur, North-Eastern Nigeria. Man, 64,
174–178.
Scheel, B. (1989). Egyptian Metalworking and Tools. Oxford: Shire Publications.
Schmidt, P. R. (2009). Tropes, materiality, and ritual embodiment of African iron smelting fur-
naces as human figures. Journal of Archaeological Method and Theory, 16(3), 262–282.
Scott, D. A. (1991). Metallography and microstructure of ancient and historic metals. Oxford:
Oxford University Press. (Getty Conservation Institute; J. Paul Getty Museum, in association
with Archtype Books).
Shaw, T. (1970). Igbo-Ukwu: An account of archaeological discoveries in Eastern Nigeria (Vol. 2).
Evanston: Northwestern University Press.
Shinnie, P. L. (1985). Iron working at Meroe. In R. Haaland & P. L. Shinnie (Eds.), African iron
working—Ancient and traditional (pp. 28–35). Oslo: Norwegian University Press.
Stanley, G. H. (1931). Some products of native iron smelting. South African Journal of Science,
28, 131–134.
Stayt, H. A. (1931). The Bavenda (No. 58). International Institute of African Languages & Cul-
tures. Oxford: Oxford University Press.
Tamari, T. (1991). The development of caste systems in West Africa. Journal of African History,
32, 223.
Thondhlana, T. (2013). Metalworkers and smelting precincts: Echnological reconstructions of sec-
ond millennium copper production around Phalaborwa, Northern Lowveld of South Africa.
Unpublished Doctoral dissertation, University College London, London.
Thondhlana, T. P., & Martinón-Torres, M. (2009). Small size, high value: Composition and manu-
facture of second millennium AD copper-based beads from northern Zimbabwe. Journal of
African Archaeology, 7(1), 79–97.
Todd, J. A. 1985. Iron production by the Dimi of Ethiopia. In R. Haaland & P. L. Shinnie (Ed.),
African iron working-ancient and traditional (pp. 88–101). Oslo: Norwegian University Press.
Vansina, J. (1969). The bells of kings. The Journal of African History, 10(02), 187–197.
Warnier, J. P., & Fowler, I. (1979). A nineteenth-century Ruhr in Central Africa. Africa, 49 (04),
329–351.
shadreck.chirikure@uct.ac.za

125
Chapter 6
The Social Role of Metals
© The Author 2015
S. Chirikure, Metals in Past Societies, SpringerBriefs in Archaeology,
DOI 10.1007/978-3-319-11641-9_6
Introduction
The utilitarian, aesthetic and ceremonial values of metals are a pivot on which not
just quotidian activities but also luxuries consumed by humanity were anchored
since the advent of metallurgy whether in Africa, Latin America or Eurasia. Around
c. 5500–6000 years ago, copper ores were used as ornaments in the Chalcolithic pe-
riod of the Middle East and adjacent regions (Hauptmann 2007). Since then, metals
have played an increasing role in the articulation of value systems, accumulation of
wealth, and innovation and knowledge transfer. As such, metals occupy a privileged
place within societies of the preindustrial period.
Diverse opinions prevail among researchers when it comes to the question of
whether the adoption and subsequent entrenchment of metallurgy in preindustrial
societies was a revolutionary change. For decades, the Childean view that the adop-
tion of metallurgy caused a social revolution in Western Eurasia was influential
(Childe 1930). This academic footing is now being challenged by positions which
argue that the impact of metallurgy was gradually felt. Whether in Egypt, West or
Southern Africa, metallurgy only increased in scale sometime after its introduc-
tion. It was only after a lengthy period—in the view of some, on a scale of millen-
nia—that humanity’s dependence on metals intensified. This anti-revolution school
further criticizes Childe and his ardent students for embracing a presentist view
which projects the all-encompassing might of metallurgy today deep into the past
(Killick and Fenn 2012). Whatever the case may be, this contestation around the
revolutionary and non-revolutionary consequences of metallurgy passes fascinat-
ing comments not just on the spatial and temporal continuities and changes across
human history, but also on the scale and evolution of techno-social systems in the
preindustrial world.
By consequence of valuation, access to metals, or lack thereof, became a tool
for creating social differentiation and inequality in society. Metallurgy became
entrenched in the evolution of craft specialization in predynastic Egypt with the
shadreck.chirikure@uct.ac.za

126 6 The Social Role of Metals
rulers and those who controlled metal production, distribution and use becoming
wealthier and concomitantly more powerful than those without (Childe 1930). After
some time, the efficiency of metal tools in domains such as agriculture made food
production comparatively easier, while metal weapons enabled various polities to
defend and expand their territories. In the long term, urbanization promoted settle-
ment aggregation in some but not all metalworking regions.
Because the distribution of ores is geologically specific and uneven, trade be-
tween resource-rich and resource-deficient areas understandably emerged. There-
fore, local, regional and long-distance trade is one of the direct consequences of
the value which metals had in society. As metal sources became depleted, societies
sought access to new sources, initiating proto-forms of globalization anchored on
not just land links but also sea links. For example, from the Early Bronze Age,
Egypt was part of the trade and exchange relationship in the Middle East. By 2000
BC, Egypt had expanded into Nubia in search of gold. This continued until Roman
times when Meroe used to supply iron to the Romans. Sub-Saharan Africa also
supplied iron and gold metal to the nascent global system of the Islamic world via
the trans-Saharan trade and the East African coast-based trade particularly after
the first-millennium AD (Nixon 2009; Nixon et al. 2011). With the advent of trade
via the Atlantic coast from the 14th-century AD onwards, African metals paid for
valuables initially in Europe and later for spices in India. Trade in metals became
invariably associated with the development of, and transfer of, diverse value and
knowledge systems. Fighting over control of metals sources spawned conflicts and
cleavages that often resulted in mobility, migration, war and even colonization.
General Impact of Metals in Society
Throughout the world, metallurgy for a long time coexisted and only gradually re-
placed stone as the major medium for manufacturing tools for a variety of purposes.
This conservatism is understandable given that for a greater part of humanity, stone
was the primary raw material for making tools. In terms of materiality, metals hold
several advantages over stone because of their ductility, ability to undergo plastic
deformation, malleability and strength (Scott 1991). When realized, these qualities
dictated that metals were used in a variety of domains, as underlined by the values
with which humanity associated them. In the pre-Columbian New World, stone re-
mained the basic raw material for tools and weapons while metals were enjoyed by
elites. The range of the values of metals varied from the more mundane ones to the
higher order ones as dictated by rarity and scarcity.
It has been argued that when agricultural communities adopted metallurgy,
the technology had a beneficial effect through the provisioning of efficient tools.
Although there are instances of cultivation with polished stone axes in West Af-
rica (for example the Kintampo complex in Ghana and the Neolithic in Gabon;
Anquandah 1993), it seems highly improbable that swidden agriculture in the
savanna woodlands (especially the miombo of Central and East Africa) would
shadreck.chirikure@uct.ac.za

127General Impact of Metals in Society
have been possible without iron (Killick 2014). Metal tools such as hoes and axes
made it far easier to cultivate the land and to clear the vegetation which in turn
opened up more land for cultivation (Phillipson 2005, p. 214). The burning of
the trees produced ash, which increased the fertility of the land. These activities
resulted in increased yields, which made it easier to sustain growing populations.
In Southern Africa, the advent of metallurgy supported agriculture, opened up the
land which drove away tsetse flies making it possible to practice cattle husbandry
(Mitchell 2002). However, it must be noted that by no means did all communi-
ties adopt metals for food production. Some Aksumites used stone implements
for cultivation until the early first-millennium AD because possession of metal
correlated with class (Phillipson 2005). Another parallel can be found in the New
World where before Columbus, various agricultural communities used stone
for agriculture and metals for luxuries (Killick and Fenn 2012). Another area in
which metallurgy supported the local economy and productive base in sub-Saha-
ran Africa is within the sphere of hunting. Metallurgy facilitated the production
of spears, arrowheads and axes that were more effective in hunting elephants for
ivory, a very significant commodity in trade and exchange relationships between
Africa and Eurasia, particularly after the mid-first-millennium AD when various
parts of Africa were integrated into long-distance trade via the Sahara and the
Indian Ocean. However, the advantages of metal over stone were only realized
over time and were thus not immediate.
Besides these utilitarian tasks, metals were highly valuable in making ceremoni-
al and decorative items. Various bronze and gold castings adorned ancient Egyptian
temples and palaces. The famous Igbo Ukwu bronzes dating between AD 900 and
1200 (Shaw 1970) and the Benin bronzes made from the 16th-century AD onwards
were used to express power and to commemorate deceased kings and were an in-
imitable expression of the Oba’s power and authority (Herbert 1984). In Great Zim-
babwe (AD 1100 to 1550; Fig. 6.1), archaeologists recovered impressive bronze
spearheads and iron gongs (Fig. 6.2) which, as elsewhere in Central and West Africa
were associated with the power of kings (Vansina 1969). Copper, bronze, gold and
iron were used to make beads and bangles for personal adornment (Miller 1996).
The values associated with colours and rarity of these metals also created a distin-
guishing factor between elites and non-elites (Smith 1981). For instance, in Ancient
Egypt and Nubia, gold was an elite metal hardly enjoyed by the commoners. Metal
was subjected to rigorous bureaucratic control. In contrast, commoners in Southern
Africa produced gold that was traded by the Rozvi state (AD 1680–1850) and even
exchanged it for glass beads which they gave to rulers are tribute. (Mudenge 1974).
As such, they were free from the bureaucratic control, which gave them some de-
grees of freedom when compared to ancient Egypt and Nubia.
The power of the hugely successful Asante kingdom (after AD 1600) in modern-
day Ghana was symbolically vested in the Golden Stool which became an emblem
of the polity (Garrard 1980, pp. 63–64). The Asante royalty also consumed sig-
nificant amounts of gold, which seemed as a high-status metal (Garrard 2011). In
Southern Africa, gold was a universal feature of centres of power such as Great
Zimbabwe, Khami and Mapungubwe (Figs. 6.3 and 6.4) but its production was
shadreck.chirikure@uct.ac.za

128
not under very strict bureaucratic control. Besides making symbolic objects and
ornaments for personal adornment, thin sheets of gold were attached to wooden
posts inside royal abodes at places such as Great Zimbabwe. The gold sheeting was
designed to enhance the aesthetic appeal of places of power. In contrast, there is al-
most no record of the presence of gold at commoner settlements in much of Africa.
It appears that copper was the preferred metal (Herbert 1984), but it must be borne
in mind that little work has been directed at commoner residences.
6 The Social Role of Metals
Fig. 6.1 Location of metal working groups and places discussed in the text
shadreck.chirikure@uct.ac.za

129General Impact of Metals in Society
Fig. 6.4 Mapungubwe
golden rhino made of gold
leaf (size, c. 10 cm). It
symbolized the “majesty” of
kingship. (Source: University
of Pretoria Mapungubwe
Collection)
Fig. 6.3 Gold leaf bowl
from Mapungubwe, Southern
Africa. It is believed that the
leaf was attached to a wooden
core which has since decayed
(Miller 2001). (Source: Uni-
versity of Pretoria Mapun-
gubwe Collection)
Fig. 6.2 Iron gong recov-
ered from Great Zimbabwe.
Natural History Museum,
Bulawayo (Photo: Author)
shadreck.chirikure@uct.ac.za

130 6 The Social Role of Metals
From predynastic Egypt to Ptolemaic Egypt and Nubia, gold foil decorated the
interior of temples, tombs and royal palaces and was thus a high-status metal from
its discovery onwards. However, there are some African communities particularly
those in Central Africa where copper and not gold was an important and high-status
metal (Herbert 1984). Indeed, Sanga burials dating from the first-millennium AD up
to the mid-second-millennium AD in the Upemba Depression had varying quanti-
ties of copper (Bisson 2000). Indeed, Cline (1937) argues that some communities in
modern-day Angola told the Portuguese that alluvial gold was abundant in some of
the major rivers but that they did not value gold when compared to copper.
The role of metals in the social nexus between societies is illustrated by
MacEachern’s study of power and political relations between the Muslim Wandala
state and populations on the margins, particularly those known as montagnards or
non-Muslims. MacEachern (1993) combined historical observations with archaeo-
logical work to document these relations. Non-Muslims occupied the Mandara mas-
sif while the Wandala lived on the plains. MacEarchern (1993) noted significant
demographic shifts marked by increased montagnard occupation of the mountain
from the early second-millennium AD onwards. These mountain groups were or-
ganized in small patrilineages with densities of between 50 and 200 people per
square km. They specialized in iron production aimed beyond local requirements.
The montagnards traded iron ore and iron blooms in exchange for foodstuffs and
other necessities brought by the Wandala. The Wandala, in turn, traded the iron
with Bornu and other leading centres of the time. Ironically, the guns and cavalry
obtained from this trade enabled Wandala people to capture montagnards as slaves,
resulting in the latter being shackled by iron they had produced (MacEachern 1993).
This example shows the dynamic and changing relationships in society that cas-
caded from iron production, its exchange and associated consequences. Indeed, it
has been suggested that in other regions of West Africa such as Bassar, the trans-
Atlantic slave trade resulted in increased iron production, although over time, as-
sociated insecurity is supposed to have decreased production levels. However, as
the montagnard and Wandala case study shows, the situation was very complex
because overall the trade could be beneficial and at other times toxic. It is the ben-
eficial aspect that fuelled iron production—basic foodstuffs and other luxuries and
necessities were important—but those who accepted iron turned the product against
its producers as they came back to raid for slaves.
Metals, Sociopolitical Complexity and Urbanism
Metallurgy is believed to be one of the key drivers of sociopolitical complexity
and concomitant urbanization (Chirikure 2007). These related ways of societal
organization immensely transformed the African sociopolitical and physical land-
scape. Sociopolitically complex societies are often ranked, have evidence of craft
specialization, division of labour, advanced subsistence and economic systems,
monumental architecture, religion, large populations and writing (Carneiro 1967;
shadreck.chirikure@uct.ac.za

131Metals, Sociopolitical Complexity and Urbanism
Renfrew and Cherry 1986). Urbanism is a form of social organization in which
specialized centres provide and receive specific services from hinterland areas
(LaViolette and Fleisher 2005). The appropriateness of some of these indicators in
non-western communities has been questioned by scholars such as Connah (2001)
and McIntosh (1999) who argue that a contextual approach is required to best un-
derstand African varieties of urbanism.
Although the physical evidence for sociopolitical complexity and urbanism dif-
fer across sub-Saharan Africa (see Mitchell and Lane 2013), the most widely cited
cases of urbanism are those of the West African Sahel, the West African rain forest
and its fringes, the middle Nile in the Sudan, the Ethiopian and Eritrean mountains,
the Swahili settlements of East Africa, the Zimbabwe plateau, Tswana towns of
18th- and 19th century Southern Africa, and the capitals of states in the Upemba
depression in Central Africa (Connah 2001). Historical evidence suggests that there
was a tendency for population to agglomerate in much of sub-Saharan Africa, such
that by the close of the 19th century, larger areas of precolonial Africa were quintes-
sentially urban in character (Fletcher 1993; Fig. 6.5).
The quintessential sub-Saharan urban centre has distinguishing characteristics
which vary from place to place, but the agglomeration of varying numbers of people
in one area seems to be the common denominator. Often, urban centres in Central
and West Africa were distinguished by settlement clusters demarcated by streets.
In Sudanic West Africa, perimeter walls of mud brick often encircled the urban
settlement (Ogundiran 2005). In the forest zone, major earthworks created a net-
work of cells in which a dispersed urban population lived. In Southern Africa, the
Zimbabwe culture urban centres such as Mapela, Mapungubwe, Great Zimbabwe
and Khami were heavily built up and consisted of impressive dry-stone walled en-
closures and platforms constructed using the precise placement method without any
binding mortar (Chirikure et al. 2013). Metallurgical specialization and division
of labour is an all pervading characteristic of most urban centres in Egypt, Nubia
and Aksum, but elsewhere in West Africa there is little evidence of specialist metal
production at urban centres (for exception see Garrard’s 2011 discussion of special-
ist gold artisans employed by the Asante kings). In Southern Africa, Tswana towns
such as Marothodi were specialist metal-producing towns that worked copper, iron
and tin bronze exchanging the metals and alloys locally and regionally (Hall et al.
2006). Although debatable, it has been argued that Great Zimbabwe was serviced
by a large-scale metallurgical industry at places such as the nearby Chigaramboni
Hill (Ndoro 1994).
Overall, the bureaucratic control of metal production and use in Egypt and Nubia
when compared to a contrasting lack of rigorous control in sub-Saharan Africa ques-
tions the depth of arguments that link monopoly over trade in gold with the rise of
urban centers and more complex state systems in the former. It seems that the power
base was rooted in something else other than control of gold trade. Control of the
land, people, cattle and general fertility are among the possibilities (Mudenge 1974).
In order to evaluate metallurgy’s role in the emergence of sociopolitical com-
plexity as broadly defined, it is important to briefly sketch the pathways to urban-
ism in Africa. It seems that different regional histories motivated for differences in
shadreck.chirikure@uct.ac.za

132 6 The Social Role of Metals
pathways to complexity across much of Africa (McIntosh 1999). In West Africa,
archaeologists have advanced the theory that the advent of food production and not
metallurgy created itinerant pastoral elites who built the megaliths of Dar Tichitt
in Southern Mauretania (MacDonald 1998). Similar processes where stone-using
food-producing elites built large monumental structures were also reported at Zilum
on the Chad plain (Magnavita et al. 2006). While these cases are a manifestation of
heterarchy or some form of social stratification, it is not clear whether the societies
Fig. 6.5 Location of prominent urban centers in Africa
shadreck.chirikure@uct.ac.za

133Metals, Sociopolitical Complexity and Urbanism
that built them can be described as fully urban (McIntosh 1999). However, the pres-
ence of gigantic megaliths implicates an ability to mobilize and control pools of
labour by one section of the population. Holl (2009) argues that, even if this early
sociopolitical complexity was not associated with metallurgy, the prevailing evi-
dence suggests that in much of Africa metallurgy became the oil that lubricated state
formation and urbanization.
Southern Africa paints a vivid contrast on the canvas of the role of metallurgy
in the beginning and flourishing of sociopolitical complexity. In this region, agri-
culture appeared simultaneously with iron and copper metallurgy early in the first-
millennium AD. Perhaps the region benefited from its “late adopter” advantage,
skipping experimenting with metallurgy in the face of a deeply entrenched stone
making technology. Yet even in Southern Africa, it took between two and three
centuries for sociopolitical complexity to emerge following settlement of the re-
gion by iron-using agriculturalists (Chirikure et al. 2013). Pwiti (1996) identified
four stages in the organization of Southern African communities between the mid-
first-millennium AD and the early second-millennium AD. In the first stage (ca.
AD 300 to 500), heterarchically organized early farming communities occupied
dispersed villages, but metalworking was a prominent feature of such societies. The
second stage, extending from the 7th- to the 8th-century AD, saw the introduction
of external trade. Archaeologically, there was a shift in production towards goods
with long-distance exchange value (Pwiti 1996). The third stage, from about the
9th-century AD, sees an increase in the volume of external trade and is character-
ized by villages which begin to show evidence of ranking and social differentiation.
Shells from the coast also start to make their appearance (Fig. 6.6). The fourth and
Fig. 6.6 Indian Ocean cowrie and sea shells made their appearance in Southern Africa from AD
700 onwards (Photograph: Author)
shadreck.chirikure@uct.ac.za

134 6 The Social Role of Metals
last stage sees the establishment of state structures such as Mapungubwe, Mapela
and Great Zimbabwe. Fagan (1969) argued for the presence of wide regional trade
networks involving ores and metals in Central and Southern Africa since AD 700.
The size of some of the early to mid-first-millennium villages such as Swart Village
(Chirikure 2007) suggests some form of ranking, but African archaeology has yet
to develop indicators of ranking where locally produced goods such as cattle are
involved. Quite interestingly, there is no evidence of large-scale metal production
in Southern Africa at any period that has implications for organization of produc-
tion. If the production was carried out at villages scattered across the landscape with
the surplus being siphoned through various mechanisms (Mudenge 1974), it is un-
likely that large-scale remains will ever be found. However, this does not negate the
fact that metals formed a critical element of local and long-distance trade. Again,
this challenges the beliefs that even centralized states such as the Rozvi (AD1680–
1850), Great Zimbabwe (AD 1100 to 1550) and Mapungubwe (AD1220–1290) mo-
nopolized trade in metals.
In sub-Saharan Africa, the story of iron smiths and metalworkers who forged
new states and brought civilization is widespread across the breadth and length of
the region from the first-millennium AD until the 19th-century AD (Bocoum 2006;
Chirikure 2007; de Maret 1985; Holl 2009; Humphris and Iles 2013; Killick 2009;
Kim and Kusimba 2008). Holl (2009) presents interesting observations relating to
the contribution of metallurgy to state formation in West Africa. There are some indi-
cations that radical change in the social status of metalworkers took shape in the first
half of the second-millennium AD in West Africa. For example, the founding dy-
nasty of Takrur, the Jaa-Ogo, was of iron-producer extraction. An exclusive control
of the craft and its concomitant esoteric knowledge, was the core reason for the Jaa-
Ogo’s accession to kingship. Takrur was invaded and conquered by a Soninke army
from the neighbouring kingdom of Ghana in the 11th-century AD, prompting the
Jaa-Ogo dynasty to lose political power in the process (Holl 2009). These observa-
tions have resonance with those made by Huysecom and Augustoni (1997) who ar-
gued that the Dogon kings of Mali were reputed ironworkers. Tamari (1991) studied
the development of endogamous occupational groups or castes in West Africa. Iron-
workers and potters belonged to the same caste who inter-married and were often on
the low levels of the social ladder. These endogamous castes were totally absent in
Southern Africa. Tamari (1991) argues that ironworkers were politically powerful in
West Africa until the formation of the state of Mali, which engineered their margin-
alization. This seems to reach some convergence with Holl’s (2009) argument that in
Senegal area, the marginalization of ironworkers was a by-product of Islamization
which took place from the 9th-century AD onwards (see also McNaughton 1993).
The positive correlation between political power and metallurgy was also a char-
acteristic of first-millennium and second-millennium AD Central and East African
communities. Urban centres such as Buganda were supported by extensive iron
production industries which produced weapons and utilitarian tools while the Bach-
wezi used iron objects during royal investiture (Reid and MacLean 1995). Perhaps
the most unequivocal case is that of the Luba and Lunda kingdoms whose ori-
gin myths make it explicit that their founders were great metal smiths (de Maret
shadreck.chirikure@uct.ac.za

135Metals, Sociopolitical Complexity and Urbanism
1985). In these kingdoms, royal inauguration involved symbolic appropriation of
the occult powers of the ironworker, with kings being symbolically transformed
into positions of authority. According to de Maret (1985), this practice is for most
of the time archaeologically detectable in Central Africa. For instance, excavations
of burials in the Upemba depression uncovered people buried with accompanying
ingots (Fig. 6.7), smiths’ tools/symbols demonstrating that metallurgy was a pres-
tige technology used as leverage to political power.
Fig. 6.7 Ingots from burials excavated from the Upemba depression in the modern-day DRC.
(Redrawn from Bisson 1982, p. 131; Fig. 6.2)
shadreck.chirikure@uct.ac.za

136 6 The Social Role of Metals
In Southern Africa, the 17th-century Njanja (part of group labelled Shona in
Fig. 6.1) presents one of the best and most remarkable examples in which renowned
metal smiths used their expertise in metal working to establish a large chiefdom.
The Njanja people migrated from Sena in Mozambique and settled in modern-
day Central Zimbabwe sometime in the 17th century (Chirikure 2006; Mackenzie
1975). Oral traditions claim that Neshangwe, the founder of this chiefdom, was a
reputed ironworker whose skills attracted followers precipitating the establishment
of a chiefdom leveraged by iron production. The Njanja reorganized their produc-
tion by employing a shift system of labour and by being open to use of female
labour when some groups seemed to shun it. The result was a distribution network
covering a 200-km radius (MacKenzie 1975). This and other examples summarized
above demonstrate that metallurgy was an important tool for social engineering.
States were made and unmade in no small part by the symbolic and practical impor-
tance of metallurgy. In fact, metal participated in relations of coercion.
Trade in metals, particularly gold, has been touted as one of the most significant
ingredients in emergence of hierarchical societies and early state systems. In South-
ern Africa, it is believed that gold production was the pulse for the florescence of
state systems based at Khami, Mapungubwe and Great Zimbabwe (Killick 2009).
Gold featured heavily in trade and exchange relationships between the Zimbabwe
Culture states and the Indian Ocean trading system. Ruling elites controlled this
trade and amassed substantial amounts of wealth, which was important in giving
them political power. Not surprisingly, gold was an elite metal, only associated with
centres of power such as Bosutswe, Mapungubwe and Great Zimbabwe to mention
but a few. The available evidence suggests that gold was produced in the hinterland
areas, but commoners were free to trade it for various commodities. In some cases
however, although production was not controlled, once in the hands of elites its ex-
ternal distribution and internal consumption was (Phimister 1974; Mudenge 1988).
Proceeds from trade in gold were then invested in building monumental architecture
and other elite displays of power. The trade in gold, iron and other resources such as
ivory brought in bronze, glass beads and other commodities which were exploited
by elites to cement their power. According to Killick (2009), the introduction of
bronze and gold represent the adoption by the new Southern African elites of an
alien value system, in which the golden colour of gold and bronze supplanted the
red of copper for personal ornamentation among emerging elite.
Although gold trade was important in state formation, the gold trade stimulus hy-
pothesis is fraught with inconsistencies that render it highly problematic (Mudenge
1974). The first is that although gold trade was important, the dominant thinking
makes it appear as if the impact of gold trade was instantaneous when in actual fact
it was gradual. This is supported by the observation that even though trade between
Southern Africa and the Indian Ocean started around AD 700, gold only becomes
visible archaeologically at the beginning of the second-millennium AD when ele-
ments of ranking and complexity had already developed (Pwiti 2005). Trade in gold
only intensified pre-existing processes involving control over iron production, fertil-
ity, land, ritual and other local ideologies (Chirikure 2007). In West Africa, the rise
and florescence of ancient Ghana, Mali and Songhai is intricately associated with
shadreck.chirikure@uct.ac.za

137Metals, Sociopolitical Complexity and Urbanism
trans-Saharan gold trade channeled to the Islamic world via entrepots such as Tewda-
goust (Levtzion 1973; Nixon et al. 2011). Although gold trade was important, it is
now generally accepted that local factors and the contributions of other metals such
as iron have been lamentably downplayed (Mudenge 1974). Therefore, the argument
that certain ranked polities seized control of the earliest gold trade and very rapidly
became states does not appear to be convincing in contexts where there is no strict
bureaucratic control over its production as in Nubia and Egypt. The problem becomes
one of enforcing the control and monopoly given the capacity challenges facing many
developing states. As Mudenge (1974) argues, the kings levied tribute in glass beads,
cloth, gold, iron and even cattle and for commoners to obtain exotic goods they had
to trade in gold and other resources, thereby weakening hypothesis that elites at Ma-
pungubwe, Mapela, and Great Zimbabwe monopolized long-distance trade.
The agglomerations of large populations at urban centres and the associated hin-
terland areas created problems of control for the rulers and elites. As such, one
of the major contributions of metallurgy to African social engineering was in the
provision of weapons for defending territorial integrity as well as for territorial
expansion (Fig. 6.8). For example, armed with mass produced (Maggs 1992) short
stabbing spears, Shaka’s 19th-century army was instrumental in building a substan-
tial urban centre as well as a state in the KwaZulu-Natal region of South Africa (see
Fig. 6.1). The army was used to gain the allegiance of different groups of people.
In West Africa, the well-respected army of Samori Toure was also armed with iron
weapons, although Samori later had gun smiths who imitated European imports
thereby demonstrating ingenuity and flexibility (Adu-Bohen 2006). Going back in
time, Portuguese records on the Zimbabwe plateau mention that Mutapa armies
were armed with iron spears which were used to maintain territorial integrity. The
Portuguese were even expelled from the plateau in the 17th century by a force of
Changamire armed with those spears (Mudenge 1988). In Central and West Africa,
the armies of the Mbanza kingdom of Kongo were also armed with spears just as
were the Luba and Lunda.
Fig. 6.8 Iron projectiles used by the Buluba of the Katanga region of the Democratic Republic of
Congo. Natural History Museum of Zimbabwe, Bulawayo
shadreck.chirikure@uct.ac.za

138 6 The Social Role of Metals
Therefore, as a socially embedded technology, metallurgy had an all pervading
influence in society. Control of the knowledge of metallurgy was important to eco-
nomic and political centralization, just as the tools and weapons were important in
subsistence, economic and defence spheres. The representation and materialization
of power in objects provided a link between royal ancestors, political authority and
the aesthetic and physical properties of metals. As a socially embedded technology,
the adoption and entrenchment of metallurgy had numerous socioeconomic and
political permutations, particularly in the emergence of state systems.
Metallurgy, Culture Contact (Interaction), Proto-
Globalization and Technology Transfer
The values attached to metals in antiquity, when coupled with their differential
availability, were some of the principal motivators for initiating local, regional, as
well as long-distance cultural contact on the land and sea. The resulting trade and
exchange network connected people of different cultures and value systems, leading
to complex intersections of local and non-local systems from early on. For instance,
the Romans exported their values while adopting those of others. During the golden
age of Islamic civilization, most of the Old World adopted some technologies and
values from the Middle East. This process was responsible for the Islamization of
West and East Africa from early on. From the 19th century onwards, Europeans suc-
cessfully exported industrial capitalism to all parts of the globe, resulting in inten-
sified cultural contact, technology transfer and biological and cultural exchanges.
This contact took place locally, regionally and internationally. Although not al-
ways emphasized in African archaeological literature, presumably because of the
difficulty in identifying items traded locally, there existed a very strong and vibrant
local and regional trade involving metals throughout sub-Saharan Africa. Processed
metals and alloys as well as raw ore were widely traded in the subcontinent and
associated with both biological and technological exchanges. Historical and ethno-
graphic sources allude to the existence of an intricate local trading system involving
copper and iron at Musina and Phalaborwa in Northern South Africa after AD 1000.
The famous Venda musuku and Phalaborwa lerale ingots served as one of the pri-
mary forms in which copper was circulated in the lowveld and beyond. Venda kings
transformed the once independent Lemba copper workers into attached specialists
who produced metal for the king (Stayt 1931). More importantly, there also existed
a thriving trade in diamond-shaped hoes produced in Phalaborwa but with a wide
circulation covering adjacent parts of Zimbabwe and Mozambique after the sec-
ond half of the second-millennium AD. The technological style of making diamond
hoes (Fig. 6.9) as well as the musuku and lerale ingots became a “brand” associ-
ated with metal workers of Northern South Africa in the second-millennium AD.
Surprisingly, there was little imitation, for the Shona across the Limpopo preferred
oval-shaped hoes, demonstrating that social differentiation can, to some extent, be
read from the typology of metal objects. The metal was exchanged for cattle, grain
shadreck.chirikure@uct.ac.za

139Metallurgy, Culture Contact (Interaction), Proto-Globalization ...
and other local commodities which leave little in terms of material remains. This
invisibility explains why the long-distance contact that brought durable and visible
commodities such as beads is emphasized in importance over local trade. In South-
ern Cameroon, a vibrant trade involving very thin iron bars known as bikies was
also practised. These strips were a form of currency and a medium of exchange with
different bundles possessing different values ranging from a wife, a cow to basic
foodstuffs (Bohannan 1955; Ringquist 2008). As with Southern Africa, it is difficult
to archaeologically study this trade owing to the poor survival rate of some of the
commodities involved.
In terms of regional trade, one of the most important, but poorly understood
cases relates to the exchange of copper and iron between Southern Zambezia and
Central Africa in the first- and second-millennium AD. The very distinctive copper
ingots called Katanga crosses which were made in this part of Africa were found at
Great Zimbabwe and other places to the north (Garlake 1973). However, little com-
positional or isotopic work has been conducted on the Katanga crosses in Southern
Africa to evaluate if they were imitations. There also existed iron gongs which
were recovered from Great Zimbabwe (AD 1100 to 1550) and certainly originated
through contact with Central Africa. Possibly, cattle, grain and other local com-
modities formed part of this exchange network.
Another important regional trade involved metallic tin in Southern Africa. Rooi-
berg in the Southern Waterberg (Fig. 6.10) is the only uncontested source of tin in
preindustrial Southern Africa. Estimates by early mining geologists indicated that
18,000 t of ore had been mined preindustrially but interestingly there is very little in
terms of settlement around Rooiberg (Chirikure et al. 2010). In contrast to this, vari-
ous forms of tin ingots were recovered from numerous places north and northeast
of Rooiberg such as Polokwane, Venda and among others Great Zimbabwe (Grant
1999) (Fig. 6.10). Geochemical and isotopic fingerprinting indicated that virtually
all these tin ingots scattered throughout Southern Africa were made using Rooiberg
tin. The mechanisms of this trade are unknown, but it is possible that it either was
Fig. 6.9 Diamond-shaped hoes produced by Phalaborwa smiths in Northern South Africa between
c. AD 1600 to 1900. The hoes were used as currency (Photo credit author)
shadreck.chirikure@uct.ac.za

140 6 The Social Role of Metals
down the line trade or a series of middleman traversed the landscape from corner to
corner exchanging the metal for various merchandize. Middlemen such as Tsonga
and Shona were famous in this regard (Mudenge 1988).
In West Africa, the increasing scale of production, particularly from the late
first-millennium AD onwards, has strong implications for regional trade. De Barros
(1986) argues that the Bassar iron trade was geared towards meeting the demands
of an external market. At Sukur in Northeastern Nigeria, there was a semi-industrial
process aimed at producing iron objects which were exported to Bornu and other
centres (David and Sterner 1997; Sasson 1964). Equally, the montagnards occupy-
ing the Mandara Mountains of Northern Cameroon produced iron blooms which
were used by Wandala to produce shackles (MacEachern 1993), thereby feeding
into the demands of the slave trade. Another important large-scale iron production
centre in Cameroon is that of the Ndop plain discussed by Warnier and Fowler
(1979). The scale of production in all these areas was staggering and exceeded that
produced at places such as Meroe (de Barros 1986). According to David and Sterner
(1997), iron was the source of power in Sukur and adjacent areas.
This regional contact was also responsible for mobility, technology transfer and
cross-borrowing. For example, contact between Central, Eastern and Southern Af-
rica was responsible for the dispersal of techniques such as wire drawing (Bisson
2000). However, there was an element of conservatism too, for various regions
continued to be associated with distinctive furnace and metal ingot types. It appears
that each group was happy to receive metal and work it using locally acceptable
Fig. 6.10 Location of Rooiberg in relation to capitals
shadreck.chirikure@uct.ac.za

141
methods. The low rate of technology transfer was most likely an outcome of spe-
cialization. Specialists produced their metal within culturally acceptable norms
which often constrained the readiness with which they could welcome techniques
from elsewhere. Without exporting ethnographic understanding to the deep past, it
is debatable whether different technological styles diachronically and synchronical-
ly became bound up in processes of identification in relation to what we understand
today as ethnicity or not. Often, metalworkers migrated to areas rich in ore sup-
ply. In the new areas, they inter-married locally resulting in genetic flows. A good
example is provided by the Njanja who migrated from Northern Mozambique and
established home in Central Zimbabwe (Mackenzie 1975). There is no reason why
the proceeds from local and regional metal trade would not have been important in
fostering societal transformations.
Besides localized and regional trade in metals, there existed a long-distance trade
network that connected together interior communities with those resident at the
coast and beyond. One of the oft-cited cases of long-distance interconnections re-
lates to the Indian Ocean-based trading system that connected Southern and Eastern
Africa and the Indian Ocean rim regions such as Persia, the Indian subcontinent,
and regions afar. The Periplus of the Erythrean Sea mentions that metal objects were
exported to East Africa by Greco-Romans c. 50AD. This trade became more visible
from the 8th-century AD and involved iron, gold, ivory and other local commodities.
From its advent until Vasco da Gama’s voyage to India, this trade involved Swahili
middlemen who were happy to travel into the interior to barter their commodities.
At the same time, Shona middlemen known as vashambadzi obtained commodities
from the Swahili and exchanged them in the interior for a profit (Mudenge 1988).
In other contexts, the Shona travelled to the coast to obtain the goods which they
exchanged locally. Al Masudi (cited in Summers 1969) reported that iron exported
from East Africa was highly valued in the Indian Ocean world, together with gold
and possibly copper (Kusimba 1999).
The success of this trade saw Greco-Roman merchants and later those from India,
Arabia and later one Chinese expedition anchoring at East African ports with mer-
chandize for trade from very early on. This brought in Indian glass beads, Persian
glass beads, Chinese porcelain as well as alloys such as bronze, brass and leaded
gun metal (Robertshaw et al. 2010; Stanley 1931). These metals were worked using
local technologies of manipulating metals. It has been argued that long-distance
trade introduced alien value systems which were appropriated and monopolized by
the elites. For example, gold became a prestige metal together with bronze and the
very rare imports of Chinese celadon and porcelain (Fig. 6.11), Persian wares and
glass beads. Of all the imports, glass beads seem to dominate as they were recov-
ered from both commoner and elite sites in large numbers. In fact the abundance of
glass beads at elite and non-elite sites poses the question whether glass beads were
status indicators or they were just an item of dress and decoration. Early on around,
AD 700 when they were introduced, beads might have been prestige goods but by
about AD900, they increased in frequency to the point of not being rare suggesting
that their value may also have decreased. This is pertinent because Wood (2012)
demonstrated that quantities of beads recovered from elite and non-elite areas in the
Metallurgy, Culture Contact (Interaction), Proto-Globalization ...
shadreck.chirikure@uct.ac.za

142 6 The Social Role of Metals
Shashi-Limpopo do not differ much. This suggests that the influx of goods was not
subject to royal control. In the historical period, many young men who were near
marriage age could travel to the coast in search of beads or chuma to give to their
prospective bride. Furthermore, vashambadzi or middlemen obtained large amounts
of beads which were exchanged internally (Mudenge 1988). During the Portuguese
period in Southern Africa (AD 1500–1900), markets or feiras were established in
the interior at places such as Massapa (Baranda), Dambarare and Luanze. Direct
trade took place at these centers while middlemen took the commodities to distant
lands (Pikirayi 2001). It is possible that this may not apply to earlier times when the
Zimbabwe state was likely more unified, but the absence of a formal bureaucracy
makes it unlikely.
The other important question regarding value transfer is why Chinese porcelain
was not popular in Southern Africa when it was present in substantive quantities at
the East African coast (Horton and Middleton 2000). Perhaps this has something to
do with the internalization of alien values. Although porcelains were highly valued,
they may not have been widely accepted locally when compared to glass beads. It
is unlikely that porcelain was so expensive to the extent of being prohibitive. Por-
celain has mostly been found at elite sites but this is also where most research has
been carried out (cf. LaViolette and Fleisher 2005). Garlake (1968) discusses the
presence of porcelain from the Marcadoni claims in the very auriferous rich West
Nicholson area of Southwestern Zimbabwe. Even at the elite sites, the porcelain
and celadon fragments are few, and if Garlake’s (1968) estimate is to be believed,
not more than 90 fragments of the pre-Portuguese porcelains were recovered on
the Zimbabwe plateau. When contrasted with the abundance of glass beads, cowrie
and alloys such as brass and gun metal, it is clear that porcelains may have failed
Fig. 6.11 Ming Dynasty porcelain from Zimbabwe housed at Iziko Museum, Cape Town. (Source:
author)
shadreck.chirikure@uct.ac.za

143
to appeal to local values in a way these other objects did. This is hardly surpris-
ing for Hall (1998) who suggested that even within a colonial frontier, imports
struggled to replace domestic pottery in many activities. It may be possible that
this was the same situation with porcelains. Although seemingly radical, this think-
ing is adequately supported by two independent but related observations. The first
is that the Portuguese complained that hinterland communities shunned European
glass beads in favour of the deeply entrenched Indo-Pacific beads (Wood 2012).
Secondly, although the Portuguese brought chocolate on white porcelain as an item
of trade, comparatively it does not seem to have gained wide local acceptance when
compared to glass beads and guns for example. As a result, it is only concentrated at
trading sites such as Dambarare (Pikirayi 2001). These points reinforce the obser-
vation that people consumed goods according to their own logics and prior prefer-
ences (Prestholdt 2008).
While these interconnections were taking place on the East African littoral, par-
allel developments were unfolding in West Africa where long-distance commercial
traffic lubricated by trade in gold and slaves was deeply rooted (Fig. 6.12). As with
Southern Africa, long-distance trade in West Africa exploited pre-existing localized
and regional trade in regions such as the inland Niger Delta of Mali. This exchange
brought alloys such as bronze and brass as well as glass beads. The 9–10th century
AD site of Igbo Ukwu in Nigeria participated in this trade as evidenced by the
recovery of burials with thousands of imported glass beads. As discussed above,
Igbo Ukwu also yielded spectacular bronze artifacts produced through the lost wax
Fig. 6.12 Regional and trans-continental connections between Southern and Western Africa and
the trans-Saharan and Indian Ocean worlds (Base map: Google Images)
Metallurgy, Culture Contact (Interaction), Proto-Globalization ...
shadreck.chirikure@uct.ac.za

144 6 The Social Role of Metals
method that were bound up in the materialization of power and ideology (Shaw
1970). Gold played a key role in the flourishing caravan-based trans-Saharan trade
that enchained the Islamic world and West Africa. The successive states of ancient
Ghana, Mali and Songhai extensively traded the gold from Bambuk, with mer-
chants based at places such as Tewdagoust. This trade brought in salt and exotic
goods such as glass beads as well as brass and other imported metals and alloys.
The story of the legendary Mansa Musa (c. 1280–c. 1337), the Mali monarch who
went on pilgrimage to Mecca carrying with him significant amounts of gold, is one
of the best-known examples involving gold trade. So strategic was the control of
the gold trade that hegemony over source areas was a question of life and death for
these entities. However, there are successful states such as Kanem-Bornu near Lake
Chad which, like many communities in Northern Nigeria and Northern Cameroon,
did not participate in the gold trade (Garrard 2011). From the 14th century, West
Africa became increasingly connected directly to Europe via the Atlantic littoral.
This trade resulted in an influx of huge amounts of European alloys such as bronze
and brass and metals such as iron. Gold, particularly that from the Gold Coast, to-
gether with other local resources such as ivory, was important in this trade.
Often, the desire to control the metal-producing regions was a source of conflict
and cleavages in society, for example internecine wars were fought to control the
gold fields in Northern Zimbabwe (Mudenge 1988). A good example is that of the
Almoravids who laid siege at Koumbi Saleh, the capital of the Soninke state of an-
cient Ghana, thereby precipitating its demise (Levtzion 1973). An ability to control
the trade routes and gold producing regions empowered Songhai to suffocate its
predecessor Mali. In Southern Africa, the Portuguese desire for gold, fuelled by un-
controlled rapacity, plunged them deep into the internal affairs of the Mutapa state
(AD 1450–1900) in Northern Zimbabwe which gradually weakened with the unin-
tended result of nearly halting gold production (Pikirayi 2001). Local agents were
not idle while the Portuguese were making their bid for fortune; in the late 17th
century, Changamire Dombo fought them and dislodged them from the Zimbabwe
plateau, leading to the demise of old and rise of new empires. The lure of African
metals also promoted the settlement on African shores by people from its trading
partners. In the late 19th century, it was partly as a result of Africa’s metal wealth
that the continent was colonized by European powers expanding under industrial
capitalism. The colonial elites continued to work the mines historically worked by
preindustrial people and introduced their own methods of production. This drove
the millennia old African techniques into extinction.
One of the most astounding observations is that although Africa was an impor-
tant cog in the development of “proto” forms of globalization, it did not participate
much in the technology transfer taking place between some of its trading partners.
According to Killick and Fenn (2012), African metals such as gold paid for exotic
commodities in India and Persia as well as the porcelain and gunpowder from China.
Yes, Africa obtained commodities from these areas but the incorporation of import-
ed goods was not matched by a technological transfer. For example, the technology
of metal production remained the bloomery process, while neither the blast furnace
nor technologies of producing wootz steel or Damascus swords were adopted in the
shadreck.chirikure@uct.ac.za

145Conclusion
continent. The same applies to West Africa, which was also in direct contact with
the Islamic world. The technology of producing glass was also not developed lo-
cally with the exception of the melting of imported glass beads to produce garden
rollers at K2 (AD 1000–1200) in the Middle Limpopo valley of Southern Africa
and the autochthonous Yoruba glass production industry of the early second-millen-
nium AD in Southwestern Nigeria (see Lankton et al. 2006). It is not that Africans
were incapable of assimilating these technologies; rather, there existed deep seated
cultural barriers that may have prevented the assimilation of exotic technologies.
Furthermore, Hopkins (1973) argues that external methods of production before
the industrialization did not have much of an advantage when compared to Afri-
can ones. For example, David Livingstone commented that African iron was of a
better quality when compared to that produced in Europe in the late 19th century
(Chirikure 2006). Bloomery iron production was still used in the USA up to the 19th
century. The only reason why colonialism managed to change African technologies
was that it introduced a different value system based on capitalism and Christianity
while directing heavy assaults on local technological practices. Therefore, Africans
had no choice, whereas in the past they exercised their freedom by sticking to what
worked for them. As such, for as long as the technologies worked, there was no need
to change them in favour of alien ones. Also, some of the technologies from Africa’s
trading partners were suited for contexts with comparatively large and concentrated
populations and may not have worked in the variably populated parts of the conti-
nent. For example, regions such as Southern Africa were not as heavily populated as
the Upper Nile or the Indian subcontinent with the result that large-scale production
methods were unlikely to achieve a similar effect.
Conclusion
In summary, it is clear that metallurgy had a significant impact on not just African
communities but also those in Eurasia. Far from being isolated, Africa was part of
the developments in the Old World. At this point, it is important to revisit the ques-
tion, did the adoption of metallurgy represent a revolution or not? This is not an
easy question to answer because inherently it implicates different scales of analyses,
different scales of sociopolitical organization and different scales of population dy-
namics. As with other places in the world, the adoption of metallurgy was gradual
with some communities sticking to the tried and tested technology of stone. How-
ever, once metal became more and more incorporated into local value systems, it
became so intensely embedded in society that its impact was widely felt.
Metallurgy affected various communities that practised it—with strong conse-
quences for food production, defence, wealth accumulation and local, regional as
well as long-distance interconnections. Although metals were known in the western
hemisphere, it was only after Columbus that metals became widely used for tool
making. However, in areas of the Old World such as Eurasia and Africa, metal-
lurgy played an important part in regional and international integration, promoting
shadreck.chirikure@uct.ac.za

146 6 The Social Role of Metals
urbanization and marked social differentiation. Because of these networks, Africa
has always been at the centre of developments in the world. It, however, seems
that the continent’s population may have been somewhat at a disadvantage because
technologies in use in other regions were best suited for sustaining large populations
and not small ones. This means that technologies in use on the African continent had
to fit this specific context.
While control over metal production and use was a feature of Egyptian, Nubian
and Aksumite states, to the extent that stock taking was adhered to religiously, it is
difficult to comprehend how rulers in other parts of Africa would have controlled
production and metal use. As Mudenge (1974) expressed it, it seems unlikely that
there was a rigorous bureaucratic control in the Mutapa and Rozvi states of second-
millennium AD Southern Africa. Equally, Levtzion (1973) makes it clear that nei-
ther the rulers of ancient Ghana nor Mali controlled the actual production of gold. In
the Mutapa and Rozvi states, the commoners based in the villages worked gold, ex-
changed it for glass beads and cloth, some of which they gave their rulers as tribute.
At a broad scale, the gradual dominance of metallurgy in every aspect of society
from the aesthetic, utilitarian, sensual, ceremonial—in fact the whole value sys-
tem—indicates that metallurgy had significant consequences for social life. If we
emphasize continuity, we realize that precious metals such as gold were valuable to
Hebrew King Solomon 3000 years ago just as they were to Cecil John Rhodes in
the late 19th century. Both sought valuable metals in lands afar showing incidences
of history repeating itself but in markedly different contexts. This provides motiva-
tion for more work into understanding preindustrial metal production and use. As
may be divined from this discussion, it is difficult to understand the beginning and
functioning of the ancient and modern worlds without studying metals—a critical
part of our sociological and technological past.
References
Adu-Boahen, K. (2006). Pawn of contesting imperialists: Nkoransa in the Anglo-Asante rivalry in
Northwestern Ghana, 1874–1900. Journal of Philosophy and Culture, 3(2), 55–85.
Anquandah, J. (1993). The Kintampo complex: a case study of early sedentisim and food produc-
tion in sub-Sahelian West Africa. In T. Shaw, P. Sinclair, B. Andah, & A. Okpoko (Eds.), The
archaeology of Africa food metals and towns (pp. 255–260). London: Routledge.
Bisson, M. S. (1982). Trade and tribute. Archaeological evidence for the origin of states in South
Central Africa (Commerce et tribute. Documents archéologiques sur l’origine des États du sud
de l’Afrique centrale). Cahiers d’Etudes africaines, 22(87/88), 343–361.
Bisson, M. (2000). Precolonial copper metallurgy: sociopolitical context. In M. Bisson, P. de Bar-
ros, T. C. Childs, & A. F. C. Holl (Eds.), Ancient African metallurgy: the sociocultural context
(pp. 83–146). Walnut Creek: AltaMira Press.
Bocoum, H. (2006). Histoire technique et sociale de la métallurgie du fer dans la vallée moyenne
du Sénégal. Doctorat d’Etat thesis. University Cheikh Anta Diop, Dakar
Bohannan, P. (1955). Some principles of exchange and investment among the Tiv. American An-
thropologist, 57(1), 60–70.
Carneiro, R. L. (1967). On the relationship between size of population and complexity of social
organization. Southwestern Journal of Anthropology, 23(3), 234–243.
Childe V. G. (1930). The bronze age. Cambridge: Cambridge University Press.
shadreck.chirikure@uct.ac.za

147References
Chirikure, S. (2006). New light on Njanja iron working: towards a systematic encounter between
ethnohistory and archaeometallurgy. South African Archaeological Bulletin, 61, 142–151.
Chirikure, S. (2007). Metals in society: Iron production and its position in Iron Age communities
of Southern Africa. Journal of Social Archaeology, 7(1), 72–100.
Chirikure, S., Heimann, R. B., & Killick, D. (2010). The technology of tin smelting in the Rooi-
berg Valley, Limpopo Province, South Africa, ca. 1650–1850 CE. Journal of Archaeological
Science, 37(7), 1656–1669.
Chirikure, S., Manyanga, M., Pikirayi, I., & Pollard, M. (2013). New pathways of sociopolitical
complexity in Southern Africa. African Archaeological Review, 30(4), 339–366.
Cline, W. B. (1937). Mining and metallurgy in Negro Africa (No. 5). George Banta Publishing
Company
Connah G (2001). African civilization. 2nd edition. Cambridge: Cambridge University Press.
(Conservation Institute; J. Paul Getty Museum, in association with Archtype Books).
David, N., & Sterner, J. (1997). Water and iron: phases in the history of Sukur. In H. Jungraithmayr,
D. Barreteau, & U. Seibert (Eds.), L’homme et l’eau dan le basin du lac Tchad (pp. 255–270).
Actes du Colloque Mega-Tchad. Paris: Colloques et Seminaires, Editions ORSTOM.
de Barros, P. (1986). Bassar: A quantified, chronologically controlled, regional approach to a tra-
ditional iron production centre in West Africa. Africa, 56(2), 152–174.
de Maret, P. (1985). The smith’s myth and the origin of leadership in Central Africa. In R. Haaland
& P. Shinnie (Eds.), African iron working—ancient and traditional (pp. 73–87). Oslo: Norwe-
gian University Press.
Fagan, B. M. (1969). Early trade and raw materials in South Central Africa. Journal of African
History, 10(1), 1–13.
Fletcher, R. J. (1993). Settlement area and communication in African towns and cities. In T. Shaw,
P. Sinclair, B. Andah, & A. Okpoko (Eds.), The archaeology of Africa food metals and towns
(pp. 732–749). London: Routledge.
Garlake, P. S. (1968). The value of imported ceramics in the dating and interpretation of the Rho-
desian Iron Age. The Journal of African History, 9(1), 13–33.
Garlake, P. S. (1973). Great Zimbabwe. London: Thames and Hudson.
Garrard, T.F. (1980). Akan weights and the gold trade. London: Longman.
Garrard, T. F. (2011) (1989). African Gold: jewellery and ornaments from Ghana, Côte d'Ivoire,
Mali andSenegal in the collection of the Barbier-Mueller Museum. Munich: Prestel.
Grant, M. R. (1999). The sourcing of Southern African tin artefacts. Journal of Archaeological
Science, 26(8), 1111–1117.
Hall, S. L. (1998). A consideration of gender relations in the Late Iron Age “Sotho” sequence of the
Western Highveld, South Africa. In S. Kent (Ed.), Gender in African prehistory (pp. 235–260).
Walnut Creek: AltaMira Press.
Hall, S. L., Miller, D., Anderson, M., & Boeyens, J. (2006). An exploratory study of copper and
iron production at Marothodi, an early 19th century Tswana town, Rustenberg District, South
Africa. Journal of African Archaeology, 4, 3–35.
Hauptmann A. (2007). The archaeometallurgy ofc Copper: Evidence from Faynan, Jordan. New
York: Springer.
Herbert, E. W. (1984). Red gold of Africa: copper in precolonial history and culture. Madison:
University of Wisconsin Press.
Holl, A. F. (2009). Early West African metallurgies: new data and old orthodoxy. Journal of World
Prehistory, 22(4), 415–438.
Hopkins, A. G. (1973). An economic history of West Africa. New York: Columbia University Press.
Horton, M. C., & Middleton, J. (2000). The Swahili: the social landscape of a mercantile society.
Oxford: Blackwell.
Humpris, J., & Iles, L. (2013). Pre-colonial iron production in Great Lakes Africa: recent research
at UCL Institute of Archaeology. In J. Humpris & Th. Rehren (Eds.), The world of iron (pp. 56–
64). London: Archetype.
Huysecom E., & Agustoni B. (1997). Inagina: l’ultime maison du fer/the last house of iron. Ge-
neva: Telev. Suisse Romande. (Videocassette, 54 min).
shadreck.chirikure@uct.ac.za

148 6 The Social Role of Metals
Killick, D. J. (2009). Agency, dependency and long-distance trade: East Africa and the Islamic
World, ca. 700–1500 C.E. In S. Falconer & C. Redman (Eds.), Polities and power: archaeo-
logical perspectives on the landscapes of early states (pp. 179–207). Tucson: University of
Arizona Press.
Killick, D., & Fenn, T. (2012). Archaeometallurgy: The study of preindustrial mining and metal-
lurgy. Annual Review of Anthropology, 41, 559–575.
Killick, D. (2014). Cairo to Cape: the spread of metallurgy through Eastern and Southern Africa. In
B. W. Roberts & C. Thornton (Eds.), Archaeometallurgy in global perspective (pp. 507–527).
New York: Springer.
Kim, N. C., & Kusimba, C. M. (2008). Pathways to social complexity and state formation in the
Southern Zambezian region. African Archaeological Review, 25(3–4), 131–152.
Kusimba, C. M. (1999). The rise and fall of Swahili states. Walnut Creek: AltaMira Press.
Lankton, J., Ige, A., & Rehren, T. (2006). Early primary glass production in Southern Nigeria.
Journal of African Archaeology, 4, 111–138.
LaViolette A., & Fleisher J. (2005) The archaeology of sub-Saharan urbanism: Cities and their
countrysides. In A. B. Stahl (Ed.), African Archaeology: A critical Introduction (pp. 327–352).
Oxford: Blackwell.
Levtzion, N. (1973). Ancient Ghana and Mali (Vol. 7). London: Methuen.
MacDonald, K. C. (1998). Before the empire of Ghana: Pastoralism and the origins of cultural
complexity in the Sahel. In G. Connah (Ed.), Transformations in Africa: Essays on Africa’s
later past (pp. 71–103). London: Leicester University Press.
MacEachern, S. (1993). Selling the iron for their shackles: Wandala-Montagnard interactionsin
Northern Cameroon. Journal of African History, 34, 247–247.
Mackenzie, J. M. (1975). A pre-colonial industry: the Njanja and the iron trade. Nada, 11(2),
200–220.
Maggs, T. (1992). ‘My father’s hammer never ceased its song day and night’: the Zulu ferrous
metalworking industry. Southern African Humanities, 4, 65–87.
Magnavita, C., Breunig, P., Amfje, J., & Posselt, M. (2006). Zilum: a mid-first millennium BC
fortified settlement near lake Chad. Journal of African Archaeology, 4, 153–169.
McIntosh, S. K. (Ed.). (1999). Beyond chiefdoms: pathways to complexity in Africa. Cambridge:
Cambridge University Press.
McNaughton, P. R. (1993). The Mande Blacksmiths: knowledge, power, and art in West Africa.
Bloomington: Indiana University Press.
Miller, D. (1996). Tsodilo Jewellery. Cape Town: University of Cape Town Press.
Miller, D. (2001). Metal assemblages from Greefswald areas K2, Mapungubwe Hill and Mapun-
gubwe Southern terrace. The South African Archaeological Bulletin, 56, 83–103.
Mitchell, P. (2002). The archaeology of Southern Africa. Cambridge: Cambridge University Press.
Mitchell, P., & Lane, P. (Eds.). (2013). The Oxford handbook of African archaeology. Oxford:
Oxford University Press.
Mudenge, S. I. (1974). The role of foreign trade in the Rozvi empire: A reappraisal. Journal of
African History, 15(3), 373–391.
Mudenge, S. I. (1988). A political history of Munhumutapa c 1400–1902. Harare: Zimbabwe Pub-
lishing House.
Ndoro, W. (1994). Natural draught furnaces south of the Zambezi River. Zimbabwean Prehistory,
14, 29–32.
Nixon, S. (2009). Excavating Essouk-Tadmakka (Mali): new archaeological investigations of
early Islamic trans-Saharan trade. Azania: Archaeological Research in Africa, 44(2), 217–255.
Nixon, S., Rehren, T., & Guerra, M. F. (2011). New light on the early Islamic West African gold
trade: coin molds from Tadmekka, Mali. Antiquity, 85(330), 1353–1368.
Ogundiran A (2005). Four millennia of cultural history in Nigeria (ca. 2000BC–AD 1900). Journal
of World Prehistory, 19(2), 133–168.
Phillipson, D. W. (2005). African archaeology. Cambridge: Cambridge University Press.
Phimister, I. R. (1974). Alluvial gold mining and trade in nineteenth-century South Central Africa.
The Journal of African History, 15(3), 445–456.
shadreck.chirikure@uct.ac.za

149References
Pikirayi, I. (2001). The Zimbabwe culture: origins and decline in Southern Zambezian states.
Walnut Creek: Altamira Press.
Prestholdt, J. (2008). Domesticating the world: African consumerism and the genealogies of glo-
balization. Berkeley: University of California Press.
Pwiti, G. (1996). Continuity and change: an archaeological study of farming communities in
Northern Zimbabwe AD 500–1700. Uppsala: Societa Archaeologica Uppsaliensis.
Pwiti, G. (2005). Southern Africa and the East African coast. In Stahl, A. B. (Ed.), African archae-
ology: a critical introduction (pp. 378–391). London: Blackwell.
Reid, A., & MacLean, R. (1995). Symbolism and the social contexts of iron production in Karag-
we. World Archaeology, 27(1), 144–161.
Renfrew, C., & Cherry, J. F. (Eds.). (1986). Peer polity interaction and socio political change.
Cambridge: Cambridge University Press.
Ringquist. J. (2008). Kongo iron: Symbolic power, superior technology and slave wisdom. African
diaspora archaeology network. Newsletter (Sept):1–20.
Robertshaw, P., Wood, M., Melchiorre, E., Popelka-Filcoff, R. S., & Glascock, M. D. (2010).
Southern African glass beads: chemistry, glass sources and patterns of trade. Journal of
Archaeological Science, 37(8), 1898–1912.
Sasson, H. (1964). Iron-Smelting in the Hill Village of Sukur, North-Eastern Nigeria. Man, 64,
174–178.
Scott, D. A. (1991). Metallography and microstructure of ancient and historic metals. Getty Con-
servation Institute; J. Paul Getty Museum, in association with Archtype Books.
Shaw, T. (1970). Igbo-Ukwu: an account of archaeological discoveries in Eastern Nigeria (Vol. 2).
Evanston: Northwestern University Press.
Smith, C. S. (1981). A search for structure. Selected essays on science, art and history. Cambridge:
MIT.
Stanley, G. H. (1931). Some products of native iron smelting. South African Journal of Science,
28, 131–134.
Stayt, H. A. (1931). The Bavenda (No. 58). International Institute of African Languages & Cul-
tures. Oxford: Oxford University Press.
Summers, R. (1969). Ancient mining in Rhodesia and adjacent areas. Salisbury: Trustees of the
National Museums of Rhodesia.
Tamari, T. (1991). The development of caste systems in West Africa. Journal of African traditional
iron production centre in West Africa. Africa, 56(2), 152–174.
Vansina, J. (1969). The bells of kings. The Journal of African History, 10(02), 187–197.
Warnier, J. P., & Fowler, I. (1979). A nineteenth-century Ruhr in Central Africa. Africa, 49(04),
329–351.
Wood, M. (2012). Interconnections: glass beads and trade in Southern and Eastern Africa and
the Indian Ocean-7th to 16th centuries AD. African and Comparative Archaeology. Uppsala:
(Department of Archaeology and Ancient History), Uppsala University.
shadreck.chirikure@uct.ac.za

151
Chapter 7
Bridging Conceptual Boundaries, A Global
Perspective
© The Author 2015
S. Chirikure, Metals in Past Societies, SpringerBriefs in Archaeology,
DOI 10.1007/978-3-319-11641-9_7
“Pre-European metalworkers are worthy of respect for the
results they achieved with primitive methods”
(Steel 1975, p. 232)
Introduction
The production and use of metals in antiquity rank as one of the humanity’s most
consequential sociocultural and technological developments, which make the study
of Africa’s preindustrial metallurgy a topic of global significance. Although Africa
is part of the Old World, its pathway to metallurgy was remarkably different to that
of its counterparts. Unlike regions such as Thailand which are believed based on
current evidence to have received metallurgy from China in exactly the same suc-
cession of copper, bronze and iron (Pryce et al. 2010), it appears as if some parts
of Africa adopted metallurgy in reverse of the situation throughout Eurasia. While
Egypt, Nubia, Ethiopia and Eritrea followed the Eurasian trajectory, most of West,
Central, Southern and Eastern Africa started metallurgy with iron and in some in-
stances iron and copper and only adopted bronze, gold and tin more than a thousand
years later. Not surprisingly, researchers are not agreed on the source of Africa’s
metallurgy. Proponents of the external origins hypothesis argue that given the im-
portance of early encounters with pyrotechnology and the radiocarbon black hole
between 800 and 400 BC, it seems that knowledge of African metallurgy diffused
from the Middle East (Alpern 2005). However, those who support local origins
maintain that vast differences in pathways between the Middle East and Africa, and
increasing numbers of dates clustered around 1000 BC indicate that African metal-
lurgy may be independent (Holl 2009). The acceptability of the two positions is
affected by poor radiocarbon dates, poor contexts of recovery and too few sites that
have been excavated, as well as the challenges created by calibration in the critical
shadreck.chirikure@uct.ac.za

152 7 Bridging Conceptual Boundaries, A Global Perspective
period between 800 and 400BC. The technical reasons supporting the external ori-
gins hypothesis, such as the difficulty of smelting iron before mastering easier met-
als, make considerable scientific sense (Alpern 2005; Clist 2013).
However, the big question is, given that Eurasia had knowledge of the materials
and materiality of iron, lead, copper, bronze and gold, during the time when African
metallurgy began, why did Africans only choose to take up iron and copper and
not the other metals and alloys? Although it is possible that people without a prior
knowledge of the materials and materiality of metals were able to exercise choices
informed by the physical properties of metals to the extent of selecting the more
durable iron, it is not clear why they would ignore gold which had a very high value
and was highly sought after. If materiality was the biggest motivator for the adop-
tion of early metallurgy (Smith 1981), then iron is certainly not the most beautiful
of all metals, implicating other reasons. Furthermore, why did sub-Saharan Africa
accept tin, bronze and gold almost a thousand years later? It has been shown that the
adoption of technology is a result of prior experiences (Prestholdt 2008). As such, it
was easier to accept gold once copper and iron were known. If Craddock (2010) is
right, then sub-Saharan iron may have evolved out of Nubian metallurgy.
However, chance may have promoted the advent of iron metallurgy; tempera-
tures for smelting iron and copper do not vary much, given that gangue minerals
such as silica in the ore were fluxed to form slags at temperatures over 1000 °C
(Bachmann 1982). The dynamics of innovation, diffusion and technology and value
transfer are neither simple nor logical and may be easily discerned through common
sense rooted to a specific cultural logic. In other words, we should endeavour to
understand these transfers in relation to local, culturally specific logics. It has been
argued that iron smelting is complicated, but there is skepticism that people without
experience with pyrotechnology can get an idea from somewhere and execute that
difficult idea with relative ease. Thus, if iron smelting is as difficult as some believe,
then it is unlikely that a mere transmission of ideas would result in a transfer that
we see in West, Central and East Africa. As Holl (2009) has argued, there is no
hard evidence indicating technology transfer between sub-Saharan Africa and the
donor regions. The copper at Akjoujt may represent a different dynamic that indi-
cates what happens in zones of contact. Had this happened in other areas, the idea
of diffusion would make sense; however, Akjoujt seems to be an exception rather
than the rule. Because some of the greatest inventions in human history were mere
accidents, what is required is more fieldwork backed up by robust assessment of
contexts of recovery and dating programs so that the conclusions are based on hard
evidence. The significance of the great controversy regarding the origins of African
metallurgy is that it forces researchers to reflect on innovation as well as value and
technology dispersal in extremely varied contexts.
By the mid-first millennium AD, sub-Saharan Africa was fully integrated into
the triangular trading system involving Africa, the Near East and South Asia. In ad-
dition to other commodities, Africa supplied gold, iron, and possibly copper and tin,
which were in demand within Islamic-controlled Saharan and Indian Ocean trade
networks. In return, Eurasia supplied glass beads, brass, bronze, silver and among
other resources cloth and later ceramics, augmented by small quantities of celadon
and porcelain. African metallurgy was, therefore, instrumental in the development
shadreck.chirikure@uct.ac.za

153African Metallurgy and the Bridging of Conceptual Boundaries Between …
of land-based as well as sea-based linkages and therefore connections internal to
the continent and externally with Eurasia. This interaction resulted in mobility and
settlement in Africa by Arabs along the East African littoral and north of the Sahara.
The Chinese under the legendary Captain Zheng Ho anchored on the East African
coast once 60 years before Vasco Da Gama. Over time, the Europeans followed
suit and built a number of castles and forts alongside the Atlantic and Indian Ocean
coasts. Such activities resulted in the inward and outward exchange of ideas, dis-
eases, animals and metals. However, strong cultural filters enabled Africa to only
incorporate commodities and practices that aligned with extant values and systems
of valuation, while ignoring those unsuited in its context. Also, simple exposure to
new practices or ideas does not necessarily lead to adoption of those practices or
ideas. There is little evidence in much of Africa of assimilation of either Western
or Asian architecture and methods of metalworking, despite the many levels of in-
teraction.
This culture contact and interaction contradicts views that have traditionally pro-
filed Africa as a cultural backwater isolated from the rest of the world (Stahl 2014a;
Mitchell 2005). Neither was Africa passive nor was it at the mercy of its trading
partners. Instead, it exercised a great deal of agency, choice and initiative that result-
ed in the continent co-opting what worked in its context while rejecting that which
did not suit. Rather than being at the mercy of other regions, an encounter with Af-
rican metallurgy indicates that Africans in large part controlled their destiny. There-
fore, it is important to put aside the ideas relating to both overwhelmed Africa and
overwhelming Eurasia as well as backward Africa and advancing Eurasia because
different dynamics involving population sizes and cultural contexts were at play.
Besides interaction and culture contact, African metallurgy is rich in symbolic
and cultural information regarding the different stages in the chaîne opératoire of
metal production and use. The success of African metal working was dependent
on technical skill as much as on the associated belief systems, an observation that
equally applies to most parts of the world with knowledge of metallurgy. There is a
great deal of versatility and diversity in furnace types, the range and quality of ores
worked and the methods of provisioning air to the furnaces.
African Metallurgy and the Bridging of Conceptual
Boundaries Between Technology, Society and Culture
Any academic excursion into Africa’s preindustrial metallurgy shows that its suc-
cess was based on the intense interaction between nature and culture to the ex-
tent that the boundary between the two became increasingly blurred as nature was
transformed into culture. For example, the raw materials for metallurgy—ore, clay
and wood—conceptually came from nature, such that metalworkers appealed to
the power of ancestors, deities and the supernatural. These raw materials were
transformed into cultural products through smelting and the subsequent smithing
and fabrication. The metal, representing a product from nature, became not just
a cultural commodity in the service of society but also a medium through which
shadreck.chirikure@uct.ac.za

154 7 Bridging Conceptual Boundaries, A Global Perspective
culture interacted with nature. Ethnographies conducted from the late 19th century
onwards demonstrate the intertwined nature of technology, science and ideology in
the African past. Therefore, a fusion of magic, science, society and belief systems
is an important element of African technological practice, repertoire and style just
as they were part of medieval European science and technology (Hansen 1986). As
such, global archaeology will only be poorer if it ignores the web of entangled and
associated beliefs that were integral to the success of metallurgy. Insoll (2008) has
demonstrated the utility of this approach with respect to ritual practice and shrines,
using insights from Northern Ghana to generate new perspectives on European
Neolithic and Bronze Age contexts.
In this regard, African metallurgy provides one of the best examples of pre-
capitalist technologies of practice and technological style. For example, historical
documents, ethnoarchaeological studies and archaeological field research show that
in many African societies, the smelting of metal was metaphorically equivalent to
copulation, gestation and birth. The furnace was metaphorically perceived to be
a woman, while smelting was analogous to intercourse and gestation. The metal
bloom growing in the furnace was akin to a fetus, while male smelters were simul-
taneously husbands and midwives (Schmidt 2009; Killick and Fenn 2012). In some
instances, this symbolism was written on the furnaces as evidenced by decorations
of female breasts, genitalia, navels and waist belts worn to enhance fertility (Childs
1991). The reduced metal was conceptually seen as a child, who passed through
several stages of life and in turn contributed to fertility of the land and society
in general. Not surprisingly, in some regions, iron hoes could be exchanged for
women’s reproductive power resulting in societal renewal and growth.
Bridging Analytical Boundaries: From Sources of
Ethnographies to Domains of Integrated Studies
Whether one considers Africa as an independent inventor of metallurgy or a receiv-
er from the Middle East and adjacent territories does not matter much. Rather, the
most germane topic that also envelopes elements of the origins issue is what is the
value of African preindustrial metallurgy in understanding global archaeology. As I
have argued elsewhere (Chirikure 2010), there are two Africas in global archaeol-
ogy: The first is the origin of humanity and centre of cultural developments for ap-
proximately 95 % of human history; the second is a more recent Africa, perceived as
backward in sociocultural developments from the mid-Holocene onward. As such,
it is a continent whose significance in knowledge production on Pleistocene human
development is acknowledged globally on the one hand, and a continent perceived
as having little to offer in the study of the last 2000 years, except perhaps as a source
of ethnographies for validating models developed elsewhere. A question that arises
is what motivated these contrasting academic positions? Obviously, this implicates
the context of knowledge production and how research topics are articulated locally
and globally.
shadreck.chirikure@uct.ac.za

155Bridging Analytical Boundaries: From Sources of Ethnographies to Domains of …
According to Robertshaw (1990), from the late 19th century, Africa was of-
ten depicted as a backward continent that still hosted metallurgical practices long
extinct elsewhere. Therefore, the continent’s recent past became a rich source of
analogues for developing interpretations in other places. According to Kense and
Okoro (1993), this “Africa as source of ethnographies” paradigm dictated that the
continent was not seen as a region possessing technological developments with po-
tential to provide different, if not independent trajectories of the evolution of met-
allurgy. Rather, it was the source of analogues for interpreting early metallurgy in
other continents using its ethnographic record. And yet, Africa’s preindustrial met-
allurgy metamorphosed in different directions, which warrant detailed comparisons
with practices elsewhere to develop a global picture on technological variation and
cross-borrowing (Chirikure 2005).
At different intervals in the 20th century, a number of smelting re-enactments
were carried out in Africa to record the process before the knowledgeable prac-
titioners passed on. Because such work was carried out by men and women who
understood and appreciated the relevance of African preindustrial metallurgy, it
was mostly comprehensive in its approach. Most of the time, the recording started
with raw material collection, through smelting and smithing to the final products
and waste materials. Schmidt (1978, 1997), for example, as discussed in Chap. 4,
combined oral histories, ethnographies and the archaeology and archaeometallurgy
to develop a long-term perspective on iron production in Tanzania. While some of
Schmidt’s insights, for example, the pre-heating hypothesis discussed above, have
been challenged, his work was important for attempting to tease out important in-
novations relating to African smelting. Overall, Schmidt’s integrated work demon-
strated that it is important to engage with early history of African metallurgy to un-
derstand its development and evolution. This is important because what seemed like
a “fossil” technology in the early 20th century was vibrant, dynamic and innovative
across the ages as human beings responded to various challenges, technological or
otherwise. This underscores the value of comparative approaches conducted across
long time scales in identifying the problematic character of projecting ethnographic
practices into the past. For example, the variability of spatial location of smelt-
ing inside recent historical settlements and those of the Early Iron Age villages in
Southern and Eastern Africa indicates the ability of long-term perspectives in teas-
ing out continuities and changes.
Although Cline (1937) sketched much in terms of ethnographic distribution of
techniques of metalworking, rarely did researchers study the different varieties
of metalworking in their context to elicit their most salient features. David et al.
(1989) combined ethnographic and archaeometallurgical study of Mafa smelting
in Cameroon discussed in Chap. 4 and revealed a technology between bloomery
and blast furnace for it could produce soft and cast iron. Within Africa, these in-
between processes contrast remarkably with those that produced usable iron from
very low grade ores and the comparatively smaller bowl furnaces that equipped
Shaka’s armies discussed in Chap. 5. The mechanisms of operating furnaces varied
from slag-tapping bowl furnaces to non-slag-tapping natural draught furnaces and
from slag-tapping natural draught furnaces to non-slag-tapping low shaft and bowl
shadreck.chirikure@uct.ac.za

156 7 Bridging Conceptual Boundaries, A Global Perspective
furnaces. The bellows too differed from area to area but in a complex mix. This
prompted Schmidt (2001) to argue that referring generically to African metallurgi-
cal processes as bloomery techniques obscures this anourishing variation and di-
versity. Therefore, Africa offers an unrivalled potential to foster understanding not
just of variability in technological practice but also diversity in the materials and
materialities of preindustrial metallurgy. There is in fact a much wider variety of
bloomery iron smelting processes documented in Africa than elsewhere. Is it pos-
sible that these were also once practiced elsewhere in Eurasia but were erased by the
spread of the blast furnace and finery? On the other hand, African copper smelting
technology seems very restricted compared to other parts of the world. Why was
this? Perhaps because copper smelting was never a major technology within the
continent and that Eurasia had a very long time to experiment with copper from c.
5000 BC to c. 1500. Had Africa started with copper, perhaps it too would exhibit
such diversity.
Global archaeology is quite creative when it comes to developing new theories.
For example, given the bifurcation of studies of preindustrial metallurgy into the sci-
entific and magical, researchers such as Lechtman (1977) propositioned, following
Mauss, that all technologies were socially constructed and embedded. The differ-
ent techno–cultural solutions gestated in different parts of the world shaped diverse
technological styles—distinctive ways of doing things that in most cases achieve
more or less the same result. By the 1990s and 2000s, the concept of chaîne opéra-
toire, borrowed from French anthropology, sought to explore the different elements
of these technological styles. Within that paradigm, scholars seek to simultaneously
consider the physical properties and cultural ethos associated with technologies.
While this is ground-breaking work in global archaeology, Africanists have long rec-
ognized and appreciated Africa’s divergent technological styles (see, for example,
Rickard 1939; Cline 1937). Therefore, the view that technology is culturally em-
bedded and is socially constructed has always been axiomatic in studies of Africa’s
preindustrial metallurgy—if only global archaeology was paying more attention!
Given that it is now universally acknowledged that Africa’s preindustrial tech-
nologies were culturally mediated, responses to a specific situation and not relics of
technologies extinct elsewhere, it is important to invest more in understanding the
process from an interdisciplinary and diachronic point of view. As an example, a
study by Heimann et al. (2010) focusing on the complex mineralogy of tin smelting
slags from Southern Africa has greatly enhanced our understanding of the intri-
cacies of preindustrial tin smelting. Furthermore, an exploratory statistical study
of slags from African archaeological sites conducted by Chirikure and Bandama
(2014) indicated that different furnaces used in preindustrial Africa had different
reduction capacities. The tall natural draught furnaces were more reducing, such
that their slag composition appeared closer to equilibrium than those from low shaft
and bowl furnaces when plotted on ternary diagrams. The decision to choose a
specific furnace type depended on many other variables such as efficiency in labour
and time, quality of ore and nature of smelting technology (Chirikure and Reh-
ren 2006). Clearly, there is more to be learned from Africa regarding technological
development beyond being a mere source of ethnography. At the same time, the
shadreck.chirikure@uct.ac.za

157Local Responses to Technology Transfer and Knowledge Dispersal
ethnoarchaeological/ethnographic insights into the importance of ritual in African
technological practice hold potential to enrich studies of metallurgy elsewhere—
less in terms of the specifics of ritual practice, and more in relation to the culturally
embedded character of what is often perceived as “purely” technological practices
(Stahl 2014a, b).
Local Responses to Technology Transfer and Knowledge
Dispersal
The subject of Africa’s preindustrial metallurgy is intricately interwoven with
broader issues related to technology transfer and knowledge transmission between
interacting peoples. As part of the Old World, Africa has always explicitly and im-
plicitly interacted with various polities in Eurasia. The resources of the Egyptian
Sudan and adjacent margins contributed to the Middle Eastern political economy.
Black Nubian monarchs also ruled Egypt as the 25th Dynasty. However, the connec-
tions between North Africa and the Middle East on the one hand and sub-Saharan
Africa on the other before the onset of the first millennium AD are at best vague and
at worst opaque. The presence of this “black hole” makes it difficult to understand
cultural and knowledge exchanges between regions to the north and south of the
Sahara. Obviously, the origins of African metallurgy features strongly in this debate
because the routes for the supposed north to south transmission of knowledge have
not been well articulated, not least because of a lack of research and the concomitant
deficiencies of radiocarbon in the interval 800–400 BC. Perhaps, the most impor-
tant issue as far as technological transfer is concerned with respect to origins ques-
tions centres on the time lag between adopting a new technology and modifying it
to suit the local situation. Surely, the outward differences between the metallurgy
of the source areas and that of sub-Saharan Africa suggest a complex and confusing
technology transfer mechanism (if such a transfer happened at all).
The issue of innovation is intimately associated with the local origins thinking
and the subsequent development of the technology. The lesson from the past is that
innovations hardly follow “common sense”. Often an accident results in a discov-
ery of huge significance. Yes, it is pyrometallurgically more difficult to smelt iron
than copper, but some furnaces used in preindustrial Africa comprised rudimentary
structures—in some cases made of banana stems—that still produced iron (Celis and
Nzikoyabanka 1976). Furthermore, temperatures for reducing copper are not that
different from those for reducing iron given the fact that typical iron and copper ores
contained gangue materials which required slag formation at temperatures above
1000 °C. The only answer lies in more research at sites with evidence of early Afri-
can metallurgy. Fundamentally, some perspectives that threaten acquired knowledge
or the orthodoxy may not be widely accepted (Holl 2009). For example, would the
Catholic Church of the Middle Ages have envisaged that Copernicus’ idea that the
world was a globe would become an unquestioned fact? Whatever the mechanism for
origins and innovation, once established, it is clear that Africa’s preindustrial metal-
shadreck.chirikure@uct.ac.za

158 7 Bridging Conceptual Boundaries, A Global Perspective
lurgy underwent regional and context-specific innovations, which produced a very
diverse range of technological styles from the large-scale production in West Africa
to the small-scale production in comparatively smaller furnaces in Southern Africa.
That Africa did not adopt metalworking traditions of the Eurasian world after in-
creased contact between these two areas in the first millennium AD underscores that
the adoption of technologies is a culturally mediated experience. In addition, some
of the Eurasian technologies were designed for very large populations, which were
not an issue in sparsely occupied areas of sub-Saharan Africa. That existing tech-
nologies could meet local demand was all that mattered. Even so the quality was
admirable and in some cases better than that produced in the so-called advanced
blast furnaces of Europe in the mid-19th century (Chirikure 2006).
African Metals: Land and Sea Links and Protoforms of
Globalization
Africa has always been part of the Old World, participating in different social and
economic relations. The varying mineral availability gradients precipitated contact
between different areas. Internally in Africa, this connected different communities.
However, it seems that technological styles were an important aspect of identity, for
it was rare for Africans to adopt methods of the others. Different regions of Africa
such as Southern and Central Africa were well networked, just as those of West and
North Africa. In Southern Africa too, contact and cultural exchanges took place be-
tween the coastal communities and hinterland communities. These land links were
vital for long-distance trade for the objects that formed the life blood of this system
came from hinterland areas.
The long-distance trade that linked Southern Africa and the Indian Ocean world
via coastal East African communities resulted in a long-distance network spanning
continents. African gold, iron and other resources such as ivory were exchanged
for Indo-Pacific and Persian glass beads, Persian ceramics and Chinese porcelain.
Eurasian alloys such as brass, leaded gun metal and bronze were also brought to Af-
rica. Despite this contact, technology transfer was minimal. The trans-Saharan trade
also introduced trade beads and brass to West Africa just as the Atlantic trade based
in the same region. The commodities from long-distance trade were incorporated
into local value systems where their possession was a source of political leverage.
Those who controlled the trade, sometimes had access to political power. However,
in some cases, the trade was very open with citizens free to participate in the trade
and the elites were not always in control. Political power was based on control over
land and ideology and not distribution and redistribution of imports (Bhila 1982).
In this contact, not all objects and technologies were locally accepted in
Africa. Brass was imported, but worked using local techniques. This shows that
the acceptance of technologies is culturally specific. Prestholdt (2004) notes that
metal manufacturers in the USA adjusted their standards of copper wire to meet the
demands of East African trade, producing gauge acceptable to groups such as Maa-
sai. Similarly, in their encounter with Southern Africa, the Portuguese were frus-
shadreck.chirikure@uct.ac.za

159Changing Contexts of Knowledge Production and the Future of African …
trated because its inhabitants shunned European glass beads, preferring those from
India. Africa supplied commodities that through trade networks enabled people in
Eurasia to have access to luxuries and necessities such as porcelains and gun pow-
der among other commodities. Therefore, the world has always depended on the
mother continent for raw materials. Underdevelopment theorists such as Rodney
(1974) argue that this position disadvantaged the continent, but it is also a result of
the disruption of the African value system at conquest.
Changing Contexts of Knowledge Production and the
Future of African Preindustrial Metallurgy
I would remark that of the woolly haired Africans, who constitute the principal part of the
inhabitants of Africa, there is no history, & there can be none. That race has remained in
barbarism from the first ages of the world; their country has never been explored very fully
by civilized man (Webster cited by Yacovone 2002, p. ii)
According to Holl (2000, p. 6) “throughout the colonial period, sub-Saharan Af-
rica was considered a backward continent on the receiving end of technological
innovations”. For example, preindustrial metal production in the subcontinent has
historically been viewed as derivative in its origins and retarded in its development
(Rickard 1939). There was a popular belief in the West that claimed that human
societies had evolved through several stages from savagery through barbarism to
civilization (see Robertshaw 1990). Whereas Europeans perceived themselves as
having reached the civilized stage with their level of technological sophistication,
Africa was perceived to be still languishing at the foot of the development ladder
in savagery. It is, therefore, not surprising that explorers, travelers and colonialists
who either spread throughout the continent or commented on Africa on the eve of
colonization argued that Africans occupied a stage in the evolutionary tree that Eu-
ropeans had passed a thousand or more years ago (Hall 1987, p. 5).
In 1843, a black American religious leader and anti-slavery campaigner resident
in New Haven, Connecticut, Amos G. Beman approached Noah Webster, the famed
propagator of the first American Dictionary for guidance on what were the best texts
about the African people. According to Yacovone (2002, p. ii), Webster responded
with the quote that opens this section. However, if Mr. Beman was interested in
northern fringes of the continent such as Egypt, Webster retorted that any good ency-
clopedia was good enough. The net result of such notions was the nurturing of long-
lasting stereotypes fuelled by biased social values and Western racial priorities (Ya-
covone 2002). Going back to Hegel in the 1820s, African societies and technologies
such as iron working were thought to be in a “deep and perpetual slumber” without
any advancement (Brown 1973, p. 3; Curtin et al. 1978; Goody 1971). On his part,
Stanley (1931) argued that “from about 1200 BC onward to the making of iron in the
present by the Negroes, the production of iron has been entirely the same that I know
no way of distinguishing it…” (Stanley cited in Caton-Thompson 1931, p. 201).
Ritual and symbolism or the materiality of African metal production was not
spared from derogatory and biased perceptions. For example, when he failed to
shadreck.chirikure@uct.ac.za

160 7 Bridging Conceptual Boundaries, A Global Perspective
understand the significance of symbolic dimensions of Venda iron workers in South
Africa, Beuster (1889, cited in Rickard 1939, p. 89) posited that, “it was the ancient
custom … for the smith to add human flesh to the ore in order that iron might make
a good hoe, and if no flesh was available the smith sought for it among the dead”. As
an expert on Venda ethnography, Van Warmelo (1935) did not find any evidence for
scavenging the dead, and thus, such statements resonate very well with notions of
a dark and savage Africa. Of course, Plug and Pistorius (1999) report the recovery
of human finger bones from pits in the floor of some furnaces at Phalaborwa, but it
is unlikely that they were scavenged from the dead who are viewed as sacred in the
area. It would, therefore, appear that the whole idea was exaggerated by Beuster, in
so doing fulfilling stereotypical views of African society, culture and technology.
However, as we have seen, if the multiple furnace types and technological styles
that dot the archaeological landscape across the African continent are considered,
there is ample evidence for considerable historical and regional diversity, innova-
tion and variation (Cline 1937; Miller et al. 2001; Okafor 1993; Prendergast 1975;
Schmidt 1997; Sutton 1985). The Mafa smelters of Northern Cameroon, for in-
stance, produced cast iron from their furnaces, a product otherwise restricted to the
blast furnace method introduced after colonization (David et al. 1989). The Fipa
of Tanzania employed both bellows-driven and natural draught furnaces. Okafor
(1993) has reported on the existence of slag tapping in the Late Iron Age of Nsukka,
eastern Nigeria, a technological development not documented in the preceding Ear-
ly Iron Age of the region. The Njanja devised ways of increasing the air to fuel to
ore ratio and reorganized their production by employing a shift system of labour to
initiate a hugely successful metal production enterprise (Chirikure 2006). This dia-
chronic and spatial diversity in the technologies of practice in preindustrial metal-
lurgy across sub-Saharan Africa unravels an important source of comparative cases
in any study of the different trajectories of metallurgy locally, regionally and glob-
ally. For example, why did one type of iron smelting furnace (the tall natural draft
furnace) spread so widely in the second millennium AD, but other technologies,
like the Mafa furnaces, have not spread beyond the Mandara Mountains? This im-
plicates the presence of historical and demographic processes at work in both cases.
Since its advent, the material and materiality of African metal production played
an important role in making items for economic, social and political ends as shown
by tools such as hoes, art that represented leaders, and trade and exchange relation-
ships that linked West Africa, North Africa, and the Middle East and Southern Africa
and the Indian Ocean rim on the other. Thus, metalworking was a major nexus that
fed the heart of local, regional and international interconnections, with the conse-
quence that it stimulated varied economic, political and economic systems spatially
and diachronically. Therefore, rather than being, isolated, desolate, and a cultural
backwater, Africa lay at the heart of the development of the world from very early
on. It is impossible to discuss the success of various generations of Eurasians, with-
out the sterling role that Africa played in supplying raw materials, finished products
and ideas in history. The Islamic world system thrived on African gold from the
mid- to late first millennium AD onward, and the same applies to the Indian subcon-
tinent. Later in history, African metals paid for European luxuries and in the 19th
century was one of the principal reasons for the scramble and partition of Africa. As
shadreck.chirikure@uct.ac.za

161Conclusion
such, the world’s history, values and attitudes are all inscribed in the development
of Africa and its interactions with the rest of the world at different temporal scales.
Conclusion
It is clear that metals played an important role in African societies and those of the
other continents. There are so many lessons for understanding technology in the
world based on the African experience. Africa offers a different technological itin-
erary as far as metallurgy is concerned. As such, one of the greatest enigmas of our
time is the origins of African metallurgy. Although this question is still debated, it
has implications for understanding technology transfer and innovation. Innovations
are technologically and culturally specific and often result from accidents.
The process of metal production is socially embedded, evoking the power of the
ancestors, magic and the deities. It is, therefore, a fallacy of late 19th-century and
20th-century science that magic had no role in science and technology (Hansen
1986). This universal was witnessed in other parts of the world such as Asia and
Latin America, showing that metallurgy participated in the production and repro-
duction of society and was therefore an integral component. The differences in value
systems may have prevented technology transfer between various regions. There is
need to consider the role of population size in technology transfer and change. Huge
populations require technologies suited for servicing them as do small populations,
and such differences often motivate variations that are visible archaeologically.
Because metallurgy participated in the production of society, it illuminates inter-
esting gender division of labour and cross-borrowing. The full production chain of
metallurgy included a special role and place for women and their material culture.
Winnowing was adopted for panning gold in Northern Zimbabwe just as grinding
was used to separate ore from waste materials. Most crucibles used in Southern
and Central Africa were domestic pots made by women. This is hardly surprising
because crucibles, pots and furnaces are all containers important for heat-mediated
transformations. Metal production represented men appropriating symbolically the
reproductive power of women. As such, while some studies are quick to make dis-
tinctions between metallurgy as a purely male domain and other pursuits which
were female, the division is not crystal clear as there were so many cross-overs,
metaphorically and in practice.
Finally, far from being isolated, Africa played an important role in the develop-
ment of the world. As such, it was an important geopolitical space in the past. The
immediate adoption of metallurgy may not have been revolutionary, but the ag-
gregation of short and long-term impacts shows that the technology has had a big
impact from value systems to subsistence and defence strategies to wealth accumu-
lation and political power. As such, in order to understand society, it is important
to study such influential technologies for they participated in societal continuity
and change. Therefore, technology is fundamentally about humans, and if we learn
about material and physical properties, we learn about the various aspects of hu-
manity. This is the message that comes from Africa’s preindustrial metallurgy as
shadreck.chirikure@uct.ac.za

162 7 Bridging Conceptual Boundaries, A Global Perspective
studied from an integrated view which develops a synergy between ethnographic,
historical, geological and archaeological and archaeometallurgical points of view.
References
Alpern, S. B. (2005) Did they or didn’t they invent it? Iron in sub-Saharan Africa. History in
Africa, 32, 41–94.
Bachmann, H. (1982). The Identification of slags from archaeological sites. (Occasional Publica-
tion no 6). London: Institute of Archaeology.
Bhila, H. H. K. (1982). Trade and politics in a Shona Kingdom: The Manyika and their African
and Portuguese neighbours (pp. 1575–1902). Harlow: Longman.
Brown, J. (1973). Early iron production. Rhodesian Prehistory, 7, 3–7.
Caton-Thompson, G. (1931). The Zimbabwe culture: ruins and reactions. London: F. Cass.
Celis, G., & Nzikobanyanka, E. (1976). La métallurgie traditionnelle au Burundi: Techniques et
croyances (Archives d’anthropologie, no. 25). Tervuren: Musée royal de l'Afrique centrale.
Childs S. T. (1991). Style, technology and iron smelting furnaces in Bantu-speaking Africa. Jour-
nal of Anthropological Archaeology, 10(4), 332–59.
Chirikure, S. (2005). Iron production in Iron Age Zimbabwe: Stagnation or innovation. Unpub-
lished doctoral dissertation, University College, London.
Chirikure, S. (2006). New light on Njanja iron working: Towards a systematic encounter between
ethnohistory and archaeometallurgy. South African Archaeological Bulletin, 61, 142–151.
Chirikure, S. (2010). On evidence, ideas and fantasy: The origins of iron in sub-Saharan Africa.
Thoughts on É. Zangato & AFC Holl’s On the iron front. Journal of AfricanArchaeology, 8(1),
25–28.
Chirikure, S., & Bandama, F. (2014). Indigenous African furnace types and slag composition—Is
there a Correlation? Archaeometry, 56(2), 296–312.
Chirikure, S., & Rehren, T. (2006). Iron smelting in pre-colonial Zimbabwe: Evidence for dia-
chronic change from Swart Village and Baranda, Northern Zimbabwe. Journal of African Ar-
chaeology, 4, 37–54.
Cline, W. B. (1937). Mining and metallurgy in Negro Africa (No. 5). Menasha: George Banta
Publishing Company.
Craddock, P. T. (2010). New paradigms for old iron: Thoughts on É. Zangato & A.F.C. Holl’s on
the Iron Front. Journal of African Archaeology, 8, 29–36.
Curtin, D., Feierman, S., Thompson, L., & Vansina, J. (1978). African history. London: Longman.
David, N., Heimann, R., Killick, D., & Wayman, M. (1989). Between bloomery and blast furnace:
Mafa iron-smelting technology in North Cameroon. African Archaeological Review, 7(1),
183–208.
Goody, J. (1971). Technology, tradition and the state in West Africa. London: Oxford University.
Hall, M. 1987. The changing past. Cape Town: David Phillip.
Hansen, B. (1986). The complementarity of science and magic before the scientific revolution.
American Scientist, 74(2), 128–136.
Heimann, R. B., Chirikure, S., & Killick, D. (2010). Mineralogical study of precolonial (1650–
1850 CE) tin smelting slags from Rooiberg, Limpopo Province, South Africa. European
Journalof Mineralogy, 22(5), 751–761.
Holl, A. (2000). Metals and precolonial African society. In M. Bisson, S. T. Childs, P. de Barros,
& A. Holl (Eds.), Ancient African Metallurgy: The sociocultural context (pp. 1–81).Walnut
Creek: AltaMira Press.
Holl, A. F. (2009). Early West African metallurgies: New data and old orthodoxy. Journal of World
Prehistory, 22(4), 415–438.
Insoll, T. (2008). Negotiating the archaeology of destiny. An exploration of interpretive possibili-
ties through Tallensi Shrines. Journal of Social Archaeology, 8(3), 380–403.
shadreck.chirikure@uct.ac.za

163References
Kense, F. J., & Okoro, J. A. (1993). Changing perspectives on traditional iron production in West
Africa. In T. Shaw, P. Sinclair, B. Andah, & A. Okpoko (Eds.), The archaeology of Africa:
Food, metals and towns (pp. 449–458). London: Routledge.
Killick, D., & Fenn, T. (2012). Archaeometallurgy: The study of preindustrial mining and metal-
lurgy. Annual review of anthropology, 41, 559–575.
Lechtman H. (1977). Style in technology—some early thoughts. In H. Lechtman & R. Merrill
(Eds.), Material culture: Styles, organization and dynamics of technology (pp. 3–20). New
York: West.
Miller, D., Killick, D., & Van der Merwe, N. J. (2001). Metal working in the Northern Lowveld,
South Africa AD 1000–1890. Journal of Field Archaeology, 28(3–4), 3–4.
Mitchell, P. (2005). African connections: Archaeological perspectives on Africa and the wider
world. Walnut Creek, CA: AltaMira.
Okafor, E. (1993). New evidence on early iron-smelting in Southeastern Nigeria. In T. Shaw, P.
Sinclair, B. Andah, & A. Okpoko (Eds.), The archaeology of Africa: Food, metals, and towns
(pp. 432–448). London: Routledge.
Plug, I., & Pistorius, J. C. C. (1999). Animal remains from industrial Iron Age communities in
Phalaborwa, South Africa. African Archaeological Review, 16(3), 155–184.
Prendergast, M. D. (1975). A new furnace type from the Darwendale Dam basin. Rhodesian Pre-
history, 7(14), 16–20.
Prestholdt, J. (2008). Domesticating the world: African consumerism and the genealogies of glo-
balization. Berkeley: University of California Press.
Prestholdt, J. (2004). On the global repercussions of East African consumerism. The American
Historical Review, 109(3), 755–781.
Pryce, T. O., Pigott, V. C., Martinón-Torres, M., & Rehren, T. (2010). Prehistoric copper produc-
tion and technological reproduction in the Khao Wong Prachan Valley of Central Thailand.
Archaeological and Anthropological Sciences, 2(4), 237–264.
Rickard, T. A. (1939). The primitive smelting of iron. American Journal of Archaeology, 43(1),
85–101.
Robertshaw, P. (Ed.). (1990). A history of African archaeology. London: James Curry.
Rodney, W. (1974). How Europe underdeveloped Africa. Dar es Salam: Dar es Salam University
Press.
Schmidt, P. R. (1978). Historical Archaeology: A structural approach in an African culture. West-
port: Greenwood Press.
Schmidt, P. R. (1997). Iron technology in East Africa: Symbolism, science, and archaeology.
Bloomington: Indiana University Press.
Schmidt, P. R. (2001). Resisting homogenisation and recovering variation and innovation in Afri-
can iron smelting. Mediterranean Archaeology, 14, 219–27.
Schmidt, P. R. (2009). Tropes, materiality, and ritual embodiment of African iron smelting fur-
naces as human figures. Journal of Archaeological Method and Theory, 16(3), 262–282.
Smith, C. S. (1981). A search for structure. Selected essays on science, art and history. Cambridge:
MIT Press.
Stahl, A. B. (2014a). Africa in the World: (Re)centering African history through archaeology.
Journal of Anthropological Research, 70(1), 5–33.
Stahl, A. B. (2014b). Metal working and ritualization: Negotiating change through improvisational
practice in Banda, Ghana, AD 1300–1650. In L. Overholtzer & C. Robin (Eds.), The material-
ity of everyday life. Archaeology papers of the American Anthropological society. Arlington:
American Anthropological Society.
Sutton, J. 1985. Temporal and spatial variability in African iron furnace. In R. Haaland & P. L.
Shinnie (Eds.), African Iron working-ancient and traditional (pp. 164–197). Oslo: Norwegian
University Press.
Van Warmelo, N. J. 1935. A preliminary survey of the Bantu tribes of South Africa. Pretoria: Gov-
ernment Printer.
Yacovone, D. (2002). Editor’s Introduction. Special issue, race & slavery. Massachusetts Histori-
cal Review, 4, i–iii.
shadreck.chirikure@uct.ac.za

165© The Author 2015
S. Chirikure, Metals in Past Societies, SpringerBriefs in Archaeology,
DOI 10.1007/978-3-319-11641-9
Index
A
Agriculture, 43, 126, 127, 133
swidden, 126
Akan, 42, 108, 110
Akjoujt, 20, 24, 29, 78, 105, 152
Amos G. Beman, 159
Annealing, 103, 115
Anthropology, 2, 10–12, 63, 88, 156
of mining, 2, 12, 35, 38, 40, 43, 44, 46, 49,
51, 56, 64
of smelting, 63, 64, 67, 75, 79, 88,
Anthropology of metal fabrication, 119
Archaeology, 3, 5, 6, 7, 10–12, 43, 49, 63, 65,
78, 85, 134, 154–159
of mining, 43
of smelting, 43
of smithing, 110
Asante, 29, 115, 121, 127, 131
B
Backfilling,47, 56
Backward Africa, 153
Bag, 67
Beliefs, 37, 40, 55, 88, 93, 120, 134, 154
Bellows, 61, 63, 67, 70, 72, 75, 81, 84, 88, 92,
102, 103
goatskin, 106, 111, 112
Bowl, 67
ceramic, 70
furnaces, 65, 70, 80, 84, 85, 155, 156
Bridging conceptual boundaries, 151
Buhaya, 5, 27, 83, 84, 91
Buhen, 19, 71
Burkina Faso, 106
C
Copper ingots, 70, 112, 118, 139
Cowrie, 142
D
Decorated furnaces, 90
Dekpassanware, 64, 75, 107
Diffusion, 50, 152
Dimi of Ethiopia, 111
E
Economy, 127, 157
El Tio, 37, 55
Ethnography of, 87
mining, 109
smelting, 104, 105, 109, 110, 184
smithing, 16
Exotic imports, 144, 145
Expressive objects, 99
F
Fire setting, 44, 49
Fitola, 26, 27
Furnace types, 7, 24, 27, 63, 67, 76, 79, 80,
153, 160
G
Glass beads, 25, 116, 127, 136, 137, 141–144,
146, 152, 158, 159
Gold weights, 108
Gold working, 115–117, 121
Great Zimbabwe, 12, 30, 38, 86, 116, 117,
119, 120, 127, 128, 131, 134, 136,
137, 139
H
Hammering, 101–104, 106, 107, 111, 116
Hoisting, 50, 53
ore, 46
Human reproduction, 88
shadreck.chirikure@uct.ac.za

166 Index
I
Igbo Ukwu, 29, 78, 108, 127, 143
Imported materials, 106, 116, 143, 144
Improvisation, 88, 90
Independent origins, 20, 27, 120
Iron and kingship, 134
K
Kansanshi, 45, 65, 81, 82
Kaonde group, 44, 64
Khami, 116, 119, 127, 131, 136
L
Leija, 26, 27
Lobola, 120
Low shaft, 65, 67, 78, 80, 84, 155, 156
Luminescence dates, 27
M
Mabveni, 23
Magic, 6, 10, 53, 55, 61, 88, 154, 161
Magnetite, 6, 8, 37, 41, 43, 52, 64, 77, 79, 87
sands, 41, 43, 64
skins, 77
Malachite, 52, 64, 70, 71, 81
Mandara Mountains, 6, 41, 107, 140, 160
Mapela, 131, 134, 137
Mapungubwe, 116, 118, 120, 127, 131, 134,
136, 137
Mastaba of Mereruka, 71
Meroe, 19, 27, 39, 67, 72, 75, 92, 121, 126,
140
Metals and socio-political complexity, 130,
131, 133, 145
Metals and urbanism, 130, 131
Metaphors, 11, 79, 89
of reproduction, 88, 90. 92
socio-cultural, 88
Mining tools, 49, 53
Molds, 81, 103, 104, 108, 112, 116, 117
soapstone, 117
steatite, 108
Musuku, 114, 115, 138
N
Natural draught, 65–67, 75, 76, 80, 82, 86,
106, 107, 155, 156, 160
Nubian Pharaohs, 21
O
Obui, 25
Old wood problem, 24, 28
Open mining, 40, 42, 44–46, 49, 50
Ores
copper, 21, 37, 38, 40, 46, 125, 157
P
Pictographs in Egyptian tombs, 70
Political power, 134–136, 158, 161
Pot, 67, 70, 72, 79, 84, 92, 108
R
Radiocarbon black hole, 17, 29, 77, 105, 151
Reproduction of society, 161
Ritual and magic, 6
Rwiyange, 23
S
Sanga, 181, 130
Scale of production
fabrication, 153
mining, 12
smelting, 39, 76
Shackles, 107, 108, 140
Shona, 5, 6, 42, 45, 51, 52, 64, 120, 136, 138,
140, 141
Smelting
iron, 152
Smelting East Africa, 92, 152
Smelting Egypt, Nubia and North Africa, 63
Smelting Southern Africa, 84, 85, 87, 88, 92
Smelting West Africa, 27, 64, 74
Smithing Egypt, Nubia and North Africa, 63
Southern Zambezia, 5, 51, 52, 112, 139
Sukur, 106, 107, 121, 140
Swahili, 12, 30, 115, 141
T
Technology transfer, 19, 29, 138, 140, 144,
152, 157, 161
Trade and exchange, 7, 121, 126, 127, 136,
138, 160
U
Underground mining, 37, 38, 40, 45–47, 49,
50, 56
Urbanism, 131
V
Value transfer, 142, 152
W
Wadi Dara, 64, 69
Waldadé, 77
Wandala, 130, 140
Wire drawing, 106, 117, 118, 140
Y
Yeke, 44, 81
shadreck.chirikure@uct.ac.za