Burials, Migration and Identity in the Ancient Sahara and Beyond 110847408X, 9781108474085

Citation preview

Burials, Migration and Identity in the Ancient Sahara and Beyond

This ground-breaking volume explores a series of interrelated key themes in Saharan archaeology and history. Migration and identity formation can both be approached from the perspective of funerary archaeology, using the combined evidence of burial structures, specific rites and funerary material culture, and integrated methods of skeletal analysis including morphometrics, palaeopathology and isotopes. Burial traditions from various parts of the Sahara are compared and contrasted with those of the Nile Valley, the Maghrib and West Africa. Several chapters deal with the related evidence of human migration derived from linguistic study. The volume presents the state of the field of funerary archaeology in the Sahara and its neighbouring regions and sets the agenda for future research on mobility, migration and identity. It will be a seminal reference point for Mediterranean and African archaeologists, historians and anthropologists as well as archaeologists interested in burial and migration more broadly. maria carmela gatto is Honorary Visiting Fellow at the University of Leicester specialising in the prehistoric and historic archaeology of the Nile Valley and the Sahara. She is the director of the Aswan-Kom Ombo Archaeological Project in Egypt and has extensively worked on the Garamantes of Central Sahara, also as Research Associate on the Trans-SAHARA Project. david j. mattingly is Professor of Roman Archaeology at the University of Leicester. He has worked in the Sahara for almost forty years and is the author of many books and articles related to Saharan archaeology, such as Farming the Desert (2 vols, 1996), which won the James R. Wiseman book award of the American Institute of Archaeology, and The Archaeology of Fazzan series (4 vols, 2003–2013). He was the principal investigator of the European Research Council–funded TransSAHARA Project (2011–2017) which created the groundwork for this volume, and he is the overall series editor of Trans-Saharan Archaeology, in which this is the second of four projected volumes.

Published online by Cambridge University Press

nick ray is an archaeologist who specialises in the Roman economy, consumption and funerary archaeology in North Africa. He is the Assistant Director of the Oxford Roman Economy Project at the University of Oxford and previously worked as a research associate on the Trans-SAHARA Project at the University of Leicester. He has co-edited several books, including De Africa Romaque (2016). martin sterry is Teaching Fellow in Roman Landscape Archaeology and GIS at Durham University. His research on the archaeology of the Sahara and North Africa makes particular use of GIS and remote sensing. He has undertaken fieldwork on various projects in Italy, Britain, Libya and most recently southern Morocco, where he is co-director of the Middle Draa Project. He has published many articles on the Libyan Fazzan, Saharan trade, urbanisation and oasis settlements.

Published online by Cambridge University Press

Burials, Migration and Identity in the Ancient Sahara and Beyond Edited by

m. c. gatto University of Leicester

d. j. mattingly University of Leicester

n. ray University of Oxford

m. sterry Durham University Trans-Saharan Archaeology Volume 2 Series Editor: D. J. Mattingly

The Society for Libyan Studies

Published online by Cambridge University Press

University Printing House, Cambridge CB2 8BS, United Kingdom One Liberty Plaza, 20th Floor, New York, NY 10006, USA 477 Williamstown Road, Port Melbourne, VIC 3207, Australia 314–321, 3rd Floor, Plot 3, Splendor Forum, Jasola District Centre, New Delhi – 110025, India 79 Anson Road, #06–04/06, Singapore 079906 Cambridge University Press is part of the University of Cambridge. It furthers the University’s mission by disseminating knowledge in the pursuit of education, learning and research at the highest international levels of excellence. www.cambridge.org Information on this title: www.cambridge.org/9781108474085 DOI: 10.1017/9781108634311 Society for Libyan Studies © Cambridge University Press 2019 This publication is in copyright. Subject to statutory exception and to the provisions of relevant collective licensing agreements, no reproduction of any part may take place without the written permission of Cambridge University Press. First published 2019 Printed and bound in Great Britain by Clays Ltd, Elcograf S.p.A. A catalogue record for this publication is available from the British Library. Library of Congress Cataloging-in-Publication Data Names: Gatto, M. C. (Maria Carmela), editor. | Mattingly, D. J., editor. | Ray, Nick, editor. | Sterry, Martin, editor. Title: Burials, migration and identity in the ancient Sahara and beyond / edited by M.C. Gatto, D.J. Mattingly, N. Ray, M. Sterry. Description: Cambridge, United Kingdom ; New York, NY : Cambridge University Press, 2019. | Includes bibliographical references. Identifiers: LCCN 2018022370 | ISBN 9781108474085 (Hardback) Subjects: LCSH: Funeral rites and ceremonies – Africa, North. | Africa, North – Emigration and immigration – History. | Archaeology – Africa, North. Classification: LCC GT3289.A35 B87 2019 | DDC 393/.930961–dc23 LC record available at https://lccn.loc.gov/2018022370 ISBN 978-1-108-47408-5 Hardback Cambridge University Press has no responsibility for the persistence or accuracy of URLs for external or third-party internet websites referred to in this publication and does not guarantee that any content on such websites is, or will remain, accurate or appropriate.

Published online by Cambridge University Press

Contents

List of Figures [page viii] List of Tables [xvi] List of Contributors [xviii] Preface [xxi] david j. mattingly

1 Burials, Migration and Identity: The View from the Sahara [1] david j. mattingly, maria carmela gatto, martin sterry and nick ray part i burial practices in the central sahara [51] 2 Dying to be Garamantian: Burial, Migration and Identity in Fazzan [53] david j. mattingly, martin sterry and nick ray 3 Identity Markers in South-Western Fazzan: Were the People of the Wadi Tanzzuft/Tadrart Akakus Region Garamantes? [108] maria carmela gatto, lucia mori and andrea zerboni 4 Human Mobility and Identity: Variation, Diet and Migration in Relation to the Garamantes of Fazzan [134] ronika k. power, efthymia nikita, david j. mattingly, marta mirazó n lahr and tamsin c. o’connell 5 The Garamantes from Fewet (Ghat, Fazzan, Libya): A Skeletal Perspective [162] francesca ricci, mary anne tafuri, francesca castorina, fabio di vincenzo, lucia mori and giorgio manzi

v

Published online by Cambridge University Press

vi

Contents

part ii looking east [193] 6 Between the Nile and the Sahara: Some Comparative Perspectives [195] david n. edwards 7 Isotopic Approaches to Mobility in Northern Africa: A Bioarchaeological Examination of Egyptian/Nubian Interaction in the Nile Valley [223] michele r. buzon, sarah a. schrader and gabriel j. bowen part iii looking north [247] 8 Numidian Burial Practices [249] joan sanmartí , irene cruz folch, jordi campillo and david montanero 9 Revisiting First Millennium BC Graves in North-West Morocco [281] emanuele papi part iv looking west [313] 10 Protohistoric and Pre-Islamic Funerary Archaeology in the Moroccan Pre-Sahara [315] youssef bokbot 11 Burial Practices in Western Sahara [341] joanne clarke and nick brooks part v looking south [373] 12 Burial and Society at Kissi, Burkina Faso [375] sonja magnavita 13 Burial Practices, Settlement and Regional Connections around the Southern Lake Chad Basin, 1500 BC–AD 1500 [399] scott maceachern part vi linguistic aspects of migration and identity [429] 14 The Linguistic Prehistory of the Sahara [431] roger blench

Published online by Cambridge University Press

Contents

15 Berber Peoples in the Sahara and North Africa: Linguistic Historical Proposals [464] christopher ehret 16 The Archaeological and Genetic Correlates of Amazigh Linguistics [495] elizabeth fentress 17 Concluding Discussion [525] martin sterry, david j. mattingly, maria carmela gatto and nick ray Index [549]

Published online by Cambridge University Press

vii

Figures

1.1. The Trans-Saharan zone, with indication of the major areas and sites covered in this volume. (made by M. Sterry) [page 8] 1.2. a) Garamantian settlements in Fazzan and key regional toponyms; b) Garamantian cemeteries in Fazzan. (made by M. Sterry and M. C. Gatto) [11] 2.1. Location map of tombs and cemeteries excavated by the DMP. (Imagery ©Esri) [55] 2.2. Typology of cairns, shafts and other simple burials. [58] 2.3. Typology of more complex tombs and monumental burials. [59] 2.4. Distribution of Type 1 and Type 3 tombs along the escarpment. [65] 2.5. Distribution of Type 2 and Type 5 tombs along the escarpment. [67] 2.6. Distribution of Classic Garamantian Type 2 and more complex tombs along the Wadi al-Ajal. [68] 2.7. Skeletal orientation and side on which the body was laid. [82] 2.8. Skeletal orientation and side compared for Jarma area and Taqallit area. [82] 2.9. Orientation of head in relation to sex/age of individual. [83] 2.10. Radar charts showing orientation of head in relation to sex/ age of individual. [83] 2.11. Mean number of vessels per tomb for different classes of vessel and types of tomb. [92] 2.12. Mean number of vessels per tomb for different time periods. [93] 2.13. a) and b) Mean number of vessels per tomb for different gendered classifications. [95] 2.14. a) and b) Relationship between presence of different bead types and tomb types. [96]

viii

Published online by Cambridge University Press

List of Figures

2.15. Relationship between presence of different bead types and time periods. [97] 2.16. Relationship between presence of different bead types and gender classifications. [97] 2.17. Distribution of stelae along the escarpment. [98] 2.18. Distribution of ceramic surface finds along the escarpment. [99] 3.1. Satellite image illustrating the main sites in the Wadi Tanzzuft and the Tadrart Akakus region (after Liverani 2006, Fig. 1.2). [109] 3.2. The so-called ‘Royal Tumulus’ at In-Aghelachem: a) in its landscape (picture taken from west); b) its plan (after di Lernia and Manzi 2002, Figs 5.42 and 5.43). [112] 3.3. a) Conical and b) drum-shaped tumuli from Fewet; (below) Plans and profiles of the tombs. [113] 3.4. Leather shroud wrapping the corpse of a burial in the Fewet cemetery. [116] 3.5. a) Pictures and plans of Aghram Nadharif citadel; b) reconstruction (drawing T. D’Este). [121] 3.6. a) The remains of the Fewet compound after the excavation; b) reconstruction of the ancient compound (drawing T. D’Este). [122] 3.7. a) Two-roomed dwelling modules in the Fewet compound after excavation; b) reconstruction (drawing T. D’Este). [123] 3.8. Vesicular bowls used as lamps, from Fewet (after Mori 2013, Fig. 8.6). [124] 3.9. Old water channel at the fringes of the Ghat oasis. [125] 4.1. Location map showing the sites mentioned in the text. [137] 4.2. Plot of tooth carbonate oxygen and carbon isotopic data from Garamantian burials. [146] 4.3. Comparison of Garamantian tooth carbonate oxygen isotopic data to other published data from Africa, the Levant and the Mediterranean. [147] 4.4. Scatterplot of δ13C/δ18O by cemetery. [150] 4.5. Scatterplot of Garamantian crania along PC1 and PC2, highlighting the population’s variation in size and shape. [151] 4.6. Visual comparison of two distinctive cranial morphologies among the Garamantes. [152]

Published online by Cambridge University Press

ix

x

List of Figures

4.7. Distribution of Garamantian crania along PC1 (size) and PC2 (broad vaults with narrow and orthognathic faces versus narrow vaults with broad and prognathic faces). [154] 5.1. Satellite image of the southern Wadi Tanzzuft with the location of the oases of Fewet, Ghat and Barkat and the Garamantian citadel of Aghram Nadharif (after Liverani 2003, Fig. 3.1). [163] 5.2. Mortality distribution according to sex. [167] 5.3. Distribution across time of the horizontal cranial shape at Fewet and 96/129. [169] 5.4. Principal component analysis of the cranial variables applied to the three skeletal samples from Fazzan. [170] 5.5. Canonical variates analysis of the cranial variables, applied to the three skeletal samples from Fazzan, with the addition of the medieval sample from Quingentole, northern Italy. [171] 5.6. BL diameter (in mm) of mandibular P4. Comparisons between Fewet and other North African samples. [173] 5.7. Side distribution of the asymmetry for the measurements of length and section. [176] 5.8. Stature estimated for Fewet series and comparisons, male samples. [178] 5.9. Jitterplot with strontium isotope ratios of human and animal dental enamel at Fewet. [182] 6.1. Map showing the location of sites and regions cited in the text (made by M. Sterry). [196] 7.1. Map showing location of Tombos and other sites used in this study. [224] 7.2. Distribution of human 87Sr/86Sr values for human samples in the Nile Valley. [229] 7.3. Distribution of δ18O values for human samples in the Nile Valley. [236] 7.4. Distribution of δ18O values for individuals identified as local and nonlocal using strontium isotope analysis. [236] 8.1. Map showing the locations mentioned in the text. [250] 8.2. Classification of North African funerary monuments. [253] 8.3. Haouanet of Sidi Latrech, Tunisia. [254]

Published online by Cambridge University Press

List of Figures

8.4. Choucha of Djebel Kharrouba, Algeria (after Payen 1863; re-drawn by M. Hawkes). [255] 8.5. Complex dolmen of Ellès (photograph by M. C. Belarte). [255] 8.6. The Thugga mausoleum, looking north (photo by J. Sanmartí). [256] 8.7. The Medracen. [257] 8.8. Bazina of Simitthus, Chemtou (Photo by J. Sanmartí). [258] 8.9. Zenithal access dolmen, Djebel Mazela (after Martin 1881, 1139, Fig. 117; re-drawn by M. Hawkes). [259] 9.1. The main centres of north-west Morocco between the eighth and third/second centuries BC. [282] 9.2. a) Tangier region cemeteries; b) the tumuli surrounding Oujda; c) Atlantic region tumuli (distribution of the ‘chefferies’ based on Hamdoune 1993). [284] 9.3. Orientation of tombs and skeletal remains, position of bodies and percentage of adults/children of the cemeteries of Aïn Dahlia el Kebiria, Djebila and Marchan/Tangier. [288] 9.4. Distribution of the tombs in the necropolis of Djebila. [289] 9.5. Dimensions of monumental tumuli in the Atlantic region. [290] 9.6. Section of selected monumental tumuli in the Atlantic region compared to the ‘Tombeau de la Chrétienne’, the Medracen and Djedar C (adapted from Camps 1961, Fig. 82) [294] 9.7. Grave goods from the necropolis of Aïn Dahlia el Kebiria. [295] 9.8. Grave goods from the necropolis of Djebila. [296] 9.9. Location of identified mines of north-west Morocco (adapted from Saadi 1975). [298] 9.10. African habitat of the Struthio camelus, probable ancient habitat in Morocco, and distribution of ostrich eggshells in first millennium BC sites. [304] 10.1. Distribution map of cemeteries from the Moroccan pre-Sahara (made by M. Sterry). [317] 10.2. Tumulus ‘à fosse’ from Erfoud (after Ruhlmann 1939, 74 redrawn by M. Hawkes). [318] 10.3. Bouïa Necropolis, tumulus Type 1 (after Margat and Camus 1959, 350, re-drawn by M. Hawkes). [319] 10.4. Bouïa Necropolis, tumulus Type 3 (after Margat and Camus 1959, 354, re-drawn by M. Hawkes). [320]

Published online by Cambridge University Press

xi

xii

List of Figures

10.5. Bouïa Necropolis, tumulus Type 2 (after Margat and Camus 1959, 352, re-drawn by M. Hawkes). [321] 10.6. Necropolis of Taouz with location of tombs cited in the text (Image © 2016 DigitalGlobe, Google Earth). [322] 10.7. Necropolis of Taouz, tumulus (T1) ‘à chapelle’ (photo Margat, Camps 1961, Pl. X.1). [322] 10.8. Necropolis of Taouz, tumulus (T4) ‘à chapelle’(after Meunié and Allain 1956, 74, re-drawn by M. Hawkes). [323] 10.9. Necropolis of Braber, stepped bazina with funerary chapel (Meunié and Allain 1956, 82, re-drawn by M. Hawkes). [324] 10.10. General map of the necropolis at Foum Larjam (yellow dots are tumuli; labelled are hilltop sites – Middle Draa Project, made by L. Rayne and M. Sterry). [325] 10.11. Adrar Zerzem, overview of the necropolis (Image © 2016 CNES/Astrium, Google Earth). [326] 10.12. Adrar Zerzem, tumulus ‘à antennes’: a) superstructure; b) inhumation. [327] 10.13. Adrar Zerzem, tumulus ‘à niche’: a) superstructure; b) inhumation. [327] 10.14. Grave goods from tombs at Adrar Zerzem. [328] 10.15. Tumuli with funerary chapels in form of a cross from North Africa. [329] 10.16. Painted panel from Jorf Torba (after Camps 1986, re-drawn by M. Hawkes). [330] 10.17. Jabal Bouïa, pre-Islamic necropolis and fortified settlement (after Margat and Camus 1959, 349, re-drawn by M. Hawkes). [335] 11.1. Map of Western Sahara showing places mentioned in the text. [342] 11.2. Map of the TF1 Study Area showing monument distribution. [343] 11.3. Tumulus, WS023 before excavation. [349] 11.4. Schematic illustration of a platform tumulus. [349] 11.5. Low tumulus bounded by kerb stones. [350] 11.6. An example of a bazina. [350] 11.7. A large stepped bazina in the Southern Sector of the Free Zone. [351] 11.8. An example of a corbeille. [351] 11.9. An example of a V-shaped monument. [352]

Published online by Cambridge University Press

List of Figures

11.10. An example of a goulet. [353] 11.11. Distribution map of goulets (Image Landsat/Copernicus © Google Earth). [354] 11.12. An example of a complex monument. [355] 11.13. An example of a ridge monument. [356] 11.14. WS001 standing stone monument with quartz scatter. [357] 11.15. Single standing stone. [358] 11.16. An example of the alignment of the corridor of a goulet referencing a flat-topped ‘mesa’ on the horizon. [363] 11.17. Viewshed from the standing stone monument WS001 showing site lines to surrounding monuments. Note that a significant number of monuments can be seen from WS001. [365] 12.1. The Eastern Niger Bend and the location of Kissi. [376] 12.2. Topography of the archaeological site of Kissi. [377] 12.3. A typical mound at Kissi, and another one near the car in the background (Photo by P. Breunig). [377] 12.4. Radiocarbon dates from Kissi, settlement areas and burials. [380] 12.5. a) Stone stele cemetery Kissi 4, used by a nearby-living Tuareg group; b) trench in Kissi 13, with two stone stele still standing. [381] 12.6. A horizontal jar burial in the north profile wall of Kissi 13, Grave 7, oriented west-east, with the mouth of the vessel towards the west. [382] 12.7. Examples of west-east oriented inhumations at Kissi. [383] 12.8. Example of a west-east oriented skeleton in prone position: Kissi 14C, Grave 17. [384] 12.9. Example of a burial with two individuals: Kissi 14C, Grave 14. [385] 12.10. Skull burials at Kissi. [385] 12.11. Examples for the position of anklets in graves. [387] 12.12. Examples of beads position in graves. [388] 12.13. The position of cowries in graves. [388] 12.14. Curved arm-dagger on left upper-arm. [389] 12.15. Examples of looped daggers. [389] 12.16. Textiles on an iron ring. [390]

Published online by Cambridge University Press

xiii

xiv

List of Figures

12.17. Bead belts or ‘cache-sèxe’: a) drawing of how Kurumba women wore them (after Schweeger-Hefel and Staude 1972); b) Kissi 14C, Grave 23; c) Kissi 14C, Grave 2. [392] 12.18. Armlets with hooks and loops: Kissi 3, Grave 10. [393] 12.19. Musical instruments (not in scale): a) Kissi 3, Grave 11; b) Kissi 3, Grave 10 (after Schweeger-Hefel and Staude 1972, 443). [393] 12.20. Looped daggers. [394] 13.1. Culture areas and sites south of Lake Chad. [402] 13.2. Sub-adult burial from the Doulo Kwovré site (PMW 631). [410] 13.3. Horse burial from Aissa Dugjé (PMW 642). [412] 13.4. Pot burial exposed in stream bank at the Thuliva Kwatcha site (PMW 768). [413] 14.1. Map of the Sahara showing the overall current linguistic situation. [434] 14.2. Map of the Sahara showing distribution of Berber languages. [435] 14.3. Saharan languages subgroup. [440] 14.4. Map with distribution of Songhay languages. [441] 14.5. Internal structure of the Songhay languages. [443] 14.6. Bilingual Latin/Berber inscription from Kef beni Fredj, Algeria (after Gsell 1933). [445] 14.7. Garamantian inscriptions in Berber characters from the Wadi al-Ajal (after Mattingly 2010, Fig. 6.8). [447] 14.8. Punic inscription from Carthage (Carthage Museum). [448] 14.9. Residual populations among modern Niger-Congo. [450] 15.1. Matrix showing relationship between Central Amazigh languages. [466] 15.2. Tree showing putative branches of Central Amazigh. [467] 15.3. Matrix for Shilh and western languages of Central Amazigh. [470] 15.4. Matrix relating to links between Matmata and Central Amazigh languages. [471] 15.5. Tree illustrating relationship of Matmata with Central Amazigh languages. [471] 15.6. Matrix tracing relationship between Wide-Central Amazigh and various Saharan groups. [473] 15.7. Full Berber matrix of cognate counts. [476]

Published online by Cambridge University Press

List of Figures

15.8. 15.9. 16.1. 16.2. 16.3.

The overall tree of East-West Amazigh relationships. [478] The overall classification of Berber languages. [481] Modern distribution of Amazigh languages. [497] The diffusion of Zenaga, c.2000–1500 BC. [499] The diffusion of Proto-Amazigh, c.1000 BC, into Fazzan and beyond. [504] 16.4. The diffusion of Proto-East-West Amazigh, c.500 BC, into the territories said to be occupied by the Gaetuli. [509] 16.5. The diffusion of Proto Central Amazigh, c.AD 400, from the Saharan oases into the Aurès Mountains and beyond. [513]

Published online by Cambridge University Press

xv

Tables

1.1. Models of different forms of mobility and migration in a Saharan context. [page 20] 2.1. Dating periods for the Garamantes with the division between Classic and Late Garamantian modified to reflect a hiatus in the imports of ceramic goods from the Roman world in the third century. [54] 2.2. Summary of tombs excavated by DMP and Daniels. [70] 2.3. Dating evidence for different tomb types. [72] 2.4. AMS dates obtained from funerary structures excavated as part of DMP. [73] 2.5. Demographic profile of DMP burials. [80] 2.6. The relationship between demographic profile and tomb type, including burials with more than one individual interred together. [80] 2.7. Traces of textile, leather, matting and ochre in shaft burials at UAT008. [84] 2.8. Occurrence of main grave goods and inclusions from excavated cemeteries/burials in Wadi al-Ajal. [88] 4.1. Summary of osteology and 3-DGM data collection. [144] 5.1. Fewet sample: list of the burials and main anthropological and archaeological data. [165] 7.1. δ18O values for human tooth samples in the Nile Valley. [234] 7.2. Average δ18O values and ranges for human samples in the Nile Valley. [235] 9.1. Summary of data from the nine tumuli with diameter/length of between 30 and 100 m, region of Larache, Meknes and Ouezzane. [291] 11.1. List of monument types and their frequency. [346] 11.2. Monuments in the Northern and Southern Sectors with quartz scatters. [360] 14.1. Arabic dialects of the Sahara. [433] xvi

Published online by Cambridge University Press

List of Tables

14.2. 14.3. 14.4. 14.5. 14.6. 15.1.

Proto-Songhay terms associated with urbanism. [443] Punic borrowings into Proto-Berber. [449] Comparative mammal lexicon in South Berber. [453] A Nilo-Saharan root for ‘crocodile’. [454] A Saharan root for ‘hippo’. [454] Proto-Berber root words showing they raised cattle, sheep and goats, and milked their animals. [484] 15.2. Proto-Berber root words for cooked grain, bread and bread dough. [486] 15.3. Proto-Chadic root words for male domestic stock. [491] 15.4. Proto-Chadic speakers tended a variety of tropical African food plants. [492]

Published online by Cambridge University Press

xvii

Contributors

Roger Blench is Visiting Fellow of the McDonald Institute for Archaeological Research, University of Cambridge. Youssef Bokbot is Professor at the National Institute for Archaeological Sciences and Heritage (INSAP), Morocco. Gabriel J. Bowen is Professor in the Geology and Geophysics Department, University of Utah. Nick Brooks is Visiting Fellow of the Tyndall Centre, University of East Anglia. Michele R. Buzon is Professor in the Department of Anthropology, Purdue University. Jordi Campillo is Researcher in the Department of Prehistory, Ancient History and Archaeology, University of Barcelona. Francesca Castorina is Associate Professor in the Department of Earth Sciences, Sapienza University of Rome. Joanne Clarke is Senior Lecturer in the School of Art, Media and American Studies, University of East Anglia. Irene Cruz Folch is Researcher in the Department of Prehistory, Ancient History and Archaeology, University of Barcelona. Fabio Di Vincenzo is Research Associate in the Department of Environmental Biology, Sapienza University of Rome. David N. Edwards is Lecturer in the School of Archaeology and Ancient History, University of Leicester. Christopher Ehret is Professor in the Department of History, University of California Los Angeles. Elizabeth Fentress is an independent researcher based in Rome. Maria Carmela Gatto is Honorary Visiting Fellow in the School of Archaeology and Ancient History, University of Leicester. Scott MacEachern is Professor in the Division of Social Sciences, Duke Kunshan University. Sonja Magnavita is Lecturer at the Institute for Archaeological Research, Ruhr-University in Bochum. Giorgio Manzi is Professor in the Department of Environmental Biology, Sapienza University of Rome. David J. Mattingly is Professor and Head of the School of Archaeology and Ancient History, University of Leicester. xviii

Published online by Cambridge University Press

List of Contributors

Marta Mirazón Lahr is Professor in the Department of Archaeology and Anthropology, University of Cambridge. David Montanero is Researcher in the Department of Prehistory, Ancient History and Archaeology, University of Barcelona. Lucia Mori is Assistant Professor in the Department of Classics, Sapienza University of Rome. Efthymia Nikita is Lecturer at the Science and Technology in Archaeology Research Center, The Cyprus Institute. Tamsin C. O’Connell is Senior Lecturer in Archaeology at the University Cambridge. Emanuele Papi is Professor at University of Siena and Director of the Italian Archaeological School in Athens. Ronika K. Power is Lecturer at the Department of Ancient History, Macquarie University. Nick Ray is Assistant Director of the Roman Economy Project, University of Oxford. Francesca Ricci is Research Associate in the Department of Environmental Biology, Sapienza University of Rome. Joan Sanmartí is Professor in the Department of Prehistory, Ancient History and Archaeology, University of Barcelona. Sarah A. Schrader is Assistant Professor in the Faculty of Archaeology, Leiden University. Martin Sterry is Teaching Fellow in the Department of Archaeology, Durham University. Mary Anne Tafuri is Researcher in the Department of Environmental Biology, Sapienza University of Rome. Andrea Zerboni is Researcher in the Department of Earth Sciences, University of Milan.

Published online by Cambridge University Press

xix

Published online by Cambridge University Press

Preface

When I was working on my PhD thesis on the Roman province of Tripolitania (north-west Libya) in the early 1980s, I became intrigued by a desert people who inhabited Fazzan, the area of the Central Sahara to the south of Tripolitania. This was my first introduction to the Garamantes. They were regularly mentioned in the ancient Greek and Roman sources, though seldom in complimentary terms – for the most part being depicted as nomadic and uncouth barbarians.1 However, some pioneering archaeological work in the 1930s and then again in the 1960s–1970s had revealed their physical traces to be considerably more sophisticated than would be assumed on the basis of the literary stereotypes.2 This volume arises out of my subsequent direct engagement across twenty years now with the archaeology of Fazzan. In 1996, I was given the chance to renew field research in what were effectively the Garamantian heartlands. Following an initial scoping visit that year, I directed the Fazzan Project across six years, carrying out excavations and survey around the capital of the Garamantes at Garama (Old Jarma), with an emphasis on tracing evidence for their settlements, but also mapping other archaeological features including cemeteries and irrigation systems.3 A notable result of this work was the clear demonstration of the sophisticated and substantial network of oasis farming settlements that lay at the heart of the Garamantian territory. Rather than being ‘nomadic barbarians’, the Garamantes now appear to have been predominantly sedentary oasis farmers, living in substantial permanent and complex settlements of mud-brick buildings. That is not to say that the Garamantes did not also incorporate pastoral elements, as will be further discussed in Chapter 1, but simply to highlight the unexpected density and sophistication of sedentary oasis settlements. 1

2 3

See in particular, Mattingly 2003, 79–81; 2011, 34–37 on the concept of ‘progressive barbarisation’ imposed by ancient authors as a factor of distance from the Mediterranean. Ayoub 1967; Daniels 1968; 1970; 1971; 1989; Fontana 1995; Pace et al. 1951. There were five seasons of fieldwork (1997–2001) and a finds study season (2002). The results are now fully published as Mattingly 2003; 2007; 2010; 2013a. Funding for the Fazzan Project came primarily from the Society for Libyan Studies, the Leverhulme Trust, the British Academy and the Arts and Humanities Research Council.

https://doi.org/10.1017/9781108634311.019 Published online by Cambridge University Press

xxi

xxii

Preface

There is strong evidence to identify the top level of their settlement hierarchy as ‘urban’ in character and their overall society as an early Saharan state.4 My work on the Garamantes has subsequently evolved through a series of further projects. Between 2007 and 2011, I directed the Desert Migrations Project, with a particular focus on Garamantian burials and funerary traditions.5 The increasing availability of high resolution satellite imagery opened a new avenue of research in 2011, the Peopling the Desert project, which extended research on the Garamantes to another of the major oasis bands in Fazzan, the Murzuq depression.6 The TransSAHARA Project (2011–2017) marks a further evolution of this body of work, seeking to place the Garamantes in their Saharan context and to address the wider implications of the results obtained in the earlier work.7 One of the major obstacles hindering understanding of the Sahara through history is that the study of the desert and the neighbouring zones of North Africa, the Nile Valley, Sudan and West Africa has tended to be compartmentalised into chronologically or regionally specific investigations. Broader synthesis across the vast Trans-Saharan zone is lacking. The term ‘Trans-Sahara’ should be understood in the context of this book as referring to the connected spaces of the Sahara and its eastern, northern and southern peripheries. The Sahara has often been likened to a great sea and no sea can be understood without reference to its adjacent shore-lands. The idea of Trans-Saharan perspectives on historical developments thus shares much in common with recent studies of the Mediterranean, which have stressed the importance of connectivity and supra-regional influences.8 The work of the Trans-SAHARA project was organised around a series of four work groups, each one supported by early career post-doctoral research associates and each dealing with a discrete group of themes: trade; migration, burial practice and identity; mobile technologies; urbanisation and state formation. As a key element of the work programme, a workshop 4 5

6

7 8

Mattingly 2013a, 530–34; Mattingly and Sterry 2013. Five planned seasons of fieldwork were completed by 2011, but the scheduled study season could not take place in 2012 because of the Libyan civil war. Interim reports have been published in Libyan Studies from 2007 to 2011, Mattingly et al. 2007; 2008; 2009; 2010a; 2010b; 2011. Funding for the Desert Migrations Project came primarily from the Society for Libyan Studies. Sterry and Mattingly 2011; 2013; Sterry et al. 2012. The Peopling the Desert Project was funded by the Leverhulme Trust. The Trans-SAHARA project was funded by the European Research Council (grant no. 269418). Abulafia 2011; Broodbank 2013; Horden and Purcell 2000. See Lichtenberger 2016 for the explicit comparison of Mediterranean and Sahara.

https://doi.org/10.1017/9781108634311.019 Published online by Cambridge University Press

Preface

was held at Leicester for each of the work groups, to which international scholars working on neighbouring areas of the Trans-Saharan zone were invited. From the outset, these workshops were conceived as offering a chance to engage a group of leading experts in the field in a high-level debate about the implications of the new information on the Garamantes for studies of the wider Trans-Saharan world. Papers were commissioned for an intended series of agenda-setting volumes on Trans-Saharan archaeology and pre-circulated so that the workshop sessions focused entirely on discussion of their content.9 This volume, the second in a projected series of four, thus presents some of the key work of the Trans-SAHARA team and an international pool of collaborators on the themes of burials and human migration, along with the implications of these topics for representations of identity in Saharan societies. The recent systematic work on Garamantian burial practices has demonstrated considerable variation and social diversity and provides a central case study on which much discussion is built.10 However, it is the wider comparisons and contrasts with other areas of the Trans-Saharan zone that serve to contextualise these data. It must be emphasised at the outset that the archaeological datasets from this region are meagre compared to Europe, for example, and a certain caution needs to be advised about the generation of detailed statistical analyses. The analyses presented here are thus more qualitative than quantitative, but hopefully provide pointers to the future direction of study of funerary archaeology. As we are asking our readers to often step outside their core areas of knowledge and expertise to engage with material from other parts of the Trans-Saharan zone, place names and their mapping have exercised us all. Systems of transliteration and spelling of place names across the TransSaharan region vary enormously and the same site can be presented in several distinct ways. We have tried to impose a measure of consistency in the transliteration of names, following the practice I adopted for the Archaeology of Fazzan series. However, for ease of recognition some 9

10

Special thanks are due to the Trans-SAHARA team: the research associates who helped organise the conference, Dr Nick Ray and Dr Martin Sterry, Dr Maria Gatto who has led the editorial team, Dr Aurélie Cuénod who translated the original French version of Chapter 10 and Dr Victoria Leitch who has been both a member of the team and the publications manager for the Society for Libyan Studies. Many maps were reworked to a consistent style by Martin Sterry and numerous other figures redrawn by Dr Mike Hawkes. We are also grateful to two anonymous Cambridge University Press referees whose comments helped us strengthen the structure and argument of the volume. Michael Sharp at Cambridge University Press gave us particularly helpful guidance. Mattingly et al. 2007; 2008; 2009; 2010a; 2010b; 2011.

https://doi.org/10.1017/9781108634311.019 Published online by Cambridge University Press

xxiii

xxiv

Preface

exceptions have been allowed for sites whose canonical spelling is so well established in the literature. We trust that the maps provided will prove helpful with the identification of places named in the text, but hope that readers will share our sense of being on a journey of discovery as they read the following contributions. David J. Mattingly

References Abulafia, D. 2011. The Great Sea: A Human History of the Mediterranean. London: Allen Lane. Ayoub, M. S. 1967. Excavations in Germa between 1962 to 1966. Tripoli: Ministry of Education. Broodbank, C. 2013. The Making of the Middle Sea. A History of the Mediterranean from the Beginning to the Emergence of the Classical World. London: Thames and Hudson. Daniels, C. M. 1968. Garamantian excavations: Zinchecra 1965–67. Libya Antiqua 5: 113–94. Daniels, C. M. 1970. The Garamantes of Southern Libya. London: Oleander. Daniels, C. M. 1971. The Garamantes of Fezzan. In F. F. Gadallah (ed.), Libya in History, Benghazi: University of Libya, 261–87. Daniels, C. M. 1989. Excavation and fieldwork amongst the Garamantes. Libyan Studies 20: 45–61. Fontana, S. 1995. I manufatti romani nei corredi funerari del Fezzan. Testimonianza dei commerci e della cultura dei Garamanti (I–III sec. d.C.). In P. Trousset (ed.), Productions et exportations africaines. Actualités archéologiques, Paris: Editions du CTHS, 405–20. Horden, P. and Purcell, N. 2000. The Corrupting Sea: A Study of Mediterranean History. Oxford: Blackwell. Lichtenberger, A. 2016. ‘Sea without water’ – conceptualising the Sahara and the Mediterranean. In M. Dabag, D. Haller, N. Jaspert and A. Lichtenberger (eds), New Horizons. Mediterranean Research in the 21st Century, Paderborn: Ferdinand Schoningh, 267–83. Mattingly, D. J. (ed.). 2003. The Archaeology of Fazzan. Volume 1, Synthesis. London: Society for Libyan Studies, Department of Antiquities. Mattingly, D. J. (ed.). 2007. The Archaeology of Fazzan. Volume 2, Site Gazetteer, Pottery and Other Survey Finds. London: Society for Libyan Studies, Department of Antiquities. Mattingly, D. J. (ed.). 2010. The Archaeology of Fazzan. Volume 3, Excavations Carried Out by C. M. Daniels. London: Society for Libyan Studies, Department of Antiquities.

https://doi.org/10.1017/9781108634311.019 Published online by Cambridge University Press

Preface

Mattingly, D. J. 2011. Imperialism, Power and Identity: Experiencing the Roman Empire. Princeton: Princeton University Press. Mattingly, D. J. (ed.). 2013a. The Archaeology of Fazzan. Volume 4, Survey and Excavations at Old Jarma (Ancient Garama) Carried Out by C. M. Daniels (1962–69) and the Fazzan Project (1997–2001). London: Society for Libyan Studies, Department of Antiquities. Mattingly, D. J. 2013b. To south and north: Saharan trade in antiquity. In H. Eckardt and S. Rippon (eds), Living and Working in the Roman World, Portsmouth (RI): JRA Suppl, 169–90. Mattingly, D. J. and Sterry, M. 2013. The first towns in the Central Sahara. Antiquity 87.366: 503–18. Mattingly, D. J., Lahr, M., Armitage, S., Barton, H., Dore, J., Drake, N., Foley, R., Merlo, S., Salem, M., Stock, J. and White, K. 2007. Desert Migrations: people, environment and culture in the Libyan Sahara. Libyan Studies 38: 115–56. Mattingly, D. J., Dore, J. and Lahr, M. M., with contributions by Ahmed, M., Cole, F., Crisp, J., Rodriguez Gonzales, M., Hobson, M., Ismayer, M., Leitch, V., Moussa, F., Nikita, E., Reeds, I., Savage, T. and Sterry, M. 2008. DMP II: 2008 fieldwork on burials and identity in the Wadi al-Ajal. Libyan Studies 39: 223–62. Mattingly, D. J., Lahr, M. and Wilson, A. 2009. DMP V: investigations in 2009 of cemeteries and related sites on the west side of the Taqallit promontory. Libyan Studies 40: 95–131. Mattingly, D. J., Abduli, H., Aburgheba, H., Ahmed, M., Ali Ahmed Esmaia, M., Baker, S., Cole, F., Fenwick, C., Gonzalez Rodriguez, M., Hobson, M., Khalaf, N., Lahr, M., Leitch, V., Moussa, F., Nikita, E., Parker, D., Radini, A., Ray, N., Savage, T., Sterry, M. and Schörle, K. 2010a. DMP IX: Summary report on the fourth season of excavations of the Burials and Identity team. Libyan Studies 41: 89–104. Mattingly, D. J., Al-Aghab, S., Ahmed, M., Moussa, F., Sterry, M. and Wilson, A. 2010b. DMP X: Survey and landscape conservation issues around the Taqallit headland. Libyan Studies 41: 105–32. Mattingly, D. J., Abduli, H., Ahmed, M., Cole, F., Fenwick, C., Gonzalez Rodriguez, M., Fothergill, B. T., Hobson, M., Khalaf, N., Lahr, M., Moussa, F., Nikita, E., Nikolaus, J., Radini, A., Ray, N., Savage, T., Sterry, M. and Wilson, A. 2011. DMP XII: Excavations and survey of the so-called Garamantian Royal Cemetery (GSC030–031). Libyan Studies 42: 89–102. Pace, B., Sergi, S. and Caputo, G. 1951. Scavi sahariani. Monumenti Antichi 41: 150–549. Sterry, M. and Mattingly, D. J. 2011. DMP XIII: reconnaissance survey of archaeological sites in the Murzuq area. Libyan Studies 42: 103–16.

https://doi.org/10.1017/9781108634311.019 Published online by Cambridge University Press

xxv

xxvi

Preface

Sterry, M., and Mattingly, D. J. 2013. Desert Migrations Project XVII: further AMS dates for historic settlements from Fazzan, South-West Libya. Libyan Studies 44: 127–40. Sterry, M., Mattingly, D. J. and Higham, T. 2012. Desert Migrations Project XVI: radiocarbon dates from the Murzuq region, Southern Libya. Libyan Studies 43: 137–47.

https://doi.org/10.1017/9781108634311.019 Published online by Cambridge University Press

1

Burials, Migration and Identity The View from the Sahara david j. mattingly, maria carmela gatto, martin sterry and nick ray

Introduction This book is the second volume of four proposed thematic books on aspects of the archaeology and history of what we term the TransSaharan zone – broadly conceived of as the vast spaces of Maghrib, Sahara and Sub-Saharan Sahel between the Atlantic in the west, the Mediterranean in the north, the Nile in the east and the equatorial African forests in the south. The territorial expanse of this zone is huge and, given the hostile climate and environment of the Sahara across the last 5,000 years, it is perhaps unsurprising that scholarly research has become regionally segmented. A good starting point for this volume is to consider to what extent the idea of a Trans-Saharan region makes sense? The chapters touch on places as far flung as the Western Sahara, the Tunisian Steppe, the Upper Nile and Lake Chad, an area of c.12,000,000 km2 within which there are significant environmental challenges to movement. This is an area so vast and, in many places, so empty of significant human habitation that many scholars have considered it to have been impassable prior to the medieval period.1 Because of the physical and environmental separation, past study has been regionally fragmented and compartmentalised. Archaeologists have most commonly self-identified with the Classical or Medieval Maghrib, with the Nilotic civilisations, or with the precocious polities of West Africa. Saharan proto-historical and historical archaeologists have been fewer in number, vastly outnumbered by prehistorians (and especially the devotees of rock art). In consequence, when considered at all, the proto-historical and historical Sahara has often been viewed from the outside, looking in. The Trans-Saharan Archaeology series seeks to explore the interconnections across this zone in new ways, bringing together archaeologists, 1

Austen 2010; Lydon 2009; Fauvelle 2016 for the latest attempts to deny or minimise the existence of Trans-Saharan trade before the Islamic era. Compare with Mattingly et al. 2017a.

https://doi.org/10.1017/9781108634311.001 Published online by Cambridge University Press

1

2

David J. Mattingly et al.

anthropologists and historians from different regions, varied academic traditions and multiple time periods and cultural phases. A novel aspect of this book is that we seek to place the Sahara more centrally in the discussion and to look out from the Sahara towards its neighbours in search of parallels and contrasts. We also see the Sahara as somewhat similar to a great sea like the Mediterranean and have been influenced in this enterprise by recent scholarly debates about the Mediterranean and the attempts to construct pan-Mediterranean histories and archaeological overviews.2 A few words are necessary at the outset concerning changes to the climate and environment of the Sahara in the past.3 At various times in prehistory, the Sahara has oscillated between wet and arid phases. The concept of a ‘green Sahara’ is now well appreciated in relation to the pluvial phases, which created substantial river systems and vast lakes.4 The last significant wet phase was in the Early-Mid Holocene period, broadly 10,000–3500 BC. During this period, the wide availability of water in the form of seasonal rivers, small lakes and a high water table supported Saharan connectivity and mobility.5 As a general trend, mobile human communities of hunter-gatherers adapted to herding of domesticated animals – primarily cattle.6 Although there is evidence for periodic climatic oscillations already within the Early-Mid Holocene phase, with a major abrupt arid spell recorded at around 6200 BC, it is apparent that with the Late Holocene, at c.3500 BC, there was a significant step in climatic change, which marked the start of the modern hyper-arid phase in the Sahara. Minor climatic oscillations are still recorded in some parts of the Sahara, such as certain areas of the mountain massifs, which receive somewhat higher rainfall than the region as a whole. However, the human experience of and interaction with the Sahara over the last 5,000 years has concerned a harsh desert environment that imposes limitations on settlement, movement and lifestyles. That is not to say, of course, that the desert denies long-range movements and contacts, but that these have necessarily become more focused along axes where water is more readily available in 2 3

4

5

Abulafia 2011; Broodbank 2013; Harris 2005; Horden and Purcell 2000; Lichtenberger 2016. For some of the most recent syntheses on the subject see: Brooks et al. 2005; Cremaschi and Zerboni 2011; Gatto and Zerboni 2015; Kuper and Kröpelin 2006; Mattingly 2003, 37–74, 327–46 with reviews of earlier literature. Barker et al. 1996a, 291–302; Cremaschi 2001; Cremaschi and di Lernia 1998; deMenocal and Tierney 2012; Larrassoaña et al. 2013; Mattingly 2003, 37–74, 327–46 for detailed discussions of climate change in the Sahara and fuller references; see also http://www.greensahara-leverhulme .com/. Drake et al. 2011; Manning and Timpson 2014. 6 di Lernia 2013.

https://doi.org/10.1017/9781108634311.001 Published online by Cambridge University Press

1 Burials, Migration and Identity: The View from the Sahara

the form of springs and a high water table. There has been progressive decline in water availability in the Sahara as non-renewable subsurface water sources have been diminished by natural and anthropogenic action, and this has had implications for both Saharan populations and the ease of movement.7 The domestication of key pack animals like donkeys, horses, mules and camels has been another crucial factor in facilitating the navigation of the arid spaces of the Sahara.8 All the beasts of burden mentioned above were present in the Sahara by the later first millennium BC, though the importance of the camel increased over time with the progressive drop in water tables increasing the distance between and the delivery capacity of wells on Saharan trails.9 It is precisely because of such constraints on movement and on habitation that the Sahara is such an interesting theatre in which to explore themes related to human connectivity across space. There are three interlinked themes in this book. Indeed, the key objective of this volume is to explore the interrelationships between human burials, migration and identity in the Trans-Saharan zone in the late prehistoric and historic periods (broadly covering the last 4,000 years, but with a core focus on the first millennia BC/AD). Burial, migration and identity are not equal categories and the connections between them are variedly reflected in the chapters that follow. Burials are a fundamental human behaviour and are commonly distinctive and reflective of profound ideological beliefs.10 As well as providing information on the body, funerary archaeology involves the study of a series of processes and material attributes – from the basic mode of treatment of the body, to the typology of formal structure of monuments and graves in which remains were placed, to the evidence of attendant funerary rites and grave furniture, to the grave goods included in the burial or offerings left alongside it, to integrated approaches to human demographics, health and diet.11 Migration, as a form of mobility, is an important factor in human history, sometimes documented and easily traceable, and more often in earlier historical periods it has to be inferred from a variety of material and biological markers.12 Burials are thus potentially a good barometer for 7 9 11

12

8 Cremaschi and Zerboni 2009; Drake et al. 2004. Lichtenberger 2016, 269. 10 Mattingly et al. forthcoming. Tarlow and Nielsson Stutz 2013, for a broad overview. Scientific techniques have greatly added to the discussion in recent years through DNA and isotope analyses, material composition and provenance studies, as well as entomology, see Gowland and Knusel 2009; Tarlow and Nielsson Stutz 2013. For good historiographical overviews on the debate about migration, see Burmeister 2000; Hackenbeck 2008.

https://doi.org/10.1017/9781108634311.001 Published online by Cambridge University Press

3

4

David J. Mattingly et al.

detecting societal change, identity affiliation and mobility among populations.13 Identity is another significant theme in modern archaeological study and frequently takes funerary archaeology as a key dataset. In modern societies, ethnicity, religion and nationality are the predominant identity types. However, in antiquity multiple identity markers were more common, linked to varying groups (and ideas of groupness) and often socially contingent on situation, audience, status, gender or age.14 In the rest of this introductory chapter, we shall offer some further commentary on these key themes and on current theoretical underpinnings of their study in archaeological work. In discussing burials, we shall review the historiography of burial archaeology in the Maghrib, the Sahara and sub-Sahara in general and provide some essential background to the Central Saharan region of Fazzan in particular. We conclude the chapter with a commentary on the structure of the book and some indications concerning how the chapters relate to the three key themes.

Approaches to Funerary Archaeology There is a long tradition in archaeology for exploring the social dimension of burial rites, with approaches becoming increasingly sophisticated over recent decades with the accumulation of large datasets and the development of new theoretical frameworks.15 As we shall see, when we turn to the Saharan and related material below, the African datasets are much smaller, especially in regards to well-excavated and analysed burials, and their study has been less influenced by the wider evolution of theoretical thinking. Environmental constraints and past colonialism, with its ‘eurocentric’ perspective, have certainly had a strong impact on their study. The mundane and inevitable event of death has, in human societies, become linked to complex processes of commemoration. Funerary evidence is often read as a sort of transcript of social complexity. Several separate issues need to be recognised here. Communities have often developed normative rules for the treatment of the corpse immediately after 13 14 15

Cool 2004 is a good attempt to merge these themes in a report on a single cemetery in Britain. For useful introductory overviews see Diaz Andreu and Lucy 2005; Insoll 2007. Barthel 1982; Chapman 1981; Duday 2009; Meskell 1999; Parker-Pearson 1999; Saxe 1970; Silverman and Small 2002; Tarlow and Nilsson Stutz 2013. See, for example, Pearce 2013; Pearce et al. 2000; Philpott 1991, for studies with a focus on Roman Britain, and which can serve as exemplars of what can be done with large datasets of well-excavated burials.

https://doi.org/10.1017/9781108634311.001 Published online by Cambridge University Press

1 Burials, Migration and Identity: The View from the Sahara

death (washing, laying out, dressing, etc.), with specific choices also regarding the mode of disposal (inhumation, cremation and excarnation) and, where buried in the ground, as regards positioning of the body or ashes within a grave cut. The selection of material to accompany the body can also be revealing about ideas of an afterlife or social status; however, this is not just a reflection of the dead but also of the living communities associated with the burials, as they are the ones who select and place the objects into the grave.16 Sacrifices may be made and ritual meals consumed at the graveside, indicating the longer-term experiential dimension of these features for the ancestors. In some societies memorialisation includes visual imagery added to the tomb or formal inscribed grave markers, which can include significant biographical detail relating to the deceased. Increasingly, archaeologists have recognised the importance of contextual approaches to ancient burial practice and the need to carefully consider the theoretical dimension when examining social implications.17 Much of this evidence relates directly to the ceremonies of disposal of the body, but additional aspects bear on the memorialisation of the individual after death.18 The creation and enhancement of social memory through funerary rituals and monuments appear to be a key motivation behind mortuary practices in many societies.19 The form of the burial structure20 and the incorporation of inscriptions or relief carvings on the monument represent long-term visible reminders of the individual.21 Indeed, the careful placement of monuments within the landscape was a common strategy to establish symbolic importance for tombs of ancestors.22 In many cultures, the grave or tomb was provided with funerary furniture, such as stelae and offering tables, indicating the maintenance of specific funerary rituals after the closure of the grave. In many societies, such memorialisation has created a significant role for ancestors in the lives of succeeding generations.23 As such, it is also important to consider the funerary landscape beyond the tombs as individual phenomena, not just in terms of single-period aggregations 16 17

18 20 22

23

See Tarlow and Nielsson Stutz 2013, in particular the contribution by Ekerngren. Ucko 1969 is a seminal work for challenging assumptions in the interpretation of funerary remains and encouraging a more theoretical approach. See, for example, Williams 2003; 2006. 19 Chesson 2001; Daróczi 2012; O’Shea 1984. Dillehay 1995; Hope 2001. 21 Carroll 2006; Carroll and Rempel 2011; Hope 1997. Arnold 2002; Barrett 1990; Pearce 2011. On funerary landscapes in general, see inter alia, Ucko and Layton 1999. An interesting aspect of the work of Williams (2003; 2006) is the way he has demonstrated that the social meaning of funerary monuments in the landscape could change over time, with reworking and reuse of such structures in later phases. Buikstra 1995; Graham 2009; Parker-Pearson 1999; Parker-Pearson and Ramilisonia 1998; Whitley 2002.

https://doi.org/10.1017/9781108634311.001 Published online by Cambridge University Press

5

6

David J. Mattingly et al.

but as complex formations that evolved over centuries or millennia and are reflective of changing (or, indeed, constant) burial customs.24 Since Saxe’s important thesis that attempted to produce cross-cultural models of funerary behaviours and mortuary ritual, there has been considerable discussion about the degree to which funerary identities correspond with the identity of the deceased when alive.25 There is certainly evidence that in many societies lower order groups attempt to emulate the funerary behaviours of more elite groups as a means of bolstering their social standing.26 At the very least, this suggests that we should be cautious about assuming that there were neat relationships (and that are easily mapped) between funerary material culture/behaviours and social identity.

Funerary Archaeology in the Sahara and Beyond Funerary archaeology in the Sahara and its neighbouring lands has a long history. There are regions such as the Nile Valley where the study of pyramids and tombs of different types have been paramount for reconstructing the ancient civilisations of Egypt and Nubia.27 However, in some other regions, like the Middle Niger Delta or Mauretania, settlement archaeology or graves relating to recent periods have received more attention in archaeological research.28 These excavation imbalances between settlement and funerary archaeology are not uncommon, but they do present some challenges for attempts to synthesise data about burial practices across the Trans-Saharan region. Another challenge comes from the timing and scope of archaeological research, which developed earlier in the Nile Valley, the Maghrib and Mediterranean Africa and at its inception had a major interest in finding evidence of cultural influence from the Near

24

25

26

27 28

Stone and Stirling 2007. The work of the UNESCO Libyan Valleys Survey demonstrates the complexity of many clusters of funerary features within sites (Barker et al. 1996a; 1996b), a theme that is continued in many of the contributions in this volume. Saxe 1970, who concluded that the post-mortem identity presented in the mortuary rituals was a selective and often composite of life identities, and often dependent on choices of the living, rather than of the deceased; cf. discussion in Chesson 2001; Parker-Pearson 1999; Silverman and Small 2002. Cannon 1989; cf. Mouritsen 2011, on Roman freedmen in Italy who sometimes constructed extravagant funerary monuments imitating those of the elite, only for the highest elite groups to modify their own behaviour and make their tombs less ostentatious. See Edwards, Chapter 6, this volume. See, for instance, the research at Dhar Tichitt and Jenné-jeno: Munson 1980; Holl 1986; MacDonald et al. 2009; McIntosh and McIntosh 1980.

https://doi.org/10.1017/9781108634311.001 Published online by Cambridge University Press

1 Burials, Migration and Identity: The View from the Sahara

East, the Mediterranean and Europe. An approach that certainly had an impact on data interpretation. The earliest known burials across the Sahara, either isolated or in clusters, open air or in shelters, lack superstructures.29 From the Middle Holocene onwards, however, the Sahara was dotted by funerary stone structures, which only in Proto-historic periods clustered around settlements to create spatially discrete cemeteries. The earliest evidence of stone monuments comes from the Eastern Sahara (the deserts on both sides of the Nile), dating to the Early Holocene (c.7000 BC). They represent ritual monuments, which initially had nothing to do with human burials.30 Later, around 5500 BC, stone structures of different size and shapes were used to cover animal burials,31 a practice that developed alongside pastoralism. It was only from the fifth millennium BC that stone tumuli were used to mark human graves.32 From the Eastern Sahara, they spread westward quite rapidly, following pastoral mobility. In the Central Sahara, animal burials, particularly of cattle, are recorded from the fifth millennium BC and human burials soon after.33 Some of those structures became very complex in shape and monumental in size, and although isolated in the vast Saharan landscape, they surely could make an impression, at least on those who participated in building such labour-intensive constructions. They have been interpreted as graves of important members of mobile groups used to mark their territory.34 In the Western Sahara, funerary stone structures, as well as other kinds of ritual megaliths, arrived later in time, again following a general westward cultural movement.35 An emphasis on the presence/ absence of stone superstructure and on its typology has become a characteristic of Saharan funerary archaeology, as thus far actual excavations of tombs have been rather limited. In the sub-sections here below, an east to west trajectory is followed across the Trans-Saharan zone to provide a brief review of the more regionalised history of research (Fig. 1.1). We want to emphasise here that one of the aims of this book is to look out from the Garamantes to wider issues of comparisons and contrasts in funerary culture and behaviour in and beyond the Saharan zone, to the north, east, south and west. Placing the Sahara at the centre of such a study is novel, but we think will provide an interesting perspective on the data. 29

30 33 35

See, for instance, di Lernia and Tafuri 2013; Garcea 2013; Petit-Maire and Riser 1983; Usai et al. 2010. Bobrowski et al. 2014. 31 Wendorf and Schild 2001. 32 Gatto 2012. di Lernia and Manzi 2002. 34 di Lernia et al. 2002, 285, with bibliographical reference. See Chapter 11 by Clarke and Brooks, this volume.

https://doi.org/10.1017/9781108634311.001 Published online by Cambridge University Press

7

8

David J. Mattingly et al.

Figure 1.1. The Trans-Saharan zone, with indication of the major areas and sites covered in this volume. (made by M. Sterry)

Egypt, the Nile Valley and the Eastern Sahara From the monumental pyramids of Giza to the hidden rock-cut tombs of the Valley of the Kings, ancient Egyptian funerary architecture and practices are celebrated worldwide. Their number and variety is overwhelming and it is difficult to synthesise in few sentences more than 200 years of research on the subject.36 The Egyptian funerary traditions (including the rise of mummification) branched out from a common Nilotic background, shared with cultures from Nubia and Central Sudan,37 only with the rise of the Pharaonic civilisation. Predynastic burials mainly consisted of a shaft dug into the ground, circular to rectangular in shape, with no superstructure. However, elite tombs, such as those found in Hierakonpolis, had complex superstructures made of perishable materials like wood, and wattle and daub.38 It was only with the Early Dynastic period that mudbrick/stone superstructures became a common element of the Egyptian funerary architecture.39 From the bench-shaped mastabas developed the pyramidal structures of the third millennium BC, which disappeared in the 36 37 39

For good recent syntheses see Dodson and Ikram 2008 and Ikram 2015. Gatto 2011; Wengrow et al. 2014. 38 Friedman et al. 2011. The earliest being the royal tombs of Dynasty 0 in Abydos, dated to the end of the fourth millennium BC, Dreyer 1992.

https://doi.org/10.1017/9781108634311.001 Published online by Cambridge University Press

1 Burials, Migration and Identity: The View from the Sahara

next millennium to be replaced by hypogea. Non-royal tombs occasionally preserved small pyramids marking their surface, but they had gone out of fashion before the Greco-Roman period. The practice of mummification and of wrapped bodies can be traced from Egypt through the oases of the Western Desert to Siwa and beyond.40 The most westerly examples of the Egyptian mummification rite appear to be at al-Jiarabub on the Libyan/Egyptian border between Siwa and Awjila.41 A few examples of mummified bodies have been found in Fazzan, but these seem to be accidental outcomes of burial in hyper-arid conditions, rather than the chemically assisted and invasive processes of handling the corpse. In this book, we particularly address the Nubian funerary tradition, which in comparison to the Egyptian one retained more Saharan connections.42 As mentioned already, the concept of a stone structure covering a grave developed among the (Nubian) nomads of the Eastern Sahara. From the beginning of the historic period, in the third millennium BC, stone superstructures became characteristic of the Nubian funerary landscape also along the Nile, and in some areas continued to be so up until the arrival of Islam. The C-Group tumuli of the second millennium BC are particularly interesting for our discussion, as they consisted of drum-shaped structures with stone slabs used as stelae and outer offering areas, sometimes in the form of built ‘chapels’.43 From the second half of the second millennium, however, with the colonisation of Nubia by Egypt, Egyptian funerary tradition was adopted in northern Nubia and by the successive Napatean and Meroitic elites. Meroitic pyramids are well-known, but, although resembling Egyptian prototypes, they differ in terms of dimensions, construction techniques and materials. Nubian tumuli of various dates have also been found in Egypt, but apart from those related to the Pan-Grave culture of the late second millennium BC, they have not received much scholarly attention.44 The relationship between Saharan and Nilotic burial traditions represents an interesting line of enquiry. There are striking structural similarities between certain Nile Valley burial monuments and those of the Garamantes. Some, like the construction of mud-brick pyramids and of rectangular, stepped tombs similar to the mastaba type appear to be distant echoes in relative isolation, rather than part of a more generalised cultural

40 41 43 44

Bahariya golden mummies: Hawas 2000; Siwa: Mattingly 2000. Mattingly 2000; Mohammed 1998. 42 See Edwards, Chapter 6, this volume. Bietak 1968. See, for instance, Friedman 2001; 2004; 2007; Gatto 2005; 2013; Gatto et al. 2014; Ralston 2002.

https://doi.org/10.1017/9781108634311.001 Published online by Cambridge University Press

9

10

David J. Mattingly et al.

adoption of the burial form within the Sahara. However, stone-built tumuli in general, and the C-Group ones in particular, which are very similar to Garamantian tumuli, are evidence of a shared Saharan tradition that goes far back in time. The laying out of the corpse on its back in an extended posture is a characteristic of Egypt, the Western Desert oases and of some of the late historic Nubian graves, whereas as we shall see for much of the rest of the Sahara, starting from Nubia, the body was placed in a contracted position, laid on one side, and covered by a leather shroud of some kind.

The Central Sahara and the Garamantes The Libyan Saharan region known as Fazzan is notable for the range and complexity of its pre-Islamic funerary archaeology. Since pioneering excavations by an Italian mission in the 1930s,45 some of the most spectacular evidence has been recognised as relating to the Garamantes, a Libyan people contemporary with the Greco-Roman civilisations of the Mediterranean (Fig. 1.2a/b for summary map of the settlement and funerary evidence relating to the Garamantes).46 The burials of the Garamantes are notable for the abundance of Roman ceramics and glass found within them and an array of distinctive monument types.47 In the early phases of research the nature of the ‘Roman’ finds somewhat overshadowed Saharan aspects of their burial context. In the last two decades, there has been important renewed work by an Italian team on the Pastoral (Neolithic) peoples of the Libyan Sahara and their burial practices.48 Although, as previously said, the earliest stone monuments in the Libyan Sahara relate to cattle-based pastoral groups of the Middle Pastoral period, these were chronologically separated by up to a millennium from the first monuments with human burials that were constructed from the third millennium BC onwards.49 Many were simple stone cairns, but increasingly complex and larger forms were created so that by the end of the period there were types such as antenna tombs, crescents, keyhole tombs and cairns up to 15 m across. Some of the larger cairns were used for multiple interments. It is within these 45 46

47 48 49

Caputo 1937; 1949; 1951. See also Bellair et al. 1953; Camps 1955; Daniels 1970; 1989; el-Rashedy 1988; Fontana 1995; Mattingly and Edwards 2003. Ayoub 1967a; 1967b; 1967c; Mattingly 2010. di Lernia and Manzi 2002; di Lernia et al. 2001; di Lernia and Tafuri 2013. di Lernia and Manzi 2002.

https://doi.org/10.1017/9781108634311.001 Published online by Cambridge University Press

1 Burials, Migration and Identity: The View from the Sahara

Figure 1.2. a) Garamantian settlements in Fazzan and key regional toponyms referred to in this volume; b) Garamantian cemeteries in Fazzan with rectangles marking the location of the main case study areas of Chapters 2–5. (made by M. Sterry and M. C. Gatto)

complex forms that we find the earliest evidence of the elaboration of a special importance for the east side of the tombs and the intersection of funerary and religious rites.50 The same Italian project has also surveyed and excavated a number of burials of Garamantian date in the Wadi Tanzzuft area of south-western Fazzan, c.400 km south-west of 50

di Lernia 2006.

https://doi.org/10.1017/9781108634311.001 Published online by Cambridge University Press

11

12

David J. Mattingly et al.

the Garamantian heartlands.51 The typical ‘Garamantian’ tombs in this area were circular cairns and drum tombs, and these forms seem to have been used across a long period. The parallel work of a British-Libyan team directed by David Mattingly since 1997 has significantly increased the published dossier on the Garamantes, making them the best studied Saharan people.52 In addition to presenting new field results, the publications attempted a fuller account of earlier research carried out by Charles Daniels and Mohamed Ayoub in the 1960s and 1970s.53 The Desert Migrations Project (or DMP) from 2007–2011 included a programme of survey and excavations of Garamantian burials.54 The DMP explicitly brought together our themes of burials, identity and migration and led to the Trans-SAHARA project, which was designed to explore these issues on a broader scale. The total excavated burials from Fazzan exceeds 400, with about half that number excavated to a modern standard. A significant number of radiocarbon dates exist to confirm the broad phases of development and use of particular aspects of the funerary tradition.55 The nature and provenance of grave goods indicate wide contacts within the Trans-Saharan region. This makes the material of particular significance within the Sahara as a whole and a benchmark for study of other regions.

The Northern Sahara and the Maghrib Much of the earliest work on pre-Roman burials in the Maghrib had speculated on external origins for key features and key burial types, such as the megalithic structures (dolmens) that are so common in Tunisia and Algeria. Camps showed that attempts to correlate the megalithic structures with the main European tradition of megalith building were mistaken and instead he drew important conclusions about how the distribution of certain types of monument correlated with the heartland territories relating to the precocious state formation of the Numidian and Mauretanian 51

52 54

55

Liverani 2006 – by Aghram Nadharif; Castelli et al. 2005; Mori 2013 – at Fewet. See further Gatto et al. and Ricci et al., Chapters 3 and 5, this volume. Mattingly 2003; 2007a; 2010; 2013. 53 Notably in Mattingly 2010, for funerary sites. Mattingly et al. 2007; 2008; 2009; 2010a; 2010b; 2011. See also an important University of Leicester PhD thesis by Gonzalez Rodriguez 2014. For all radiocarbon dates discussed in this volume we have reproduced basic information where available – lab references, radiocarbon age and calibrated dates at 2σ (95.4 per cent probability). Where original uncalibrated dates were available these have been calibrated to the IntCal13 curve (Reimer et al. 2013) using OxCal 4.3 (Bronk Ramsey 2009).

https://doi.org/10.1017/9781108634311.001 Published online by Cambridge University Press

1 Burials, Migration and Identity: The View from the Sahara

kingdoms.56 The fundamental importance of this last observation for our understanding of proto-historic societies was for long neglected, though recent work is at last addressing this crucial issue.57 Something that remains striking about Camps’s typology is the range and complexity of burial types and of associated ritual structures – niches, chapels and enclosures for offerings.58 This was not some straightforward cultural koine, but a highly varied and regionally specific repertoire, that nonetheless adhered to some common or recurrent features, such as a penchant for circular burial monuments, crouched inhumation, eastern orientations, the use of red ochre for ritual and decoration, and so on. While Camps recognised the importance of considering the protohistoric tombs and funerary rituals of the Maghrib in their African context, he was hampered in taking this as far as it could have gone by the lack of well-excavated Saharan parallels and by some of the ostensible differences between Saharan and Maghribian monuments. Given the vast repertoire of monuments he recorded in the Maghrib alone, such diversity is to be expected, but the question of broader Saharan influences on North African funerary rituals clearly merits further exploration. Key issues to be investigated across the Trans-Saharan zone include the practice of multiple versus single interments, persistent patterning in the positioning of the body in the tomb or orientation of the monuments themselves,59 changes in monument type and shape (in particular, patterns of circularity as opposed to rectilinearity), the provision of funerary furniture (stelae, offering tables and so on), inclusion of grave goods (and the impact of imported goods from outside the region on practice). For many parts of the Maghrib, the work of Camps remains unsurpassed due to lack of new excavations. The few major overview studies of funerary monuments that have been attempted in the last 20 years remain rather under-developed for want of extensive new datasets.60 The groups of funerary monuments that have continued to attract the most attention comprise the highly prestigious mausolea61 and other elite burial types 56

57

58 59 60

61

Camps 1961, 29–62, 565–71. Discussions and concerns about the ‘European Perspective’ on African monumental architecture are still evident in modern scholarship: see Thomas 2013. See Sanmartí et al. and Papi, Chapters 8 and 9, this volume; Stone 2016 for a different approach, based on volumetric analysis of large earthen tumuli. Camps 1961, 181–87; 1986. Belmonte et al. 2002; compare with Mattingly and Edwards 2003, 226–27. Important contributions include, Déroche and Leclant 2010; Stone and Stirling 2007; Trousset 1995. Brogan and Smith 1985; Camps 1981, 199–205; Di Vita 1976; Hitchner 1995; Moore 2007; Rakob 1979.

https://doi.org/10.1017/9781108634311.001 Published online by Cambridge University Press

13

14

David J. Mattingly et al.

associated with or imitative of the Mediterranean civilisations of the littoral zone – Carthaginians/Phoenicians in the Maghribian countries62 and western Libya, Greeks in Cyrenaican East Libya.63 Rock-cut tombs of northern Tunisia and Algeria known as haouanet and their Punic equivalents at the Libyphoenician coastal centres have tended to dominate the literature on pre-Roman burial alongside the contentious issue of the child burials at tophet sites.64 Among a number of important excavations of Roman-era cemeteries, the work at Leptiminus is significant for the detailed publication of burial structures, human remains (including isotopic analyses), associated artefacts and cemetery layouts.65 However, insofar as such studies focus on monument types or practices where external (Mediterranean) influences were strong, there has been a tendency to prioritise that perspective over indications of African traditions and practices.66 While progress in general on more generically African burial types such as the megalithic and drum tombs of the Maghrib has been hampered by the lack of detailed survey and modern excavations, there has been important new work by a Spanish-Tunisian team led by Sanmartí and Kallala at Althiburos67 and by an Italo-Tunisian team around Mdidi (near Makhtar) in western Tunisia.68 An important aspect of the Tunisian work is that the surveys of funerary zones have shown major concentrations of these burials around suspected proto-urban settlements that are part of a broader body of evidence now emerging to support a new understanding of the formation of the first indigenous polities.69 Excavated burials have also yielded radiocarbon dates, clarifying aspects of their chronology. For the northern Sahara, in the Libyan pre-desert zone, the Libyan Valleys Survey produced a useful dossier of data on burials, though most notable for the local engagement with Roman architectural forms of 62

63 64

65

66

67 68

69

Ben Younes 1995; 1996; 1997; 2007; Ben Younes and Krandel-Ben Younes 2014; Krandel-Ben Younes 2013; Longerstay 1995; 2011; Stone 2007a; 2007b. Cherstich 2008; Thorn 2005. Haouanet: Stone 2007a; 2007b; Punic tombs: Ben Younes 1995; 2007; tophets: Benichou Safar 1995; Ferjaoui 2007. Ben Lazreg and Mattingly 1992, 177–259, 301–33; Stirling et al. 2007, 107–201; Stone et al. 2011, 120–204. For the isotopic analysis, see Keenleyside et al. 2009. Another important cemetery excavation with exemplary publication of at least part of its material is Pupput, Ben Abed-Ben Khader and Griesheimer 2004. See Mattingly 2007b, arguing for greater recognition of African traits in Roman burial practices in the Maghrib. Kallala and Sanmartí 2011; Kallala et al. 2014; 2018. Tanda et al. 2009; Sanmartí et al. 2012 and Sanmartí, Chapter 8, this volume; also Camps 1961; 1995; Camps and Camps-Fabrer 1964. Sanmartí et al. 2012.

https://doi.org/10.1017/9781108634311.001 Published online by Cambridge University Press

1 Burials, Migration and Identity: The View from the Sahara

mausolea.70 Short notes on individual examples of distinctive inner Saharan burial types regularly appear in the relevant journals, though most of these have not added greatly to knowledge or understanding.71 A number of regionally focused studies have highlighted the local characteristics of monument types, such as Brooks and others in the Western Sahara.72 Less common have been attempts to demonstrate an influence of Saharan burial traditions on the Maghrib.73

The Western Sahara The most detailed work on funerary archaeology in Western Algeria and Morocco has been by Youssef Bokbot, building on colonial era observation and sporadic excavation.74 There are several distinct zones to consider. In the first place, there is the region north of the Atlas mountains, which constituted the western sector of the territory of the Mauretanian kingdom and the Roman province of Mauretania Tingitana.75 A second band of preIslamic funerary sites is located on the southern flanks of the Atlas and Saharan Atlas ranges, running from western Algeria through to the Atlantic coast. This encompasses a number of key sites including Jorf Torba on the Wadi Gir in Algeria, several significant pre-Islamic cemeteries in the Tafilalat area (Erfoud, Bouïa, Taouz, Braber) and in the Wadi Draa.76 There is considerable range in types of funerary monuments across these cemeteries, but also some commonalities. A particular feature is the large scale of the biggest tombs, which can be 10 m and more in diameter. Most tombs were circular in form, though some square and rectangular monuments have also been recorded in the Tafilalat. The main types of monument are mound cairns, corbelled cairns and some hybrid tumuli/ drum tombs with a low vertical external wall, but a mounded upper structure. Currently, only a small number of these western pre-Saharan tombs have been excavated and even fewer of those can be well dated. There are some hints that the date of the most elaborate and largest monuments may correspond to the period from the third to seventh centuries AD. That some elements of the mortuary culture date back much earlier is indicated by the identification of a number of crescentshaped monuments of presumed Neolithic date and some human remains 70 71 72 74 75

Barker et al. 1996a; 1996b; Mattingly 1999; Nikolaus 2016. Gauthier and Gauthier 1999; Milburn 1981; 1983, 1993; 1996. See, for example, Brooks et al. 2006. 73 Fentress and Wilson 2016; Mattingly 2007b. Bokbot 1991; Chapter 10, this volume. On this area, see also Papi, Chapter 9, this volume. 76 Bokbot, Chapter 10, this volume.

https://doi.org/10.1017/9781108634311.001 Published online by Cambridge University Press

15

16

David J. Mattingly et al.

retrieved from robbed tumuli at Foum Larjam that have been dated to early in the first millennium BC.77 A particularly noteworthy feature of a number of the large tombs is the construction of elaborate ‘chapel’ structures against or built into their east side.78 These chapel structures also resemble examples found further east in the region of the Roman frontier, as at Besseriani, where they appear to be a late Roman or late antique feature.79 Painted and engraved panels of horses, other animals and people have been recovered from at least three examples of these chapel structures at the site of Jorf Torba.80 The two scenes with painted human figures are particularly celebrated and stylistically appear to have similarities with Late Roman art. However, the iconographical details and clothing appear to relate to Saharan people and one of these scenes has accompanying Libyan inscriptions. Recently, further examples of painted decoration depicting humans have been recorded near Foum Larjam in the Wadi Draa.81 The depictions of horses at Jorf Torba are closely comparable with a number of engraved rock art depictions in the Wadi Draa, including two recently discovered concentrations from hillfort settlements that again seem to date to the first half of the first millennium AD.82 The lifestyle of these people is an open question in the absence of detailed investigation of settlements of this age. However, it is notable that the largest concentrations of pre-Islamic tombs have been found alongside the main oasis locations in southern Morocco – notably the Tafilalat and Draa. The funerary archaeology of the desert zone in the Moroccan Sahara/ Mauritania and Western Sahara overlaps to some extent with the archaeology of the pre-Saharan zone just described. A number of similar monument types occur, including mound cairns, some drum tombs and even stepped drum tombs. But there are also larger numbers of the more peculiar ‘Saharan’ forms like antennae tombs (including v-monuments and crescents as variants on a theme).83 The lack of excavated monuments in this region leaves their dating overly reliant on comparanda, but the common 77

78 80

81

82

83

For radiocarbon dates, see Mattingly et al. 2017b; Delibrias et al. 1982, date Gif-2912, 2360 ± 250 BP. Camps 1986. 79 Fentress and Wilson 2016. Lihoreau 1993, 109–15 and 123–26 for the most complete descriptions of the engraved and painted stelae. See also Camps 1984; Reygasse 1950, 104, 107–08. See also front cover, inset image. Zaïnabi 2004, 40. The tomb was relocated during fieldwork of the Trans-SAHARA team in 2016 and a neighbouring tomb also shown to have a similar painted funerary chapel. Publications on these important new finds are in preparation. The most recent work in the Wadi Draa is directed by Youssef Bokbot and David Mattingly with Martin Sterry, see Mattingly et al. 2017b. Brooks et al. 2006; Clarke and Brooks 2018 and Chapter 11, this volume.

https://doi.org/10.1017/9781108634311.001 Published online by Cambridge University Press

1 Burials, Migration and Identity: The View from the Sahara

alignments (especially with an eastward orientation) of many monuments strongly supports their cultural congruity.84

The Southern Sahara Burial studies from the southern Sahara have commonly focussed upon the pastoral societies of the Neolithic.85 However, an important body of work has also been achieved on Proto-historic and historic era tombs, notably by Reygasse at several sites in the Ahaggar massif (Tit, Silet, Abalessa, Tin Hinan) and by Paris and others in the Aïr mountains.86 Arguably the single most famous excavation of a Saharan burial concerns the tomb of Tin Hinan near Aballessa in the Algerian Ahaggar, where the body of a woman was unearthed, laid out on her back on an ornate wooden bed with a range of elaborate gold and silver jewellery and items attesting to contact with the Late Roman world, including Roman glass tableware.87 Intriguingly, the site had long been revered by the local Tuareg clan as the burial site of its maternal founder, the first queen of Ahaggar, Tin Hinan.88 Among the jewellery were seven gold bracelets worn on her upper left arm and seven silver bracelets on the upper right arm, a necklace (with gold beads and decorative elements alongside, large polished stone beads) and sundry rings and other beadwork items. The body had been dressed in textile garments (poorly preserved) and covered in a leather shroud coated in ochre and a small stone bowl containing ochre was found with the burial, along with a biconical incense burner similar to the standard Late Garamantian type. Other ritual objects included a small female statuette with exaggerated sexual organs. Coin impressions of Constantine the Great (but not the actual coins) were also found.89 The site is highly atypical, as quite apart from the wealth represented by the finds, the burial was actually made within a fortified building. Finds from other rooms within the building included additional late Roman items such as a moulded lamp, long and deeply barbed iron arrowheads 84 85 86 87

88

89

Gauthier 2015; Gauthier and Gauthier 2007. See, for example Sudan: Salvatori and Usai 2008, Niger: Garcea 2013; Sereno et al. 2008. Paris 1984; 1990; 1995; 1996; Reygasse 1950, 32–117. Brett and Fentress 1997, 208; Camps 1997; de Prorok 2001, 347–67; Hachid 2006; Reygasse 1950, 88–106. The first excavations in 1926 were carried out by a combined French-American expedition led by Reygasse, Gauthier and the American adventurer and chancer de Prorok, whose account is by no means to be relied on! Camps 1997 and Reygasse 1950, 88–97 for versions of the oral history of Tin Hinan collected at the time of the excavations in 1928. The ultimate origin of Tin Hinan as related in these histories was the Tafilalat oasis in Morocco. On the finds, see Reygasse 1950, 92–99; see also Camps 1965, for the incense burner.

https://doi.org/10.1017/9781108634311.001 Published online by Cambridge University Press

17

18

David J. Mattingly et al.

and a curved knife blade of distinctive types. On the interior walls of the building were an engraving of a camel and a number of Libyan inscriptions, some already broken when incorporated into the building. The anthropological examination of the skeleton identified it as a middleaged woman of Berber phenotype and above average height, with some osteological damage to her lower vertebrae and right leg.90 Around the outside of the fortified building were 14 high-sided drum tombs, several of which Reygasse also excavated, but which did not yield diagnostic finds. The skeletons within these tombs were laid in a crouched or contracted position on the right side and, in the majority of cases, with head pointing to the south-west and feet pointing to the north-east. Some of the drum tombs had stelae alongside, with rough Libyan inscriptions in at least two cases.91 Lhote also recorded examples of these drum tombs with protostelae that closely parallel examples found in the Garamantian heartlands in the Ahaggar.92 Further extraordinary Saharan discoveries have been made at Iwelen in Niger, where large tumuli have been excavated alongside a settlement and a rock art site.93 The biconical mode of depiction in the rock art of people sporting feathered headdresses, armed with metal weapons and associated with horses and chariots, shares much in common with the Garamantes and a Trans-Saharan connected world. A series of radiocarbon dates from Iwelen shows that the burial activity spanned a long period from the second millennium BC to the late first millennium AD.94 Finds from the burials included leather shrouds (often coloured red with ochre), patterned textiles, copper alloy and bone bracelets, handmade pottery and saddle querns.95 The predominant tomb types here were cairns (especially with a deliberate depression in the top, ‘tumulus à cratère’) and drum tombs with a well-built outer wall and flat top. The burial placement seems to normally have been in a crouched or flexed position, on the right side for men and left side for women (in 18 cases where the sex could be determined with certainty). Orientation was more mixed, with some preference for heads pointing to the east or south, but given the relatively small sample size and the long period of use of the burial area it is difficult to draw firm conclusions beyond the fact that there was possibly some distinction between women/children and men.96

90 93 96

Reygasse 1950, 99–100. Paris 1990; Roset 1986. Paris 1990, 50–62.

91 94

Reygasse 1950, 109–16. 92 Lhote 1984, 104–14. Paris 1990, 67–71. 95 Paris 1990, 62–67.

https://doi.org/10.1017/9781108634311.001 Published online by Cambridge University Press

1 Burials, Migration and Identity: The View from the Sahara

The complexity and wealth of some of these tombs are far from the ‘normative’ Saharan burials.97 What were the antecedents and inspirations for such extraordinary funerary assemblages? What has been lacking hitherto in discussions of these exceptional discoveries at Abalessa and Iwelen is an appreciation of how these can be understood in relation to Saharan burial traditions in general. The flexed burial postures, the recurrent use of leather shrouds and ochre with the burials, the similarities in the basic forms of monument are suggestive of underlying Saharan traits. As we shall see, the story of Garamantian burials provides significant information to help us contextualise Tin Hinan and Iwelen within a connected Trans-Saharan world.

Chad and West Africa Comparisons between the Sahara and the Sub-Saharan zones have not hitherto been made, but need to be explored as part of the new Trans-Saharan debate – if only to draw out the contrasts. Some of the key literature related to Chad and the Niger region of West Africa is summarised in the chapters in this volume by Sonja Magnavita and Scott MacEachern. There are a number of difficulties with the exploration of burials in the Sub-Saharan zone. Individual burials were not always marked with distinctive surface markers or necessarily clustered in recognisable funerary landscapes – though there are important exceptions to this, such as the funerary cliff sites in the Dogon region or the protohistoric megalithic burials in Senegal/Gambia.98 Surveys of Niger valley landscapes often record primarily the large, low mounds of settlement sites, though some of these also incorporated burials.99 There was clearly a high level of regional variability between the Chad basin and the Atlantic coast of West Africa and much diachronic development.100

Migration Ideas about population movement have a long history in archaeological study, in a variety of guises, of which migration is just one manifestation. 97 98 99

100

See Camps 1961; di Lernia and Manzi 2002; Reygasse 1950; Savary 1966; Souville 1959. Mayor et al. 2014; Laporte et al. 2012. Khalaf 2016 is an excellent example of the extent to which settlement mounds dominate the survey record and of the relative invisibility of the funerary landscape. See for example the variation present in Kiethéga et al. 1993; Kote 2000; Magnavita 2014; Sidibé 1980.

https://doi.org/10.1017/9781108634311.001 Published online by Cambridge University Press

19

20

David J. Mattingly et al.

Table 1.1. Models of different forms of mobility and migration in a Saharan context. Type of Migration or Mobility Evidence Mobile Pastoralism Colonisation/ Large-scale Migration Exogamy

Location

Isotopes, faunal data, All Sahara rock art inscriptions Linguistics, material Northern and culture Central Sahara Isotopic, osteological Central Sahara and funerary data Indirect All Sahara Roman and early Arab North Africa sources, architecture Indirect from material Central Sahara (Jarma, Kissi?) culture and early Southern Sahara Arab sources (Tadmakka) Roman sources and North Africa inscriptions Roman and early Arab Central Sahara sources

Slavery Conquest Trading Communities

Wage Labour Refugees

1000–1 BC AD 1–500 AD 500–1000 X

X

X

X?

X

X

X?

X?

X?

X?

X? X

X? X

X

X

X

X?

X?

X

Other terms include: dispersal, conquest, diaspora, refugee displacement, diffusion, translocation, inter-regional trade and other forms of economic migration, pastoral mobility (nomadic and transhumant), and so forth (Table 1.1). While the absolute scale of movement is rarely clear, we regularly encounter the effects from both historical sources and archaeological evidence. Starting with the first dispersals of Homo sapiens ‘out of Africa’, one of the defining characteristics of human communities has been their ability to adapt to new environments and their inclination towards restlessness.101 Early archaeological studies were very much affected by ideas of diffusionism, often equating the spread of certain objects with distinct cultural groups who were highly mobile.102 The tendency is particularly strong in the Germanic tradition.103 There was a concerted backlash against such ideas in the New Archaeology of the 1970s,104 but, while 101 102

103

Lahr 2010; Rouse 1986. See for example, Childe 1950. An alternative emphasis on indigenous development had already begun by the mid-1960s, Clarke 1966; see further Trigger 1997. Burmeister 2000, 539; Härke 1998; 2004. 104 See, for example, Adams et al. 1978.

https://doi.org/10.1017/9781108634311.001 Published online by Cambridge University Press

1 Burials, Migration and Identity: The View from the Sahara

simple culture historical models have long been discredited, the fact remains that there is much evidence to demonstrate the reality of past population movements. A revived debate has ensued in the last 25 years.105 It refers to mobility as a definition for any kind of human movement,106 while addressing the importance of distinguishing between forms of mobility, as they are related to different socio-economic and political behaviours and agendas.107 The difficulties centre on the feasibility of achieving this level of analysis through using archaeological evidence alone.108 A recent theoretical focus on issues like connectivity and globalisation has also highlighted the importance of other sorts of human movement – for instance, forced migrations within big imperial systems such as the Roman Empire through institutions like the garrison pattern and slavery, accompanied by higher levels of economic migration or political resettlement that the imperial situation stimulated.109 Current scientific studies of human remains, especially isotopic studies, from Roman sites have been providing confirmation of a high level of migration and social mobility.110 In Britain, for example, study of a number of Roman urban cemeteries have revealed that 40–60 per cent of skeletons sampled were individuals who had grown up in other regions of Britain or further afield in the empire.111 This sort of work has also revealed a disjuncture between the ethnic assumptions made by archaeologists based on material culture and the evidence of migrants and locals traced through the isotopic study of their teeth. A classic instance is the Lankhills cemetery at Winchester, where distinctive metalwork assemblages in select graves have long been considered to represent Germanic incomers, but where the situation in fact proves to be more complex with a mix of people born and raised in the locality and others coming from external locations associated with the brooches and belt fittings.112 In a Saharan context, Boozer has drawn

105

106 107 108

109 110

111

112

Antony 1990; 1992; Chapman 1990; Chapman and Hamerow 1997; Chapman and Dolukhanov 1992; Härke 1998. Barnard and Wendrich 2008; Hackenbeck 2008; Leary 2014. Barnard and Wendrich 2008. Burmeister 2000 uses a modern case study from the Americas to reassess the evidence of Anglo-Saxon migration; Hackenbeck 2008 stresses the importance of new scientific approaches to osteology, DNA and isotopes. Pearce 2010. Chenery et al. 2010; Eckardt 2010b; Leach et al. 2009; 2010, for isotopic studies from Gloucester, York. Although the researchers were specifically targeting individuals who they suspected had moved, Eckardt et al. 2014; Pearce 2015, 156. Eckardt 2010b; Eckardt et al. 2009.

https://doi.org/10.1017/9781108634311.001 Published online by Cambridge University Press

21

22

David J. Mattingly et al.

attention to the migration of exiles and dissidents of the Roman Empire in relation to the Egyptian oases (Oasis Magna).113 Alternative approaches to migration focus on linguistic evidence and human genetics – both areas where the evidence of human mobility seems unassailable in broad perspective, but where the fine detail remains elusive.114 Nonetheless, the pendulum has certainly swung back in favour of migration as a central theme in archaeology, as a number of recent studies illustrate.115 In early medieval Europe, while there have been revisionist accounts of the numbers involved in the so-called Barbarian migrations, the range and impacts of population movements in this era cannot be denied, whether in terms of the eruption of Asiatic Huns into Western Europe, or the long migration of people like the Vandals ahead of them from Germany, through France and Spain and ultimately across North Africa.116

Migration and Mobility in the Trans-Saharan Zone There are many practical difficulties with tracing human migration in the remote past related to demonstrating the ethnic composition of society and of making connections across wide spaces, evaluating the scale of movement of people and the precise chronology of such migrations, and understanding the motivation for such exchanges. In this section, we review the evidence for these forms in the Sahara and how we might characterise migration in different periods (Table 1.1). Ethnic migration is a notable feature of both Berber and later Arab oral histories of Saharan peoples,117 and while large-scale migration is one aspect of a wider pattern of mobility, not all mobility was necessarily ‘migration’. However, other forms of mobility can contribute in similar ways to migration to produce archaeological evidence of cultural and ethnic mixing. Our contributors consider migration broadly but as a specific form of mobility wherein the movement is generally long distance, permanent and involves moving from one cultural group to another. This still encapsulates a broad range of migrations including those related to conquest, slavery, trading communities, colonisation and refugees. 113 115 116

117

Boozer 2012. 114 Renfrew 1987; 2000. Burmeister 2000; Chapman and Hamerow 1997; Hackenbeck 2008; Härke 2004. On the Barbarian migrations, see inter alia, Halsall 2008; Heather 2009; Pohl and Reimitz 1998. On the ‘Germanic’ (Vandal) invaders of the Maghrib, see Merrills and Miles 2010; Modéran 2003; Modéran and Perrin 2014. Camps 1981; Ibn Khaldun 1958, Muqaddimah (translated by F. Rosenthal).

https://doi.org/10.1017/9781108634311.001 Published online by Cambridge University Press

1 Burials, Migration and Identity: The View from the Sahara

However, this is not to say that we exclude transhumance, wage labour and nomadism from the archaeological record; indeed these are integral to the Trans-Saharan past. For societies across the Sahara, even when climatic conditions were better than nowadays, mobility was a constant way to cope with environmental stress. During the Neolithic period, mobility was mainly related to the need of herding, particularly cattle. In historic times, when the climate dried out and oases dotted the desert landscape, mobility became a skill that was put to good use, for instance, by developing a complex TransSaharan trade system.118 Exogamy and raiding were other sources of mobility well known throughout the Sahara. There has also been a significant engagement in migration in African historiography, perhaps most famously relating to the idea of Bantu migration.119 In the absence to date of profound genetic studies, much debate focuses on language groups and linguistic evidence in general.120 In the context of the Maghrib, Sahara and sub-Sahara, there are several main attested migrations, which had potentially profound social consequences. The spread of Berber (Amazigh) speaking peoples is generally accepted as a phenomenon, but its scale and chronology remain controversial.121 Phoenician/Punic, Greek and Roman colonial adventures along the Mediterranean coastline have all had an impact and important consequences for the societies of the Maghrib, as to an even greater extent did the Arab conquest from the AD 640s. In medieval and early modern times, there have been periodic incidents of long-range political control crossing the desert – whether from the south (the Almoravids and the kingdom of Kanim’s control of Fazzan) or the north (the Ottoman outpost in Fazzan, the Moroccan suzerainty over Timbuktu). In addition, there have been repeated instances of mobile populations within the Trans-Saharan zone and the medieval and early modern oral and written histories are built up around migration mythologies.122 This has in part been a competition for resources and mobile pastoral peoples have often played a key role in these population movements. But another aspect of migration relates to the consequences of imperial or state power, especially in relation to the forced migration related to Saharan and Trans-Saharan slave trading. One consequence of that trade has been the involvement of Sub-Saharan states in Saharan 118 120 121

122

For an overview on the Trans-Saharan trade, see Mattingly et al. 2017a. 119 Ashley 2013. MacEachern 2013; Blench 2013. Brett and Fentress 1997; Camps 1981; 1987; Hachid 2000; see also chapters by Blench, Ehret and Fentress, Chapters 14–16, this volume. See numerous papers in Lewicki 1976/1983 and Thiry 1995.

https://doi.org/10.1017/9781108634311.001 Published online by Cambridge University Press

23

24

David J. Mattingly et al.

spaces – as with the kingdom of Kanim in Fazzan or Ghana and Songay further west. The traditional models of Saharan history have thus tended to assume a significant level of population movement. This is in part represented by a presumed natural predilection of pastoral peoples for long-range movements (though as we shall see, this is possibly more complicated than once believed due to the probable early emergence of oasis communities). The historiography of such ideas is ancient – there are already hints of this sort of thing in the Greco-Roman sources and the Islamic accounts of tribal origins were regularly built on ideas of migration and population displacement/replacement.123 The idea of migration is enshrined in the literature on North Africa from an early date, but primarily developed as part of a specific and simple discourse of nomadism. As an extension of this, some scholars have recently stressed the importance of mobility and migration for strengthening social relationships – contrary to the European perspective of it creating disorder – especially in African contexts where a high level of mobility is considered normal.124 The model of migration to be explored in this book builds on this movement away from aetiological (push-and-pull) factors, acknowledging and demonstrating greater complexity and mobility. The movement of people is shown to involve the evolution of oasis communities, their fluctuating relations with pastoral groups, the development of Saharan and Trans-Saharan trade, the extension of the globalising reach of the Mediterranean world and the impact of Sub-Saharan states.125 Globalisation theory is clearly of relevance to the operation of such long-distance interactions on economic and social structures – even when the Sahara has tended to provide an extreme test for the globalising reach of early states.126

Identity The complexity that has emerged in relation to the interpretation of burial traditions has obliged archaeologists to move away from simplistic cultural 123

124 125

126

Mattingly 2003, 76–90; 2011, 30–37, for analysis of Greco-Roman sources on Saharan peoples. On Islamic sources, see Ibn Khaldun 1958, Muqaddimah (translated by F. Rosenthal). Bilger and Kraler 2005; De Bruijn et al. 2001; Klute and Hahn 2008. The mobility of ideas and objects is instead addressed respectively in the ‘Mobile Technologies’ and ‘Trade’ volumes of this series (Cuénod et al. Forthcoming; Mattingly et al. 2017a). For some excellent examples of how the globalisation concept can be employed in an archaeological context, see Hingley 2005; Pitts and Versluys 2015; Versluys 2014; Witcher 2000; Boivin et al. 2014; Croucher 2003; Sindbaek 2007.

https://doi.org/10.1017/9781108634311.001 Published online by Cambridge University Press

1 Burials, Migration and Identity: The View from the Sahara

historical models that neatly link the material record with ethnic or large cultural groupings. It has also demanded more sophisticated approaches to issues of social identity and how this is manifested and replicated across communities.127 Debate about ethnic identity has been particularly intense, though the dangers of back-projecting modern notions of ethnicity and race onto past societies have long been recognised.128 The importance of unitary identity affiliations in modern society, dominated by nation states and global religions, is much less apparent for earlier periods. As Sen has demonstrated, the potential for multiple affiliations allows groups and individuals to construct a multiplicity of identities.129 The important work of Rogers Brubaker has focused attention on different types of identity expressions, again emphasising the multiple and diverse ways in which individuals and groups highlight their similarities and differences with others in society.130 A clear and present danger with current archaeological applications of identity theory is the very wide range of approaches to and understanding of the term ‘identity’.131 Not all work clearly articulates how identity is being defined, with inevitable ambiguities in the arguments advanced and their reception. Within this, we cannot explore the Trans-Saharan world without considering the role and nature of Berber (Amazigh) identity. Berbers, or more specifically Berber speakers, are found in almost all the areas outlined at the start of this chapter from the oasis of Siwa in Egypt to Zenaga on the Atlantic coast and from the Mediterranean coast as far as the Niger River and beyond; although in many of these areas they are today a minority. Three chapters in this book (by Ehret, Blench and Fentress) engage explicitly with the linguistic evidence for widely dispersed Berber-speaking communities within the Trans-Saharan region in antiquity. Berber identity has a strong cultural resonance today, but although rooted in the past it is ill-defined, having been used rather unsatisfactorily as a shorthand for indigenous North Africans through terms such as Proto-Berber and Libyco-Berber (although scholars have used these terms to portray a broad spectrum

127

128

129

130 131

For some key discussions, see Gardner 2004; Hoffman 2011; Insoll 2007; Diaz Andreu et al. 2005; Meskell 2001. See, notably, Jones 1997; also Derks and Roymans 2009; Gardner et al. 2013; Hall 1997; Hu 2013; Isaac 2004; Lucy 2005; Voss 2008. Sen 2006. There are important echoes here of the conclusions of Anderson 1983 in his work on imagined communities. Brubaker 2004; Brubaker and Cooper 2000. As Pitts 2007 emphasises. Some of the variability can be appreciated by comparing papers in Brody and Hoffman 2011; 2014; Hales and Hodos 2010; Laurence and Berry 1998.

https://doi.org/10.1017/9781108634311.001 Published online by Cambridge University Press

25

26

David J. Mattingly et al.

of societies).132 In the north, we have the historical specificity of the Hellenistic Numidian and Mauretanian kingdoms, established states near the Mediterranean coast; in the Sahara, the ‘timeless’ pastoralism of the Tuareg. And in the south, the Berbers are portrayed as agents of destruction, for example, in causing the collapse of the Tichitt tradition.133 Nor can Berber-speakers claim to be the only inhabitants of the Trans-Saharan region. Today and historically there have been a wide range of languages and groups with both greater and lesser political autonomy (including but not limited to speakers of Hausa, Mandé, Chadic and Arabic languages). A single Berber identity is clearly not recoverable from the past, nor desirable. Despite the potential problems associated with the lack of a unified understanding or definition of identity, there is no doubt that the study of past social identities has made significant progress. To take the example of the Roman Empire, the old emphasis on Romanisation and the prioritisation of a reified notion of Roman identity are now increasingly under challenge as a result of broader identity approaches.134 Two related theoretical positions will be highlighted here. Structuration theory, as first outlined by Giddens, has attracted many adherents and proved an effective tool for exploring the variations in material culture that characterise the archaeological record.135 The key aspect of structuration theory is that it brings together consideration of human agency and the social structures of society that inform and constrain those choices. The second approach is what the lead author of this chapter has dubbed ‘discrepant identity’.136 This again emphasises the importance of human agency, but always operating within social frameworks that direct and constrain choices. The idea of discrepant identity highlights the plural possibilities for identity expressions, noting also that an individual’s identity was often socially contingent with varied manifestations in different social groupings. As such, it has been successful in dismantling reified notions of identity such as those encapsulated by Romanisation and other ‘-isations’.137 Movement between

132 133 134

135 136

137

Bates 1970; Brett and Fentress 1997; Camps 1980; 1987; Hachid 2000. Munson 1980; MacDonald et al. 2009. Wallace-Hadrill 2008; Woolf 1997 for more narrowly focused departures from traditional models of Romanisation. Giddens 1984; Gardner 2007 for an excellent example of its application. Mattingly 2004 and 2011, 213–45, for explanations of discrepant identity; Mattingly 2006 for a developed case study relating to Britain in the Roman Empire. Mattingly 2004; 2011; 2014; Versluys 2014. See also, Dietler 2010; Prag and Quinn 2013 for approaches to Hellenisation and the impact of colonial settlement.

https://doi.org/10.1017/9781108634311.001 Published online by Cambridge University Press

1 Burials, Migration and Identity: The View from the Sahara

groups is potentially much easier and more fluid and this has particular implications for consideration of the evidence of potential changes. A further crucial aspect of how we should seek to address identity in the archaeological record is that we need to look beyond mere patterns in the material culture record. In relation to burial archaeology – we can analyse identity not just in terms of material culture in and around tombs, but in funerary landscapes, ideas about the place and role of the dead and in the physical bodies of the dead. Modern scientific techniques permit new insights into biological identities and highlight that these do not always neatly correspond with archaeologically defined cultural identities. Artefacts are of course enormously important to the construction and referencing of identity and the powerful agency of artefacts themselves in this regard should not be underestimated.138 However, it is also important to look at the way the material culture was employed – that is, to seek evidence for the human behaviours that lie behind the physical and artefactual record.139 Current scholarship on material culture focuses on a technological approach that combines the theoretical concept of ‘habitus’ with the ‘chaîne opératoire’ analytical framework.140 The latter takes into consideration the whole operational sequence behind an object: the decision-making behaviour, its material outcome and its spatial distribution. The identity marker in this case is defined as ‘technological style’: a combination of multiple choices made during the manufacturing sequence.141 According to Osborne, choices that are made consciously to engage the agent’s current actions with past actions creates an active relationship between the agent and their past that is better defined as tradition.142 In this perspective, a ‘technological style’ may also be defined as a ‘technological tradition’. Style or tradition, in and of themselves, do not equate with ethnic or cultural identity; they become markers of identity when deployed in the kind of ritual contexts, sacred or secular, familiar or public, that reproduce that identity.143 The diachronic perspective of a tradition is of particular significance because it identifies elements that represent a group identity through time, going beyond the mere markers of an archaeological culture. This has clear salience with burial archaeology and may help us to interpret its complexities.

138 139 140

141

Appudarai 1986; Gosden 2005; Gosden and Marshall 1999. See Pitts 2007; Mattingly 2014 for two articles that emphasise the importance of behaviour. On habitus and chaîne opératoire see Bourdieu 1977, for recent scholarship includes Pfaffenberger 1992; Stark 1999; 2006; Stark et al. 2008; Sellet 1993. Lechtman 1977. 142 Osborne 2008. 143 Porter 2012.

https://doi.org/10.1017/9781108634311.001 Published online by Cambridge University Press

27

28

David J. Mattingly et al.

There have been a number of recent studies of identity in relation to ancient African societies.144 In a Trans-Saharan context, the interconnections between the separate cultural zones of the Mediterranean lands, the Nilotic belt, the Sudanic zone and West Africa have been little considered in relation to the pre-Islamic era. Yet as the linguistic studies in this volume show, the spread of the Berber language group suggests a more interconnected space in the proto-historic and early historic period and this can be linked also to aspects of the rock art featuring horses and Libyan inscriptions (some of considerable antiquity) across the Sahara.145 One of the most interesting things about the Garamantes of the Central Sahara is precisely that their cultural identity was not imported ready-made from a distant homeland. Rather, it seems to have been forged in situ, drawing on very diverse artefactual and technological traditions as well as distinctive cultural reference points. One of the challenges of this book is to demonstrate the interrelationships of the three themes delineated in this opening chapter: burials, migration and identity. To anticipate one of the main conclusions of the volume, this is by no means straightforward and linear. It is apparent nonetheless from numerous contributions that these themes do overlap in interesting ways. There are a number of markers of the Sahara’s interconnectedness in the first millennium BC and first half of the first millennium AD, including language, the extension of oasis communities alongside pastoral societies, the importance of horses and chariots, shared burial rites and funerary structures across vast distances. Just as the late Neolithic Saharan pastoral societies can be viewed as widely dispersed, but operating within a common cultural framework, so their Proto-historic successors from the first millennium BC seem to have been part of a Saharan koine. That there were also profound differences between these proto-historical Saharan societies and in turn between them and their neighbours to the east, north and south is unsurprising, in light of modern work that downplays the significance of singular identity affiliations in the past.146 An awareness of the operation of a plurality of identities within societies can help us interpret changes in things like burial practice in more subtle ways. 144

145

For pre-Roman North Africa, see Krandel-Ben Younès 2002; Roman Africa, see, for example, Briand-Ponsart 2005; Briand-Ponsart and Crogiez 2002; Marshall 1998; Mattingly 1999; 2011, 236–68; Christian North Africa, see Conant 2012; Rebillard 2012; on Nilotic identities, Smith 2003; Edwards 2004; Boozer 2012; Hubschmann 2010; East Africa: Wynne-Jones 2013; West Africa: Gosselain 2000; Haour and Rossi 2010. Gauthier and Gauthier 2011; Dupuy 2010. 146 Sen 2006.

https://doi.org/10.1017/9781108634311.001 Published online by Cambridge University Press

1 Burials, Migration and Identity: The View from the Sahara

The Structure of This Book In this volume, the mortuary evidence from different regions of the Sahara is for the first time critically discussed and compared, trying to identify markers of identity and behaviours in the archaeological record so to detect trajectories of mobility. As outlined at the start of this chapter, our intention is to make the Sahara central to the discussion and the volume is structured accordingly. In a book of this scale, it is not possible to be comprehensive in coverage of the entirety of the Trans-Saharan zone and every example of funerary traditions. We hope at least to have included enough regional case studies to allow a baseline assessment of similarities and differences to be made. Contributors have been encouraged to engage with the three key themes of burial, migration and identity. In the first section of the book, paired chapters on the archaeological evidence of burial traditions and the stable isotope analysis of buried individuals are presented for two key case studies from the Central Sahara. Chapter 2, co-authored by members of the Trans-SAHARA team, deals with the Garamantian heartlands around Jarma in the Wadi alAjal. The following chapter by Gatto et al. focuses on the Italian work in the Wadi Tanzzuft/Tadrart Akakus area, at the south-west fringes of Garamantian territory, and reveals a substantially different trajectory of changing funerary rites and traditions compared to the Wadi al-Ajal area. The scientific exploration of these two populations is then taken up by both Power et al. and Ricci et al. in their respective chapters. Both teams have used a combination of osteological, 3-D geometric morphometric and isotopic (oxygen and carbon isotopes in the former and carbon, nitrogen and strontium in the latter) analyses to present a richer investigation of biological identities in these Central Sahara societies. Although the datasets are not directly comparable, the results presented here indicate the potential importance of migrants within both the examined burial assemblages and the complex interplay between biological and cultural identities. The second section, Looking East, turns back to the civilisations of the Nile valley. Chapter 6 by Edwards explores the similarities and contrasts between Nilotic and Saharan societies and their burial traditions. In the following chapter, Buzon et al. explore the potential of strontium and oxygen isotope analysis to examine mobility between Egypt and Nubia and in North Africa more generally. They conclude that although isotopic values overlap between different regions, possible immigrants can be identified. However, stronger and more persuasive interpretations can

https://doi.org/10.1017/9781108634311.001 Published online by Cambridge University Press

29

30

David J. Mattingly et al.

only come by combining the results with aspects of skeletal remains and burial ritual, again demonstrating a complex relationship between burials and identity. The third section, Looking North, assesses the possibility of Saharan influences on two important indigenous peoples of the Maghrib, the Numidi and the Maures. The studies by Sanmartí et al. and Papi are important re-evaluations of the contribution of these indigenous population in evolving funerary practices in the Maghrib in the first millennium BC. These chapters relate to the same landscapes that were the focus of Camps’ great book. As noted already, Camps identified evidence of increasing hierarchy and complexity in the first millennium BC, culminating in the emergence of the Numidian and Mauretanian kingdoms by the latter centuries BC. Although this at one time seemed very different to the Saharan world, the more detailed understanding we now have of Garamantian society suggests much greater parallelism. It is apparent that, in addition to Mediterranean cultural influences on indigenous burials in the Maghrib during the first millennium BC, there was also some overlap with prevailing Saharan traditions. The next section, Looking West, links a pair of papers that explore further these potential links with Saharan traditions in the Western Sahara. Youssef Bokbot considers a wide range of pre-Islamic funerary monuments from southern Morocco, many located south of the Atlas Mountains, but combining a range of influences from the Mediterranean and Saharan zones. The overview of funerary evidence from the ‘Free Zone’ of the Western Sahara by Clarke and Brooks has many elements in common with the evidence from Fazzan, suggesting that there was indeed some sort of Saharan koine, though they also insist on the potential significance of Atlantic connections. Again, it is a matter of debate whether these came independently from Atlantic Europe to the Western Sahara, or represent a degree of intersection with Proto-historic peoples of northern Morocco. However, it does once again indicate that long-range movement of identity markers need to be carefully considered, even if these do not always correspond with actual movements of people. The final regional comparisons involve Looking South to two studies of Sub-Saharan burial traditions. Sonja Magnavita provides a summary on her important work on the key site of Kissi in Burkina Faso, while Scott MacEarchern provides an overview of burial traditions in the Lake Chad area. Burials in both these case studies reveal pronounced differences in terms of structures, burial rite and memorialisation from those of the

https://doi.org/10.1017/9781108634311.001 Published online by Cambridge University Press

1 Burials, Migration and Identity: The View from the Sahara

Sahara and Maghrib, though Kissi is noteworthy for the incorporation into burials of a number of artefacts that demonstrate long-range interconnections, perhaps reflecting external trade contacts. Whether or not the Mediterranean-sourced copper alloys attested at Kissi were traded via the Garamantes cannot be proven at present, nor whether the Garamantes were potential suppliers of the woollen textiles recorded.147 It is at any rate interesting that, as with the Garamantes, the availability of external goods created new patterns of social display and identity presentation in this society. The final section deals with the issue of migration through the evidence of language groups and linguistic history of the TransSaharan region. The chapters by Roger Blench and Chris Ehret both primarily rely on the linguistic evidence to propose contrasting reconstructions and chronologies. The third chapter by Lisa Fentress offers a more archaeologically informed response to their views and in particular attempts to show the implications of Ehret’s model when measured against the (admittedly still meagre) archaeological data. Although there is no overall consensus between these analyses, what they all recognise is that the Trans-Saharan zone shows clear indications of some major language changes that may have been ultimately linked to the migration or movement of people. As Fentress puts it, because of the lacunose nature of the evidence ‘reasonable people can give very different reconstructions of the history of language diffusion’ and the fact that we have three overlapping accounts with different emphases and conclusions should not occasion surprise or dismay. The issue of a potential major Berber diaspora to account for the spread of Berber languages in Sahara and Maghrib is a key one to resolve. Blench favours a date around 6000 BC for this and associates it with cattle pastoralists, while Ehret and Fentress suggest a slightly later date 1000 BC, which would link with the spread of the first oasis cultivators, including the Garamantes. It is clear from both Blench and Ehret that the linguistic evidence suggests a relatively early spread of a Proto-Berber language right across the Sahara. However, a problem with reconstructing the direction of travel is that the evidence of surviving language groups is heavily biased towards the Western Sahara and Western Maghrib. One consequence of this is Ehret’s argument, against prevailing archaeological evidence, for a west-to-east direction of travel. 147

See Fenn et al. 2009; Magnavita 2003; 2008; 2009. There does not appear to have been local production of textiles in the Niger area at so early a date.

https://doi.org/10.1017/9781108634311.001 Published online by Cambridge University Press

31

32

David J. Mattingly et al.

For such a vast area and covering such a long timescale, it is perhaps inevitable that a volume such as this will raise as many questions as it can answer. We hope at least that the following chapters will reinvigorate debate about the meaning of linguistic and burial data in relation to the bigger issues of migration, social complexity, identity and behaviours. Some concluding thoughts are provided at the end by the editors.

References Abulafia, D. 2011. The Great Sea: A Human History of the Mediterranean. London: Allen Lane. Adams, W. Y., Van Gerven, D. P. and Levy, R. S. 1978. The retreat from migrationism. Annual Review of Anthropology 7: 483–532. Anderson, B. 1983. Imagined Communities: Reflections on the Origin and Spread of Nationalism. London: Verso. Anthony, D. W. 1990. Migration in archaeology: the baby and the bathwater. American Anthropologist 92: 895–914. Anthony, D. W. 1992. The bath refilled: migration in archaeology again. American Anthropologist 94: 174–76. Appadurai, A. (ed.). 1986. The Social Life of Things: Commodities in Cultural Perspective. Cambridge: Cambridge University Press. Arnold, B. 2002. A landscape of ancestors. The space and place of death in Iron Age West-Central Europe. In Silverman and Small 2002: 129–44. Ashley, C. 2013. Archaeology and migration in Africa. In Mitchell and Lane 2013: 77–85. Austen, R. A. 2010. Trans-Saharan Africa in World History. Oxford: Oxford University Press. Ayoub, M. S. 1967a. Excavations in Germa between 1962 to 1966. Tripoli: Ministry of Education. Ayoub, M. S. 1967b. The Royal cemetery at Germa. A preliminary report. Libya Antiqua 3–4: 213–19. Barker, G., Gilbertson, D., Jones, B. and Mattingly, D. J. 1996a. Farming the Desert. The UNESCO Libyan Valleys Archaeological Survey. Volume 1, Synthesis. (Principal Editor, G. Barker). Paris, London: UNESCO, Society for Libyan Studies. Barker, G., Gilbertson, D., Jones, B. and Mattingly, D. J. 1996b. Farming the Desert. The UNESCO Libyan Valleys Archaeological Survey. Volume 2, Gazetteer and Pottery (Principal Editor, D. J. Mattingly). Paris, London: UNESCO, Society for Libyan Studies. Barnard, H. and Wendrich, W. 2008. The Archaeology of Mobility: Old and New World Nomadism. Los Angeles: Cotsen Institute of Archaeology at UCLA.

https://doi.org/10.1017/9781108634311.001 Published online by Cambridge University Press

1 Burials, Migration and Identity: The View from the Sahara

Barrett, J. C. 1990. The monumentality of death: the character of Early Bronze Age mortuary mounds in southern Britain. World Archaeology 22.2: 179–89. Bartel, B. 1982. A historical review of ethnological and archaeological analysis of mortuary practice. Journal of Anthropological Archaeology 1: 32–58. Bates, O. 1970. The Eastern Libyans. London: Frank Cass. Bellair, P., Gobert, E. G., Jodot, P. and Pauphillet, D. 1953. Mission au Fezzan (1949). Tunis: Institut des Hautes Études de Tunis. Belmonte. J. A., Esteban, C., Perera Betancourt, M. and Marrero, R. 2002. Archaeoastronomy in the Sahara: the tombs of the Garamantes at Wadi al-Ajal, Fazzan, Libya. Archaeoastronomy, 27 (JHA suppl. 33): S1–S19. Ben Abed-Ben Khader, A. and Griesheimer, M. (eds). 2004. La nécropole romaine de Pupput. Rome: CollEFR 323. Benichou-Safar, H. 1995. Tophets et nécropoles puniques. In Trousset 1995: 91–102. Ben Lazreg, N. and Mattingly, D. J. (eds). 1992. Leptiminus (Lamta): A Roman Port City in Tunisia, Report no. 1. Ann Arbor: JRA Suppl. Series. Ben Younes, H. 1995. L’architecture funéraire punique au Sahel. Etat et perspectives. In Trousset 1995: 73–90. Ben Younes, H. 1996. Le Sahel et l’île de Djerba: les charactérisiques del’architecture funéraire punique. Africa 14: 7–12. Ben Younes, H. 2007. Interculturality and the Punic funerary world. In Stone and Stirling 2007: 32–42. Ben Younes, H. and Krandel-Ben Younes, A. 2014. Punic identity in North Africa: the funerary world. In J. Crawley Quinn and N. Vella (eds), The Punic Mediterranean. Identities and Identification from Phoenician Settlement to Roman Rule, Cambridge: Cambridge University Press, 148–67. Bietak, M. 1968. Studien zur Chronologie der Nubischen C-Gruppe. Ein Beitrag zur Frühges-chichte Unternubiens zwischen 2200 und 1550 vor Chr. Wien: VÖAW. Bilger, V. and Kraler, A. 2005. Introduction: African migrations. Historical perspectives and contemporary dynamics. Wiener Zeitschrift für kritische Afrikastudien; Special Issue – African Migrations. Historical Perspectives and Contemporary Dynamics 8: 5–21. Blench, R. 2013. Language, linguistics and archaeology: their integration into the study of African prehistory. In Mitchell and Lane 2013: 49–64. Bobrowski P., Czekaj-Zastawny, A. and Schild, R. 2014. Gebel el-Muqaddas (Site E-06-4). The early Neolithic tumuli from Nabta Playa (Western Desert, Egypt). In J. R. Anderson and D. A. Welsby (eds), The Fourth Cataract and Beyond, Leuven: Peeters, 293–301. Boivin, N., Crowther, A., Prendergast, M. and Fuller, D. Q. 2014. Indian Ocean food globalisation and Africa. African Archaeological Review 31.4: 547–81. Bokbot, Y. 1991. Habitats et monuments funéraires du Maroc protohistoriques. Thèse de Doctorat. Université de Aix-en-Provence.

https://doi.org/10.1017/9781108634311.001 Published online by Cambridge University Press

33

34

David J. Mattingly et al.

Boozer, A. 2012. Globalizing Mediterranean identities: the overlapping spheres of Egyptian, Greek and Roman worlds at Trimithis. Journal of Mediterranean Archaeology 25.2: 93–116. Bourdieu, P. 1977. Outline of a Theory of Practice. Cambridge: Cambridge University Press. Brett, M. and Fentress, E. 1997. The Berbers. Oxford: Wiley Blackwell. Briand-Ponsart, C. (ed.). 2005. Identités et cultures dans l’Algérie antique. Rouen: Publications de l’Université de Rouen. Briand-Ponsart, C. and Crogiez, S. (eds). 2002. L’Afrique du nord antique et médiévale: mémoire, identité et imaginaire. Rouen: Publications de l’Université de Rouen. Brody, L. R. and Hoffman, G. L. (eds). 2011. Dura-Europos: Crossroads of Antiquity. Chestnut Hill: McMullen Museum of Art, Boston College. Brody, L. R. and Hoffman, G. L. (eds). 2014. Roman in the Provinces. Art on the Periphery of Empire. Chestnut Hill: McMullen Museum of Art, Boston College. Brogan, O. and Smith, D. J. 1985. Ghirza: A Romano-Libyan Settlement in Tripolitania. Libyan Antiquities Series 1, Tripoli. Broodbank, C. 2013. The Making of the Middle Sea. A History of the Mediterranean from the Beginning to the Emergence of the Classical World. London: Thames and Hudson. Brooks, N., di Lernia, S., Drake, N., Chiapello, I., Legrand, M., Moulin, C. and Prospero, J. 2005. The environment-society nexus in the Sahara from prehistoric times to the present day. The Journal of North African Studies 304: 253–92. Brooks, N., Clarke, J., Crisp, J., Crivellaro, F., Jousse, H., Markiewicz, E., Nichol, M., Raffin, M., Robinson, R., Wasse, A. and Winton, V. 2006. Funerary sites in the ‘Free Zone’: report on the second and third seasons of fieldwork of the Western Sahara Project. Sahara 17: 73–94. Bronk Ramsey, C. 2009. Bayesian analysis of radiocarbon date. Radiocarbon 51.1: 337–60. Brubaker, R. 2004. Ethnicity without Groups. Cambridge: Cambridge University Press. Brubaker, R. and Cooper, F. 2000. Beyond ‘identity’. Theory and Society 29: 1–47. Buikstra, J. E. 1995. Tombs for the living . . . or for the dead: the Osmore ancestors. In T. Dillehay (ed.), Tombs for the Ancestors, Washington, D.C.: Dumbarton Oaks Research Library and Collection, 229–80. Burmeister, S. 2000. Archaeology and migration – approaches to an archaeological proof of migration. Current Anthropology 41.4: 539–67. Camps, G. 1955. De nouveau sur l’archéologie du Fezzan. Travaux de l’Institut de Recherches Sahariennes 13: 189–98. Camps, G. 1961. Aux origines de la Berbérie. Monuments et rites funéraires protohistoriques. Paris: AMG. Camps, G. 1965. Le tombeau de Tin Hinan à Abalessa. Travaux de l’Institut de Recherches Sahariennes 24 : 65–83.

https://doi.org/10.1017/9781108634311.001 Published online by Cambridge University Press

1 Burials, Migration and Identity: The View from the Sahara

Camps, G. 1980. Berbères. Aux marges de l’histoire. Toulouse. Camps, G. 1981. L’origine des Berbères. Les Cahiers du CRESM 12: 9–33. Camps, G. 1984. Rex maurorum et romanorum, recherches sur les royaumes de la Maurétanie des VIème et VIIème siècles. Antiquités africaine 20: 183–218. Camps, G. 1986. Funerary monuments with attached chapels from the Northern Sahara. The African Archaeological Review 4: 151–64. Camps, G. 1987. Les Berbères. Mémoire et identité. Paris: Errance. Camps, G. 1995. Les nécropoles mégalithiques de l’Afrique du nord. In Trousset 1995: 17–31. Camps, G. 1997. Tin Hinan et sa légende. A propos du tumulus princier d’Abalessa (Ahaggar, Algérie). Bulletin archéologique du comité des travaux historiques et scientifiques ns 24: 173–95. Camps, G. and Camps-Fabrer, H. 1964. La nécropole mégalithique du Djebel Mazela à Bou Nouara. Paris: Arts et métiers graphiques. Cannon, A. 1989. The historical dimension in mortuary expressions of status and sentiment. Current Anthropology 30.4: 437–58. Caputo, G. 1937. Archeologia. In Sahara Italiano 1937: 301–30. Caputo, G. 1949. Scavi archeologici nel Sahara libico. Annali dell’Istituto Orientale di Napoli ns 3: 13–33. Caputo, G. 1951. Parte II. In Pace, Sergi and Caputo 1951: 200–442. Carroll, M. 2006. Spirits of the Dead: Roman Funerary Commemoration in Western Europe. Oxford: Oxford University Press. Carroll, M. and Rempel, J. 2011. Living with the Dead: Burial and Commemoration in the Classical World. Oxford: Oxbow. Castelli, R., Cremaschi, M., Gatto, M. C., Liverani, M. and Mori, L. 2005. A preliminary report of excavations in Fewet, Libyan Sahara. Journal of African Archaeology 3: 69–102. Chapman, R. 1990. Emerging Complexity: The Later Prehistory of South-Eastern Spain, Iberia and the West Mediterranean. Cambridge: Cambridge University Press. Chapman, J. and Dolukhanov, P. M. 1992. The baby and the bathwater: pulling the plug on migrations. American Anthropologist 94: 169–74. Chapman, J. and Hamerow, H. (eds). 1997. Migration and Invasion in Archaeological Explanation. British Archaeological Reports International Series 664. Oxford: Archaeopress. Chapman, R. 1981. The Archaeology of Death. Cambridge: Cambridge University Press. Chenery, C., Müldner, G. H., Evans, J., Eckardt, H., Leach, S. and Lewis, M. E. 2010. Strontium and stable isotope evidence for diet and mobility in Roman Gloucester, UK. Journal of Archaeological Science 37.1: 150–63. Cherstich, L. 2008. The Southern Necropolis of Cyrene. Oxford: University of Oxford.

https://doi.org/10.1017/9781108634311.001 Published online by Cambridge University Press

35

36

David J. Mattingly et al.

Chesson, M. S. 2001. Social memory, identity, and death: anthropological perspectives on mortuary rituals. Archaeological Papers of the American Anthropological Association 10.1: 1–10. Childe, V. G. 1950. Prehistoric Migrations in Archaeology. Oslo: H. Aschehoug and Co. Clark, G. 1966. The invasion hypothesis in British archaeology. Antiquity 11: 172–89. Clarke, J. and Brooks, N. (eds). 2018. The Prehistory of Western Sahara: A Synthesis of Fieldwork, 2002 to 2009. Oxford: Oxbow Books. Conant, J. 2012. Staying Roman: Conquest and Identity in Africa and the Mediterranean, 439–700. Cambridge: Cambridge University Press. Cool, H. 2004. The Roman Cemetery at Brougham, Cumbria: Excavations 1966–67. London: Roman Society. Cremaschi, M. 2001. Holocene climatic changes in an archaeological landscape: the case study of Wadi Tanezzuft and its drainage basin (SW Fezzan, Libyan Sahara). Libyan Studies 32: 3–27. Cremaschi, M. and di Lernia, S. (eds). 1998. Wadi Teshuinat. Palaeoenvironment and Prehistory in South-western Fezzan (Libyan Sahara). Survey and Excavations in the Tadrart Acacus, Erg Uan Kasa, Messak Sattafet and Edeyen of Murzuq, 1990–1995. Milan: CNR. Cremaschi, M. and Zerboni, A. 2009. Early to Middle Holocene landscape exploitation in a drying environment: two case studies compared from the Central Sahara (SW Fezzan, Libya). C.R. Geoscience 341: 689–702. Cremaschi, M and Zerboni, A. 2011. Human communities in a drying landscape. Holocene climate change and cultural response in the central Sahara. In I. P. Martini and W. Chesworth (eds), Landscape and Societies, Selected Cases, Springer, London, 67–89. Croucher, S. L. 2003. Globalization and Belonging: The Politics of Identity in a Changing World. Lanham, Boulder, New York: Rowman and Littlefield. Cuénod, A., Duckworth, C. and Mattingly, D. J. Forthcoming. Mobile Technologies in the Ancient Sahara and Beyond. Trans-Saharan Archaeology Volume IV, Series editor D. J. Mattingly. Cambridge: Cambridge University Press and The Society for Libyan Studies. Daniels, C. M. 1970. The Garamantes of Southern Libya. London: Oleander. Daniels, C. M. 1989. Excavation and fieldwork amongst the Garamantes. Libyan Studies 20: 45–61. Daróczi, T. T. 2012. Death, disposal and social memory – towards a definition of funerary landscapes. Journal for Ancient Studies 3: 199–207. De Bruijn, M., van Dijk, R. and Foeken, D. 2001. Mobile Africa. Changing Patterns of Movement in Africa and Beyond. Leiden: Brill. Delibrias, G., Guillier, M. T. and Labeyrie, J. 1982. GIF Natural Radiocarbon Measurements IX. Radiocarbon 24.3, 291–343.

https://doi.org/10.1017/9781108634311.001 Published online by Cambridge University Press

1 Burials, Migration and Identity: The View from the Sahara

DeMenocal, P. B. and Tierney, J. E. 2012. Green Sahara: African humid periods paced by Earth’s orbital changes. Nature Education Knowledge 3.10: 12. De Prorok, B. K. 2001. In Quest of Lost Worlds: Five Archaeological Expeditions 1925–1934. Santa Barbara: Narrative Press (reprint of 1935 book). Derks, T. and Roymans, N. (eds). 2009. Ethnic Constructs in Antiquity: The Role of Power and Tradition. Amsterdam: Amsterdam University Press. Déroche, F. and Leclant, J. (eds). 2010. Monuments et cultes funéraires d’Afrique du Nord. Paris: Académie des inscriptions et belles-lettres. Diaz Andreu, M., Lucy, S., Babic, S. and Edwards, D. 2005. The Archaeology of Identity: Approaches to Gender, Age, Status, Ethnicity and Religion. London: Taylor and Francis. Dietler, M. 2010. Archaeologies of Colonialism: Consumption, Entanglement, and Violence in Ancient Mediterranean France. Berkeley: University of California Press. di Lernia, S. 2006. Building monuments, creating identity: cattle cult as a social response to rapid environmental change in the Holocene Sahara. Quaternary International 151.1: 50–62. di Lernia, S. 2013. The emergence and spread of herding in Northern Africa: a critical reappraisal. In Mitchell and Lane 2013, 527–40. di Lernia, S. and Manzi, G. (eds). 2002. Sands, Stones and Bones. The Archaeology of Death in the Wadi Tanezzuft Valley (5000–2000 BP). AZA 3. Firenze: All’Insegna del Giglio. di Lernia, S., and Tafuri, M. A. 2013. Persistent deathplaces and mobile landmarks. The Holocene mortuary and isotopic record from Wadi Takarkori (SW Libya). Journal of Anthropological Archaeology 32: 1–15. di Lernia, S., Manzi, G. and Merighi, F. 2002. Cultural variability and human trajectories in later prehistory of the Wadi Tanezzuft. In di Lernia and Manzi 2002, 281–302. di Lernia, S., Bertolani, G. B., Merighi, F., Ricci, F. R., Manzi, G. and Cremaschi, M. 2001. Megalithic architecture and funerary practices in the late prehistory of Wadi Tenezzuft (Libyan Sahara). Libyan Studies 32: 29–48. Dillehay, T. (ed.). 1995. Tombs for the Ancestors. Washington, DC: Dumbarton Oaks Research Library and Collection. Di Vita, A. 1976. Il mausoleo punico-ellenistico B di Sabratha. RM 83: 273–85. Dodson, A. and Ikram, S. 2008. The Tomb in Ancient Egypt: Royal and Private Sepulchres from the Early Dynastic Period to the Romans. London: Thames and Hudson. Drake, N., Wilson, A., Pelling, R., White, K., Mattingly, D. J. and Black, S. 2004. Water table decline, springline desiccation and the early development of irrigated agriculture in the Wadi al-Ajal, Libyan Fazzan. Libyan Studies 35: 95–112. Drake, N. A., Blench, R. M., Armitage, S. J., Bristow, C. S. and White, K. H. 2011. Ancient watercourses and biogeography of the Sahara explain the peopling of the desert. Proceedings of the National Academy of Sciences of America 108: 458–62.

https://doi.org/10.1017/9781108634311.001 Published online by Cambridge University Press

37

38

David J. Mattingly et al.

Dreyer, G. 1992. Zur Rekonstruktion der Oberbauten der Königsgräber der 1. Dynastie in Abydos. Mitteilungen des Deutschen Archäologischen Instituts Abteilung Kairo 47: 93–104. Duday, H. 2009. The Archaeology of the Dead. Lectures in Archaeothanatology. Oxford: Oxbow. Dupuy, C. 2010. Les apports archéologiques des gravures rupestres de l’Aïr (Niger) et de l’Adrar des Iforas (Mali). Les Nouvelles d’archéologie 120–121: 29–37. Eckardt, H. (ed.). 2010a. Roman Diasporas: Archaeological Approaches to Mobility and Diversity in the Roman Empire. Portsmouth, Rhode Island: Journal of Roman Archaeology Supplement series. Eckardt, H. 2010b. A long way from home: diaspora communities in Roman Britain. In: Eckardt 2010a: 99–130. Eckardt, H., Booth, P., Chenery, C., Müldner, G. H., Evans, J. A. and Lamb, A. 2009. Oxygen and strontium isotope evidence for mobility in Roman Winchester. Journal of Archaeological Science 36.12: 2816–25. Eckardt, H., Muldner, G. and Lewis, M. 2014. People on the move in Roman Britain. World Archaeology 46.4: 534–50. Edwards, D. N. 2004. The Nubian Past. An Archaeology of Sudan. London: Routledge. el-Rashedy, F. 1988. Garamantian burial customs: their relation to those of other peoples of North Africa. In Libya Antiqua. Report and papers of the symposium organized by UNESCO in Paris, 16 to 18 January 1984. Paris: 77–105. Fauvelle, F. 2016. What is trade about? A political ecology of Trans-Saharan trade and a taphonomic approach to its remains. In the 23rd Biannual Meeting of the Society of Africanist Archaeologists, What Past for Africa? Abstract Volume, Toulouse, Université Jean Jaurès, 26 June–2 July 2016, 194–95. Fenn, T. R., Killick, D. J., Chesley, J., Magnavita, S., and Ruiz, J. 2009. Contacts between West Africa and Roman North Africa: archaeometallurgical results from Kissi, northeastern Burkina Faso. In Magnavita et al. 2009: 119–46. Fentress, E. and Wilson, A. 2016. The Saharan Berber diaspora and the southern frontiers of Byzantine North Africa. In S. T. Stevens and J. P. Conant (eds), North Africa under Byzantium and Early Islam, Dumbarton Oaks Byzantine Symposia and Colloquia 7, Washington, DC: Dumbarton Oaks Research Library and Collection, 41–63. Ferjaoui, A. 2007. Le sanctuaire de Henchir el-Hami de Ba’al hammon au Saturn Afriicaine 1er s. av J. C. – Ive s. ap. J. C. Tunis: INP. Fontana, S. 1995. I manufatti romani dei corredi funerari del Fezzan. Testimonianza dei commerci e della cultura dei Garamanti (I–III sec.d. C.). In Productions et exportations Africaines. Actualités archéologiques en Afrique du nord Antique et Médiévale. Roma: Edition du CTHS, 405–20. Friedman, R. F. 2001. Excavations in the Nubians Cemeteries. Sudan & Nubia 5: 29–37.

https://doi.org/10.1017/9781108634311.001 Published online by Cambridge University Press

1 Burials, Migration and Identity: The View from the Sahara

Friedman, R. F. 2004. Excavation of the C-Group cemetery at HK27C, Hierakonpolis. Sudan & Nubia 8: 47–51. Friedman, R. F. 2007. The C-Group cemetery at Locality HK27C: Results of the 2007 Season. Sudan & Nubia 11: 57–62. Friedman, R. F., Van Neer, V. and Linseele, V. 2011. The elite Predynastic cemetery at Hierakonpolis: 2009–2010 update. In R. F. Friedman and P. N. Fiske (eds), Egypt at Its Origins 3. Proceedings of the Third International Conference Origin of the State. Predynastic and Early Dynastic Egypt. London, 27th July–1st August 2008, Peeters: Leuven, 157–91. Garcea E. A. A. 2013. Gobero: The No-Return Frontier. Archaeology and Landscape at the Saharo-Sahelian Borderland. Frankfurt: Africa Magna Verlag. Gardner, A. (ed.). 2004. Agency Uncovered: Archaeological Perspectives on Social Agency, Power, and Being Human. London: University College London Press. Gardner, A. 2007. An Archaeology of Identity: Soldiers and Society in Late Roman Britain. Walnut Creek, CA: Left Coast Press. Gardner, A., Herring, E., Lomas, K. (eds). 2013. Creating Ethnicities and Identities in the Roman World. London: Institute of Classical Studies. Gatto, M. C. 2005. Nubians in Egypt: Survey in the Aswan-Kom Ombo Region. Sudan & Nubia 9: 73–76. Gatto, M. C. 2011. The Nubian pastoral culture as link between Egypt and Africa: a view from the archaeological record. In K. Exell (ed.), Egypt and Its African Context, Oxford: Archaeopress, 21–29. Gatto, M. C. 2012. The Holocene prehistory of the Nubian Eastern Desert. In H. Barnard and B. Duistermaat (eds), History of the People of the Eastern Desert from Prehistory to Present, Los Angeles: UCLA/CIOA Press, 42–57. Gatto, M. C. 2013. Late Prehistoric sites in the area between Aswan and Kom Ombo. In D. Raue, S. J. Seidlmayer and P. Speiser (eds), The First Cataract of the Nile–One Region, Diverse Perspectives, Berlin: De Gruyter, 59–65, pl. 11–13. Gatto, M. C. and Zerboni, A. 2015. Holocene supra-regional environmental crises as motor for major socio-cultural changes in Northeastern Africa and the Sahara. African Archaeological Review 32: 301–333. Gatto, M. C., Curci, A. and Urcia, A. 2014. Nubian evidence in the Egyptian First Nome: Results of the 2013–2014 field seasons of the Aswan-Kom Ombo Archaeological Project (AKAP). Journal of Ancient Egyptian Interconnections 6.4: 38–41. Gauthier, Y. 2015. Pre-Islamic dry-stone monuments of the central and western Sahara. In C. L. N. Ruggles (ed.), Handbook of Archaeoastronomy and Ethnoastronomy. Heidelberg: Springer-Verlag GmbH, 1059–77. Gauthier, Y. and Gauthier, C. 1999. Orientation et distribution de divers types de monuments lithiques du Messak et des régions voisines (Fezzan, Libye). Sahara 11: 87–108.

https://doi.org/10.1017/9781108634311.001 Published online by Cambridge University Press

39

40

David J. Mattingly et al.

Gauthier, Y. and Gauthier, C. 2007. Monuments funéraires sahariens et aires culturelles. Les Cahiers de l’AARS 11: 65–78. Gauthier, Y. and Gauthier, C. 2011. Des chars et des Tifinagh: étude aréale et corrélations. Les Cahiers de l’AARS 15: 91–118. Giddens, A. 1984. The Constitution of Society. Outline of the Theory of Structuration. Cambridge: Polity Press. Gonzalez Rodriguez, M. 2014. Death Is Another Country. Mortuary Rituals and Cultural Identity in Fazzan, Libya. Unpublished PhD thesis, University of Leicester. Gosden, C. 2005. What do objects want? Journal of Archaeological Method and Theory 12: 193–211. Gosden, C. and Marshall, Y. 1999. The cultural biography of objects. World Archaeology 31.2: 169–78. Gosselain, O. P. 2000. Materializing identities: an African perspective. Journal of Archaeological Method and Theory 7.3: 187–217. Gowland, R. and Knusel, C. 2009. The Social Archaeology of Funerary Remains. Oxford: Oxbow Books. Graham, E. J. 2009. Becoming persons, becoming ancestors. Personhood, memory and the corpse in Roman rituals of social remembrance. Archaeological Dialogues 16.1: 51–74. Hachid, M. 2000. Les premiers berbères. Entre Méditerranée, Tassili et Nil. Aix-enProvence: Edisud. Hachid, M. 2006. Du nouveau sur le monument d’Abalessa (Ahaggar, Algérie). De la date de l’introduction du dromadaire au Sahara central, du personnage d’Abalessa et des inscriptions rupestres dites ‘libyco-berbères’. Sahara 17: 95–120. Hackenbeck, S. 2008. Migration in archaeology: are we nearly there yet? Archaeological Review from Cambridge 23.2: 9–26. Hales, S. and Hodos, T. (eds). 2010. Material Culture and Social Identities in the Roman World. Cambridge: Cambridge University Press. Hall, J. 1997. Ethnic Identity in Greek Antiquity. Cambridge: Cambridge University Press. Halsall, G. 2008. The Barbarian Migrations and the Roman West, 376–568. Cambridge: Cambridge University Press. Haour, A. and Rossi, B. 2000. Being and Becoming Hausa: Interdisciplinary Perspectives. Leiden and Boston: Brill. Härke, H. 1998. Archaeologists and migrations. Current Anthropology 39.1: 19–45. Härke, H. 2004. The debate on migration and identity in Europe. Antiquity 78: 453–56. Harris, W. V. 2005. Rethinking the Mediterranean. Oxford: Oxford University Press. Hawas, Z. 2000. Valley of the Golden Mummies. Cairo: American University in Cairo Press.

https://doi.org/10.1017/9781108634311.001 Published online by Cambridge University Press

1 Burials, Migration and Identity: The View from the Sahara

Heather, P. 2009. Empire and Barbarians: Migration, Development and the Birth of Europe. London: Macmillan. Hingley, R. 2005. Globalising Roman Culture: Unity, Diversity and Empire. London: Routledge. Hitchner, R. B. 1995. The culture of death and the invention of culture in Roman Africa. Journal of Roman Archaeology 8: 493–98. Hoffman, G. L. 2011. Theory and methodology: study of identities using archaeological evidence from Dura-Europos. In L. R. Brody and G. L. Hoffman (eds), Dura-Europos: Crossroads of Antiquity, Chestnut Hill: McMullen Museum of Art, Boston College, 45–69. Holl, A. 1986. Economie et société néolithique du Dhar Tichitt (Mauritanie). Paris: Editions Hope, V. M. 1997. Words and pictures: the interpretation of Romano-British tombstones. Britannia 28: 245–58. Hope, V. M. 2001. Constructing Identity: The Roman Funerary Monuments of Aquileia, Mainz and Nimes. Oxford: British Archaeological Reports. Horden, P. and Purcell, N. 2000. The Corrupting Sea: A Study of Mediterranean History. Oxford: Blackwell. Hu, D. 2013. Approaches to the archaeology of ethnogenesis: past and emergent perspectives. Journal of Archaeological Research 21.4: 371–402. Hubschmann, C. 2010. Igai: a little-known deity from Dakhleh Oasis. Rosetta 8: 42–61. Ibn Khaldūn. 1958. The Muqaddimah: An introduction to history. Translated from the Arabic by Franz Rosenthal. 3 vols. New York: Princeton. Ikram, S. 2015. Death and Burial in Ancient Egypt. Cairo: American University in Cairo Press. Insoll, T. 2007. The Archaeology of Identities: A Reader. London: Routledge. Isaac, B. 2004. The Invention of Racism in Classical Antiquity. Princeton: Princeton University Press. Jones, S. 1997. The Archaeology of Ethnicity: Constructing Identities in the Past and Present. London: Routledge. Kallala, N. and Sanmartí, J. 2011. Althiburos I. La fouille dans l’aire du capitole et la nécropole méridionale. Taragona: Institut Català d’Arqueologia Clàssica. Kallala, N., Sanmartí, J., Jornet, R., Belarte, M. C., Canela, J., Chérif, S., Campillo, J., Montanero, D., Miniaoui, S., Bermúdez, X., Fadrique, T., Revilla, V., Ramon, J. and Ben Moussa, M. 2014. La nécropole mégalithique de la région d’Athiburos, dans le massif du Ksour (Gouvernorat du Kef, Tunisie). Fouille de trois monuments. Antiquités Africaines 50: 19–60. Kallala, J., Sanmartí, J. (directors) and Belarte, M. C. (ed.). 2018. Althiburos III. La nécropole protohistorique de Althiburos-Massif du Ksour. Taragona: Institut Català d’Arqueologia Clàssica.

https://doi.org/10.1017/9781108634311.001 Published online by Cambridge University Press

41

42

David J. Mattingly et al.

Keenleyside, A., Schwarcz, H., Stirling, L. and Ben Lazreg, N. 2009. Stable isotopic evidence for diet in a Roman and late Roman population from Leptiminus, Tunisia. Journal of Archaeological Science 36: 51–63. Kiethéga, J. B., Sidibé, S. and Bedaux, R. M. A. 1993. Les pratiques funéraires. In J. Devisse (ed.), Vallées du Niger, Paris: Editions de la Réunion des Musées Nationaux, 425–40. Khalaf, N. 2016. A Field Survey and Geographical Information System (GIS) Based Investigation of the Archaeological Landscape in the Niger River Valley, Republic of Benin. Unpublished PhD thesis, University of East Anglia, UK. Klute, G. and Hahn, H. P. (eds). 2008. Cultures of Migration: African Perspectives. Münster: Lit Verlag. Kote, L. 2000. L’archéologie funéraire. In R. Vernet (ed.), L’archéologie en Afrique de l’Ouest. Sahara et Sahel, Nouakchott: Editions Sepia, 51–58. Krandel-Ben Younes, A. 2002. La presence punique en pays numide. Tunis: INP. Krandel-Ben Younes, A. 2013. Les Numides quelle identité Libyque ou punique? Africa 23: 45–59. Kuper, R. and Kropelin, S. 2006. Climate-controlled Holocene occupation in the Sahara: motor of Africa’s evolution. Science 313: 803–07. Lahr, M. M. 2010. Saharan corridors and their role in the evolutionary geography of ‘Out of Africa I’. In J. J. Fleagle, J. J. Shea, F. E. Grine, A. L. Baden and R. E. Leakey (eds), Out of Africa I: The First Hominin Colonization of Eurasia, New York: Springer, 27–46. Laporte, L., Bocoum, H., Cros, J. P., Delvoye, A., Bernard, R., Diallo, M., Diop, M., Kane, A., Dartois, V., Lejay, M., Bertin, F. and Quesnel, L. 2012. Megalithic monumentality in Africa: from graves to stone circles at Wanar, Senegal. Antiquity 86.332: 409–27. Larrasoaña, J. C., Roberts, A. P. and Rohling, E. J. 2013. Dynamics of Green Sahara periods and their role in hominin evolution. PLoS ONE 8.10: e76514. Laurence, R. and Berry, J. 1998. Cultural Identity in the Roman Empire. London: Routledge. Leach, S., Lewis, M. E., Chenery, C., Müldner, G. H. and Eckardt, H. 2009. Migration and diversity in Roman Britain: a multidisciplinary approach to immigrants in Roman York, England. American Journal of Physical Anthropology 140.3: 546–61. Leach, S., Eckardt, H., Chenery, C., and Lewis, M. 2010. A Lady of York: migration, ethnicity and identity in Roman Britain. Antiquity 84.323: 131–45. Leary, J. 2014. Past Mobilities: Archaeological Approaches to Movement and Mobility. Aldershot: Ashgate. Lechtman, H. 1977. Style in technology: some early thoughts. In H. Lechtman and R. S. Merrill (eds), Material Culture: Style, Organization and Dynamics of Technology, New York: West Publishing Co., 3–20. Lewicki, T. 1976/1983. Etudes maghrebiennes et soudanaises, 2 vols. Warsaw: Éditions scientifiques de Pologne.

https://doi.org/10.1017/9781108634311.001 Published online by Cambridge University Press

1 Burials, Migration and Identity: The View from the Sahara

Lhote, H. 1984. Le Hoggar. Espace et temps. Paris: Armand Colin. Lichtenberger, A. 2016. ‘Sea without water’ – Conceptualising the Sahara and the Mediterranean. In M. Dabag, D. Haller, N. Jaspert and A. Lichtenberger (eds), New Horizons. Mediterranean Research in the 21st Century, Paderborn: Ferdinand Schoningh, 267–83. Lihoreau, M. 1993. Djorf Torba, nécropole saharienne antéislamique. Paris: Karthala. Liverani, M. (ed.). 2006. Aghram Nadarif. A Garamantian citadel in the Wadi Tannezzuft. AZA 5. Firenze: All’Insegna del Giglio. Longerstay, M. 1995. Les haouanet. Etat de la question. In Trousset 1995: 33–53. Longerstay, M. 2011. La présence de haouanet definit-elle une identité régionale dans le nord de la Proconsulaire. In C. Briand-Ponsart and Y. Modéran. (eds), Provinces et identités provincials dans l’Afrique romaine, Caen: CRAHM, 189–200. Lucy, S. 2005. Ethnic and cultural identities. In Díaz Andreu et al. 2005: 86–109. Lydon, G. 2009. On Trans-Saharan Trails: Islamic Law, Trade Networks and Crosscultural Exchange in Nineteenth-century Western Africa. Cambridge: Cambridge University Press. MacDonald, K. C., Vernet, R., Martinon-Torres, M. and Fuller, D. Q. 2009. Dhar Ne´ma: From early agriculture to metallurgy in southeastern Mauritania. Azania: Archaeological Research in Africa 44: 3–48. MacEachern, S. 2013. Genetics and archaeology. In Mitchell and Lane 2013: 65–76. McIntosh, S. K. and McIntosh, R. J. 1980. Prehistoric Investigations in the Region of Jenne, Mali: Part I. Oxford: Archeopress. Magnavita, S. 2003. The beads of Kissi, Burkina Faso. Journal of African Archaeology 1.1: 127–38. Magnavita, S. 2008. The oldest textiles from sub-Saharan West Africa: woolen facts from Kissi, Burkina Faso. Journal of African Archaeology 6.2: 243–57. Magnavita, S. 2009. Sahelian crossroads: Some aspects on the Iron Age sites of Kissi, Burkina Faso. In Magnavita et al. 2009: 79–104. Magnavita, S. 2014. 1500 Jahre am Mare de Kissi. Eine Fallstudie zur Besiedlungsgeschichte des Sahel von Burkina Faso. Frankfurt: Africa Magna. Magnavita, S., Koté, L., Breunig, P. and Idé, O. A. (eds). 2009. Crossroads/Carrefour Sahel. Cultural and Technological Developments in First Millennium BC/AD West Africa. Journal of African Archaeology Monograph Series 2. Frankfurt am Main: Africa Magna Verlag. Manning, K. and Timpson, A. 2014. The demographic response to Holocene climate change in the Sahara. Quaternary Science Reviews 101: 28–35. Marshall, E. 1998. Constructing the self and the other in Cyrenaica. In Laurence and Berry 1998: 49–64. Mattingly, D. J. 1999. The art of the unexpected: Ghirza in the Libyan predesert. In S. Lancel (ed.), Numismatique, langues, écriture et arts du livre,

https://doi.org/10.1017/9781108634311.001 Published online by Cambridge University Press

43

44

David J. Mattingly et al.

spécificité des arts figures Afrique du Nord antique et médiévale, Paris: CTHS, 383–405. Mattingly, D. J. 2000. Ammon. In R. Talbert (ed.), Barrington Atlas of the Greek and Roman World, Princeton: Princeton University Press, map 73 and Map-byMap Directory, 1108–16. Mattingly, D. J. (ed.). 2003. The Archaeology of Fazzan. Volume 1, Synthesis. London: Society for Libyan Studies, Department of Antiquities. Mattingly, D. J. 2004. Being Roman: expressing identity in a provincial setting. Journal of Roman Archaeology 17: 5–25. Mattingly, D. J. 2006. An Imperial Possession. Britain in the Roman Empire. London: Penguin. Mattingly, D. J. (ed.). 2007a. The Archaeology of Fazzan. Volume 2, Site Gazetteer, Pottery and other Survey Finds. London: Society for Libyan Studies, Department of Antiquities. Mattingly, D. J. 2007b. The African way of death: burial rituals beyond the Roman Empire. In Stone and Stirling 2007: 138–63. Mattingly, D. J. (ed.). 2010. The Archaeology of Fazzan. Volume 3, Excavations Carried Out by C. M. Daniels. London: Society for Libyan Studies, Department of Antiquities. Mattingly, D. J. 2011. Imperialism, Power and Identity: Experiencing the Roman Empire. Princeton: Princeton University Press. Mattingly, D. J. (ed.). 2013. The Archaeology of Fazzan. Volume 4, Survey and Excavations at Old Jarma (Ancient Garama) Carried Out by C. M. Daniels (1962–69) and the Fazzan Project (1997–2001). London: Society for Libyan Studies, Department of Antiquities. Mattingly, D. J. 2014. Identities in the Roman world: discrepancy, heterogeneity, hybridity and plurality. In Brody and Hoffmann 2014: 32–61. Mattingly, D. J. and Edwards, D. 2003. Religious and funerary structures. In Mattingly 2003: 177–234. Mattingly, D. J., Lahr, M., Armitage, S., Barton, H., Dore, J., Drake, N., Foley, R., Merlo, S., Salem, M., Stock, J. and White, K. 2007. Desert Migrations: people, environment and culture in the Libyan Sahara. Libyan Studies 38: 115–56. Mattingly, D. J., Dore, J. and Lahr, M. M., with contributions by Ahmed, M., Cole, F., Crisp, J., Rodriguez Gonzales, M., Hobson, M., Ismayer, M., Leitch, V., Moussa, F., Nikita, E., Reeds, I., Savage, T. and Sterry, M. 2008. DMP II: 2008 fieldwork on burials and identity in the Wadi al-Ajal. Libyan Studies 39: 223–62. Mattingly, D. J., Lahr, M. M. and Wilson, A., with contributions by Abduli, H., Ahmed, M., Baker, S., Cole, F., Rodriguez Gonzales, M., Hobson, M., Leitch, V., Moussa, F., Nikita, E., Radini, A., Reeds, I., Savage, T. and Sterry, M. 2009. DMP V: investigations in 2009 of cemeteries and related sites on the west side of the Taqallit promontory. Libyan Studies 40: 95–131.

https://doi.org/10.1017/9781108634311.001 Published online by Cambridge University Press

1 Burials, Migration and Identity: The View from the Sahara

Mattingly, D. J., Abduli, H., Aburgheba, H., Ahmed, M., Ali Ahmed Esmaia, M., Baker, S., Cole, F., Fenwick, C., Gonzalez Rodriguez, M., Hobson, M., Khalaf, N., Lahr, M. M., Leitch, V., Moussa, F., Nikita, E., Parker, D., Radini, A., Ray, N., Savage, T., Sterry, M. and Schörle, K. 2010a. DMP IX: Summary report on the fourth season of excavations of the Burials and Identity team. Libyan Studies 41: 89–104. Mattingly, D. J., al-Aghab, S., Ahmed, A., Moussa, F., Sterry, M. and Wilson, A. 2010b. DMP X: Survey and landscape conservation issues around the Taqallit headland. Libyan Studies 41: 105–32. Mattingly, D. J., Abduli, H., Ahmed, M., Cole, F., Fenwick, C., Fothergill, B. T., Gonzalez Rodriguez, M., Hobson, M., Khalaf, N., Lahr, M. M., Moussa, F., Nikita, E., Nikolaus, J., Radini, A., Ray, N., Savage, T., Sterry, M. and Wilson, A. 2011. DMP XII: excavations and survey of the so-called Garamantian Royal Cemetery (GSC030-031). Libyan Studies 42: 89–102. Mattingly, D. J., Leitch, V., Duckworth, C. N., Cuenod, A., Sterry, M. and Cole, F. (eds). 2017a. Trade in the Ancient Sahara and Beyond. Trans–Saharan Archaeology Volume I. Series Editor D. J. Mattingly. Cambridge: Cambridge University Press, Society for Libyan Studies. Mattingly, D. J., Bokbot, Y., Sterry, M., Cuénod, A., Fenwick, C., Gatto, M., Ray, N., Rayne, L., Janin, K., Lamb, A., Mugnai, N. and Nikolaus, J. 2017b. Long-term history in a Moroccan oasis zone: the Middle Draa Project 2015. Journal of African Archaeology 15.2: 141–72. Mattingly, D. J., Sterry, M. and Fothergill, B. T. Forthcoming. Animal traffic in the Sahara. In V. Blanc-Bijon (ed.), XIe Colloque international “Histoire et Archéologie de l’Afrique du Nord”, Aix-en-Provence. Mayor, A., Huysecom, E., Ozainnea, S. and Magnavita, S. 2014. Early social complexity in the Dogon Country (Mali) as evidenced by a new chronology of funerary practices. Journal of Anthropological Archaeology 34: 17–41. Merrills, A. and Miles, R. 2010. The Vandals. Blackwell Peoples of Europe. Oxford, Malden, MA: Wiley-Blackwell. Meskell, L. 1999. Archaeologies of life and death. American Journal of Archaeology 103.2: 181–99. Meskell, L. 2001. Archaeologies of identity. In I. Hodder (ed.), Archaeological Theory Today, Cambridge: Polity Press, 187–213. Milburn, M. 1981. Multi-arm tombs of the Central Sahara. Antiquity 55: 210–14. Milburn, M. 1983. On the keyhole tombs (‘monuments en trou de serrure’) of Central Sahara. Libya Antiqua 13–14 (1976–1977) [1983]: 385–90. Milburn, M. 1993. Saharan stone monuments, rock pictures and artefact contemporaneity: some suggestions. In G. Calegari (ed.), L’arte el’ambiente del Sahara preistorico, Milan: Memorie della Società Italiana di Scienze Naturali, 363–74. Milburn, M. 1996. Some recent dates for central and southern Sahara including monuments. Sahara 8: 99–103.

https://doi.org/10.1017/9781108634311.001 Published online by Cambridge University Press

45

46

David J. Mattingly et al.

Mitchell, P. and Lane, P. 2013. The Oxford Handbook of African Archaeology. Oxford: Oxford University Press. Modéran, Y. 2003. Les Maures et l’Afrique romaine IVe-VIIe siècle. Bibliothèque des Ecoles françaises d’Athènes et de Rome 314. Rome: Ecole française de Rome. Modéran, Y. and Perrin, M. Y. 2014. Les Vandales et l’Empire romain. Paris: Errance – Actes Sud. Mohammed, F. A. 1998. El Jaghbub. In E. Catani and S. Maria Marengo (eds), La Cirenaica in età antica, Macerata: 263–73. Moore, J. P. 2007. The ‘mausoleum culture’ of Africa Proconsularis. In Stone and Stirling 2007: 75–109. Mori, L. (ed.). 2013. Life and Death of a Rural Village in Garamantian Times. Archaeological Investigations in the Oasis of Fewet (Libyan Sahara). AZA 6. Firenze: All’Insegna del Giglio. Mouritsen, H. 2011. The Freedman in the Roman World. Cambridge: Cambridge University Press. Mugnai, N., Nikolaus, J. and Ray, N. 2016. De Africa Romaque. Merging Cultures across North Africa. London: Society for Libyan Studies. Munson, P. J. 1980. Archaeology and the prehistoric origins of the Ghana Empire. Journal of African History 21: 457–66. Nikolaus, J. 2016. Beyond Ghirza: Roman-period mausolea in Tripolitania. In Mugnai et al. 2016: 199–214. O’Shea, J. M. 1984. Mortuary Variability: an Archaeological Investigation. Orlando: Academic Press. Osborne, R. 2008. Introduction: for tradition as an analytical category. World Archaeology 40.3: 281–94. Paris, F. 1984. La région d’In Gall – Tegidda-n Tesemt (Niger). Programme archéologique d’urgence 1977–1981. III. Les sépultures du néolothique final à l’Islam. Niamey : Institut de Recherches en Sciences Humaines. Paris, F. 1990. Les sépultures monumentales d’Iwelen (Niger). Journal des Africanistes 60.1: 47–75. Paris, F. 1995. Essai de classification des monuments funéraires sahariens. Bulletin de la Société Préhistorique Française 92.4: 549–53. Paris, F. 1996. Les sépultures du Sahara nigérian, du Néoloithique à l’Islamisation. 2 vols, Paris: Bondy. Parker-Pearson, M. 1999. The Archaeology of Death and Burial. Stroud: Sutton. Parker-Pearson, M. and Ramilisonina 1998. Stonehenge for the ancestors: the stones pass on the message. Antiquity 72.276: 308–26. Pearce, J. 2010. Burial, identity and migration in the Roman provinces. In Eckardt 2010a: 79–98. Pearce, J. 2011. Marking the dead: tombs and topography in the Roman provinces. In Caroll and Rempel 2011: 134–58. Pearce, J. 2013. Contextual Archaeology of Burial Practice: Case-Studies from Roman Britain. Oxford: Archaeopress.

https://doi.org/10.1017/9781108634311.001 Published online by Cambridge University Press

1 Burials, Migration and Identity: The View from the Sahara

Pearce, J. 2015. Urban exits: commercial archaeology and the study of death rituals and the dead in the towns of Roman Britain. In M. Fulford and N. Holbrook (eds), The Towns of Roman Britain. The Contribution of Commercial Archaeology since 1990, London: Society for Roman Studies, 138–66. Pearce, J., Millett, M. and Struck, M. 2000. Burial, Society and Context in the Roman Empire. Oxford: Oxbow. Petit-Maire, N. and Riser, J. (eds). 1983. Sahara ou Sahel? Quaternaire récent du Bassin de Taoudenni (Mali). Marseille: Centre National de la Recherche Scientifique. Pfaffenberger, B. 1992. Social Anthropology of Technology. Annual Review of Anthropology 21: 491–516. Philpott, R. 1991. Burial Practices in Roman Britain: a Survey of Grave Treatment and Furnishings, AD 43–410. Oxford: Archaeopress. Pitts, M. 2007. The Emperor’s new clothes? The utility of identity in Roman archaeology. American Journal of Archaeology 111.4: 693–713. Pitts, M. and Versluys, M. J. (eds). 2015. Globalisation and the Roman World. World History. Connectivity and Material Culture. Cambridge: Cambridge University Press. Pohl, W. and Reimitz, G. (eds). 1998. Strategies of Distinction. The Construction of Ethnic Communities 300–800. Leiden: Brill. Porter, A. 2012. Mobile Pastoralism and the Formation of Near Eastern Civilizations. Cambridge: Cambridge University Press. Prag, J. and Quinn, J. 2013. The Hellenistic West. Rethinking the Ancient Mediterranean. Oxford: Oxford University Press. Rakob, F. 1979. Numidische Königsarchitektur in Nordafrika. In H. G. Horn and C. B. Ruger (eds), Die Numider: Reiter und Könige nördlich der Sahara, Bonn, 119–71. Ralston, I. 2002. Exploring the secret of the gebel. Nekhen News 14: 19–20. Rebillard, E. 2012. Christians and Their Many Identities in Late Antiquity: North Africa 200–450 CE. Ithaca: Cornell University Press. Reimer, P. J., Bard, E., Bayliss, A., Beck, J. W., Blackwell, P. G., Bronk Ramsey, C., Grootes, P. M., Guilderson, T. P., Haflidason, H., Hajdas, I., Hatté, C., Heaton, T. J., Hoffmann, D. L., Hogg, A. G., Hughen, K. A., Kaiser, K. F., Kromer, B., Manning, S. W., Niu, M., Reimer, R. W., Richards, D. A., Scott, E. M., Southon, J. R., Staff, R. A., Turney, C. S. M., and van der Plicht, J. 2013. IntCal13 and Marine13 Radiocarbon Age Calibration Curves 0–50,000 Years cal BP. Radiocarbon 55.4: 1869–87. Renfrew, C. 1987. Archaeology and Language: The Puzzle of Indo-European Origins. London: Jonathan Cape. Renfrew, C. 2000. Archaeogenetics: towards a population prehistory of Europe. In K. Boyle (ed.), Archaeogenetics: DNA and the Population Prehistory of Europe, Cambridge: McDonald Institute for Archaeological Research, 3–11.

https://doi.org/10.1017/9781108634311.001 Published online by Cambridge University Press

47

48

David J. Mattingly et al.

Reygasse, M. 1950. Monuments funéraires préislamiques de l’Afrique du Nord. Paris: Arts et Métiers Graphiques. Roset, J. P. 1986. Iwelen – an archaeological site of the chariot period in Northern Aïr, Niger. In Libya Antiqua Paris: UNESCO, 13–46. Rouse, I. 1986. Migrations in Prehistory. New Haven, London: Yale University Press. Salvatori, S. and Usai, D. (eds). 2008. A Neolithic Cemetery in the Northern Dongola Reach. Excavations at Site R12. SARS 16. Oxford: Archaeopress. Sanmartí, J., Kallala, N., Belarte, M. C., Ramon, J., Telmini, B. M., Jornet, R. and Miniaoui, S. 2012. Filling gaps in the protohistory of the eastern Maghreb: the Althiburos Archaeological Project (el Kef, Tunisia). Journal of African Archaeology 10.1: 21–44. Savary, J. P. 1966. Monuments en pierres sèches du Sadnoun (Tassili n’Ajjer). Paris: Mémoires du centre de recherches anthropologiques préhistoriques et ethnographiques. Saxe, A. 1970. Social Dimensions of Mortuary Practices. Unpublished PhD thesis. University of Michigan. Sellet, F. 1993. Châine opératoire: the concept and its applications. Lithic Technology 18.1–2: 106–12. Sen, A. 2006. Identity and Violence: The Illusion of Destiny. New York: Norton. Sereno, P. C., Garcea, E. A. A., Jousse, H., Stojanowski, C. M., Saliège, J. F., Maga, A., Ide, O. A., Knudson, K. J., Mercuri, A. M., Stafford, T. W. Jr., Kaye, T. G., Giraudi, C., N’siala, I. M., Cocca, E., Moots, H. M., Dutheil, D. B. and Stivers, J. P. 2008. Lakeside cemeteries in the Sahara: 5000 years of Holocene population and environmental change. PLoS ONE 3.8: e2995. doi:10.1371/journal.pone.0002995. Sidibé, S. 1980. Archéologie funéraire de l’ouest africain: sépultures et rites. Ph. D. thesis, Université de Paris (Sorbonne), Paris. Silverman, H. and Small, D. B. (eds). 2002. The Space and Place of Death. Arlington: Archaeological Papers of the American Anthropological Association Number 11. Sindbaek, S. M. 2007. The small world of the Vikings: networks in early medieval communication and exchange. Norwegian Archaeological Review 40.1: 59–74. Smith, S. T. 2003. Wretched Kush: Ethnic Identities and Boundaries in Egypt’s Nubian Empire. London, New York: Routledge. Souville, G. 1959. Principaux types de tumulus marocains. Bulletin de la Société Préhistorique de France 56: 394–402. Stark, M. T. 1999. Social dimensions of technical choice in Kalinga ceramic traditions. In E. Chilton (ed.), Material Meanings: Critical Approaches to Interpreting Material Culture, Salt Lake City: University of Utah Press, 24–43. Stark, M. T. 2006. Glaze ware technology, the social lives of pots, and communities of practice and in the Late Prehistoric Southwest. In J. Habicht–Mauche, S. Eckert and D. Huntley (eds), The Social Life of Pots: Glaze Wares and

https://doi.org/10.1017/9781108634311.001 Published online by Cambridge University Press

1 Burials, Migration and Identity: The View from the Sahara

Cultural Transformation in the Late Pre-contact Southwest, Tucson: University of Arizona, 17–33. Stark, M. T., Bowser, B. J. and Horne, L. 2008. Why breaking down boundaries matters for archaeological research on cultural transmission: an introduction. In M. Stark, B. J. Bowser and L. Horne (eds), Cultural Transmission and Material Culture: Breaking Down Boundaries, Tucson: University of Arizona Press, 1–16. Stirling, L., Mattingly, D. J. and Ben Lazreg, N. (eds). 2001. Leptiminus (Lamta): Report no. 2, The East Baths, Cemeteries, Kilns, Venus Mosaic, Site Museum and Other Studies. Portsmouth, RI: JRA Suppl. Series. Stone, D. L. 2007a. Monuments on the margins: interpreting the Protohistoric rock–cut tombs (Haouanet) of North Africa. In Stone and Stirling 2007: 43–74. Stone, D. L. 2007b. Burial, identity, and local culture in North Africa. In P. Van Dommelen and N. Terrenato (eds), Articulating Local Cultures: Power and Identity under the Expanding Roman Republic, Journal of Roman Archaeology Supplementary Series 63, 126–44. Stone, D. L. 2016. Burial mounds and state formation in North Arica: a volumetric and energetic approach. In Mugnai et al. 2016: 39–51. Stone, D. L. and Stirling, L. (eds). 2007. Mortuary Landscapes of North Africa. Toronto: University of Toronto. Stone, D. L., Mattingly, D. J. and Ben Lazreg, N. (eds). 2011. Leptiminus (Lamta): A Roman Port City in Tunisia, Report no. 3, the Urban Survey. Portsmouth, RI: JRA Suppl. Series. Tanda, G., Ghaki, M. and Cicilloni, R. (eds). 2009. Storia dei paesaggi preistorici e protostorici nell’alto tell tunisino. Missioni 2002–2003. Cagliari: Edizioni AV. Tarlow, S. and Nilsson Stutz, L. 2013. The Oxford Handbook of the Archaeology of Death and Burial. Oxford: Oxford University Press. Thiry, J. 1995. Le Sahara libyen dans l’Afrique du nord médiéval. Leuven: Peeters. Thomas, J. 2013. Monumental architecture in sub-Saharan Africa: a European perspective. Azania: Archaeological Research in Africa 48.2: 315–22. Thorn, J. C. 2005. The Necropolis of Cyrene: Two Hundred Years of Exploration. Rome: L’Erma di Bretschneider. Trigger, B. G. 1997. A History of Archaeological Thought. Cambridge: Cambridge University Press. Trousset, P. (ed.). 1995. Monuments funéraires, institutions autochtones en Afrique du nord antique et médiévale. Actes du VIe colloque international sur l’histoire et l’archéologie de l’Afrique du Nord (Paris), Paris: Edition du CTHS. Ucko, P. 1969. Ethnography and archaeological interpretation of funerary remains. World Archaeology 1.2: 262–80 Ucko, P. J. and Layton, R. 1999. The Archaeology and Anthropology of Landscape: Shaping Your Landscape. London: Routledge. Usai, D., Salvatori, S., Iacumin, P., Di Matteo, A., Jakob, T. and Zerboni, A. 2010. Excavating a pre-Mesolithic cemetery in Central Sudan. Antiquity 84.323 Project Gallery: https://www.antiquity.ac.uk/projgall/usai323/.

https://doi.org/10.1017/9781108634311.001 Published online by Cambridge University Press

49

50

David J. Mattingly et al.

Versluys, M. J. 2014. Understanding objects in motion. An archaeological dialogue on Romanization. [and responses]. Archaeological Dialogues 21.1: 1–64. Voss, B. 2008. Domesticating imperialism: sexual politics and the archaeology of empire. American Anthropologist 110.2: 191–203. Wallace-Hadrill, A. 2008. Rome’s Cultural Revolution. Cambridge: Cambridge University Press. Wengrow, D., Dee, M., Foster, S., Stevenson, A. and Bronk Ramsey, C. 2014. Cultural convergence in the Neolithic of the Nile Valley: a prehistoric perspective on Egypt’s place in Africa. Antiquity 88.339: 95–111. Wendorf, F. and Schild, R. (eds). 2001. Holocene Settlements of the Egyptian Sahara, vol. 1, The Archaeology of Nabta Playa. New York: Kluwer Academic, Plenum. Whitley, J. 2002. Too many ancestors. Antiquity 76: 119–26. Williams, H. 2003. Archaeologies of Remembrance: Death and Memory in Past Societies. New York, London: Kluwer Academic. Williams, H. 2006. Death and Memory in Early Medieval Britain. Cambridge: Cambridge University Press. Witcher, R. 2000. Globalisation and Roman imperialism: perspectives on identities in Roman Italy. In E. Herring and K. Lomas (eds), The Emergence of State Identities in Italy in the First Millennium BC. London: Accordia Research Institute, University of London, 213–25. Woolf, G. 1997. Becoming Roman: The Origin of Provincial Civilisation in Gaul. Cambridge: Cambridge University Press. Wynne-Jones, S. 2013. Material culture, space and identity. In Mitchell and Lane 2013: 177–88. Zaïnabi, A. T. (ed.). 2004. Trésors et merveilles de la vallée du Drâa. Zagora: Ed Marsam.

https://doi.org/10.1017/9781108634311.001 Published online by Cambridge University Press

part i

Burial Practices in the Central Sahara

Published online by Cambridge University Press

Published online by Cambridge University Press

2

Dying to Be Garamantian Burial, Migration and Identity in Fazzan david j. mattingly, martin sterry and nick ray

Introduction The previous chapter introduced the Garamantes, an ancient Saharan people, whose story can be traced archaeologically through the first millennium BC and first millennium AD (see also Figs 1.1 and 1.2 for location maps). This chapter presents evidence on the funerary practices of the Garamantian heartlands in the Wadi al-Ajal in the Libyan Fazzan (Central Sahara). The Wadi al-Ajal is a long and thin oasis depression running for c.150 km from al-Abyad (to the south-west of Fazzan’s modern capital at Sabha) to Ubari. Our work has been particularly focused on the area around Jarma (ancient Garama, the Garamantian capital about 40 km east of Ubari). The burials in these Garamantian heartlands differ in certain significant ways from those recorded by the Italian mission at Aghram Nadharif and Fewet, and an interesting aspect of the discussion we shall develop below seeks to explain this difference.1 Our chronological framework for the Garamantes is presented in Table 2.1. One important point to note here in relation to this table is that previously we have set the division between the Classic and Late Garamantian periods around AD 400, but consideration of the pattern of ceramic imports to Fazzan – which saw something of a hiatus in the later third century and a revival in the early fourth century – suggests that a start date for the late Garamantian period c.AD 300 may be more appropriate.2 This also corresponds with the start of a phase of significant construction of fortified sites in the Garamantian heartlands.3 The first phase of the British-Libyan work (the Fazzan Project or FP, 1997–2002) focused on baseline survey of the cemetery sites, leading to the creation of a detailed typology of burial monuments and a proposed morphology of funerary zones and the compilation of a site gazetteer 1

2

Compare Mattingly 2003; 2007a; 2007b; 2010; 2013a; with Castelli et al. 2005; Liverani 2006; Mori 2013; Gatto et al., Chapter 3 and Ricci et al., Chapter 5, this volume. See most recently Mattingly 2013a, 130–34. 3 Sterry and Mattingly 2013.

https://doi.org/10.1017/9781108634311.002 Published online by Cambridge University Press

53

54

David J. Mattingly, Martin Sterry and Nick Ray

Table 2.1. Dating periods for the Garamantes with the division between Classic and Late Garamantian modified to reflect a hiatus in the imports of ceramic goods from the Roman world in the third century. Code

Period

Dates BC/AD

LPAST EGAR PUGAR CGAR LGAR POSTGAR

Late Pastoral Early Garamantian Proto-Urban Garamantian Classic Garamantian Late Garamantian Post-Garamantian

3000–1000 BC 1000–500 BC 500–1 BC AD 1–300 AD 300–700 AD 700–1100

covering the Wadi al-Ajal, with some outlying burials known in neighbouring areas.4 As part of this project, earlier excavations of burials by Charles Daniels at a number of Garamantian sites were brought to publication – notably Zinkekra, Saniat bin Huwaydi and the so-called Royal Cemetery.5 Subsequently, more detailed survey of cemeteries and targeted excavations have taken place as one element of the Desert Migrations Project (DMP), 2007–2011.6 This has seen c.11,000 tombs surveyed in the Wadi al-Ajal area, and over 150 have been excavated or had material sampled from them, dating from the Proto-Urban Garamantian period through to the Post-Garamantian period (c.500 BC–AD 900). The locations of the cemeteries where excavation took place are mapped in Figure 2.1. Across five seasons of fieldwork, the DMP excavated 134 tombs and surveyed many distinct cemetery zones in four areas along a 20-km stretch of escarpment from Jarma to Taqallit (ZIN–Zinkekra, UAT–Watwat, GSC–Jarma escarpment and TAG–Taqallit peninsular); these predominantly date to between the fourth century BC and the tenth century AD. The tombs excavated and surveyed by the DMP add significantly to the 77 previously excavated by Daniels and Ayoub, and the 102 excavated by Caputo in the Wadi al-Ajal.7 Of particular importance among these, on account of the unrobbed state of a group of richly furnished tombs, is the cemetery of Saniat bin Huwaydi (GER011) in the oasis centre close to Jarma.8

4 5 7

8

Mattingly 2007a (gazetteer); 2007b; Mattingly and Edwards 2003 (for the burial typology). Mattingly 2010. 6 Mattingly et al. 2007; 2008; 2009; 2010a; 2010b; 2011. Ayoub 1967a; 1967b; Caputo 1937; 1949; el-Rashedy 1988; Mattingly 2010; Pace et al. 1951, 241–406. Mattingly 2010, 213–342.

https://doi.org/10.1017/9781108634311.002 Published online by Cambridge University Press

2 Dying to Be Garamantian: Burial, Migration and Identity in Fazzan

Figure 2.1. Location map of tombs and cemeteries excavated by the DMP. (Imagery ©Esri)

Throughout this chapter, all tombs are referred to in a standard way consisting of a three-letter code indicating the archaeological region, a three-digit code indicating the site or cemetery within that region and a further three-digit code prefaced by ‘T’ indicating the tomb or burial within that cemetery.9 9

Thus GSC031.T008 is the eighth tomb of cemetery GSC031 on the Jarma Escarpment. Tomb numbers were assigned on an arbitrary basis as and when they were required – all excavated tombs were assigned a number as were all surveyed tombs with notable features (for example, associated stelae or offering tables). For explanation of the site codes and T numbers, see Mattingly 2007a, 3–6; Mattingly et al. 2007, 137–45.

https://doi.org/10.1017/9781108634311.002 Published online by Cambridge University Press

55

56

David J. Mattingly, Martin Sterry and Nick Ray

In this chapter, we report on our analysis of the funerary structures, ritual practices and rites, grave assemblages and funerary furniture. We also present some results from our analysis of the spatial patterning of the Wadi al-Ajal data. The mapping of funerary behaviour has helped identify several patterns linking different elements, such as the strong association between tomb furniture with more monumental tomb types. Furthermore, the distributions of types of grave goods have been mapped, demonstrating, for example, that glass items are most heavily concentrated close to Jarma, while African Red Slip fine ware is more widespread. The ceramic evidence also suggests a west-to-east shift in burial practice, with Late Roman pottery being more prevalent at the eastern end of the Wadi. A key aim of our recent work has been to explore the issues of human migration and identity formation/presentation within Garamantian and post-Garamantian oasis-farming communities. This chapter, complemented by Chapter 4 which looks at the biomarkers of identity, addresses primarily those aspects of the burial record that relate to cultural identity in the Garamantian heartlands.10 We are seeking to understand the burial practices in relation to a longestablished Saharan cultural context, but with significant changes evident in the Garamantian heartlands that seem to relate in part to increase in migration and trade contact with other regions. In a recent paper, Fentress and Wilson have suggested that there was a significant upturn of Saharan cultural impact in the Maghribian region in late antiquity.11 We think that one of the key issues to discuss is whether such influences also occurred much earlier and more recurrently across the first millennium BC and first millennium AD, as the evidence of trade suggests.12 The evidence assembled demonstrates variation in tomb architecture and changes in associated material culture, suggesting a culturally and ethnically diverse population in Fazzan. Despite this, some elements of the burial practices remained constant, indicating the integration of different groups or individuals into Garamantian culture. The mix of Saharan and exterior identity traits is a notable feature of Garamantian ‘identity’ and this appears to correlate with both a degree of migration and long-range contacts linked to trade.

10 11

For the work of the Cambridge group, see Power et al., Chapter 4, this volume. Fentress and Wilson 2016. 12 Mattingly et al. 2017a.

https://doi.org/10.1017/9781108634311.002 Published online by Cambridge University Press

2 Dying to Be Garamantian: Burial, Migration and Identity in Fazzan

Typologies The DMP work has broadly confirmed the robustness of the typologies of tombs, cemeteries and funerary furniture devised by the FP, with only minor additions to these being made.13 Understanding the broader implications of the complex typology for chronology and social identity provided one of the main objectives of the DMP. The majority of the most visible cemeteries and funerary zones were located on the fringes of the oasis landscape – in the Wadi al-Ajal this led to their dense placement along the steep escarpment that marked the southern edge of the oasis depression. Burials in this location often made use of the natural stone clutter on the escarpment for construction materials. There were additional cemeteries in the oasis centre close to the Garamantian settlements, but these have often been somewhat obliterated by later oasis cultivation and are more difficult to identify. Tombs in the valley centre were mainly built in mudbrick. There is overlap in the styles of monument found in both zones, with the main difference being the construction materials. Some differences may also have a chronological significance as the earliest cemeteries seem to have been constructed only along the escarpment. Our burial typology recognised nine main categories of monument based on shape, construction and complexity (Figs 2.2–2.3): Type 1 comprised different types of cairn from simple mounds of stones (1a); more complex domed structures, often with a corbelled superstructure (1b); and simple cairns defined by a low outer kerb of larger blocks (1c). The FP Type 1d ‘crater cairn’ with an indented top was suspected on further examination to be an artefact of robbing of Type 1a cairns and was not employed in the DMP work. In addition, the DMP identified another type of simple cairn, where stone was piled up against a large rock (Type 1e, crevice cairn). From the initial FP survey work, these types were in general assumed to be early in date (broadly first millennium BC, with some simple cairns and antenna tombs – see below – being potentially earlier still in the Late Pastoral era). In the Wadi Tanzzuft, c.400 km south-west of Jarma, the Late Pastoral simple cairns were often quite large in diameter and contained multiple interments, whereas by the Garamantian era the norm appears to have been single interments in each monument (though with some exceptions, see below).14 Some examples of the corbelled cairn (Type 1b) occurred in denser 13

14

Mattingly 2003, 187–217; 2007a, 6–7; compare to Mattingly et al. 2008, 230–33; 2009, 97–101; 2010b, 109–11, for minor modifications to the typology. Di Lernia and Manzi 2002.

https://doi.org/10.1017/9781108634311.002 Published online by Cambridge University Press

57

58

David J. Mattingly, Martin Sterry and Nick Ray

1a

1b

1c

1d

1a/d

1b

1e

1c

2a

2b

2c

2b

6a

6c

Figure 2.2. Typology of cairns, shafts and other simple burials.

https://doi.org/10.1017/9781108634311.002 Published online by Cambridge University Press

6b

6c

2 Dying to Be Garamantian: Burial, Migration and Identity in Fazzan

3a

3b

4a

3a

59

4b

4b

3b 5a

5a

5b

5b 5b

Pyramid tomb

Mausoleum

Figure 2.3. Typology of more complex tombs and monumental burials.

https://doi.org/10.1017/9781108634311.002 Published online by Cambridge University Press

60

David J. Mattingly, Martin Sterry and Nick Ray

clusters and were associated with imported Roman pottery of the early centuries AD, and we now think the form probably continued into at least the mid-first millennium AD (see below). Type 2 comprised shaft burials, generally circular in shape, either simple holes (2a) or with stone slab linings (2b) and occasionally with traces of some sort of original stone capping. These tend to be organised in very tight-packed cemeteries and have frequently been robbed, with relatively abundant Roman pottery indicating that many date to the Classic or later Garamantian periods. Some examples at Zinkekra, excavated by Daniels, did not have associated pottery and were thus thought to perhaps originate a little earlier in the late Proto-Urban Garamantian period. The end date for the shaft burial has not hitherto been established, but can now be put near the end of the first millennium AD (see below). Type 3 comprised drum-shaped monuments with vertical-built outer walls. The drum cairn (3a) was generally fairly low, with only three to four courses of rough blockwork ( −8‰), indicating a high intake of C4 plants in childhood. A high dietary input of C4 could result from an individual moving to Fazzan from areas with a higher preponderance of C4 grasses, such as Sub-Saharan Africa. Alternatively, it could also result from the consumption of locally cultivated C4 crops within Fazzan, such as 39

40 41

Brothwell 1981; Buikstra and Ubelaker 1994; Howells 1973, 1989; Krogman and İșcan 1986; Lahr 1992, 1996; Schwartz 1995. Nikita et al. 2012b. Also noted for the Fewet population; see Ricci et al., Chapter 5, this volume.

https://doi.org/10.1017/9781108634311.004 Published online by Cambridge University Press

145

146

Ronika K. Power et al.

–2

δ13OCO3-PDB (‰)

–4

–6

–8

–10

–12

–6

–4

–2

0

2

δ18OCO3-PDB (‰)

Figure 4.2. Plot of tooth carbonate oxygen and carbon isotopic data from Garamantian burials.

sorghum. The assessment of the diversity of both carbon and oxygen isotopic values in the analysed population is critical to discerning between these possibilities. We can compare the oxygen isotopic results of the Garamantes to other contemporaneous populations across Europe and Africa,42 with the strong caveat that some of these are so geographically and temporally distant as to make direct comparisons tenuous. Figure 4.3 shows a comparison of the Garamantian individuals’ oxygen isotopic values to other data from Africa, the Levant and the Mediterranean obtained from various published studies. Figure 4.3a is a comparison of data from bone and tooth enamel, whilst Figure 4.3b presents data obtained solely from tooth enamel, and Figure 4.3c those from tooth enamel from individuals who lived between 1500 BC and AD 1300. Comparisons with other tissues, such as hair, have been avoided due to the complications of isotopic ‘conversions’ of one tissue to another. Comparative samples span a 20,000-year period, from the Epipalaeolithic to modern times. Whilst the assessment of data that cover such a wide timeframe and based on categorisation of modern 42

Lightfoot and O’Connell 2016.

https://doi.org/10.1017/9781108634311.004 Published online by Cambridge University Press

4 Human Mobility and Identity in Relation to the Garamantes

147

(a)

δ18OCO3-SMOW (‰)

36

32

28

ly Ita

ke

y

el

Tu r

Jo

G

C

Is

rd

da Su

ra

an

n

t yp

n

Eg

oo

am er

ge Al

ar

am an

te

s

ria

24

(b)

δ18OCO3-SMOW (‰)

32

28

ly Ita

el ra Is

Jo

Su

rd

da

an

n

n er

oo

ria ge

am

G

C

ar

am

Al

an

te

s

24

(c)

ly

Ita

el

ra

Is

an

rd

Jo

da n

Su

G

ar

am

an

te

s

δ18OCO3-SMOW (‰)

Figure 4.3. Comparison of Garamantian tooth carbonate oxygen isotopic data to other published 32 data from Africa, the Levant and the Mediterranean: a) data are from bone and tooth enamel, converted to the carbonate equivalent, and reported on the 28 VSMOW scale; b) data from tooth enamel only, converted to the carbonate equivalent, and reported on the VSMOW scale; c) data from tooth enamel 24 only, from sites dating between 1500 BC to AD 1300, converted to the carbonate equivalent, and reported on the VSMOW scale. Phosphate oxygen isotopic values were converted to carbonate oxygen isotopic equivalents using the equation of Iacumin et al. 1996. Carbonate oxygen isotopic values on the VPDB scale were converted to the VSMOW scale using the equation in Coplen et al. 1983 (data are taken from Buzon and Bowen 2010; Daux et al. 2008; Diaz et al. 2012; Levinson et al. 1987; Mitchell and Millard 2009; Perry et al. 2009; Prowse et al. 2007 and White et al. 2004).

https://doi.org/10.1017/9781108634311.004 Published online by Cambridge University Press

148

Ronika K. Power et al.

geopolitical states makes it necessarily broad, some interesting features emerge. In Figure 4.3a, the range of δ18O values observed among the Garamantes is large, and covers most of that found within Egypt (Daklah Oasis, Asyut and Gebelein), but also in Italy (predominantly data from the classical Roman site of Isola Sacra), Jordan and Israel. The eight Garamantian teeth with the highest oxygen isotopic values (δ18Ocarb PDB > -1‰, equivalent to δ18Ocarb SMOW > 30‰) fall within the range of contemporaneous individuals from Sudan (Fig. 4.3a/b/c). Comparison of the oxygen isotopic range within the data (9.5‰) to that seen in other populations shows that despite being large, it is comparable to what is observed in other African regions. There are two main explanations for this variation – diachronic differences and geography. While diachronic isotopic variation in water intake is a possibility, we have limited information as to how this applies to our analysed population, both in terms of the timespan represented by the individuals, and the rainfall regime over the relevant time period; alternatively, that particular group of individuals may represent people from a different geographic origin. Differing access to different water sources, such as shallow wells, artesian springs and foggaras, may have an effect, as has been explored in other locations.43 Figure 4.3c highlights the lacuna of comparative contemporaneous tooth enamel analyses from across North Africa, a gap which we are in the process of addressing. Further nuanced interpretation is contingent on the comparative samples currently being analysed in the context of variation in cranial shape and size.

How Heterogenous Was the Garamantes Population from the Wadi al-Ajal? The stable carbon and oxygen isotopic data suggest that some individuals buried in the cemeteries of the Wadi al-Ajal were exposed to a different childhood diet than the majority of the Garamantes people. This raises two questions: firstly, were these individuals who may, in terms of their diet, be called ‘Garamantian outliers’, concentrated in particular cemeteries? Secondly, did they look sufficiently different from other Garamantes to suggest that they originated from a different population? These two questions are explored here.

43

Lightfoot et al. 2014.

https://doi.org/10.1017/9781108634311.004 Published online by Cambridge University Press

4 Human Mobility and Identity in Relation to the Garamantes

Among the Garamantian samples studied, several different cemeteries are represented from several distinct locations – Zinkekra (ZIN), Taqallit (TAG), Watwat (UAT), as well as a few individuals from the so-called Royal Cemeteries (GSC030 and GSC031) and other sites in the Jarma escarpment area (GSC042) (Fig. 4.4). The distribution of δ13C and δ18O results per cemetery reveals an interesting pattern. The majority of the sample, including all the individuals buried at the cemeteries of Zinkekra, Taqallit, as well as three individuals buried at Watwat, cluster tightly, while all individuals whose δ13C signature indicates a childhood diet rich in C4 plants, different from the majority of the sample, were buried at the Watwat cemeteries. These burials date to the second century AD or later, at a time when the cultivation of sorghum and pearl millet as summer crops was well established at Jarma, while a large proportion, but by no means all, of the ‘normative’ Wadi al-Ajal burials of lower C4 plant intake date to the late first millennium BC. We thus cannot exclude the possibility that the pattern reflects chronological change in local diet, but nor can we ignore the alternative interpretation that the people buried in the Watwat shaft burials included a group of migrants from the Sub-Saharan area. The group of individuals whose δ18O values suggest a different water intake in childhood were mostly, but not exclusively buried in cemeteries close to Jarma. Several of these burials were of more elite status than the Watwat shaft burials (GSC042, GSC030, GER011 and UAT004.T1). This would suggest that not all the outliers in the isotope results (potential incomers) were of low status. Unfortunately, only a small number of the individuals (n = 9) sampled for isotopic signatures had crania that were sufficiently complete to be included in a multivariate morphometric analysis. Of these, three correspond to individuals whose δ18O indicates a different water intake in childhood, but those whose childhood diet differed from the majority are not represented. With this caveat in mind, a preliminary analysis of the morphometric variation within the sample was carried out. For this purpose, cranial landmark coordinates were used to calculate pairwise linear distances between these landmarks and subsequently these distances were analysed by means of Principal Components Analysis. A PCA of 28 crania obtained eight significant factors (PCs) that together explain 85.3 per cent of the variance. The clustering of cases along the first two PCs (Fig. 4.5) explains nearly 50 per cent of the variance in the entire matrix. The first vector (PC1, X axis) is defined morphometrically as size, matching the recurrent cranial definition of size as consisting of measures of length of the cranium and breadth of the upper face; breadth

https://doi.org/10.1017/9781108634311.004 Published online by Cambridge University Press

149

150

Ronika K. Power et al.

0

–8

–7

–6

–5

–4

–3

–2

–1

0

1

2

3

UAT008.T87

–2 High C4 grasses in childhood

UAT008.T5

–4 UAT008.T16

UAT010.T5

δ13O

–6 GER011.T24

UAT008.T82

–8

Different childhood water intake

TAG001.T231

GSC030.T4

–10 UAT008.T32

–12

UAT004.T1

TAG063.T20 TAG001.T239

TAG001.T258

TAG006.T2 ZIN019.T4 ZIN220.T18 ZIN13.T170 GSC031.T30 UAT008.T22 ZIN13.T204 ZIN220.T2 ZIN013.T44 UAT008.T38 TAG006.T1 ZIN350-T25 UAT010.T3

–14

GSC042.T1

ZIN218.T1 ZIN220.T3 ZIN220.T1

δ18O

Figure 4.4. Scatterplot of δ13C/δ18O by cemetery.

of the vault does not contribute to ‘size’.44 In homogenous (or very large) biological populations, this PC normally correlates with sexual dimorphism. In the case of the Garamantes sample, this is not so; while all the small crania are females, the four largest individuals include two males and two females. These individuals are not only large but statistically significantly larger than the rest of the sample (PC1 xmajority = −0.3279 ± 0.704; PC1 xlarge = +1.6397 ± 0.471; t = −5.304, p < 0.001). The large individuals are from Ayoub’s excavations of the Royal Cemetery in the 1960s [GSC031 (SM-K4 (F)), GSC030 (SM-K3 (M))],45 from elite burials close to Jarma [GER011.T24 (SBH24 (F))], and from a shaft burial at Watwat [UAT008.T86-SK410 (M)]. The graph also distributes individual Garamantian crania along the vertical axis (PC2), which contrasts skulls with broad vaults and narrow maxillae and noses, with those of the opposite shape (narrow vaults, with broad maxillae and noses). Most Garamantian crania are neither particularly broad nor narrow. However, three crania form a tight cluster at the top of the chart (ZIN013.T171, ZIN013.T170, TAG006.T1). These were all burials dating to the latter 44 45

See Lahr and Wright 1996, for example, for more complete discussions. See Nikita et al. 2010, 396.

https://doi.org/10.1017/9781108634311.004 Published online by Cambridge University Press

4 Human Mobility and Identity in Relation to the Garamantes

Isotopes

Broad crania with narrow maxillae and noses

3.00000

Garamantes (major) Garamantes (outliers)

ZIN013.T171

2.00000

TAG006.T1 ZIN013.T170

PC2

1.00000 SM.K3

.00000

CLF010.T1

ZIN013.T209 GSC031.T30 UAT008.T86 GER011.T24

–1.00000

–2.00000

UAT008.T35

–2.00000

–1.00000

Narrow crania with broad maxillae and noses

.00000

1.00000

2.00000

3.00000

PC1

Figure 4.5. Scatterplot of Garamantian crania along PC1 and PC2, highlighting the population’s variation in size and shape (as defined by PC2).

part of the first millennium BC and may be typical of the local population at that time. A single individual (UAT008.T35) departs from the majority in the opposite direction, with narrow vault and broad maxilla and nose. The difference in value of PC2 between the three individuals (PC2 average = −0.1881 ± 0.650) noted and the majority (PC2 average = +1.8769 ± 0.316) is significant (t = −5.328, p < 0.001). However, the patterns identified in size (PC1) and shape (PC2) differences are not associated with variation in δ18O, and indeed two of the three δ18O outliers in the sample are within the central cluster. The third (SMK1 from GSC030.T4) has a low value of PC2, and approximates the shape that characterises UAT008.T35 and which can be broadly considered typical of some Sub-Saharan African crania (narrow vaults with broad noses and mouths). A visual comparison of the face of ZIN013.T171 and GSC030.T4 (SM-K1) clearly illustrates the differences identified here (Fig. 4.6). If GSC030 is correctly identified as a Late Garamantian royal cemetery,46 the presence of individuals with features that are typically Sub-Saharan alongside individuals with features more typical of the 46

Mattingly 2010, 359–69; Mattingly et al. 2011; Nikita et al. 2010.

https://doi.org/10.1017/9781108634311.004 Published online by Cambridge University Press

151

152

Ronika K. Power et al.

Figure 4.6. Visual comparison of two distinctive cranial morphologies among the Garamantes: a) ZIN013.T171, displaying the narrow and orthognathic face and broad vault; b) GSC030.T4 (SM-K1), displaying the broad and prognathic face and narrow vault.

overall al-Ajal sample is an interesting reflection on the possibilities of Garamantian social structures.47 To explore further this group of individuals with a markedly different cranial shape, the PCA was repeated including published craniometric data on Garamantes/Fazzan crania from two different areas in Fazzan – from the site of Barkat, south of Ghat, kept at the Musée de Bardo (Algeria) and published by Leschi, and from the later site of Murzuq, south-east of the Massak Sattafat, kept in Firenze (Italy) and studied by Parenti in 1945.48 The analysis obtained five factors, which altogether explain 72.5 per cent of the variance in the matrix. The larger sample size (49 crania) shows the distribution of males and females along PC1 (size) – the female cranium of SM-K4 (GSC031.T30) is no longer the largest, although it is still part of a cluster of very large individuals. The shape differences identified in the analysis of the crania from the Wadi al-Ajal (broad vaults with narrow and orthognathic faces versus narrow vaults with broad and prognathic faces) is also observed in the larger sample. To the cluster of individuals with broad 47 48

Sergi 1951, for earlier reflections on the mixed character/physiognomy of Garamantian society. Leschi 1945; Parenti 1945, the published information is a reduced dataset of 2-D linear measurements.

https://doi.org/10.1017/9781108634311.004 Published online by Cambridge University Press

4 Human Mobility and Identity in Relation to the Garamantes

crania and narrow faces from the Wadi al-Ajal is added a cranium from Barkat (TIII, n25). Altogether, a group of six individuals form this distinct cluster – three crania from Zinkekra (ZIN013.T170, ZIN013.T171, ZIN013. T204), two crania from one of the Taqallit cemeteries (TAG006.T1, T2) and the cranium from Barkat. The Zinkekra and Taqallit burials date to the Proto-Urban Garamantian phase (latter centuries BC), so there may be some chronological differentiation being expressed. The crania from Murzuq extend markedly the range of the opposite range, showing a more pronounced expression of the Sub-Saharan African features observed in SMK1 (GSC030.T4) (Fig. 4.7). Although the above results cannot be directly associated to the δ13C and 18 δ O data obtained from the individuals buried in the Wadi al-Ajal, they represent further evidence of the complex phylogeography and dietary diversity of the people who lived at least the later portion of their lives as part of Garamantian society and were buried in Fazzan. Both the isotopic and craniometric data reveal clusters of individuals whose childhood diets differed from that of the majority of Garamantes in that it included a significant amount of C4 plants, such as Sub-Saharan African grasses. These individuals, all buried in the Watwat cemeteries, are not represented in the craniometric dataset; however, another individual from these cemeteries shows distinct Sub-Saharan African features. The individual mentioned above (SM-K1–GSC030.T4) also has an δ18O value suggesting a different source of water in childhood. Altogether, these results could indicate that this group of people may have been Trans-Saharan migrants. It is interesting to note that the morphology observed in this group characterises the majority of the sample from Murzuq. This hypothesis has to be treated with caution because of the lack of direct association between the two datasets, the presence of individuals at the Watwat cemeteries who do not differ in their diet or morphology from the rest of the Garamantes, and the possible existence of different dietary habits within subgroups of the same society without implications towards region of origin. Nevertheless, the data are suggestive and should be explored further. The results of the craniometrics analyses also identified another subgroup within the Garamantes buried in the Wadi al-Ajal. The morphology is observed widely among Mediterranean people. The crania that show this set of features most distinctively are part of a tight cluster of δ13C and δ18O that includes most of the Garamantes from the Wadi al-Ajal, and can thus be considered as having had the same childhood diets and drinking water as most other Garamantes. Their morphological distinctiveness may therefore reflect predominant ancestry, rather than migration.

https://doi.org/10.1017/9781108634311.004 Published online by Cambridge University Press

153

154

Ronika K. Power et al.

2.00000

Narrow vaults with broad upper faces (bizygomatic breadth)

majority broad, with narrow maxillae Firenze M du Bardo

1.00000

Narrow and short crania, with broad noses, maxillae and interorbital width, prognathic

PC2

.00000

Broad and tall crania, with narrow noses, maxillae and interorbital width, orthognathic

–1.00000 ZIN013.T204 TAG006.T1 TAG006.T2 Tombe II,n25

–2.00000

ZIN013.T171 ZIN013.T170

Broad vaults with narrow upper faces (bizygomatic breadth)

–3.00000

–3.00000 –2.00000 –1.00000

.00000

1.00000

2.00000

3.00000

PC1

Figure 4.7. Distribution of Garamantian crania along PC1 (size) and PC2 (broad vaults with narrow and orthognathic faces versus narrow vaults with broad and prognathic faces), showing a cluster of individuals along PC2 of distinct morphology that include proto-urban Garamantian remains from Zinkekra (ZIN013), Taqallit (TAG006) and Barkat (M. du Bardo); the opposite extreme of shape, represented by crania, is significantly extended by the remains from Murzuq (Firenze).

Conclusions Results presented here provide the first steps towards reconstructing a biocultural profile of Fazzan communities that were seemingly characterised by simultaneous social stability and flux. Both craniometric and isotopic data support the proposition that a core ‘local’ community existed in Fazzan. That community had relatively homogeneous anatomical and biochemical signatures, and its members were buried in the scarp’s edge cemeteries of the Wadi al-Ajal through time. This community provided a social substrate, beside whom migrants to the area may have settled, and/ or above whom more transient individuals may have simply ‘passed through’. The isotopic results identify slightly less than a third of the sample with distinctly different carbon and oxygen isotopic values to the core local population, suggesting that these individuals had access to different food types and/or water sources during their childhood. Craniometric data support these findings by identifying a few individuals with features broadly characterised as typical of Sub-Saharan morphology that differ from the majority of the Garamantian sample, while also

https://doi.org/10.1017/9781108634311.004 Published online by Cambridge University Press

4 Human Mobility and Identity in Relation to the Garamantes

highlighting a subgroup of individuals with more typically ‘Mediterranean’ characteristics. When viewed in light of their isotopic similarity to ‘local’ Garamantes, these latter individuals may indicate ancestral differences in geographical affinity as opposed to more recent migration. The individuals whose carbon isotopic values suggests they consumed a different diet in childhood, one rich in C4 grasses, were all buried in a single cemetery – Watwat – and date to the second century AD or later. Their clustering at Watwat may suggest incipient social boundaries within Garamantian society. However, while all the individuals with a signature of a different childhood diet were buried at Watwat, not all those buried there do, indicating the shared used of burial space by all members of society. Equally, those individuals whose oxygen isotopic values suggest a different source of water in childhood were buried in several cemeteries, although mostly close to the Garamantian capital of Jarma, most of which were of higher social status than those at Watwat. The combination of morphometric and isotopic work further reinforces the view that Garamantian society included individuals of diverse geographical origin, some of whom may have been first generation Trans-Saharan migrants. These findings are reinforced by the discovery of the interment of a young woman of Sub-Saharan physiognomy wearing a distinctive lipplug of Sahelian type excavated during the Desert Migrations Project, dating to the later first millennium BC.49 This ornament demonstrates that some Garamantian individuals shared aspects of their material culture with Sahelian societies more broadly, either through migration or contact, while their burial within Garamantian cemeteries shows their integration into the normative funerary rituals of contemporary Garamantian society as suggested by the results of the isotopic and craniometrics analyses. In combination, these results support the hypothesis of a vivid trading community that maintained a resident population through centuries, and one which was enriched by multiple-sourced Trans-Saharan migrations, as offered throughout this volume. Further integration of the physical and biochemical signatures of the individuals buried in Fazzan with all aspects of the extraordinary cemeteries from the Wadi al-Ajal has the potential of revealing the distinctive composition of what must have been a unique society at the heart of the Sahara desert. Further analysis, bolstered through the procurement of additional contemporaneous local and regional data, should enable us to refine considerably the picture of migration already apparent in the data from Fazzan. 49

Mattingly et al., Chapter 2, this volume.

https://doi.org/10.1017/9781108634311.004 Published online by Cambridge University Press

155

156

Ronika K. Power et al.

This work is vital if we are to understand the significance of migration in the revealed identity of individuals and groups in Garamantian society. With a total sample of several hundred excavated burials, this promises to be a ground-breaking study of the linkages between cultural and biological identity in a notably mixed society.50

References Bentley, R. A., Price, T. D., Luning, J., Gronenborn, D., Wohl, J. and Fullagar, P. D. 2002. Prehistoric migration in Europe: Strontium isotope analysis of Early Neolithic skeletons. Current Anthropology 43: 799–804. Boutton, T. W., Lynott, M. J. and Bumsted, M. P. 1991. Stable carbon isotopes and the study of prehistoric human diet. Critical Reviews in Food Science and Nutrition 30: 373–85. Brett, M. and Fentress, E. 1996. The Berbers. Oxford: Blackwell. Brothwell, D. 1981. Digging Up Bones. New York: Cornell University Press. Buikstra, J. E. and Ubelaker, D. H. 1994. Standards for Data Collection from Human Skeletal Remains. Fayetteville: Arkansas Archaeological Survey Report 44. Buzon, M. R. and Bowen, G. J. 2010. Oxygen isotope analysis of migration in the Nile Valley. Archaeometry 52: 855–68. Camps, G. 1980. Berbères. Aux marges de l’histoire. Toulouse: Editions des Hespérides. Chamberlain, A. 2000. Minor concerns. A demographic perspective on children in past societies. In J. Sofaer Derevenski (ed.), Children and Material Culture, London: Routledge, 206–12. Chamla, M. C. 1968. Les populations anciennes du Sahara et des régions limitrophes. Etude des restes osseux humains néolithique et protohistoriques. Mémoire Centre Recherche Anthropologique Préhistorique Ethnologique. Paris: Arts et Métiers. Chenery, C., Müldner, G., Evans, J. A., Eckardt, H. and Lewis, M. 2010. Strontium and stable isotope evidence for diet and mobility in Roman Gloucester, UK. Journal of Archaeological Science 37: 150–63. Chisholm, B. S. 1989. Variation in diet reconstructions based on stable isotopic evidence. In T. D. Price (ed.), The Chemistry of Prehistoric Human Bone, Cambridge: Cambridge University Press, 10–37. Coplen, T. B., Kendall, C. and Hopple, J. 1983. Comparison of stable isotope reference samples. Nature 302: 236–38. Daux, V., Lécuyer, C., Héran, M. A., Amiot, R., Simon, L., Fourel, F., Martineau, F., Lynnerup, N., Reychler, H. and Escarguel, G. 2008. Oxygen isotope fractionation 50

Thanks are due to Emma Lightfoot, University of Cambridge, for her assistance in assembling comparative isotopic datasets, and the IN-AFRICA Project (http://in-africa.org/) that supported the collection of data from comparative North African samples.

https://doi.org/10.1017/9781108634311.004 Published online by Cambridge University Press

4 Human Mobility and Identity in Relation to the Garamantes

between human phosphate and water revisited. Journal of Human Evolution 55.6: 1138–47. Diaz, A. L., O’Connell, T. C., Maher, L. A. and Stock, J. T. 2012. Subsistence and mobility strategies in the Epipalaeolithic: a stable isotope analysis of human and faunal remains at ‘Uyun al-Hammam, northern Jordan. Journal of Archaeological Science 39.7: 1984–92. Eckardt, H., Booth, P., Chenery, C., Müldner, G., Evans, J. A. and Lamb, A. 2009. Isotope evidence for mobility at the late Roman cemetery at Lankhills, Winchester. Journal of Archaeological Science 36: 2816–25. Gat, J. R. 1996. Oxygen and Hydrogen isotopes in the hydrologic cycle. Annual Review of Earth and Planetary Sciences 24: 225–62. Harvati, K. and Weaver, T. D. 2006. Human cranial anatomy and the differential preservation of population history and climate signatures. The Anatomical Record Part A 288: 1225–33. Hennessy, R. J. and Stringer, C. B. 2002. Geometric morphometric study of the regional variation of modern human craniofacial form. American Journal of Physical Anthropology 117: 37–48. Hewlett, B. S. 1991. Demography and childcare in pre-industrial societies. Journal of Anthropological Research 47: 1–37. Houghton, P. 1996. The People of the Great Ocean: Aspects of Human Biology in the Early Pacific. New York: Cambridge University Press. Howells, W. W. 1973. Cranial Variation in Man. A Study by Multivariate Analysis of Patterns of Difference among Recent Human Populations. Cambridge, Mass.: Harvard University, Papers of the Peabody Museum 67. Howells, W. W. 1989. Skull Shapes and the Map: Craniometric Analyses in the Dispersion of Modern Homo. Cambridge, Mass.: Harvard University, Papers of the Peabody Museum 79. Iacumin, P., Bocherens, H., Mariotti, A. and Longinelli, A. 1996. Oxygen isotope analyses of co-existing carbonate and phosphate in biogenic apatite: a way to monitor diagenetic alteration of bone phosphate? Earth and Planetary Science Letters 142: 1–6. Kellner, C. M. and Schoeninger, M. J. 2007. A simple carbon isotope model for reconstructing prehistoric human diet. American Journal of Physical Anthropology 133: 1112–27. Krogman, W. M. and M. Y. İșcan, M. Y. 1986. The Human Skeleton in Forensic Medicine. Springfield: Charles C. Thomas. Lahr, M. M. 1992. The Origins of Modern Humans: A Test of the Multiregional Hypothesis. Cambridge: Cambridge University Press. Lahr, M. M. 1996. The Evolution of Modern Human Cranial Diversity: A Study in Cranial Variation. Cambridge: Cambridge University Press. Lahr, M. M. 2010. Saharan corridors and their role in the evolutionary geography of ‘Out of Africa I’. In J. G. Fleagal, J. J. Shea, F. E. Grine, A. L. Baden &

https://doi.org/10.1017/9781108634311.004 Published online by Cambridge University Press

157

158

Ronika K. Power et al.

R. E. Leakey (eds), Out of Africa I. The First Hominin Colonization of Eurasia, New York, Springer, 27–46. Lahr, M. M. and Foley, R. A. 1998. Towards a theory of modern human origins: geography, demography and diversity in recent human evolution. Yearbook of Physical Anthropology 41: 137–76. Lahr, M. M. and Foley, R. A. 2001. Genes, fossils and behaviour: when and where do they fit? In P. Donnelly and R. A. Foley (eds), Genes, Fossils and Behaviour, Brussels: NATO, 13–48. Lahr, M. M. and Wright, R. V. S. 1996. The question of robusticity and the relationship between cranial size and shape in Homo sapiens. Journal of Human Evolution 31: 157–91. Lee-Thorp, J. A. 2008. On isotopes and old bones. Archaeometry 50: 925–50. Leschi, L. 1945. La mission scientifique du Fezzân. Archéologie. Travaux de l’Institut de Recherches Sahariennes 3: 183–86. Levinson, A. A., Luz, B. and Kolodny, Y. 1987. Variations in Oxygen isotopic compositions of human teeth and urinary stones. Applied Geochemistry 2: 367–71. Lewis, M. E. and Gowland, R. 2007. Brief and precarious lives: infant mortality in contrasting sites from Medieval and Post-Medieval England (AD 850–1859). American Journal of Physical Anthropology 134: 117–29. Lieberman, D. E., Ross, C. R. and Ravosa, M. 2000a. The primate cranial base: ontogeny, function, and integration. Yearbook of Physical Anthropology 43: 117–69. Lieberman, D. E., Mowbray, K. M. and Pearson, O. M. 2000b. Basicranial influences on overall cranial shape. Journal of Human Evolution 38: 291–315. Lightfoot, E, Slaus, M. and O’Connell, T. C. 2014. Water consumption in Iron Age, Roman and Early Medieval Croatia. American Journal of Physical Anthropology 154.4: 535–43. Lightfoot, E. and O’Connell, T. C. 2016. On the use of biomineral oxygen isotope data to identify human migrants in the archaeological record: intra-sample variation, statistical methods and geographical considerations. PLoS ONE 11.4: e0153850. Longinelli, A. 1984. Oxygen isotopes in mammal bone phosphate – a new tool for paleohydrological and paleoclimatological research. Geochimica et Cosmochimica Acta 48.2: 385–90. Luz, B., Kolodny, Y. and Horowitz, M. 1984. Fractionation of oxygen isotopes between mammalian bone-phosphate and environmental drinking water. Geochimica et Cosmochimica Acta 48: 1689–93. MacPhee, R. D. E. and Cartmill, M. 1986. Basicranial structures and primate systematics. In D. R. Swindler and J. Erwin (eds), Comparative Primate Biology, Systematics, Evolution and Anatomy, Volume 1, New York: Alan R. Liss Inc., 219–75.

https://doi.org/10.1017/9781108634311.004 Published online by Cambridge University Press

4 Human Mobility and Identity in Relation to the Garamantes

Manzi, G. and Ricci, F. 2003. Population of the Roman era in the central Sahara: skeletal samples from the Fezzan (south-western Libya) in a diachronic perspective. In M. Liverani (ed), Arid lands in Roman times, Papers from the International Conference (Rome, 9th–10th July 2001), AZA 4, Firenze: All’Insegna del Giglio, 15–22. Mattingly, D. J. (ed.). 2010. The Archaeology of Fazzan, Volume 3. Excavations of C. M. Daniels. London: Society for Libyan Studies, Department of Antiquities. Mattingly, D. J. (ed.). 2013. The Archaeology of Fazzan. Volume 4, Survey and Excavations at Old Jarma (Ancient Garama) Carried Out by C. M. Daniels (1962–69) and the Fazzan Project (1997–2001). London: Society for Libyan Studies, Department of Antiquities. Mattingly, D. J. and Sterry, M. 2013. The first towns in the Central Sahara. Antiquity 87.366: 503–18. Mattingly, D. J., Dore, J. and Lahr, M. 2008. DMP II: 2008 fieldwork on burials and identity in the Wadi al-Ajal. Libyan Studies 39: 223–62. Mattingly, D. J., Lahr, M. and Wilson, A. 2009. DMP V: investigations in 2009 of cemeteries and related sites on the west side of the Tāqallit promontory. Libyan Studies 40: 95–131. Mattingly, D. J., Abduli, H., Ahmed, M., Cole, F., Fenwick, C., Fothergill, B. T., Gonzalez Rodriguez, M., Hobson, M., Khalaf, N., Lahr, M., Moussa, F., Nikita, E., Nikolaus, J., Radini, A., Ray, N., Savage, T., Sterry, M. and Wilson, A. 2011. DMP XII: excavations and survey of the so-called Garamantian Royal Cemetery (GSC030–031). Libyan Studies 42: 89–102. Mitchell, P. D. and Millard, A. R. 2009. Migration to the Medieval Middle East with the Crusades. American Journal of Physical Anthropology 140: 518–25. Moore, W. J. and Lavelle, C. L. B. 1974. Growth of the Facial Skeleton in the Hominoidea. London: Academic Press. Nikita, E. 2010. The Garamantes of Fazzān: Bioarchaeological evaluation of desertinduced stress and Late Holocene human migration through the Sahara. Unpublished PhD thesis, University of Cambridge, Cambridge. Nikita, E., Crivellaro, F., Stock, J., Foley, R. and Lahr, M. M. 2010. Human skeletal remains. In Mattingly 2010: 375–408. Nikita, E., Siew, Y. Y. Stock, J., Mattingly, D. and Lahr, M. M. 2011. Activity patterns in the Sahara Desert: an interpretation based on cross-sectional geometric properties. American Journal of Physical Anthropology 146: 423–34. Nikita, E., Mattingly, D. and Lahr, M. M. 2012a. Sahara: barrier or corridor? Nonmetric cranial traits and biological affinities of North African Late Holocene populations. American Journal of Physical Anthropology 147: 280–92. Nikita, E., Mattingly, D. and Lahr, M. M. 2012b. Three-dimensional cranial shape analyses and gene flow in North Africa during the Middle to Late Holocene. Journal of Anthropological Archaeology 31: 564–72.

https://doi.org/10.1017/9781108634311.004 Published online by Cambridge University Press

159

160

Ronika K. Power et al.

Nikita, E., Mattingly, D. and Lahr, M. M. 2013. Evidence of trephinations among the Gramantes, a Late Holocene Saharan population. International Journal of Osteoarchaeology 23: 370–77. O’Higgins, P. 2000. The study of morphological variation in the hominid fossil record: biology, landmarks and geometry. Journal of Anatomy 197: 103–20. Olson, T. R. 1981. Basicranial morphology of the extant hominoids and Pliocene hominids: the new material from the Hadar Formation, Ethiopia and its significance in early human evolution and taxonomy. In C. B. Stringer (ed.), Aspects of Human Evolution, London: Taylor & Francis, 99–128. Pace, B., Sergi, S. and Caputo, G. 1951. Scavi sahariani. Monumenti Antichi 41: 150–549. Parenti, R. 1945. Contributo alla conoscenza della craniologia del Fezzan. Archivio per l’antropologia e l’etnologia 75: 5–116. Perry, M. A., Coleman, D. S., Dettman, D. L. and al-Shiyab, A. H. 2009. An isotopic perspective on the transport of Byzantine mining camp labourers into south-western Jordan. American Journal of Physical Anthropology 140: 429–41. Price, T. D., Knipper, C., Grupe, G. and Smrcka, M. V. 2004. Strontium isotopes and prehistoric human migration: the Bell Beaker period in Central Europe. European Journal of Archaeology 8: 9–40. Prowse, T. L., Schwarcz, H. P., Garnsey, P., Knyf, M., Macchiarelli, R. and Bondioli, L. 2007. Isotopic evidence for age-related immigration to Imperial Rome. American Journal of Physical Anthropology 132: 510–19. Relethford, J. H. 1994. Craniometric variation among modern human populations. American Journal of Physical Anthropology 95: 53–62. Ricci, F., Manzi, G., Fornai, C., Vecchi, F. and Passarello, P. 2002. The human skeletal remains: inventory and inferences. In S. di Lernia and G. Manzi (eds), Sand, Stones and Bones. The Archaeology of Death in the Wadi Tanezzuft Valley (5000–2000 BP), AZA 3, Firenze: All’Insegna del Giglio, 217–50. Ricci, F., Fornai, C., Tiesler Blos, V., Rickards, O., di Lernia, S. and Manzi, G. 2008. Evidence of artificial cranial deformation from the later prehistory of the Acacus Mountains (south-western Libya, Central Sahara). International Journal of Osteoarchaeology 18: 372–91. Ricci, F., Tafuri, M. A., Di Vincenzo, F. and Manzi, G. 2013. The human skeletal sample from Fewet. In L. Mori (ed.), Life and Death of a Rural Village in Garamantian Times. The Archaeological Investigation in the Oasis of Fewet (Libyan Sahara), AZA Monograph 6, Firenze: All’Insegna del Giglio, 319–62. Richards, M., Harvali, K., Grimes, V., Smith, C., Smith, T., Hublin, J. J., Karkanas, P. and Panagopoulos, E. 2008. Strontium isotope evidence of Neanderthal mobility at the site of Lakonis, Greece using laser ablation PIMMS. Journal of Archaeological Science 35: 1251–56. Roseman, C. C. 2004. Detecting interregionally diversifying natural selection on modern human cranial form by using matched molecular and morphometric data. Proceedings of the National Academy of Sciences USA 101: 12824–29.

https://doi.org/10.1017/9781108634311.004 Published online by Cambridge University Press

4 Human Mobility and Identity in Relation to the Garamantes

Schwartz, J. H. 1995. Skeleton Keys. An Introduction to Human Skeletal Morphology, Development and Analysis. Oxford: Oxford University Press. Sergi, S. 1951. I resti scheletrici delle antiche popolazioni del Fezzan ed il tipo dei Garamanti. In Pace et al. 1951: 443–542. Smith, H. F. 2009. Which cranial regions reflect molecular distances reliably in humans? Evidence from three-dimensional morphology. American Journal of Human Biology 21: 36–47. Smith, H. F., Terhune, C. E. and Lockwood, C. A. 2007. Genetic, geographic, and environmental correlates of human temporal bone variation. American Journal of Physical Anthropology 134: 312–22. Spencer, M. A. and Demes, B. 1993. Biomechanical analysis of masticatory system configuration in Neanderthals and Inuits. American Journal of Physical Anthropology 91: 1–20. Spencer, M. A. and Ungar, P. S. 2000. Craniofacial morphology, diet and incisor use in three Native American populations. International Journal of Osteoarchaeology 10: 229–41. Ubelaker, D. H. 1997. Human Skeletal Remains: Excavation, Analysis, Interpretation. 3rd Edition. Washington, DC: Taraxacum. von Cramon-Taubadel, N. 2009. Congruence of individual cranial bone morphology and neutral molecular affinity patterns in modern humans. American Journal of Physical Anthropology 140: 205–15. Waldron, T. 2001. Shadows in the Soil: Human Bones in Archaeology. Stroud: Tempus. White, C., Longstaffe, F. J. and Law, K. R. 2004. Exploring the effects of environment, physiology and diet on oxygen isotope ratios in ancient Nubian bones and teeth. Journal of Archaeological Science 31.2: 233–50.

https://doi.org/10.1017/9781108634311.004 Published online by Cambridge University Press

161

5

The Garamantes from Fewet (Ghat, Fazzan, Libya) A Skeletal Perspective francesca ricci, mary anne tafuri, francesca castorina, fabio di vincenzo, lucia mori and giorgio manzi

Introduction1 The distribution of funerary stone structures in the Saharan landscape has been a subject of interest for the Italian-Libyan Archaeological Mission in the Tadrart Akakus and Massak since the early ’90s.2 This archaeological evidence gave witness to an enduring human settlement, lasting at least from the Pastoral period to Proto-historical times,3 and played an important role in the definition of the cultural identities of the local groups,4 while representing a source of information about population features and dynamics.5 In the twentieth century, only a few excavations were carried out in cemeteries located in Fazzan.6 In 1997, as part of an interdisciplinary project focused on Holocene environment and human settlement until the rise of the Garamantian civilisation, our mission started a systematic survey and excavation of funerary structures in the Wadi Tanzzuft.7 The 2004–06 investigation of Fewet necropolis needs to be viewed in this framework.8 1

2 3 4 5 6

162

7

Our warmest thank to Savino di Lernia, director of the Italian-Libyan Archaeological Mission in the Akakus and Messak. We are particularly indebted to Mario Liverani for being a constant guide and inspiration. We are also grateful to all the participants to the fieldwork activities at Fewet, as well as to Paul Fullagar and Antonello Pelosi for help with the isotopic investigation. Sincere thanks to all the participants of the Trans-Saharan Conference ‘Burials, Migration and Identity’, held in Leicester in 2014 (see the Preface by David Mattingly, in this volume, for further details) and, particularly, to the reviewers of our original draft for comments and suggestions. di Lernia and Manzi 1998; 2002; Liverani 2006; 2007. di Lernia and Manzi 2002; Liverani 2003; 2006; Mattingly 2003; 2007. di Lernia 1999; di Lernia and Tafuri 2013; Stone and Stirling 2007. di Lernia and Manzi 2002; Manzi and Ricci 2003. Leschi 1945; Pace et al. 1951; Daniels 1989; Mattingly 2010. di Lernia and Manzi 1998; 2002; Liverani 2006; 2007. 8 Mori 2013a.

https://doi.org/10.1017/9781108634311.005 Published online by Cambridge University Press

5 The Garamantes from Fewet (Ghat): A Skeletal Perspective

Figure 5.1. Satellite image of the southern Wadi Tanzzuft with the location of the oases of Fewet, Ghat and Barkat and the Garamantian citadel of Aghram Nadharif (after Liverani 2003, Fig. 3.1).

Fewet (Fig. 5.1) is a large cemetery comprising 1,329 tumuli of various typologies used by the inhabitants of the nearby villages.9 The necropolis was used over a long time, from about the Late/Final Pastoral to the whole Garamantian period.10 The excavated structures (N = 24) were sampled from among those showing no signs of disturbance, while covering all the architectural typologies. The excavation provided a sample of 33 human skeletons. A broad bioanthropological study and a preliminary isotope investigation have been conducted so far. In this chapter, we offer a synthesis of these investigations with reference to data reported in the volume edited by Mori, where descriptions of the methods and detailed analytical data are provided.11 This study includes paleodemographic observations, analyses of cranial, dental and postcranial remains, as well as the evaluation of pathologies and markers of stress; with the isotopic 9

10 11

For further information on the Fewet sites, and more in general on the settlement pattern in the Wadi Tanzzuft, refers to Gatto et al., Chapter 3, this volume. Liverani et al. 2013; Mori and Ricci 2013. Mori 2013a; compare in particular: Ricci et al. 2013; Tafuri et al. 2013.

https://doi.org/10.1017/9781108634311.005 Published online by Cambridge University Press

163

164

Francesca Ricci et al.

analysis performed to investigate food practices and patterns of residential mobility. For comparative purposes, we used data from two skeletal series from the same region: the Late Pastoral 96/129 (Wadi Tanzzuft: N = 28),12 and from earlier Italian work on the Garamantes in Wadi al-Ajal and near Ghat,13 which will be collectively referred to as ‘Sergi’s sample’ (N = 29).

The Sample In general, the skeletal remains pertaining to the 33 individuals from Fewet show a good state of preservation. By contrast, the teeth are generally badly preserved: a pattern observed at other sites of the same area, probably a consequence of fractures due to extreme temperature variations.14 A systematic description of the burials is provided in Mori and Ricci and is summarised in Table 5.1.15 The position of the corpses is consistent with other contexts in Western and North Africa, including Fazzan.16 A systematic relationship has been observed between the sex of the individuals and the orientation of the corpse inside the grave: males were always aligned east-west (with the skull to the east), whereas females were aligned west-east (with the skull to the west). This correlation seems to represent an intentional funerary ritual within the cemetery, which does not appear to have been followed as systematically in the Wadi al-Ajal.17

Mortality Profiles The Fewet skeletal sample is composed of 15 infants and juveniles (45.4 per cent) and 18 adults (54.6 per cent). Concerning the sex, 8 are males, 12 females and 13 not determinable (Table 5.1). Our sex diagnosis was based on pelvic and cranial morphology.18 Estimation of subadult age at death was obtained from skeletal and dental developmental stages, while for adults we observed cranial suture closure, pubic symphysis morphology and occlusal dental wear.19 12 15 16 17 19

di Lernia and Manzi 2002. 13 Sergi 1951. 14 Ricci et al. 2002. Mori and Ricci 2013. See also Gatto et al., Chapter 3, this volume. Camps 1979; Chamla 1968; Dutour 1989; Paris 1984; Ricci et al. 2002; Sergi 1951. Mori and Ricci 2013. 18 Acsàdi and Nemeskèri 1970; Ferembach et al. 1979. Subadult: Fazekas and Kósa 1978; France and Horn 1988; Stloukal and Hanakova 1978; Ubelaker, 1989. Adult: Meindl and Lovejoy 1985; Todd 1920; Lovejoy 1985.

https://doi.org/10.1017/9781108634311.005 Published online by Cambridge University Press

5 The Garamantes from Fewet (Ghat): A Skeletal Perspective

165

Table 5.1. Fewet sample: list of the burials and main anthropological and archaeological data. ID

Sex

Age at Death

C14 uncal. BP (collagen)

T40 H1

F

35–40

−

T41 H1

M

30–35

2390 ± 40

T120 H1

F

30–35

−

T399 H1

M

25–30

−

T593 H1

I

−

T661 H1

I

Foetal/ Neonatal 7–8

T662 H1

I

10–11

T669 H1

F

20–25

2220 ± 45

T710 H1

M

30–40

−

T712 H1

F

35–40

−

T715 H1

F

20–25

1962 ± 40

T716 H1

M

20–30

1740 ± 25

T719 H1

I

13–15

2360 ± 70

T914 H1 T954 H1

F F

15–20 20–25

1940 ± 25 1920 ± 40

T976 H1 M T1191 H1 M

35–40 25–30

1960 ± 40 2200 ± 40

T1197 H1 F

30–40

2230 ± 40

T1197 H2 I T1197 H3 I

0–6 months about 3

− −

−

Position of the Body

Orientation

Tightly contracted left side Tightly contracted left side Tightly contracted left side Tightly contracted left side −

W-E (skull at W)

Tightly contracted right side Tightly contracted left side Tightly contracted left side Tightly contracted left side Tightly contracted right side Slightly contracted right side Tightly contracted left side Contracted dorsal decubitus Contracted right side Tightly contracted right side Contracted right side Tightly contracted left side Tightly contracted ventral decubitus − Tightly contracted left side

https://doi.org/10.1017/9781108634311.005 Published online by Cambridge University Press

E-W (skull at E) W-E (skull at W) E-W (skull at E) − E-W (skull at E) W-E (skull at W) W-E (skull at W) E-W (skull at E) W-E (skull at W) SW-NE (skull at SW) E-W (skull at E) E-W (skull at E) W-E (skull at W) W-E (skull at W) E-W (skull at E) E-W (skull at E) W-E (skull at W) W-E (skull at W) W-E (skull at W)

166

Francesca Ricci et al.

Table 5.1. (Cont.) Age at Death

C14 uncal. BP (collagen)

Position of the Body

Orientation

−

−

−

− 1900 ± 25

T1223 H1 I

12–14

2060 ± 25

T1226 H1 M

40–45

2290 ± 40

T1261 H1 F

35–40

2160 ± 40

T1261 H2 I

0–6 months

−

T1287 H1 T1287 H2 T1287 H3 T1287 H4

18–20 3–4 >40 20–30

− − − 3590 ± 40

Probably contracted left side Tightly contracted left side Tightly contracted left side Tightly contracted left side Tightly contracted left side Tightly contracted right side Contracted right side − − Tightly contracted left side _ _ _

W-E (skull at W)

T1210 H1 F

Foetal/ Neonatal Foetal/ Neonatal 30–35

ID

Sex

T1197 H4 I T1197 H5 I

F I M F

T1287 H5 I 5–6 T1287 H6 I 2–3 T1178 Cenotaph − (?)

− − −

SW-NE (skull at SW) E-W (skull at E) E-W (skull at E) SW-NE (skull at SW) E-W (skull at E) W-E (skull at W) − −7 NW-SE (skull at NW) − − −

The sex ratio is 0.7; this moderate sex equilibrium and the fact that all the age classes are represented suggest that our sample is a good representation of a time-distributed segment of the population; this also disproves any sex/age segregation in the spatial deposition of the corpses. Mortality distribution at Fewet is shown in Figure 5.2. It follows the general trend expected for ancient human populations and appears very similar to that of the Late Pastoral site 96/129 located in the Wadi Tanzzuft,20 which is characterised by high infant mortality followed by a gradual decrease after three years of life up to adolescence. The risk of death increases again in relation to the reproductive/active phases of life, with no significant differences between the sexes. The peak of deaths is registered in the age class 30-40 years, whereas individuals 20

Ricci et al. 2002.

https://doi.org/10.1017/9781108634311.005 Published online by Cambridge University Press

5 The Garamantes from Fewet (Ghat): A Skeletal Perspective

30

25

20

Indeterminate

% 15

Females Males

10

5

0

0–3

3–6

6–12

12–20

20–30

30–40

40+

Age classes

Figure 5.2. Mortality distribution of the Fewet sample according to sex.

older than 40 are a very small fraction of the population (6.1 per cent), suggesting high mortality in the first four decades of life. We registered a high frequency of neonatal deaths, which represent more than two-thirds (71.4 per cent) of the children who died between 0-3 years. In addition, the juvenile-index is 0.28,21 which falls within the range expected for ancient populations, hence there is not an under-representation of children/adolescents. The decrease of deaths in the age interval 5-9 years could have a specific biological meaning, suggesting amelioration of life conditions for those children who passed the first years of life, including the crucial passage of weaning. For the adolescent/adult segment of the sample, we recorded a gradual increase of mortality, particularly in the age classes over 20 years. Concerning the sex distribution, in the age class 12-20 years, half of the individuals are females and half are sexually indeterminate. But if we consider the sex-related orientation of the corpses (see above), the indeterminate may actually be all males, giving an equal number for both sexes in this age class. We registered a slight majority of male deaths between 20-30 years and, on the contrary, a large prevalence of female deaths in the class 30-40 years. The mortality peak (27.3 per cent) documented in the age class 30-40 years, composed mostly of females, could suggest that life for females was as hard 21

Bocquet and Masset 1977.

https://doi.org/10.1017/9781108634311.005 Published online by Cambridge University Press

167

168

Francesca Ricci et al.

as that of males at Fewet, and that females possibly continued to get pregnant at a mature age, when the risk of death at parturition became higher. A final point to note is that all the individuals older than 40 years were males.

The Cranium Cranial Morphometrics Cranial measurements have been recorded on adult specimens, according to Martin and Saller, and some indices have been calculated for the neurocranium and the face.22 Both males and females display modest degrees of variation for the measurements of the vault (as suggested by the coefficients of variation),23 and slightly higher values for facial and mandible measurements. Sexual dimorphism appears generally modest, being lower than 10 for most of the measurements.24 It is worth noting that the female subsample shows the highest coefficient of variation for the variable ‘dacrion-dacrion’ (that is, upper nasal breadth), which is clearly greater than that of the males (18.0 versus 3.2). Given the north-south polarity in Africa for this trait,25 this evidence suggests the possible occurrence of greater mobility of the females in the marriage network as in many populations.26 It is of interest that the same result was found for site 96/129.27 Consistently, at Fewet, the range of the Sr isotope ratios (see below) is slightly higher among the females. As for the shape of the cranial vault, individuals with long and narrow braincases largely prevail; moreover, crania are always high and narrow, so are faces and orbits, while nasal shape is frequently discordant. These cranial features seem to depict a quite homogeneous morphotype, similar to that observed at 96/129 and to the description provided by Sergi for the Garamantes of the ancient Fazzan.28 Yet, a difference is recorded in the morphology of the nasal aperture when Fewet is compared to 96/129, where intermediate and narrow/high noses (completely absent at Fewet) prevail. Such a similarity suggests biological continuity between populations that occupied neighbouring areas in the Fazzan/Tanzzuft Valley from the Late Pastoral to Roman times and beyond. Nevertheless, the different 22 25 26 28

Martin and Saller 1957. 23 See Table 17.III in Ricci et al. 2013. Biasutti 1967; Krogman and Iscan 1986. Destro-Bisol et al. 2004; Nikita et al. 2012; Seielstad et al. 1998. Ricci et al. 2002; Sergi 1951.

https://doi.org/10.1017/9781108634311.005 Published online by Cambridge University Press

24

27

Hall 1982.

Ricci et al. 2002.

5 The Garamantes from Fewet (Ghat): A Skeletal Perspective

85

75

70

96/129 M

65

horizontal cranial index

80

96/129 F 60 FW M FW F 4000

3500

3000

2500

2000

1500

55 1000

years uncal BP

Figure 5.3. Distribution across time of the horizontal cranial shape at Fewet and 96/129. Values < 75: long/narrow braincases; values = 75–80: intermediate shapes (indicated by dotted lines); values>80: short/broad braincases.

characterisation of the nasal shape (broad/low at Fewet, prevailing narrow/ high at 96/129, narrow in Sergi’s Group I) as well as the high variability in the interorbital breadth in the females from Fewet apparently indicate a certain amount of gene flow along a north-south African axis, similar to what Sergi suggested when he considered a possible ‘hybridisation’ between his Groups I and IV.29 In order to test the hypothesis of a possible diachronic trend among samples coming from 96/129 and Fewet, in Figure 5.3 we show the dispersion of the horizontal cranial index (breadth/length of the vault) and time, using dated specimens from the two sites; the use of 14C uncalibrated years BP allowed the comparison of Fewet with site 96/129,30 while Sergi’s sample was not considered since it lacks radiocarbon dates. The graph shows that, across time (namely at Fewet, after 2,500 years BP) and in possible relationship with gene flow, the cranial shape was more variable as far as this transversal section is considered, with an increment of individuals with long/ narrow cranial vaults. Craniometric data were also used to examine the degree and pattern of intra- and inter-population variability for the three skeletal series of ancient Fazzan available to us (Fewet, 96/129 and Sergi’s sample). 29

Sergi 1951.

30

di Lernia and Manzi 2002; Mori and Ricci 2013.

https://doi.org/10.1017/9781108634311.005 Published online by Cambridge University Press

169

170

Francesca Ricci et al.

Figure 5.4. Principal component analysis of the cranial variables applied to the three skeletal samples from Fazzan (96/129, Fewet and material studied by Sergi). The first three principal components (PC) are displayed in different orientation in order to better observe the distribution of the samples. Dimension of the dots is proportional to the maximum cranial width (eu-eu) used as a proxy of the cranial size.

A Principal Component Analysis (PCA) was performed (using SPSS 13, SPSS Inc., Chicago, IL, and PAST 2.17)31 on a matrix composed by 24 cranial measurements. A synthetic result is shown in Figure 5.4, where the three main PCs explain cumulatively 67.35 per cent of the variance, and respectively 45.32 per cent (PC1), 12.48 per cent (PC2) and 9.55 per cent (PC3). The result is interesting because of the polarisation of the samples, which occupy different positions in the multivariate space, thus suggesting a different phenetic composition peculiar to each population. Given this result, we then performed a Canonical Variate Analysis (CVA; Fig. 5.5) aimed at stressing the distinction between the samples; as an outgroup¸ we added the Medieval sample (N = 21) from Quingentole, northern Italy.32 The first axis, which explains 60.12 per cent of the variance, cleanly 31

Hammer et al. 2001.

32

Dal Poz et al. 2001.

https://doi.org/10.1017/9781108634311.005 Published online by Cambridge University Press

5 The Garamantes from Fewet (Ghat): A Skeletal Perspective

Fewet Sergi Lambda-opistion arch (28) Basion-bregma height (17)

CV 2 (26.92%)

Quingentole

Nasion-bregma arch (26)

Bregma-lambda arch (27) Max. frontal diameter (10)

Max. width (8)

Facial total height (47)

Max. occipital width (12)

Facial length (40)

Min. frontal diameter (9)

96/129

–6

–4

–2

0

2

4

CV 1 (60.12%)

Figure 5.5. Canonical variates analysis of the cranial variables, applied to the three skeletal samples from Fazzan (96/129, Fewet and material studied by Sergi), with the addition of an outgroup: the medieval sample from Quingentole, northern Italy. Main vectors of change are shown.

separates the Italian sample from the Libyan ones. Quingentole is characterised by cranial variables: 13 (maximum frontal diameter), 8 (maximum cranial width) and 12 (maximum width of the occipital bone), while the three Libyan samples by variables: 17 (basion-bregma height), 26 (nasionbregma arch), 47 (facial total height) and others. In sum, Italian Medieval crania appear larger than the Libyan ones, with high cranial vaults and faces. The three Libyan samples completely overlap along the CV1, but are clearly separated along the CV2, which explains 26.92 per cent of the variance. Essentially, the second axis separates the Late Pastoral 96/129 from the two more recent Garamantian samples of Sergi and Fewet, possibly in accordance with the diachronic trends observed above and the possible effects of gene flow.

https://doi.org/10.1017/9781108634311.005 Published online by Cambridge University Press

171

172

Francesca Ricci et al.

Cranial Non-Metric (or ‘Epigenetic’) Traits Discrete cranial traits usually referred to as non-metric or ‘epigenetic’33 have been recorded on adult and subadult crania according to Manzi and Vienna;34 data were converted and analysed in terms of presence/absence. Looking at the distinction between hypostotic and hyperostotic features,35 many traits of the vault that are considered as hypostotic – such as supernumerary ossicles and persistence of foetal cranial sutures – are nearly always present among the Fewet crania; conversely, almost all the hyperostotic traits of the face – such as supraorbital notch/foramen and zygomaxillary tubercle – show high scores. It is known that such non-metric traits have been generally used in skeletal biological studies as a sort of ‘genetic marker’ at individual or population level,36 but an alternative perspective emphasises their proper ‘epigenetic’ nature, interpreting them as the outcome of contrasting pressures applied to the cranium before its complete ossification.37 In this light, the relatively high frequency of many hypostotic traits in the cranial vault at Fewet seems to suggest a shared genetic background of the population that favoured the occurrence of ‘ontogenetic stress’.38 This, in turn, may indicate the presence of some degree of endogamy inside the population: a condition that leads to the expression of unusual rare traits with epigenetic meaning.39 This is in accordance with the archaeological data describing the inhabitants of Fewet as a group with strong family ties, such as an extended family or small clan.40 In the general paucity of data on these features registered in North African prehistoric populations, Nikita et al. recently used cranial non-metric traits to calculate biological distances among a certain number of populations.41 Their results documented an isolation of the Garamantes, probably caused by the desert acting as a barrier, which is consistent with the results of our metric and non-metric observations on cranial shape variation at Fewet and other nearby penecontemporaneous samples.42

Dental Dimensions In the Fewet sample, mesio-distal (MD) and bucco-lingual (BL) crown diameters of permanent teeth have been recorded, according to 33 35 36 37 40

Berry and Berry 1967; Hauser and De Stefano 1989. 34 Manzi and Vienna 1997. Ossenberg 1970. Hanihara et al. 2003; Kellock and Parsons, 1970; Wijsman and Neves 1986. Manzi 2003; Ricci et al. 2008. 38 See Manzi et al. 1996. 39 Sjøvold 1984. Mori et al. 2013a. 41 Nikita et al. 2012. 42 See also Power et al., Chapter 4, this volume.

https://doi.org/10.1017/9781108634311.005 Published online by Cambridge University Press

5 The Garamantes from Fewet (Ghat): A Skeletal Perspective

10.5 10

mm

9.5 9 8.5 8 7.5

N

ub

ia n

U

C

M

es ol ith

ic s

kyrBP 0 m gH 2 ol u m H as na si ta -e l-A bi od W es Im te n H rn U 2 a Af N ric n K W ub as a es ia N te n eo a rn lit Sa Agr hi ic cs ha ul ra tu ra Pr lis ot ts oh is to ric s Tn z2 00 1 In 96 te /1 ns 29 iv e ag Fe ric w M et ul od tu er ra n l Eu ists ro pe an s

12

7

Figure 5.6. BL diameter (in mm) of mandibular P4. Comparisons between Fewet and other North African samples: Uan Muhuggiag (Umg H2), Imenennaden (Imn H2), Uan Kasa, 96/129, Tnz 2001 from the Akakus Massif and Wadi Tanzzuft (di Lernia and Manzi 1998; Ricci et al. 2002; Manzi and Ricci 2003; di Lernia and Cremaschi pers. comm.); Nubian Mesolithics (Calcagno 1986); Columnata (Chamla 1970); Hassi-elHabiod (Dutour 1989); Western Africa Neolithics (Chamla 1968); Nubian Agriculturalists (Calcagno 1986); Western Sahara Protohistorics (Chamla 1968); Nubian Intensive Agriculturalists (Calcagno 1986); Modern Europeans (Irish 1998). Samples are ordered according to their indicative chronology, from about 12 kyr BP up to the present.

Goose.43 A trait of particular interest is the BL diameter of the second lower premolar (mandibular P4), since it is considered representative of the general dental dimensions of a given individual.44 Figure 5.6 shows the distribution of this diameter in Fewet and other skeletal samples/individuals from various regions of North Africa divided by different chronological and cultural horizons.45 The sample from Fewet (and those from the Wadi Tanzzuft: 96/129 and Tnz 2001)46 falls within the variability of Nubian and West African groups of the Late Holocene. The trend of dental reduction registered through time may be the expression of the general skeletal gracilisation that followed the emergence of food production.47 In this perspective, the dental dimensions of the Fewet population, which appear greater than those of 96/129 and Tnz 2001, suggests the retention of archaic dental features, in accordance with the 43 46

Goose 1963. 44 Calcagno 1986. 45 Compare Manzi and Passarello 1999; Ricci et al. 2002. Ricci et al. 2002. 47 Ammerman and Cavalli-Sforza 1984; Barbujani et al. 1994.

https://doi.org/10.1017/9781108634311.005 Published online by Cambridge University Press

173

174

Francesca Ricci et al.

results obtained from the analysis of cranial morphology and, again, with some degree of gene flow with populations from other regions.

Postcranial Skeleton Postcranial Measurements and Indices Measurements of length and diameters were recorded on the long bones (according to Martin and Saller) with main postcranial indices calculated.48 Of interest are the values of the ‘diaphyseal index’ of the humerus that are indicative of flattening (platibrachy L 3

R=L L>R

2 1 0 Lower limb Upper limb Males 14

Upper limb Lower limb Females sections

N

12 10 8 R>L 6

R=L L>R

4 2 0 Upper limb

Lower limb

Males

Upper limb

Lower limb

Females

Figure 5.7. Side distribution of the asymmetry for the measurements of length and section: R = right; L = left.

suggests that the population was comprised of semi-sedentary herders, probably performing seasonal movements; this is also supported by ethnographic data as well as strontium isotope ratios.63 A similar conclusion could be suggested for Fewet, at least for the males, who were probably involved in the tending of livestock and in other tasks that required a certain level of mobility. Females, on the contrary, were probably occupied in more sedentary tasks probably related to oasis farming only.64 In this respect, the analyses of functional stress markers we performed on Sergi’s sample is also of interest. These analyses documented the existence of a clear division of

63

Cremaschi and di Lernia 2001; Puccioni 1967; Tafuri et al. 2006.

https://doi.org/10.1017/9781108634311.005 Published online by Cambridge University Press

64

Mori 2013a.

5 The Garamantes from Fewet (Ghat): A Skeletal Perspective

labour between the sexes.65 Specifically, males were significantly more affected than females by osteoarticular lesions in both the limbs, above all the legs, while females showed only a higher incidence of osteoarticular alterations in the cervical spine, probably due to the habit of carrying heavy loads on the head. The higher workload among males found in Sergi’s sample may suggest a stratified social organisation with a differentiation of roles.

Stature Regardless of the regression formula used, the mean stature at Fewet is remarkably low: for males it is 165.5 cm or 165.4 cm and for females it is 157.3 cm or 158 cm. Using data reported in the literature,66 Figure 5.8 shows statures (male means only) at Fewet and in other North African series, including those from Fazzan. All the comparative samples have a stature included in the range 165–180 cm, apart from Fewet (164.2 cm), which is actually characterised by the lowest stature values. It is well known that stature is a multifactorial phenotype, being a combination between polygenic inheritance, health status, nutrition, etc.67 In view of the other data we collected, it is reasonable to hypothesise that the extremely low stature of the people living at Fewet could also be the consequence of generally bad life conditions, including severe nutritional stress (see below) probably in relation to chronic malnutrition and/or metabolic diseases suffered during growth and development.68

Stress Markers and Dental Pathologies Life and health conditions at Fewet have also been explored by means of aspecific markers of stress (porotic hyperostosis and enamel hypoplasia) and dental pathologies (caries and ante-mortem tooth loss).

Porotic Hyperostosis Porotic hyperostosis has been recorded on adult (20 years+) and subadult crania (0–20 years).69 In particular, cribra orbitalia were recorded in seven degrees of severity, distinguishing between active and non-active (healed) 65 66 67 68

The data are unpublished; but see also Colaruotolo 2007. Calculated according to Manouvrier 1893. Formicola and Giannecchini 1999; Raxter 2011; Zakrzewski 2003. Bogin et al. 2007; Vercellotti et al. 2011. 69 Henschen 1961.

https://doi.org/10.1017/9781108634311.005 Published online by Cambridge University Press

177

178

Francesca Ricci et al.

185 180

cm

175 170 165 160

Fewet

Sergi’s

96/129

RT (H2)

RT (H1)

T1 (H1)

00/195 bis

00/195

UmgH2

Columnata

Taforalt

Hassi-elHabiod

Jebel Sahaba

Afalou

155

Figure 5.8. Stature estimated (Manouvrier 1893) for Fewet series and comparisons, males samples (values in cm). (Comparative data from Anderson 1968; Dutour 1989; Sergi 1951; Ricci et al. 2002.) Single individuals in grey, skeletal series in black.

lesions, while cribra cranii were recorded in three degrees of severity.70 The total frequency of cribra orbitalia at Fewet is 43.5 per cent (excluding degree one, 10 of of 23 individuals). The frequency is higher among subadults (71.4 per cent, 5 of 7), while adults show rather low prevalence. Fully healed and fully active lesions are completely absent in all the age classes, as well as lesions of degree six and seven. Amongst the subadults were the only examples of the highest degree of severity (five) and of lesions which had less than 50 per cent remodelling. By contrast, the total frequency of cribra cranii is 20 per cent (5 of 25), only one example of which was in a subadult, but this had the highest level of severity and the most active lesion (with less than 50 per cent remodelling). Among North African ancient populations, the prevalence of cribra orbitalia registered at Fewet is comparable to those documented for the roughly contemporary Garamantian sample excavated by C. M. Daniels in the Wadi al-Ajal (37.5 per cent of adults and adolescents n = 16) and for the Egyptians of the Dynastic period of Tarkhan (52.2 per cent), and remarkably higher than the total prevalence documented at Tombos, Nubia (11 per cent) and Jabal Sahaba, Paleolithic Nubia (12.5 per cent).71

70 71

Hengen 1971; Mittler and Van Gerven 1994; Ricci et al. 1997; Salvadei et al. 2001. Nikita et al. 2010; Starling 2005; Buzon 2006; Starling 2005.

https://doi.org/10.1017/9781108634311.005 Published online by Cambridge University Press

5 The Garamantes from Fewet (Ghat): A Skeletal Perspective

Porotic hyperostosis is generally considered a non-specific indicator of anaemic conditions induced by haematologic disorders (genetic or acquired), primarily iron deficiency anaemias suffered during early childhood as a consequence of poor nutritional conditions and/or gastro-intestinal infections.72 For other scholars, these lesions are rather indicative of chronic health problem/stressful conditions, or suggestive of other metabolic disorders such as scurvy and rickets.73 At any rate, given the complex nature of porotic hyperostosis, a multifactorial model is the most appropriate one when interpreting these lesions. At Fewet, the facial and postcranial skeleton does not show any lesions and bone deformities that may be referred to genetic anaemias.74 Moreover, the pattern observed of decreasing prevalence with the age of the individuals combined with the progressive healing are consistent with iron deficiency anaemia suffered during childhood; cribra orbitalia are then gradually reabsorbed during adolescence/adult life. At the same time, the higher frequency of remodelled cribra cranii among the adults could be the result of severe cases of anaemia suffered during infancy, leaving traces up to adulthood. In conclusion, it is possible to hypothesise that the occurrence of porotic hyperostosis at Fewet is related to general problems of malnutrition caused by insufficient intake of meat proteins/iron, particularly affecting the individuals during growth and development. Bad sanitary conditions, leading to high incidence of chronic diseases and parasitic infections, should also be considered as potential etiological sources.75

Enamel Hypoplasia and Dental Lesions Due to the poor state of preservation of the teeth, only a basic analysis of the oral health was performed. We recorded the presence of caries,76 antemortem tooth loss (AMTL),77 and enamel hypoplasia.78 The occurrence of caries on surviving permanent dentition is 78.6 per cent. The majority of the individuals examined had caries present on more than one tooth, and in the majority of cases, the degree of severity was high (penetrating or, more often, destructive); three children had caries on their deciduous teeth. 72 73

74 75 76 77 78

Benuš et al. 2010; Salvadei et al. 2001; Stuart-Macadam 1985; 1987; Walker et al. 2009. Carli-Thiele and Schultz 1997; Schultz 1993; Wapler et al. 2004; Ortner et al. 2001; Pitre et al. 2016. Hershkovitz et al. 1997; Ortner 2003. Holland and O’Brien 1997; Robledo et al. 1995; Sullivan 2005. Hillson 1996; Kelley and Larsen 1991. Recorded according to Benagiano 1977. Belcastro et al. 2004; Clarke et al. 1986. Recorded as presence/absence. Hillson 1996; Suckling 1989. Recorded in two degrees: light or strong.

https://doi.org/10.1017/9781108634311.005 Published online by Cambridge University Press

179

180

Francesca Ricci et al.

Teeth affected by caries were recorded in all the age classes and the prevalence is very high in all age classes (75–100 per cent). AMTL of permanent teeth is 72.2 per cent. It appeared in the age class 12–20 years (33.3 per cent of individuals affected) and strongly increased in the adult age classes 20-30 years (100 per cent) and 30–40 (87.5 per cent). The majority of individuals showed AMTL of more than one tooth and, unsurprisingly, the posterior teeth were particularly affected. The frequency of caries documented at Fewet (78.6 per cent) is higher than that registered in the Garamantian skeletal sample of C. M. Daniels (41.67 per cent), although both are quite high, and the prevalence of AMTL is similar (72.2 per cent versus 72.72 per cent).79 The high level of caries in Daniels’s sample was interpreted as the result of the farming Garamantian economy, which involves the cultivation of many cariogenic products,80 and the same conclusion may be valid for Fewet.81 This is also related to the poor oral hygiene as demonstrated by the high occurrence of AMTL.82 The prevalence of linear enamel hypoplasia in the Fewet sample is 75 per cent (permanent teeth); it occurred in all the age classes. The majority of the individuals examined show presence of hypoplasia on more than one tooth, although of light severity. As with the other observed pathologies, this suggests the presence of metabolic and nutritional disorders in the population during growth.83 These data strengthen, in particular, the interpretations based on porotic hyperostosis since these two markers have overlapping etiologies. Thus, it is probable that life at Fewet was harsh, particularly during infancy and childhood (as is also proved by the high infant mortality), with diffuse problems of malnutrition (scarce iron intake), high consumption of carbohydrates (even in the diet of children) and bad sanitary/hygienic conditions that caused chronic diseases and parasitic infections, especially at young age. This, in turn, presumably affected the overall growth process, which might be confirmed by the low mean stature registered in the sample.

Isotopic Data A preliminary isotopic investigation has been performed by measuring stable carbon (δ13C) and nitrogen (δ15N) isotopes (on human and animal

79 82

Nikita et al. 2010. 80 Nikita et al. 2010. 81 Nikita et al. 2010; Van der Veen 1992; 1995. Armelagos and Rose 1972; Clarke et al. 1986. 83 Corruccini et al. 1985; Suckling 1989.

https://doi.org/10.1017/9781108634311.005 Published online by Cambridge University Press

5 The Garamantes from Fewet (Ghat): A Skeletal Perspective

bone collagen) to examine food practices, and strontium isotope ratios (87Sr/86Sr; in human and animal enamel apatite) to explore residential mobility within the population. We report here a synthesis of our investigation, which is fully discussed elsewhere.84 Of the 30 samples (26 humans and four animals) prepared for collagen extraction, only 20 yielded good quality collagen (above 1 per cent collagen preserved) that could allow elemental analysis. Only 2 out of 20 specimens produced reliable stable carbon and nitrogen isotope data. δ13C values are indicative of a relatively high contribution of C4 plants to the diet, while high δ15N values might reflect the consumption of vegetal species thriving in dry environments. We obtained strontium isotope ratios on dental enamel for 13 humans and 11 (with five repeats) faunal samples (10 of Ovis vel Capra and 1 of Bos taurus). All humans came from the tumuli excavated at the cemetery, with the addition of one sample (FW 04) recovered at the nearby village of Fewet, located only a few hundred metres north of the cemetery. Faunal samples were also collected from the village excavations as no faunal specimens were recovered from the cemetery excavations. The Mean 87Sr/86Sr for the human enamel samples is 0.711027±0.000267, while the mean faunal enamel value is 0.711197±0.000171. The 87Sr/86Sr is consistent with the geological background of the area, mainly characterised by sandstone formations of the basement. We have compared the Sr signatures obtained at Fewet with that of other sites studied in the area: in particular, the (Early to Late) Pastoral site of Takarkori and the Late to Final Pastoral burials from site 96/129.85 Interestingly, there is a clear difference in the mean values from the three areas, with the earlier site of Takarkori showing the most radiogenic values, and the latest site of Fewet the least radiogenic values. Any interpretation of such isotopic variability requires a robust isotopic baseline built with samples of soils, vegetation, water and geologic materials. Looking at the Fewet data only, it is of interest to observe that there is greater variability in the ranges of females (0.001111) to males (0.0006). Considering that enamel signatures in the tooth enamel reflect Sr intake at early life, one could argue that some women at the village grew up at different locations (Fig. 5.9). This is coherent with the skeletal data (see above), which seem to suggest greater genetic variability in the female subset. 84 85

Tafuri et al. 2013. Further samples for investigation of Strontium Isotopes are being analysed. di Lernia and Tafuri 2013; Tafuri et al. 2006.

https://doi.org/10.1017/9781108634311.005 Published online by Cambridge University Press

181

182

Francesca Ricci et al.

0.7115

87Sr/86Sr

0.7110

0.7105

0.7100

0.7095

Infant Female Male Ovis Bos

Figure 5.9. Jitterplot with strontium isotope ratios of human (circles) and animal (squares and triangles) dental enamel at Fewet. The distributions and mean values clearly show the two subgroups of humans, with the faunaal data overlapping the distribution of the subgroup showing more radiogenic data.

The most interesting evidence is the clear patterning of human strontium isotope ratios in two separate clusters of individuals (Fig. 5.9). Mean values of the two subgroups of the population appear to differ significantly (Kruskal-Wallis test, p5 m) shafts of the pyramid tombs, the majority of the skeletal collection from the New Kingdom burials comes from the mudbrick chamber and shaft tombs. Intermixed with the New Kingdom pyramid tombs are additional Egyptian-style chamber and pyramid tombs with multiple interments 9 11 13

Smith 1998. 10 Morkot 1995; O’Connor 1993. Kendall 1999; Shinnie 1996; Török 1995. 12 Smith 1998; Smith and Buzon 2014. Smith 2003.

https://doi.org/10.1017/9781108634311.007 Published online by Cambridge University Press

7 Isotopic Approaches to Mobility in the Nile Valley

that date to the Napatan period (as well as reuse of some New Kingdom tombs during later times). Additionally, Nubian-style tumulus graves (mounded areas over vertical shafts, often with a chamber) with one or two individuals were also present in another section of the cemetery. Excavations concentrated on both types of burials, though the tumulus burials tended to be undisturbed and intact while the burials from the pyramid and chamber tombs were generally disturbed and commingled. The time span of Tombos from the New Kingdom to Napatan periods allows for the direct examination of individuals present during significant sociopolitical changes and population movements in the Nile Valley. As mentioned above, the adoption of Egyptian features by Nubians complicates the identification of colonists and locals who lived in Nubia during the New Kingdom period. Possible population composition changes that occurred during the development and rule of the Napatan leaders can also be examined. As one of the few sites occupied continuously during these periods, Tombos offers a unique opportunity to investigate human mobility in the past during important governmental transitions.

Isotopic Analysis and Bioarchaeological Research The application of isotopic analysis to the field of bioarchaeology has advanced tremendously in the past few decades.14 By measuring the ratio of isotopes of different elements (for example, carbon, nitrogen, lead, strontium, oxygen), we have the ability to address questions regarding diet, climate, mobility, status, gender and lifeways of ancient populations. These data frequently compliment traditional methods of archaeological research and can be used to help address broader questions about the ancient past (as discussed in the section Integrated Bioarchaeological Approach). Moreover, isotopic studies offer a unique perspective because skeletal material reflects a biological identity; this differs from a more socially influenced ethnic identity.15 The compilation of bioarchaeological isotopic data, such as this edited volume,16 is crucial to understanding isotopic values on a regional scale (for example in North Africa). In sum, isotopic analysis of human skeletal remains provides unique data that have the potential to greatly benefit future archaeological endeavours.

14 16

Ambrose and Krigbaum 2003. 15 Buzon 2006a. Power et al., Chapter 4, and Ricci et al., Chapter 5, this volume.

https://doi.org/10.1017/9781108634311.007 Published online by Cambridge University Press

227

228

Michele R. Buzon, Sarah A. Schrader and Gabriel J. Bowen

Strontium Isotope Analysis (87Sr/86Sr) The identification of first-generation immigrants in a burial sample is a common goal for isotope analyses of human skeletal and dental tissues. For strontium isotope analysis to be successful there must be sufficient geological variability between the site being investigated and the hypothesised origin locations of immigrants. Strontium that is present in food and water consumed by people is reflective of the geological age and bedrock composition of the region in which they lived during the formation of their skeletal and dental tissues if local foods have been ingested. The importation of nonlocal foods as well as the utilisation of sea salt (87Sr/86 Sr = sea water 0.7092) can mask local strontium signatures.17 The decay of 87Rb produces the radiogenic 87Sr isotope. Thus, 87Sr/86Sr reflects the 87Rb/87Sr of parent rocks and the amount of time over which radiogenic Sr has accumulated (i.e. rock age). Lower 87Sr/86Sr values generally come from younger rocks and rocks with low Rb content such as basalt (1 billion years old) are characterised by higher 87Sr/86Sr values (>0.7200).18 Strontium substitutes for calcium in the hydroxyapatite of bone and teeth. Adult dental enamel is thought to be more resistant to contamination than other skeletal tissues, due to its density and lack of porosity.19 Measurements obtained from dental enamel reflect strontium ingested during the time period of foetal growth or childhood when the tooth is developing (in utero until c.16 years).20 In contrast, data from bone reflect the last years of life (variable depending on which bone is used). Thus, it is possible to compare one’s strontium isotope signature from childhood and/or recent residence (if no diagenetic alteration has occurred) with that of one’s burial location; dental enamel and bone measurements from a single individual can also be compared to investigate movement during one’s life.21 Geological differences in the Nile Valley provide the basis for the usefulness of this method in the region. The sedimentary rock formations of Dakhla chalk, Esna shale and Theban limestone (with values c.87Sr/86 Sr = 0.70907, deposited 35–65 million years ago) characterise Egypt.22 Nubia consists primarily of pre-Cambrian Basement Complex with igneous, metamorphic and sedimentary rocks covered with the Nubian Sandstone Complex. The major cataracts of the Nile River are outcrops of granite.23 Establishing the local strontium isotope signature for a site is usually done 17 20

Fenner and Wright 2014. 18 Faure 1986. 19 Horn et al. 1994; Montgomery 2010. Hillson 1996. 21 Montgomery 2010. 22 Burke et al. 1982. 23 Whiteman 1971.

https://doi.org/10.1017/9781108634311.007 Published online by Cambridge University Press

7 Isotopic Approaches to Mobility in the Nile Valley

Memphis EGYPT Qurneh Shellal NUBIA C-Group Pharaonic Amara West Tombos Napatan Tombos New Kingdom Kerma .7065

.707

.7075

.708

.7085

.709

87Sr/86Sr

Figure 7.2. Distribution of human 87Sr/86Sr values for human samples in the Nile Valley; bar indicates median value.

using soil, geological and faunal samples; however, small animals with limited home ranges appear to be best for characterising the biologically available strontium in a particular place. While modern samples can be used, archaeological dental enamel samples are best for establishing the strontium signature in antiquity (modern chemicals change the local value).24 For the Nile Valley region, few small animal samples were available to use. Measurements taken from larger animals such as sheep/goat, pig, dog and cattle expectedly resulted in a wide range of values (0.70649–0.71248).25 The examination of human samples in the Nile Valley indicates that there is overlap between sites (Fig. 7.2). However, when examining the median 87 Sr/86Sr values for all of the samples,26 the Egyptian (mean/median 87Sr/86 Sr = 0.70777) sites have statistically significantly higher values than the Nubian sites (mean/median 87Sr/86Sr = 0.70757); additionally, the median values decrease from north to south (the exceptions of Tombos New Kingdom and Amara West will be further discussed in this section). Thus, while concretely identifying Egyptians and Nubians is not straightforward, careful examination of the data distributions and median values can offer some information about possible human mobility in the region. Continued testing of animals, with small home ranges such as rodents (like the ongoing research at Amara West) will aid in our understanding of the local ranges for specific sites. 24

Price et al. 2002.

25

Buzon and Simonetti 2013.

26

Buzon and Simonetti 2013.

https://doi.org/10.1017/9781108634311.007 Published online by Cambridge University Press

229

230

Michele R. Buzon, Sarah A. Schrader and Gabriel J. Bowen

Modern rodent samples tested from Tombos provided a more narrow range for that site of 87Sr/86Sr = 0.70710–0.7078327 using the standard method of the mean of the faunal samples ±2 standard deviations.28 Due to diagenetic contamination of bone samples, only dental enamel samples were used: 53 from the New Kingdom burials and 32 from the Napatanperiod burials. As previously detailed, the main questions being asked of the Tombos material involve sociopolitical activities during these time periods: Did Egyptian colonists move to Tombos to participate in the administration of the Egyptian empire in Nubia? Did local Nubians work with the Egyptians? Because many Nubians are Egyptianised during this time, can we use isotope data to distinguish Egyptians from Nubians at Tombos? Did the movement of individuals from farther south in Nubia or from Egypt contribute to the population at Tombos during the Napatan Twenty-fifth Dynasty of Egypt? Eighteen individuals from the New Kingdom sample had 87Sr/86Sr values above the established local range at Tombos. These higher values could be attributed to immigrants from Egypt who moved to Tombos to participate in the colonial governance.29 There appears to be a larger proportion of Egyptian immigrants in the early years of the Tombos colony, suggesting that migration waned over time. Examination of burial practices, craniometric indications of ancestry and isotope data indicate that after colonists arrived, Egyptians and local Nubians created a culturally and biologically entangled community.30 None of the New Kingdom individuals buried in local Nubian style have 87Sr/86Sr values outside of the local range. For the Napatan-period sample, all are within the local range except for two individuals. Unlike the New Kingdom ‘nonlocal’ individuals, these two samples have 87Sr/86Sr values that are below, rather than above, the local range. It is suggested that these lower values could be reflective of upstream Nubian locales, as they are within the preliminary local range determined for a site (Ginefab School Site) in the Fourth Cataract region near the royal Napatan burial places.31 Thus, the Napatan sample represents the descendants of the Egyptian immigrants and local Nubians that interacted in the New Kingdom. After the fall of the New Kingdom Empire, Tombos did not disintegrate. On the contrary, it appears to have continued to be a prominent site within the broader Napatan polity as evidenced by the size and types of tomb structures present in the cemetery.32 These data suggest that new immigrants from Egypt were not a part of the Napatan 27 30

Buzon and Simonetti 2013. 28 Price et al. 2002. 29 Buzon et al. 2007. Buzon et al. 2016. 31 Masoner et al. 2011; Buzon and Simonetti 2013.

https://doi.org/10.1017/9781108634311.007 Published online by Cambridge University Press

32

Smith 2006.

7 Isotopic Approaches to Mobility in the Nile Valley

governance at Tombos, though some individuals from further south may have been present.

Oxygen Isotope Analysis (δ18O) Researchers have investigated ancient migration by measuring and studying the abundance of two stable isotopes of oxygen in human tissues: 18 O and 16O (the ratio between expressed as δ18O [‰, all values reported relative to V-SMOW]). Oxygen in the human body comes from consumed water, water and organic-bound O found in consumed plants and animals, and inhaled oxygen.33 Hydrological, geographical and climatological factors impact oxygen isotope ratios of water sources that supply water utilised by plants, animals and humans. Temperature, humidity, rainfall, latitude, elevation and distance from the coast affect environmental (rivers, springs, lakes) and meteoric (rain, snow) waters,34 and isotope ratios of oxygen in these waters can be further modified as oxygen from water is incorporated and transferred through food webs. Hydroxyapatite carbonate and phosphate in human dental enamel and bone can be measured to determine oxygen isotope ratios. These measurements reflect the environment in which a person lived while the tooth (childhood) or bone (approximately the last decade) was forming, if the majority of ingested oxygen came from local water sources directly or via consumption of locally sourced foods.35 Several researchers have presented equations that can be used to approximate the δ18O value of a person’s local water source from human tissue samples.36 However, a recent publication indicates that these equations introduce an unacceptable magnitude of error; these authors recommend that δ18O values from human tissues should not be used as an approximation of local water but should only be compared with other human samples.37 This issue complicates our ability to identify the potential water source used by individuals. Another complicating factor relates to the effects of metabolic processes, such as those associated with disease and exercise, on oxygen isotope ratios. Recently, Reitsema and Crews presented data that suggest lower oxygen isotope ratios in mice with sickle-cell disease.38 In the Mediterranean region, there is a well-known relationship between high 33 36

37

Longinelli 1984. 34 Bowen et al. 2007; Dansgaard 1964; Gat 1996. 35 Luz et al. 1984. On phosphate: Daux et al. 2008; Iacumin et al. 1996a; Longinelli 1984; Luz et al. 1984. On carbonate-phosphate relationship: Iacumin et al. 1996b. Pollard et al. 2011. 38 Reitsema and Crews 2011.

https://doi.org/10.1017/9781108634311.007 Published online by Cambridge University Press

231

232

Michele R. Buzon, Sarah A. Schrader and Gabriel J. Bowen

frequencies of the sickle-cell trait and the prevalence of malaria, as having one copy of the sickle-cell allele provides partial resistance to malaria.39 While the parasite (P. falciparum) that causes malaria has been identified in some mummies,40 many of the human remains excavated from archaeological sites do not have this level of preservation. Identifying sickle-cell disease in skeletal remains can be problematic, though other general indications of anaemia (skeletal conditions such as cribra orbitalia and porotic hyperostosis) can be observed.41 Given the likelihood that malaria and thus sickle-cell disease were present in the ancient Mediterranean, these issues may affect the oxygen isotope ratios measured in archaeological human tissues. Previous oxygen isotope research in Egypt has indicated that it is possible to distinguish individuals who originated in the Nile Valley from those with desert oasis origins.42 In our earlier study,43 we hypothesised that δ18O values would increase downstream (heading north) along the Nile due to the preferential loss of 16O through evaporation as the river flows away from its source in central Africa to the Mediterranean Sea;44 modern values reflect this trend (δ18O at Nile source −5‰,45 δ18O at Nile delta +2.0–4.0‰46). However, a survey of published human δ18O values from various Nile Valley sites indicates a different pattern.47 Some of the highest values occur at Nubian sites, around 4‰ higher than those at Egyptian sites to the north. We suggested that one explanation may relate to different than expected regional variation in the isotope values of water used at distinct points along the river. This could be due to factors such as the addition of low δ18O groundwater to river discharge (Nubian aquifer) or shallow groundwater reservoirs of low-δ18O flood water returned to the river in the dry season. Alternatively, variable water-use strategies by residents in different locations, such as their use of wells (such as at Amarna)48 or techniques involving evaporation such as irrigation, storage and brewing of beer,49 might overprint the expected downstream increase in Nile water δ18O values and dominate the oxygen isotope signal in the human samples. Additional samples from various sites in the Nile Valley were measured for oxygen isotope ratios in an attempt to further our understanding of the 39 42 44 45 46

47

Wiesenfeld 1967. 40 Bianucci et al. 2008. 41 Hershkovitz et al. 1997; Walker et al. 2009. Dupras and Schwarcz 2001. 43 Buzon and Bowen 2010. Iacumin et al. 1996a; White et al. 2004. International Atomic Energy Agency/World Meteorological Organization 2006. Research Institute for Groundwater/International Workshop on Aliasing, Confinement and Ownership 1990. Buzon and Bowen 2010. 48 Kemp 1987. 49 Butzer 1976; Kemp 1989.

https://doi.org/10.1017/9781108634311.007 Published online by Cambridge University Press

7 Isotopic Approaches to Mobility in the Nile Valley

isotope variability in the region (Table 7.1).50 Some trends are apparent in the summary data (Table 7.2, Fig. 7.3).51 The Kruskall-Wallis test of all samples indicates highly significant differences (p = 0.000) between the sites for δ18O values. Variability is significantly higher in the Nubian samples in comparison with the Egyptian samples (ANOVA, p = 0.000). Pairwise comparison of the means/medians between Nubian (δ18O = 33.1‰) and Egyptian (δ18O = 31.7‰) samples also reveals significant differences (t-test, Mann-Whitney, p = 0.000). Three regions have samples that represent more than one temporal period (Table 7.2). In the Second Cataract area, the mean δ18O values remain relatively consistent over time (C-Group 32.5‰, Pharaonic 33.4‰, Wadi Halfa 32.7‰). Further south, Amara West (35.2‰) and Kulubnarti (34.0‰) also have similar values. At the Third Cataract, values appear to increase slightly through time (Kerma 30.8‰, Tombos New Kingdom 31.5‰, Tombos Napatan 33.5‰). These additional data support our earlier findings of higher values and more variability at Nubian sites. Yet, it is evident that differentiating between Egyptian and Nubian individuals is complex, despite the differences in variability and mean/median δ18O values. While individuals from Egyptian sites tend to have lower δ18O values and the highest values appear exclusively in Nubia, there is clear overlap in the values from Egyptian and Nubian sites. The explanations previously presented continue to be reasonable in light of the additional Egyptian and Nubian samples presented here. The lower δ18O values in the Egyptian sites could be at least partially accounted for by the mixing of low δ18O water from the Nubian Aquifer (average δ18O value of enamel samples from humans drinking aquifer water at Dakhleh Oasis is 28.8‰, range 24.3–31.6‰),52 either through direct extraction and use of aquifer water from wells or springs, or through natural discharge of this water to the Nile River.53 Only the sites north of the Wadi Halfa would have direct access to aquifer water (see aquifer boundary in Fig. 7.1). The aquifer again becomes accessible to the Nile Valley just south of the Third Cataract, which may provide an explanation for the lower δ18O values found at Kerma. Moreover, the sites with lower δ18O values (Egyptian sites and Kerma in Nubia) are located in areas that are very fertile and productive with broad flood-plains that allow for large 50

51

52

Our methodology followed Buzon and Bowen 2010, with analysis at the Purdue Stable Isotope Facility. The complete Nile Valley δ18O dataset can be found at: http://wateriso.utah.edu/waterisotopes /pages/data_access/data_tables.html. Dupras and Schwartz 2001. 53 Buzon and Bowen 2010.

https://doi.org/10.1017/9781108634311.007 Published online by Cambridge University Press

233

234

Michele R. Buzon, Sarah A. Schrader and Gabriel J. Bowen

Table 7.1. δ18O values for human tooth samples in the Nile Valley. Sample

Tooth

δ18O ‰ VSMOW

AMR-1 AMR-2 AMR-3 AMR-4 AMR-5 AMR-7 AMR-8 AMR-9 AMR-10 CGR-1 CGR-2 CGR-3 CGR-4 CGR-5 CGR-6 CGR-8 CGR-9 CGR-10 CGR-11 CGR-12 CGR-13 CGR-14 KER-7 KER-8 KER-9 KER-10 KER-11 KER-12 KER-13 KER-14 KER-15 MEM-1 MEM-3 MEM-4 MEM-6 MEM-7 MEM-8 MEM-9 MEM-10

P M1 I2 M3 M2 P P4 P3 M1 P3 P4 P4 P3 P4 P3 P4 P4 P3 P4 P4 P3 P3 P4 P3 P4 P4 P4 P3 P3 P4 P3 P4 P3 P4 P4 P4 P3 P4 P3

34.2 34.0 34.0 34.0 34.0 33.0 39.1 39.4 35.3 31.7 32.0 32.6 34.8 33.5 32.4 31.6 32.9 32.8 32.9 32.6 30.3 31.8 31.2 31.4 29.8 30.2 30.4 32.3 29.5 30.3 31.6 30.5 31.5 32.7 32.9 31.8 32.3 31.0 31.6

Sample

Tooth

δ18O ‰ VSMOW Sample

MEM-11 MEM-12 MEM-14 MEM-15 PHR-1 PHR-4 PHR-6 PHR-7 PHR-8 PHR-9 PHR-12 PHR-14 PHR-15 QUR-1 QUR-2 QUR-3 QUR-4 QUR-6 QUR-7 QUR-8 QUR-9 QUR-10 QUR-11 QUR-12 QUR-13 QUR-14 QUR-15 SHL-1 SHL-2 SHL-3 SHL-4 SHL-6 SHL-7 SHL-8 SHL-9 SHL-10 SHL-11 SHL-12 SHL-13

P3 P3 P3 P3 P3 P4 P3 P4 P3 P3 P4 P3 P4 P4 P3 P3 P4 P3 P4 P3 P3 P3 P4 P3 P4 P4 P3 P3 C P3 C P3 P3 P3 P3 P4 P3 P3 P3

33.2 32.7 31.0 32.7 32.2 32.8 33.4 33.0 32.6 33.0 33.6 36.2 33.9 32.4 33.1 33.0 30.5 32.0 32.6 30.7 31.3 34.1 32.2 30.2 32.5 32.8 32.3 33.2 33.1 31.9 32.7 32.8 33.3 32.4 32.8 32.5 33.0 32.4 33.1

https://doi.org/10.1017/9781108634311.007 Published online by Cambridge University Press

SHL-15 TOM-100 TOM-101 TOM-102 TOM-103 TOM-104 TOM-105 TOM-106 TOM-107 TOM-108 TOM-109 TOM-110 TOM-111 TOM-112 TOM-113 TOM-114 TOM-115 TOM-116 TOM-117 TOM-118 TOM-119 TOM-120 TOM-122 TOM-123

Tooth

δ18O ‰ VSMOW

P3 M3 M2 M3 M1 M1 P3 M3 P3 M1 M1 P4 P3 I1 M1 P3 P3 M1 P3 P4 P3 P4 M2 P3

32.7 33.9 34.7 33.6 32.2 34.7 34.3 32.4 33.7 33.5 33.1 33.7 35.3 31.7 33.7 33.2 33.0 34.6 33.5 33.9 33.8 31.3 31.7 34.0

7 Isotopic Approaches to Mobility in the Nile Valley

235

Table 7.2. Average δ18O values and ranges for human samples in the Nile Valley. Notes: 1. Sites are arranged from north to south in the Nile Valley; 2. Only permanent teeth are included in these calculations; 3. Curated collection dated to the New Kingdom Period, dates approximate; 4. Estimated average latitude of sites located in a 60 km area. Site Egypt

Nubia

1

N

Mean δ18O ‰ VSMOW (range)

Std Dev

2

Latitude Time period 3

Memphis Asyut

29.85 27.18

1540–1070 BC 2150–2050 BC

12 8

32.0 (30.5–33.2) 31.0 (30.2–32.1)

0.77 0.34

Qurneh Gebelein

25.90 25.48

1540–1070 BC3 3500–2600, 2150–2050 BC 1540–1070 BC3 350–1400 AD

14 11

32.1 (30.2–32.1) 31.5 (30.6–32.6)

1.18 0.24

13 74

32.8 (31.9–33.3) 32.7 (29.3–35.8)

0.16 1.78

1650–1350 BC 2000–1600 BC 500–800 AD

9 13 85

33.4 (32.2–36.2) 32.5 (30.3–34.8) 34.0 (32.5–36.3)

1.39 1.16 0.70

Amara 20.83 West Tombos –A 19.71

1540–650 BC

9

35.2 (33.0–39.4)

5.50

1400–1050 BC

23

31.3 (29.2–35.3)

2.09

Tombos –B 19.71 Kerma 19.66

1050–650 BC 1680–1550 BC

28 9

33.5 (31.3–35.3) 30.8 (29.5–32.4)

1.06 0.86

Shellal 24.05 Wadi Halfa 21.78 Pharaonic C-Group Kulubnarti

21.784 21.784 21.08

cultivation areas. In contrast, Lower Nubia has an extremely narrow floodplain confined by rocky crests and slopes resulting in much lower agricultural yields.54 This environmental variation may have required different water and agricultural practices, such as the use of ponds and dry-land crops that would have resulted in evaporative differences affecting isotopic values. Strontium and oxygen isotope data are both available for 26 New Kingdom and 23 Napatan samples (Fig. 7.4). Ten of the New Kingdom samples and none of the Napatan samples were hypothesised to be of Egyptian origin. These ten samples have a mean/median value of δ18O = 31.5/31.3‰. The 39 samples considered to be within the local 54

Adams 1977; Butzer 1976.

https://doi.org/10.1017/9781108634311.007 Published online by Cambridge University Press

Reference Present study Iacumin et al. 1996a Present study Iacumin et al. 1996a Present study White et al. 2004 Present study Present study Turner et al. 2007 Present study Buzon & Bowen 2010 Present study Present study

236

Michele R. Buzon, Sarah A. Schrader and Gabriel J. Bowen

Memphis Asyut aquifer access

EGYPT

Qurneh Gebelein Shellal Wadi Halfa CGP/Pharaonic NUBIA Kulubnarti Amara West Tombos Kerma 30

32

34

36

38

40

δ18O ‰ VSMOW

Figure 7.3. Distribution of δ18O values for human samples in the Nile Valley. 36

δ180 ‰ VSMOW

34

32

30

28 Local (87Sr/86Sr)

Non-Local (87Sr/86Sr)

Figure 7.4. Distribution of δ18O values for individuals identified as local and nonlocal using strontium isotope analysis.

range have a mean/median value of δ18O = 32.5‰. Using the parametric t-test, these means are statistically different. However, the possible Egyptian group is not normally distributed, which necessitates the nonparametric Mann-Whitney test for which the medians are not significantly

https://doi.org/10.1017/9781108634311.007 Published online by Cambridge University Press

7 Isotopic Approaches to Mobility in the Nile Valley

different. If the outlier with a very high value of 35.3‰ is removed, the medians (new nonlocal δ18O = 31‰) are found to significantly differ.

Mobility in Northern Africa and the Mediterranean As a case study, isotopic analyses of the Tombos burials provide an example of how these chemical markers can be used to investigate human mobility during various sociopolitical circumstances in the Nile Valley. These techniques have certainly been utilised to answer related questions at many ancient sites around the world, but they are still relatively underdeveloped for Africa. Nonetheless, the studies that have been completed show promise for these methods. Over the last several thousand years, changes in food production and subsistence along with the rise and fall of numerous polities in the region of northern Africa and the Mediterranean provide excellent opportunities to investigate the role that human mobility played in these transitions. As with all types of analyses, there are some difficulties involved in the reconstruction of human movements using isotopes. For instance, Stojanowski and Knudson highlight the issue that strontium isotope signatures in the Sahara are affected by Aeolian dust transport, resulting in homogenous values.55 While this is not the case for investigations of ancient mobility before the desertification of the region, the use of modern faunal samples can be biased as Saharan dust usually has high values (87Sr/86Sr = 0.72256). Nonetheless, the Sahara possesses various sediments of numerous ages from older Pre-Cambrian basement to young volcanic rocks, differences that suggest identifying past human movement may be possible. Strontium isotope studies of human samples by a few researchers reveal potential in the variability of 87Sr/86Sr values across the Trans-Saharan region. Ricci et al. report human 87Sr/86Sr values from Fewet (Fazzan) of 0.711027 ± 0.000267.57 Similarly, the work by Tafuri and colleagues in the Tadrart Akakus and Massak (also Fazzan) reveals a range of human values from 87Sr/86Sr = 0.70975–0.71206, consistent with geological values reported for the region (sandstone).58 Stojanowski and Knudson report human values from central Niger that range 87Sr/86Sr = 0.7112–0.7116.59 55 57 58 59

Stojanowski and Knudson 2011; Krom et al. 1999. 56 Weldeab et al. 2002. Ricci et al., Chapter 5, this volume. Tafuri et al. 2006; cf. Schaaf and Müller-Sohnius 2002. Stojanowski and Knudson 2011; 2014.

https://doi.org/10.1017/9781108634311.007 Published online by Cambridge University Press

237

238

Michele R. Buzon, Sarah A. Schrader and Gabriel J. Bowen

As reported above, Nile Valley human 87Sr/86Sr values range from 0.70658–0.70912.60 Furthermore, additional regional isotopic studies, including Power et al., are currently awaiting results.61 These studies reveal, at minimum, some apparent 87Sr/86Sr differences in north-east Africa versus central northern Africa. With broader interests in mind focusing on human mobility that occurred in relation to subsistence and sociopolitical changes, it is useful to consider a larger region for the origin of possible immigrants to northern Africa. For instance, the Roman Empire encompassed Britain, much of continental Europe and the Mediterranean regions of northern Africa, Europe, Asia Minor and the Levant; mobility was common among the military, free citizens and non-citizens.62 Many of these areas have produced human values that could be differentiated from those in central North Africa (87Sr/86Sr = 0.711–0.712). Rome and its suburbia has a local 87Sr/86Sr range ~0.7079–0.7103.63 Values in Britain overlap in the upper ranges with northern Africa, with some differentiation in the low end of local values at 87Sr/86Sr = ~0.707.64 Bavaria is characterised by 87Sr/86Sr = ~0.708–0.710.65 Sites in the Aegean also have comparatively low 87Sr/86Sr values ranging 0.708–0.709.66 A study from Jordan produced similarly low values ranging 87Sr/86Sr = 0.707–0.708.67 Clearly, the local 87Sr/86Sr signatures of many areas overlap, though it appears that individuals originating outside central North Africa would be visible using this technique. In Europe and North Africa, δ18O precipitation values become more positive closer to the equator. As a result, North Africa has δ18O values that are more positive than southern Europe, which are more positive than northern Europe.68 Measured human samples from Rome range δ18O = 24.7–27.1‰.69 Similarly, human values from Greece range δ18O = 25.4–28.0‰.70 Nile Valley human samples encompass a range of δ18O = ~29.2–39.4‰.71 West of the Nile Valley region, δ18O values are likely lower, due to the use of aquifer versus river water. Dupras and Schwarcz found the local δ18O at Dakhlah Oasis in the Western Desert of Egypt to range δ18O = ~24–31‰.72

60 62 64 65 68 71

Buzon and Simonetti 2013. 61 Power et al., Chapter 4, this volume. See for example, Noy 2000; Scheidel 2004. 63 Killgrove 2010. Chenery et al. 2010; 2011; Eckhardt et al. 2009; Evans et al. 2006. Schweissing and Grupe 2003. 66 Nafplioti 2008; 2011. 67 Perry et al. 2008. Bowen and Revenaugh 2003. 69 Killgrove 2010; Prowse et al. 2007. 70 Garvie-Lok 2009. Buzon and Bowen 2010. 72 Dupras and Schwarcz 2001.

https://doi.org/10.1017/9781108634311.007 Published online by Cambridge University Press

7 Isotopic Approaches to Mobility in the Nile Valley

The variability in 87Sr/86Sr and δ18O values in the Mediterranean region provides an indication that these techniques may be useful for tracing ancient human migration. It is clear that these values overlap for many locations in the larger region. However, used in conjunction, they offer a more robust way to identify possible immigrants. It may also prove useful to include analyses of additional elements such as lead 207Pb/206Pb, 208 Pb/206Pb, as another marker of locale, sulfur (34S/32S) as an indicator of marine resource use, and carbon (12C/13C) and nitrogen (15N/14N) in order to discern different dietary practices. As suggested above, other recoverable aspects of burial ritual (material indications of ethnic identity) and skeletal remains (demographic and paleopathological variables) would also enhance the investigation of immigrant individuals and human mobility in the past.

Integrated Bioarchaeological Approach The research that has been conducted at Tombos has benefitted greatly from taking an integrated bioarchaeological approach to the excavations and project as a whole (Directors Stuart Tyson Smith, University California Santa Barbara, and Michele R. Buzon, Purdue University). The study of human remains is integrated with theory, methods and data from history, archaeology, cultural anthropology, biological anthropology, medical sciences, geography and other relevant fields through an interdisciplinary bioarchaeological perspective.73 Starting from a balanced partnership between archaeology and osteology during the formulation of research questions and site planning, along with the treatment of skeletal data as a form of material culture on par with analyses of other artefacts such as ceramics, results in a more fruitful outcome that utilises all of the information available to address research topics.74 Participation of osteological experts during excavation can ensure the careful recording of the skeleton and its relationship to other features and taphonomic changes as well as other skeletal material (especially when commingled), a practice consistent with excavation procedures put forth by Duday’s ‘l’anthropologie de terrain’ approach.75 Over the last decade or so, several different types of analyses have been integrated in order to address questions of sociopolitical transition, identity (ethnic/cultural, geographical, biological), health, diet and 73

Buzon et al. 2005; Larsen and Walker 2010.

74

Buzon 2012.

https://doi.org/10.1017/9781108634311.007 Published online by Cambridge University Press

75

Duday 2006.

239

240

Michele R. Buzon, Sarah A. Schrader and Gabriel J. Bowen

activity at Tombos in the New Kingdom and Napatan periods.76 With regard to the isotopic data presented here, ethnic identity via burial ritual and biological identity via craniometric analyses have been examined in conjunction with the strontium and oxygen isotope results. For the New Kingdom samples, all of the 18 individuals with 87Sr/86Sr values above the local range were buried using Egyptian rituals. Within the remaining 35, the only four burials buried using Nubian rituals in this sample are within the local range, with δ18O values ranging 29.2–30.6‰ (three samples). The 31 burials that have a local strontium isotope signature and are buried using Egyptian features could be local Nubians who have been Egyptianised or the progeny of Egyptian colonists born in Nubia.77 Craniometric data portray a more complicated picture. While there are some shape differences between Egyptian and Nubian cranial morphology, admixture over time in the Nile Valley has resulted in many similarities.78 Individuals with both predicted Egyptian and Nubian cranial morphology appear in both the local and non-local strontium groups.79 This ethnically and biologically mixed group of people who lived at Tombos were likely administrative workers and artisans supporting the colonial government. Examinations of the skeletal conditions reveal that there are very few traumatic injuries,80 little evidence of hard physical labour81 and few signs of ill health.82 Overall, the multiethnic community at Tombos appears to have been peaceful, cooperative and prosperous. Within the Napatan sample, all individuals except for two individuals with low values fall within the local range. In the aftermath of the New Kingdom Empire, burials from this time period at Tombos displayed a sense of multivocality in their burial traditions. Both Egyptian- and Nubian-style pottery was recovered. Some were buried in Nubian-style tumulus graves, though usually with Egyptian extended burial position (all but one), while others were found in Egyptian-style pyramid and chamber tombs, despite isotopic indication that the vast majority of the individuals were local. It seems likely that the Napatan community at Tombos descended from the people who lived there during the New Kingdom.83 In comparison with the New Kingdom sample, the Napatan period 76

77 80 82 83

Buzon 2006a; 2006b; 2014; Buzon and Bombak 2010; Buzon and Bowen 2010; Buzon and Richman 2007; Buzon et al. 2016; Schrader 2012; 2013; Schrader et al. 2014; Smith 1998; 2003; 2007; Smith and Buzon 2014. Buzon and Simonetti 2013; Buzon et al. 2007. 78 Buzon 2006a. 79 Buzon et al. 2007. Buzon and Richman 2007. 81 Schrader 2012. Buzon 2006b; Buzon and Bombak 2010. Buzon et al. 2016; Schrader et al. 2014; Smith and Buzon 2014.

https://doi.org/10.1017/9781108634311.007 Published online by Cambridge University Press

7 Isotopic Approaches to Mobility in the Nile Valley

individuals at Tombos are larger in size,84 show signs of increased recovery from nutritional deficiency and infection85 and appear to have been engaged in more strenuous physical labour.86 After the fall of the Egyptian New Kingdom administrative framework, communities in Upper Nubia, such as Tombos, continued to endure and may have adapted to the changing sociopolitical environment altering their daily activities through the inclusion of increased subsistence activities and granite quarrying.87

References Adams, W. 1977. Nubia: Corridor to Africa. Princeton: Princeton University Press. Ambrose, S. H. and Krigbaum, J. 2003. Bone chemistry and bioarchaeology. Journal of Anthropological Archaeology 22.3: 193–99. Bentley, R. 2006. Strontim isotopes from Earth to the archaeological skeleton: a review. Journal of Archaeological Method and Theory 13:235–87. Bianucci, R., Mattutino, G., Lallo, R., Charlier, P., Jouin-Spriet, H., Peluso, A., Higham, T., Torre, C. and Rabino Massa, E. 2008. Immunological evidence of Plasmodium falciparum infection in an Egyptian child mummy from the Early Dynastic Period. Journal of Archaeological Science 35.7:1880–85. Bowen, G. and Revenaugh, J. 2003. Interpolating the isotopic composition of modern meteoric precipitation. Water Resources Research 39: 1299–311. Bowen, G., Ehleringer, J., Chesson, L., Stange, E. and Cerling, T. 2007. Stable isotope ratios of tap water in the contiguous USA. Water Resources Research 43: W03419. Budd, P., Millard, A., Chenery, C., Lucy, S., and Roberts, C. 2004. Investigating population movement by stable isotope analysis: a report from Britain. Antiquity 78:127–41. Burke, W., Denison, R., Hetherington, E., Koepnick, R., Nelson, N. and Otto, J. 1982.Variation of seawater 87Sr/86Sr throughout Phanerozoic time. Geology 10: 516–19. Butzer, K. 1976. Early Hydraulic Civilization in Egypt: A Study in Cultural Ecology. Chicago: University of Chicago Press. Buzon, M. 2006a. Biological and ethnic identity in New Kingdom Nubia: a case study from Tombos. Current Anthropology 47: 683–95. Buzon, M. 2006b. Health of the non-elites at Tombos: nutritional and disease stress in New Kingdom Nubia. American Journal of Physical Anthropology 130: 26–37.

84 87

Gibbon and Buzon 2016. 85 Buzon 2014. 86 Schrader 2013. Buzon et al. 2016; Smith and Buzon 2014; Schrader 2013.

https://doi.org/10.1017/9781108634311.007 Published online by Cambridge University Press

241

242

Michele R. Buzon, Sarah A. Schrader and Gabriel J. Bowen

Buzon, M. 2012. The bioarchaeological approach to paleopathology. In A. Grauer (ed.), A Companion to Paleopathology, West Sussex: Wiley-Blackwell, 58–75. Buzon, M. 2014. Tombos during the Napatan period (c.750–660 BC): Exploring the consequences of sociopolitical transitions in ancient Nubia. International Journal of Paleopathology 7: 1–7. Buzon, M. and Bombak, A. 2010. Dental disease in the Nile Valley during the New Kingdom. International Journal of Osteoarchaeology 20.4: 371–87. Buzon, M. and Bowen, G. 2010. Oxygen isotope analysis of migration in the Nile Valley. Archaeometry 52: 855–68. Buzon, M., Eng, J., Lambert, P. and Walker, P. 2005. Bioarchaeological methods. In H. D. G. Maschner and C. Chipindale (eds), Handbook of Archaeological Methods, Lanham: AltaMira Press, 871–918. Buzon, M. and Richman, R. 2007. Traumatic injuries and imperialism: the effects of Egyptian colonial strategies at Tombos in Upper Nubia. American Journal of Physical Anthropology 133: 783–91. Buzon, M. and Simonetti, A. 2013. Strontium isotope (87Sr/86Sr) isotope variability in the Nile Valley: identifying residential mobility during ancient Egyptian and Nubian sociopolitical changes in the New Kingdom and Napatan periods. American Journal of Physical Anthropology 151: 1–9. Buzon, M., Simonetti, A. and Creaser, A. 2007. Migration in the Nile Valley during the New Kingdom period: a preliminary strontium isotope study. Journal of Archaeological Science 34: 1391–401. Buzon, M., Smith, S. T. and Simonetti, A. 2016. Entanglement and the formation of the ancient Nubian Napatan state. American Anthropologist 118: 284–300. Chenery, C., Eckardt, H., and Müldner, G. 2011. Cosmopolitan Catterick? Isotopic evidence for population mobility on Rome’s northern frontier. Journal of Archaeological Science 38.7: 1525–36. Chenery, C., Müldner, G., Evans, J., Eckardt, H. and Lewis, M. 2010. Strontium and stable isotope evidence for diet and mobility in Roman Gloucester, UK. Journal of Archaeological Science 37: 150–63. Dansgaard, W. 1964. Stable isotopes in precipitation. Tellus 16: 436–68. Daux, V., Lécuyer, C., Héran, M., Amiot, R., Simon, L., Fourel, F., Martineau, F., Lynnerup, N., Reychler, H. and Escarguel, G. 2008. Oxygen isotope fractionation between human phosphate and water revisited. Journal of Human Evolution 55: 1138–47. Duday, H. 2006. L’archéothanatologie ou l’archéologie de la mort (archaeothanatology or the archaeology of death). In R. Gowland and C. Knüsel (eds), Social Archaeology of Funerary Remains, Oxford: Oxbow, 30–56. Dupras, T. and Schwarcz, H. 2001. Strangers in a strange land: stable isotope evidence for human migration in the Dakhleh Oasis, Egypt. Journal of Archaeological Science 28: 1199–208.

https://doi.org/10.1017/9781108634311.007 Published online by Cambridge University Press

7 Isotopic Approaches to Mobility in the Nile Valley

Eckardt, H., Chenery, C., Booth, P., Evans, J., Lamb, A. and Müldner, G. 2009. Oxygen and strontium isotope evidence for mobility in Roman Winchester. Journal of Archaeological Science 36.12: 2816–25. Edwards, D. 2004. The Nubian Past: An Archaeology of the Sudan. London: Routledge. Evans, J., Stoodley, N. and Chenery, C. 2006. A strontium and oxygen isotope assessment of a possible fourth century immigrant population in a Hampshire cemetery, southern England. Journal of Archaeological Science 33.2: 265–72. Faure, G. 1986. Principles of Isotope Geology. New York: Wiley-Liss. Fenner, J. and Wright, L. 2014. Focus: revisiting the strontium contribution of sea salt in the human diet. Journal of Archaeological Science 44: 99–103. Garvie-Lok, S. 2009. Population mobility at Frankish Corinth: evidence from stable oxygen isotope ratios of tooth enamel. In L. Schepartz, S. Fox and C. Bourbou (eds), New Directions in the Skeletal Biology of Ancient Greece, Athens: American School of Classical Studies, 245–56. Gat, J. 1996. Oxygen and hydrogen isotopes in the hydrologic cycle. Annual Review of Earth and Planetary Sciences 24: 225–62. Gibbon, V. and Buzon, M. 2016. Morphometric assessment of the appendicular skeleton in the New Kingdom and Napatan components from Tombos in Upper Nubia. International Journal of Osteoarchaeology 26:324–36. Hafsaas, H. 2007. Pots and people in an anthropological perspective: the C-Group people of Lower Nubia as a case study. Cahier de recherches de l’Institut de papyrologie et égyptologie de Lille 26: 163–71. Hershkovitz, I., Rothschild, B., Latimer, B., Dutour, O., Léonetti, G., Greenwald, C., Rothschild, C. and Jellema, L. 1997. Recognition of sickle cell anemia in skeletal remains of children. American Journal of Physical Anthropology 104: 213–26. Hillson, S. 1996. Dental Anthropology. Cambridge: Cambridge University Press. Horn, P., Hölzl, S. and Storzer, D. 1994. Habitat determination on a fossil stag’s mandible from the site of Homo erectus heidelbergensis at Mauer by use of 87Sr/86 Sr. Naturwissenchaften 81: 360–62. Iacumin, P., Bocherens, H., Mariotti, A. and Longinelli, A. 1996a. An isotopic palaeoenvironmental study of human skeletal remains from the Nile Valley. Palaeogegraphy, Palaeoclimatology, Palaeoecology 126: 15–30. Iacumin, P., Bocherens, H., Mariotti, A. and Longinelli, A. 1996b. Oxygen isotope analyses of co-existing carbonate and phosphate in biogenic apatite: a way to monitor diagenetic alteration of bone phosphate? Earth and Planetary Science Letters 142: 1–6. International Atomic Energy Agency/World Meteorological Organization 2006. Global Network of Isotopes in Precipitation. The GNIP Database. Kemp, B. 1987. The Amarna workmen’s village in retrospect. The Journal of Egyptian Archaeology 73: 21–50. Kemp, B. 1989. Ancient Egypt: Anatomy of a Civilization. London: Routledge.

https://doi.org/10.1017/9781108634311.007 Published online by Cambridge University Press

243

244

Michele R. Buzon, Sarah A. Schrader and Gabriel J. Bowen

Kendall, T. 1999. The origin of the Napatan State: El Kurru and the evidence for the royal ancestors. Studien zum Antiken Sudan 15: 3–17. Killgrove, K. 2010. Migration and Mobility in Imperial Rome. Ph.D. Dissertation, University of North Carolina, Chapel Hill. Krom, M., Cliff, R., Eijsink, L., Herut, B. and Chester, R. 1999. The characterization of Saharan dusts and Nile particulate matter in surface sediments from the Levantine basin using Sr isotopes. Marine Geology 155.3: 319–30. Larsen, C. S. and Walker, P. 2010. Bioarchaeology: health, lifestyle and society in recent human evolution. In C. S. Larsen, (ed.), A Companion to Biological Anthropology, Oxford: Wiley-Blackwell, 379–94. Longinelli, A. 1984. Oxygen isotopes in mammal bone phosphate: a new tool for paleohydrological and paleoclimatological research? Geochimica et Cosmochimica Acta 48: 385–90. Luz, B., Kolodny, Y. and Horowitz, M. 1984. Fractionation of oxygen isotopes between mammalian bone-phosphate and environmental drinking water. Geochimica et Cosmochimica Acta 48: 1689–93. Masoner, M., Baker, B. and Knudson, K. 2011. Preliminary isotopic analysis of Post-Meroitic to Christian period Nubians from the Ginefab School Site, Sudan. Poster presented at the Annual Meeting of the American Association of Physical Anthropologists, Minneapolis. Montgomery, J. 2010. Passports from the past: investigating human dispersals using strontium isotope analysis of tooth enamel. Annals of Human Biology 37.3: 325–46. Morkot, R. 1987. Politics, economies and ideology: Egyptian imperialism in Nubia. Wepwawet 3: 29–49. Morkot, R. 1995. The foundations of the Kushite State. Cahier de recherches de l’institute de papyrology et égyptologie 17: 229–42. Morkot, R. 2000. The Black Pharaohs: Egypt’s Nubian Rulers. London; Rubicon Press. Morkot, R. 2001. Egypt and Nubia. In S. Alcock, T. D’Altroy, K. Morrison and C. Sinopoli (eds), Empires: Perspectives from Archaeology and History, Cambridge: Cambridge University Press, 227–51. Nafplioti, A. 2008. ‘Mycenaean’ political domination of Knossos following the Late Minoan IB destructions on Crete: negative evidence from strontium isotope ratio analysis (87Sr/86Sr). Journal of Archaeological Science 35.8: 2307–17. Nafplioti, A. 2011. Tracing population mobility in the Aegean using isotope geochemistry: a first map of local biologically available 87Sr/86Sr signatures. Journal of Archaeological Science 38.7: 1560–70. Noy, D. 2000. Foreigners at Rome: Citizens and Strangers. London: Duckworth. O’Connor, D. 1993. Ancient Nubia: Egypt’s Rival in Africa. Philadelphia: The University Museum, University of Philadelphia.

https://doi.org/10.1017/9781108634311.007 Published online by Cambridge University Press

7 Isotopic Approaches to Mobility in the Nile Valley

Perry, M., Coleman, D. and Delhopital, N. 2008. Mobility and exile at 2nd century A.D. khirbet edh-dharih: strontium isotope analysis of human migration in western Jordan. Geoarchaeology 23.4: 528–49. Pollard, A., Pellegrini, M. and Lee-Thorp, J. 2011. Technical note: some observations on the conversion of dental enamel δ18op values to δ18ow to determine human mobility. American Journal of Physical Anthropology 145.3: 499–504. Price, T. D., Burton, J. and Bentley, R. 2002. The characterization of biologically available strontium isotope ratios for the study of prehistoric migration. Archaeometry 44: 117–35. Prowse, T., Schwarcz, H., Garnsey, P., Knyf, M., Macchiarelli, R. and Bondioloi, L. 2007. Isotopic evidence for age-related immigration to imperial Rome. American Journal of Physical Anthropology 132: 510–19. Reitsema, L. and Crews, D. 2011. Brief communication: oxygen isotopes as a biomarker for sickle-cell disease? Results from transgenic mice expressing human hemoglobin S genes. American Journal of Physical Anthropology 145.3: 495–98. Research Institute for Groundwater/International Workshop on Aliasing, Confinement and Ownership in object-oriented programming 1990. Western Nile Delta Region, Main Report, Inventory and Groundwater Development Plan. Schaaf, P. and Müller-Sohnius, D. 2002. Strontium and neodymium isotopic study of Libyan Desert glass: inherited pan-African age signatures and new evidence for target material. Meteoritics and Planetary Science 37: 565–76. Scheidel, W. 2004. Human mobility in Roman Italy, I: the free population. Journal of Roman Studies 94: 1–26. Schrader, S. 2012. Activity patterns in New Kingdom Nubia: an examination of entheseal remodeling and osteoarthritis at Tombos. American Journal of Physical Anthropology 149.1: 60–70. Schrader, S. 2013. Bioarchaeology of the Everyday: Analysis of Activity Patterns and Diet in the Nile Valley. PhD dissertation, Purdue University, USA. Schrader, S. A., Buzon, M. and Irish, J. 2014. Illuminating the Nubian ‘Dark Age’: a bioarchaeological analysis of dental non-metric traits during the Napatan Period. Journal of Comparative Human Biology 65.4: 267–80. Schweissing, M. and Grupe, G. 2003. Stable strontium isotopes in human teeth and bone: a key to migration events of the late Roman period in Bavaria. Journal of Archaeological Science 30: 1373–83. Shinnie, P. L. 1996. Ancient Nubia. London: Paul Kegan International. Smith, S. T. 1998. Nubia and Egypt: interaction, acculturation, and secondary state formation from the Third to First Millenium B.C. In J. Cusick (ed.), Studies in Culture Contact: Interaction, Culture Change and Archaeology, Carbondale: Southern Illinois University, 256–87. Smith, S. T. 2003. Wretched Kush: Ethnic Identities and Boundaries in Egypt’s Nubian Empire. London: Routledge.

https://doi.org/10.1017/9781108634311.007 Published online by Cambridge University Press

245

246

Michele R. Buzon, Sarah A. Schrader and Gabriel J. Bowen

Smith, S. T. 2006. A new Napatan cemetery at Tombos. Cahier de recherches de l’Institut de papyrologie et égyptologie 26: 347–52. Smith, S. T. 2007. Kirwan memorial lecture. Death at Tombos: pyramids. Sudan and Nubia 11: 2–14. Smith, S. T. and Buzon, M. 2014. Identity, commemoration and remembrance in colonial encounters: burials at Tombos during the Egyptian New Kingdom Empire and its aftermath. In B. Porter and A. Boutin (eds), Remembering and Commemorating the Dead: Recent Contributions in Bioarchaeology and Mortuary Analysis from the Ancient Near East, Boulder: University Press of Colorado, 187–217. Stojanowski, C. and Knudson, K. 2011. Biogeochemical inferences of mobility of early Holocene fisher-foragers from the Southern Sahara Desert. American Journal of Physical Anthropology 146.1: 49–61. Stojanowski, C. and Knudson, K. 2014. Changing patterns of mobility as a response to climatic deterioration and aridification in the middle Holocene southern Sahara. American Journal of Physical Anthropology 154.1: 79–93. Tafuri, M., Bentley, R., Manzi, G. and di Lernia, S. 2006. Mobility and kinship in the prehistoric Sahara: Strontium isotope analysis of Holocene human skeletons from the Acacus Mountains (southwestern Libya). Journal of Anthropological Archaeology 25: 390–402. Turner, B. L., Edwards, J. L., Quinn, E. A., Kingston, J. D. and Van Gerven, D. P. 2007. Age-related variation in isotopic indicators of diet at medieval Kulubnarti, Sudanese Nubia. International Journal of Osteoarchaeology 17.1: 1–25. Török, L. 1995. The emergence of the kingdom of Kush and her myth of state in the first millennium BC. Cahier de recherches de l’Institute de papyrologie et égyptologie 17: 243–63. Walker, P., Bathurst, R., Richman, R., Gjerdrum, T. and Andrushko, V. 2009. The causes of porotic hyperostosis and cribra orbitalia: a reappraisal of the iron-deficiency-anemia hypothesis. American Journal of Physical Anthropology 139: 109–25. Weldeab, S., Emeis, K., Hemleben, C. and Siebel, W. 2002. Provenance of lithogenic surface sediments and pathways of riverine suspended matter in the Eastern Mediterranean Sea: evidence from 143ND/144ND and 87Sr/86Sr ratios. Chemical Geology 186: 139–49. White, C, Longstaffe, F. and Law, K. 2004. Exploring the effects of environment, physiology and diet on oxygen isotope ratios in ancient Nubian bones and teeth. Journal of Archaeological Science 31: 233–50. Whiteman, A. 1971. The Geology of the Sudan Republic. Oxford: Clarendon Press. Wiesenfeld, S. L. 1967. Sickle-cell trait in human biological and cultural evolution. Science 157: 1134–40.

https://doi.org/10.1017/9781108634311.007 Published online by Cambridge University Press

part iii

Looking North

Published online by Cambridge University Press

Published online by Cambridge University Press

8

Numidian Burial Practices joan sanmartí, irene cruz folch, jordi campillo and david montanero

Introduction The Maghrib has a rich tradition of archaeological studies. However, this research has been biased from the chronological and cultural points of view. This is due to different reasons, relating to both the nature of the archaeological record and the ideological and political circumstances that developed successively under colonial rule and, from the mid-twentieth century, in the modern independent states of the area. In general, comparatively little attention has been paid to the indigenous cultures of the pre-Roman period, in particular as regards dwelling sites, which generally underlie thick stratigraphic deposits of later periods. The situation is somewhat different with regard to funerary archaeology, as North Africa, especially its eastern portion, is characterised by the existence of a surprising number and diversity of pre-Roman sepulchral monuments.1 Owing to their high visibility, these monuments constitute the best-known aspect of North Africa’s pre-Roman archaeological record; in spite of some research that has been recently carried out in Althiburos, an important habitation site,2 they still remain the most important available source of information on Numidian culture and society, and they can certainly provide a wealth of fresh knowledge if adequately investigated. This chapter provides a synthesis of the autochthonous burial practices in pre-Roman times in the eastern Maghrib, drawing heavily on several important publications by Gabriel Camps.3 To a considerable extent, this paper is a summary of his work on the subject, updated with data derived from several excavations and surveys carried out in the last three decades. After delimiting the area under study, we shall briefly describe the research history of this field, which is necessary to understand the still very limited state of knowledge. We then provide a general outline of the typology of funerary monuments and the structure of cemetery areas. We conclude with this contribution’s central matter, the funerary rituals. 1 3

Camps 1961. 2 Kallala and Sanmartí 2011; Sanmartí et al. 2012. Camps 1961; 1996; 2001a; 2001b. Camps et al. 1991.

https://doi.org/10.1017/9781108634311.008 Published online by Cambridge University Press

249

250

Joan Sanmartí et al.

Figure 8.1. Map showing the locations mentioned in the text: 1) Gabès; 2) El Hkayma; 3) Smirat; 4) Enfida; 5) Sidi M’hamed Latrech; 6) Carthage; 7) Teboursouk; 8) Thugga; 9) Djebba; 10) Simitthus; 11) Tiffech; 12) Ellès; 13) Makthar; 14) Henchir Mided; 15) Althiburos; 16) Kalaat es-Snam; 17) Gastel; 18) Tebessa; 19) Fedj el Koucha; 20) Medracen; 21) Bou Chène; 22) Sigus; 23) Sila; 24) Djebel Mazela; 25) Roknia; 26) El Khroub; 27) Aïn el Bey; 28) Cirta; 29) Tiddis; 30) Sétif; 31) Beni Messous; 32) Kbour erRoumia; 33) Siga.

Geographical Delimitation: ‘Numidia’, an Ambiguous Concept There is no straightforward geographical meaning for the term ‘Numidia’. In its widest sense, it designates the area controlled by its most famous ruler, Massinissa, which stretched from the western limits of the territory dominated by Carthage from the fifth century BC to the Moulouya River (Fig. 8.1). This includes all northern Algeria, a small portion of northeastern Morocco and a 30 km wide north-south strip in north-western Tunisia. Some scholars, however, believe that the Numidian territory stricto sensu was actually limited to eastern Algeria and a portion of the Tunisian High Tell.4 This is also, more or less, the extent of the later Roman province of Numidia. It is not our aim here to discuss this complex issue, in which several ethnic and historical-political matters are involved. The term ‘Numidian’ is used here to designate the autochthonous populations that lived during the first millennium BC in central and eastern Algeria and in Tunisia (including those that fell under the Carthaginian rule from the fifth century BC), but kept a strong local character in many aspects of their culture, including some important traits of their funerary behaviour. 4

Walsh 1965; Lassère 1977.

https://doi.org/10.1017/9781108634311.008 Published online by Cambridge University Press

8 Numidian Burial Practices

Research History Current knowledge about pre-Roman settlement sites in this area is still very limited. This is due to two factors: firstly, most of these sites persisted in occupation during the Roman imperial period, and often even further, frequently to the present day, which seriously hinders access to the oldest remains; secondly, and perhaps even more importantly, colonial-period archaeological research has almost exclusively been oriented towards the study of the Roman imperial era, and to some extent Punic civilisation, resulting in little or no excavation of the previous phases of these settlements. Consequently, the state of knowledge of pre-Roman societies of the eastern Maghrib is still poor, especially when compared with the continually increasing information we possess about the great cultures of the first millennium BC on the European side of the western Mediterranean. The situation is somewhat different if we consider the funerary record, as there are tens – perhaps even hundreds – of thousands of pre-Roman funerary monuments, which are a characteristic element of the landscape in many areas of this region. This archaeological wealth is reflected in the existence of a rich popular terminology to describe these funerary monuments (bazina, redjem, djedar, kerkour, hanoun, chouchet, biban, ghorfa, maklouba, gsar, etc.). These monuments quickly attracted the interest of European researchers. Indeed, after the mid-nineteenth century, in the case of Algeria, and following the constitution of the French protectorate in Tunisia (1881), there was considerable excavation activity of different types of funerary monuments, which was developed mainly by military personnel, doctors, teachers and other members of the colonial elites. Fieldwork conducted by professional archaeologists, using modern methods of excavation, did not really start until the 1950s. This was mainly due to Gabriel Camps, who was based at Algiers as the director of both the ‘Centre de recherches anthropologiques, préhistoriques et ethnologiques’ and the ‘Musée National d’Ethnographie et de Préhistoire du Bardo’. However, previous work, which is of variable – and sometimes remarkable – quality, had furnished a considerable volume of data. These constituted the base of Camps’ masterwork, Aux origines de la Berbérie. Monuments et rites funéraires protohistoriques.5 This is a very valuable synthesis that, even today, constitutes the reference work on this subject. However, the fact that a fifty-year-old book is still so influential demonstrates not only Camps’ abilities and thorough knowledge of the field, but also the low intensity of 5

Camps 1961.

https://doi.org/10.1017/9781108634311.008 Published online by Cambridge University Press

251

252

Joan Sanmartí et al.

research after the independence of the North African states, especially in Algeria. Since the 1980s, though, there has been some intensification of studies in the Sahel and, especially, in the Tunisian High Tell, with the excavation of two important necropolises of Punic-type shaft graves (El Hkayma and Djebba) and a large complex dolmen at Makthar.6 This is in addition to major survey and (more limited) excavation work in the area of Makthar-Ellès-Henchir Mided, as well as our own work with N. Kallala in the El Ksour massif necropolis.7

Types of Funerary Monuments The basic principles of the typology of funerary monuments were developed by Camps. They were based on criteria of complexity (from elementary to evolved forms), origin (local versus imported forms) and presence or absence of complementary elements (chapels, niches, antennas, porches, etc.).8 While recognising the validity of his basic types, we propose a different classification based on strictly formal criteria (Fig. 8.2). This classification includes three levels of discrimination: At a first level of analysis, we distinguish monuments that do not have an architectural character (Type 1 – Fig. 8.2) from those that involve some kind of construction covered with a horizontal roof (Type 2), although they might be very small. The mounds without a burial chamber, or those possessing an underground excavated one, are an exception to this rule; although this is obviously an inconsistency, it is preferable to creating a specific category for these kinds of tumuli, of which there are comparatively few in Numidia.9 At a second level of analysis, we distinguish, among non-architectural monuments, simple, open-air burial enclosures (Type 1.1) from caves and tombs that were excavated into the natural ground (Type 1.2). Among the architectural monuments, we set apart the free-standing, wholly visible buildings (Type 2.1) from those that were, or included in their structure, some kind of mound, although the latter did not always completely hide the built part of the monument. 6

7 9

For shaft graves, see Ben Younès 1986; 1987; Krandel-Ben Younès, 1993. For Makthar, see Ghaki 1997. Tanda et al. 2009; Kallala et al. 2014; Sanmartí et al. 2015. 8 Camps 1961, 60–62. At the El Ksour necropolis, for example, tumuli amount to only 3 per cent (23) of the 779 funerary monuments that can be classified. At Djebel Mazela, there are some bazinas (in Camps’ classification, an architecturally rather elaborate kind of tumulus) that represent only a small fraction of this site’s 3,000–4,000 funerary monuments; Camps and Camps-Fabrer 1964.

https://doi.org/10.1017/9781108634311.008 Published online by Cambridge University Press

8 Numidian Burial Practices

Numidian Funerary Monuments

1. Non architectural monuments

1.1 Enclosures

1.1.1 Simple 1.1.2 Multiple

1.2 Carved Tombs

1.2.1 Haouanet 1.2.2 Hypogea 1.2.3 Punic-type shaft graves 1.2.4 Caves 1.2.5 Silos 1.2.6 Pits

2. Architectural monuments

2.1 Free standing building

2.1.1 Chouchet 2.1.2 Free-standing dolmens 2.1.3 Complex megalithic monuments 2.1.4 Kabylian covered alleys 2.1.5 Multilevel towers 2.1.6 Large conical mausolea

2.2 Mounds or buildings incorporating mounds

2.2.1 Tumuli 2.2.2 Bazinas 2.2.3 Corridor dolmens 2.2.4 Zenital acces dolmens

Figure 8.2. Classification of North African funerary monuments.

The funerary enclosures are the most elementary form. They were generally defined by a single circle of stones or slabs (Type 1.1.1) but the existence of several concentric circles is also well attested (Type 1.1.2). In some cases, the area was covered with slabs or a bed of stones. Funerary enclosures are attested all over the Maghrib; in Numidia, they are quite common in the Sétif region and the high plains of central Algeria. The haouanet (singular hanoun) (Type 1.2.1) are chamber tombs carved into rocky escarpments or in isolated rocky outcrops (Fig. 8.3). They are typical of the northern Tunisian Tell and the interior of Cap Bon but are also attested in eastern Algeria, both in the Gastel-Tebessa and the Cirtean regions.10 Hypogea (Type 1.2.2) differ from haouanet because they are underground, accessible via a ramp or stairs down. The chamber is always lined by stone walls, a trait that separates them from the Punic-type shaft graves.11 The latter type is also included in our typology (Type 1.2.3) because specifically autochthonous rituals are frequently attested in such shaft tombs in the territory under Carthage’s control (such as at El Hkayma and Djebba). Although they can hardly be labelled as ‘monuments’, we have also included in this category modified natural caves (usually enlarged and/or carved to form low benches) (Type 1.2.4), silo-shaped tombs (Type 1.2.5) and simple pits (Type 1.2.6). The former are well attested in Algeria, while the silo type is characteristic of Morocco and eastern Algeria. Large pits have been found at Tiddis, in the Cirtean region, and are also attested in Fazzan.12 10 12

Longerstay 1995; Stone 2007. 11 Camps 1961, 111. Bussière 1998; Mattingly 2003, 197–98, Type 2.

https://doi.org/10.1017/9781108634311.008 Published online by Cambridge University Press

253

254

Joan Sanmartí et al.

Figure 8.3. Haouanet of Sidi Latrech, Tunisia (after M. Rais—Travail personnel, CC BY-SA 3.0, https://commons.wikimedia.org/w/index.php?curid=12154089).

As regards category Type 2.1, free-standing structures, chouchet (singular choucha) (Type 2.1.1) are tower-shaped cylindrical monuments (about 2.5–3 m high), generally covered with large slabs (Fig. 8.4).13 This type of monument, which is frequent in the Aurès and Hodna regions in Algeria, is similar to some extent to the drum tombs of Fazzan, though these are smaller (about 1 m in height) and uncovered.14 In central Tunisia, there is an important group of dolmens, often of considerable size, formed by large stone slabs. They do not appear to have been covered by a tumulus (Type 2.1.2).15 Sometimes, two or three monuments are immediately juxtaposed or share common walls. Camps designated such a series of chambers, which are typical of the central Tunisian High Tell, as ‘complex megalithic monuments’ (Type 2.1.3). Some may have developed from the aforementioned juxtaposed dolmens. In particular, the monuments of the necropolis of Ellès (Fig. 8.5), which comprise multiple burial chambers arranged on either side of a central corridor from which they are accessible.16 These tombs also have a sort of small side porch supported by large vertical slabs. Other complex forms 13 15

Camps 1961, 170–72. 14 Mattingly 2013, 198–99, Type 3. Tanda et al. 2009; Marras et al. 2009. 16 Camps 1961; Camps 1996; Tanda et al. 2009.

https://doi.org/10.1017/9781108634311.008 Published online by Cambridge University Press

8 Numidian Burial Practices

Figure 8.4. Choucha of Djebel Kharrouba, Algeria (after Payen 1863; re-drawn by M. Hawkes).

Figure 8.5. Complex dolmen of Ellès (photograph by M. C. Belarte).

https://doi.org/10.1017/9781108634311.008 Published online by Cambridge University Press

255

256

Joan Sanmartí et al.

Figure 8.6. The Thugga mausoleum, looking north (photo by J. Sanmartí).

have been attested at Makthar.17 A further group in the typology consists of the covered galleries of Kabylia, about 8 to 15 m long and exceeding 2 metres in height (Type 2.1.4).18 We also include among the free-standing funerary monuments two types that, owing to their architectural features and decorative patterns, clearly follow Hellenistic models. For this reason, and given their huge size and their geographical location, they must be considered as royal mausolea.19 The first one is constituted by multilevel funerary towers with a pyramidal crowning (Type 2.1.5; Fig. 8.6). This was originally a Phoenician type that was widely spread in the eastern Mediterranean and introduced into the Numidian area by Punic architects during the second century BC. Three monuments of this kind are known: 1. at Siga, the capital of the Masaesyli Numidians; 2. at Thugga; and 3. at El Khroub, 14 km from Cirta, the capital of the Massylii Numidians.20 They are all dated to the second century BC – the last one probably being the tomb of Micipsa and Hiempsal, who died in 118 BC.

17 19

Pauphilet 1953; Camps 1961, 191–94; Ghaki 1997. 18 Camps 1995, 24–26. Rakob 1979; 1983; Coarelli and Thébert 1988. 20 Rakob 1979; 1983.

https://doi.org/10.1017/9781108634311.008 Published online by Cambridge University Press

8 Numidian Burial Practices

257

Figure 8.7. The Medracen (after Reda Kerbouche, Wikimedia Commons – Travail personnel, CC BYSA 3.0, https://commons.wikimedia.org/w/index.php?curid=28647236).

A very different type (Type 2.1.6) is constituted by the Medracen (imedghasen in Amazigh spelling), near Batna, eastern Algeria,21 and the Kbour Roumia (the ‘Christian’s Tomb’, usually designated by the French translation ‘Tombeau de la Chrétienne’), Sidi Rached, 60 km west of Algiers (Fig. 8.7).22 Both are very large circular monuments (diameters of 58.7 m and 60.9 m, respectively), constituted by a comparatively low cylindrical body decorated with engaged columns, above which stands a large stepped cone (rising 18.5 m and 32.4 m, respectively) and ending in a platform that was originally crowned with a slim pyramid or a sculptural group. Despite the fact that they somehow recall certain types of autochthonous tumuli (the bazinas, of which more will be said below), Rakob, and later Coarelli and Thébert, have argued persuasively that these large funerary monuments were inspired by Hellenistic funerary architecture.23 As for their dating, Camps proposed the early third century BC for the Medracen, drawing on both stylistic comparanda and 14 C samples obtained from the cedar beams still preserved in the building.24 Rakob, however, thinks that it should be dated to the early second century BC.25 The Kbour Roumia is clearly inspired by the Medracen, and was probably erected in the second or, more probably, the first century BC.26 The tumuli and other monuments that include some kind of mound (Type 2.2) constitute the most important category from a quantitative point of view. Simple tumuli that frequently – but not always – hid funerary chambers that were completely invisible from the outside constitute a very important type (Type 2.2.1), which is found all over the Maghrib. Several subtypes may be distinguished in relation to the funerary chamber: (i) the 21 24

Camps 1973. 22 Christofle 1951. 23 Coarelli and Thébert 1988. Camps 1973, 510–12. 25 Rakob 1983, 331–32. 26 Camps 1961, 205; Rakob 1983, 331.

https://doi.org/10.1017/9781108634311.008 Published online by Cambridge University Press

258

Joan Sanmartí et al.

Figure 8.8. Bazina of Simitthus, Chemtou (Photo by J. Sanmartí).

existence of a chamber (if it is absent, the bodies lay directly under the tumulus); (ii) its location (below or inside the tumulus); (iii) its accessibility (that is, whether it was possible to access it for successive depositions without having to remove a part of the tumulus). In the Numidian area, as defined above, variants with an accessible chamber are attested in the region of Constantine, but tumuli are not overall very frequent. On the contrary, simpler subtypes existed further south in the northern Sahara, between the Wadi Draa in Morocco, southern Tunisian and Libyan Fazzan.27 Camps discriminated between stone tumuli from those made basically of earth (for which he reserved the term ‘tertre’), but we do not think this distinction is significant enough to define two different types. The name ‘bazina’ is applied in the popular language to different circular monuments. Camps, however, limited the meaning of this word to tumuli bordered by more or less substantial circular walls, lacking megalithic elements in their construction (Type 2.2.2; Fig. 8.8). He also distinguished many variants within this type, including stepped specimens (‘bazinas à degrés’). These are very frequent in Numidia, well attested in Morocco and bear some resemblance to an important class (FP Type 5) found in Fazzan.28 27

28

Camps 1961, 65–84. See also Mattingly 2013, 197, Type 1, as well as the contributions by Papi, Chapter 9 and Bokbot, Chapter 10, this volume. Camps 1961, 162; Mattingly 2013, 200.

https://doi.org/10.1017/9781108634311.008 Published online by Cambridge University Press

8 Numidian Burial Practices

Figure 8.9. Zenithal access dolmen, Djebel Mazela (after Martin 1881, 1139, Fig. 117; re-drawn by M. Hawkes).

Following Camps, we apply the term ‘corridor dolmens’ (Type 2.2.3) to a group of monuments consisting of a built chamber that is covered with a megalithic slab and is laterally accessible through a corridor crossing the mound that covers the whole building.29 Generally small, they are found in great numbers in the littoral area between Algiers and the Enfida region, in the eastern Tunisian shores. Another type consists of generally small-built chambers, surrounded by circular walls and covered by megalithic slabs (Type 2.2.4; Fig. 8.9); these were also labelled ‘dolmens’ by Camps (and others before and after him), despite the fact that they do not have any side access. ‘Zenithal access dolmens’ might be an appropriate name for this kind of monument. In Camps’s opinion, the mound delimited by the circular wall did not always rise to the top of the chamber, but recent work at the El Ksour necropolis, close to Althiburos, suggests that it generally did (though some exceptions seem nevertheless possible).30 Type 2.2.4 is typical of the Tunisian High Tell and the massifs of eastern Algeria. Dolmenic monuments of types 2.2.3 and 2.2.4 may thus be considered as a specifically 29

Camps 1961, 125–39; Camps 1995.

30

Kallala et al. 2014.

https://doi.org/10.1017/9781108634311.008 Published online by Cambridge University Press

259

260

Joan Sanmartí et al.

Numidian type, in spite of the existence of some small specimens in northern Morocco.31 It is worth noting that the distribution areas of all these different types frequently overlap at the same site or in close proximity. For example, at Kalaat es Snam, in Tunisia, haouanet, zenithal access dolmens and hypogea are found side by side. At Djebel Mazela, bazinas and dolmens are attested, though the latter outnumbered the former.32 Similar observations have been made at the El Ksour necropolis. At Tiddis, bazinas, zenithal access dolmens and pits were used.33 The reasons for this diversity are unclear. Given that most monuments have not been dated, it is impossible to decide whether it is due to chronological differences, or rather to social, cultural or ethnic reasons. The latter possibilities, however, must hold true in many instances. For example, at Djebel Mazela the pottery found in Bazina XXII is quite similar to the vessels from the dolmens, so probably precluding the possibility of important chronological differences. At Tiddis, as well, the same kind of very distinctive painted pottery has been found in pit burials and bazinas.

Structure of Cemeteries The spatial structure of the Numidian cemeteries is markedly different from that of the necropolises of the contemporary cultures of the Mediterranean world. In the latter, indeed, tombs are mostly concentrated in more or less discrete areas, often, in the case of urban cultures, around the cities, or alongside the roads leading to them. This type of organisation is attested in two late cemeteries (third to first centuries BC) located within the territory of Carthage (El Hkayma and Djebba); they both display a strong mix of indigenous and Punic cultural elements. Such discrete cemeteries may have also existed around the indigenous towns, but in general Numidian tombs tended to be scattered over broad areas, which was also the case in Fazzan34 and, at least to some extent, in the western Maghrib. This is particularly true of the large dolmenic ‘necropolises’ of eastern Algeria and the Tunisian High Tell. For example, the estimated number of funerary monuments at Djebel Mazela is about 3,000–4,000, over an area of 400 ha.35 This is not a unique instance: the El Ksour massif 31 33 34

35

Camps 1961, 124–25. 32 Camps and Camps-Fabrer 1964. Berthier 1956; Bussière 1998. Mattingly 2003, 213–17. See Bokbot, Chapter 10, in this volume, in particular the necropolises of Jebel Bouïa (composed of approximately 1,200 tumuli) and Foum Larjam. Camps and Camps-Fabrer 1964, 8.

https://doi.org/10.1017/9781108634311.008 Published online by Cambridge University Press

8 Numidian Burial Practices

survey (still not completed) has recorded about 800 funerary structures (plus a number of unclassifiable structures, most of which must be the remains of severely damaged dolmens); 292 are known to exist at Henchir Mided; 72 at Ellès (but these were complex structures with several burial chambers);36 about 200 to 300 at Beni Messous;37 and several thousand at Roknia and Bou Chène.38 Smaller groups, of as little as some tens or even just a few tombs, are also attested but even these do not seem to form dense groups. Such a distribution pattern tends to suggest that these large groupings were related to tribal groups rather than to large settlement sites. However, since the development of urban life must have started by the mid-first millennium (or even before this date), and given that a dispersed settlement pattern has not as yet been clearly attested, some kind of connection with towns that acted as tribal centres must also be envisaged. In this sense, the existence of a group of dolmens very close to Thugga and the bazinas beneath the forum at Simitthus (Fig. 8.8) are hints to the existence of stricto sensu ‘urban’ Numidian cemeteries.39 We should add to this the necropolises of El Hkayma, which is related to the site of Henchir Merbess (ancient Tegea?),40 and Djebba (ancient Thigibba Bure).41 In spite of the loose distribution of the graves within the above-mentioned large sepulchral areas, a degree of organisation has sometimes been detected, but such cases are still poorly understood. The clearest evidence is the existence of low walls or, rather, relatively substantial stone alignments. In some cases, these structures delimitate empty areas, suggested to have had some kind of ritual meaning, although none of these has, to our knowledge, been excavated yet. In other instances, they link different funerary monuments to define spaces around specific groups of tombs. This has been taken to indicate the existence of family or clan relations between the deceased within these tombs.42

Funerary Rituals Following Camps’ approach, we shall successively consider three different aspects of the funerary ritual: the treatment of the bodies, the nature and volume of the funerary offerings, and the funerary cults, including funerary banquets.43 36 39 41

Tanda et al. 2009. 37 Camps et al. 1991. 38 Camps 1995. Icard 1905; Khanoussi 1993. 40 Ben Younès 1986, 31–32. Krandel-Ben Younès 1993, 179. 42 Camps 1961, 535–36.

https://doi.org/10.1017/9781108634311.008 Published online by Cambridge University Press

43

Camps 1961: 461–546.

261

262

Joan Sanmartí et al.

The Treatment of Bodies: Cremation, Primary and Secondary Inhumation Although the bodies could have been either cremated or inhumed, the former was clearly an unfamiliar rite to the Numidians, except for the deeply Punicised elites, as attested in the El Khroub mausoleum. It was, however, quite popular among the Phoenicians.44 In fact, in ‘normal’ tombs it is only found in the Sahel in some third-century BC burials of the El Hkayma necropolis (Tombs 3, 13 and 17), within an area under direct Carthaginian rule.45 Inhumation was, then, the characteristic Numidian funerary rite, as it also was in the western Maghrib and the Saharan region.46 Nevertheless, an important divide exists regarding the integrity of the bodies at the point when they were definitively interred. The first possibility (primary deposition) is that the bodies were placed in the tomb in a complete and articulated state. In that case, they are generally found in a crouched position, which is frequently (but certainly not always) extremely forced, with the legs drawn up to the chest and the knees close to the face. This is also the largely dominant practice in Fazzan and Morocco.47 When the body is found in this position, one must conclude that it was tied with ropes, or even that the lower limbs were dislocated prior to inhumation (probably when the soft tissues had already decomposed to some extent). The bodies may be found lying either on the right or the left side; in some instances they were laid on their backs. Supine or extended lateral positions are poorly attested. Camps thought that these extended positions were introduced at a rather late date by Punic or even Roman influence, but this is seemingly in contradiction with the fact that they are mostly attested in eastern and central Algeria (11 out of 14 sites). Moreover, flexed and crouched positions, which are typically Saharan, are extremely frequent in necropolises like Djebba and El Hkayma, which were located within the limits of the Carthaginian territory and otherwise attest a very strong Punic influence. The second possibility (secondary deposition) is the placement within the tomb of either only some selected bones or the whole skeleton, but disarticulated, which obviously implies previous decomposition or intentional stripping of the soft tissues. It is far from clear how excarnation was carried out. Camps suggested two non-mutually exclusive possibilities: (i) the body was exposed to the elements, or (ii) that there was a temporary 44 46 47

Rakob 1979; 1983. 45 Ben Younès, 1986; 1988. Mattingly 2003. See also Papi and Bokbot, Chapters 9 and 10, this volume. Mattingly 2003, 223. See chapters by Bokbot, Chapter 10, and Papi, Chapter 9, this volume.

https://doi.org/10.1017/9781108634311.008 Published online by Cambridge University Press

8 Numidian Burial Practices

grave where the body was left to decompose before the remains were transferred to the tomb.48 In spite of some hints of the existence of post-mortem manipulation of the bodies in Fazzan and some instances of excarnation in eastern Algeria and Morocco, this seems to be an essentially Numidian practice, not frequently attested contemporaneously, to our knowledge, in the rest of the Maghrib, and certainly absent in the European shores of the Western Mediterranean.49 At first glance, primary and secondary burial seem roughly equal in number, as each are present in 32 locations respectively. However, it seems clear that secondary burial was particularly prevalent in the territories that were never directly subjected to the power of Carthage, since its quantitative importance increases visibly as we progress westwards.50 Thus, 57 per cent of the locations in Tunisia (eight instances, excluding, of course, the strictly Punic necropolises) have only primary burials, 29 per cent (four occurrences) attest both forms of bodies’ treatment, and only 14 per cent (two cemeteries) have exclusively secondary burials. In addition, at least at Djebba, the number of secondary burials is very low, as it represents less than 10 per cent of the total. In eastern Algeria, on the contrary, both forms are attested in 44 per cent of localities (10 instances), and the exclusive presence of primary burial is reduced to 30 per cent (seven occurrences), while secondary burial is the only attested form in the remaining 26 per cent (six cases). In central Algeria, finally, secondary burial alone is largely dominant (80 per cent, but only four instances), and it is also attested along with primary burials in another location (20 per cent), while primary burial alone seems to be completely absent. There is no straightforward explanation for these differences. While Punic influence may have had some role in the popularity of primary inhumation in Tunisia (but this can only be verified when we can count on a larger body of chronological data), it still remains that both rites were concomitantly practised in wide areas that were never subjected to Carthage’s rule, even in the same localities, and sometimes in the same tomb (Bazina XXII at Djebel Mazela).51 Again, one must seek for social or cultural reasons to account for this apparently unexplainable diversity. 48 49

50

51

Camps 1961, 481–501; Camps 2001b. Fazzan: Mattingly 2003, 226. Algeria and Morocco: Camps 1961, 482; see also Bokbot, Chapter 10, in this volume. The following absolute figures and consequent percentages are mainly based on the data contained in Camps 1961. Camps and Camps-Fabrer 1964, 43.

https://doi.org/10.1017/9781108634311.008 Published online by Cambridge University Press

263

264

Joan Sanmartí et al.

It is also impossible, at the moment, to know to what extent the popularity of these two forms of treatment of the bodies varied over time, since the vast majority of tombs have not been dated. We can say, at any rate, that two zenithal access dolmens of the El Ksour necropolis with undisputed secondary burial remains are dated by 14C to the sixth/fifth centuries BC.52 Ceramic typology also indicates a chronology of eighth to fifth centuries BC for five zenithal access dolmens of the Djebel Mazela necropolis.53 Still at Djebel Mazela, Bazina XXII, which contained one primary deposition and human remains of at least five individuals resulting from secondary depositions, may also be dated to the eighth to fifth centuries BC, though a later chronology cannot be completely excluded. The three cases of secondary burial attested at Djebba are dated to the third (Tomb VI) and the second century BC (Tombs II and IX).54 A date of around the third century BC has also been proposed for the pits and bazinas at Tiddis (Algeria), but other instances from Sila (also in the Cirtean region) may be dated to the first century AD.55 In short, despite the limitations of our data, there are elements to suppose that secondary burial was practised throughout the first millennium BC, and even later. Regarding the dating of primary inhumations, data are even more limited. The best excavated and published sites are El Hkayma, in the Sahel, and Djebba, in the Tunisian High Tell. The first is dated between the third century BC and the first half of the second century BC, with a total number of 24 primary inhumations (in addition to the two previously mentioned cremations). The second site, where more than 30 primary inhumations have been attested, is dated between the mid third and the first centuries BC. Another relatively precise date has been provided by recent excavation of a simple dolmen at Henchir Mided, also in the Tunisian High Tell. This grave contained two bodies, which have been dated by 14C within the Hallstatt Plateau, that is, to the eighth to fifth centuries BC.56 In other sites where primary inhumation is attested, chronological data are too inaccurate to suggest any precise dating. At any rate, these data indicate that primary burial was also practised throughout the first millennium BC. The presence of children has been recorded in different types of graves, and using both kinds of treatment of the remains. In Dolmen 53 of the El Ksour necropolis, which, as already said, contained secondary burials, children actually constituted half (three) of the total number of individuals. 52 55

Kallala et al. 2014, 29 and 41. 53 Kallala et al. 2014, 55. Camps 1975. 56 Marras 2009.

https://doi.org/10.1017/9781108634311.008 Published online by Cambridge University Press

54

Krandel-Ben Younès 1993.

8 Numidian Burial Practices

The presence of a large number of infants’ bones in secondary position has also been observed in hypogea at Sila, in the Cirtean area.57 At this site, the remains of children were often collected in vessels containing bones of different individuals, sometimes only infants (24 cases), sometimes with mixed remains of children and adults (20 cases); 10 of these vessels contained skulls or infants’ facial bones.58 A similar practice has been observed at Tiddis, in different bazinas (some of which seemingly contained only the bones of children).59 Infant burials have also been noted in the main cemeteries of Fazzan, as well as on settlement sites.60 The deposition of children in primary position is also well attested. The clearest cases have been observed in the necropolises of El Hkayma and Djebba, both located within the territory dominated by Carthage. At Djebba, a significant (but unspecified) number of childrens’ bodies was found, their ages ranging from the early months to adolescence. They were buried with adults in the same graves and, insofar as it is possible to say (since the remains were considerably disturbed), the bodies were placed in a foetal position. It is also said that similar observations were made in the necropolis of Chemtou (Simitthu).61 In the necropolis of El Hkayma, in the Tunisian Sahel, there was only one instance of infant burial, which was in a very bad state, and therefore it has not been possible to specify the position of the body.62 In spite of their limited number, these data indicate that there were no significant differences in the funerary treatment of children and adults, which is in accordance with the funerary record from north-eastern Morocco63 but contrasts sharply with what has been observed contemporaneously in other cultural areas of the western Mediterranean, including the Punic world and Iberian cultures. In short, what has been said so far indicates that primary and secondary burials are contemporary practices that are attested, albeit in very different proportions, in the same geographical area, frequently in the same cemetery, and even in the same tomb, and that they were used for both children and adults. That being said, it is still worth discussing the existence of peculiar practices within either of these main kinds of treatment of bodies. Regarding the different positions of the body in the case of primary inhumation (that is, supine decubitus, flexed and foetal), they all may be found in the same necropolis. This has been attested at El Hkayma, where 57 60 61 62

Logeart 1936. 58 Logeart 1936; Camps 1961, 492–93. 59 Camps 1961, 495. Mattingly 2003, 226. Krandel-Ben Younès 1993, 188–91; the excavation remains unpublished. Ben Younès 1988, 56. 63 See Papi, Chapter 9, this volume.

https://doi.org/10.1017/9781108634311.008 Published online by Cambridge University Press

265

266

Joan Sanmartí et al.

seven are of the first type, while 13 were buried in flexed position and four in foetal position.64 The same association (but lacking precise numbers) is attested in two eastern Algerian sites (Roknia and Aïn el Bey).65 The association of just the second and the third types is only attested at Djebba and the Enfida area in Eastern Tunisia. The simultaneous presence of crouched and extended bodies (that is, bodies not in the foetal position) has been observed at Gastel and Bou Chène, both in eastern Algeria. The remaining combination (bodies in foetal and extended position) is attested at Tiffech and Sigus, again in eastern Algeria. In other cases, though, there is only one type attested in a single location. Flexed lateral decubitus, for example, is the only form of deposition noticed at Teboursouk, Makthar, Gabès and Fedj el Koucha, while foetal position on its own is extremely frequent, since it is attested in t12 locations scattered over the whole eastern Maghrib (two in central Algeria, six in eastern Algeria, four in Tunisia). Finally, stretched bodies alone have also been attested over the whole area, but in much lesser numbers (two cases in central Algeria, three in eastern Algeria, three in Tunisia). Again, diversity seems to be the rule, and the available data provide no clues to explain it. Thorough analyses of sex, age and even ethnic traits, might perhaps provide some indication to account for these cultural practices. There are also considerable differences in secondary inhumations. In particular, there is no rule regarding the number, size and kind of bones deposited in the graves. Sometimes, they even do not contain any bones at all, or just a few small amorphous fragments. In other instances, a few fragmentary bones belonging to a single individual were deposited, sometimes carefully scattered over the whole surface of the tomb, or gathered inside a small vessel. This and the previous case are seemingly typical of inland dolmenic necropolises like Djebel Mazela and El Ksour.66 An intact sealed layer at the bottom of the chamber of Dolmen 53 at El Ksour also contained a few fragmentary bones, which belonged to a minimum number of six individuals (three children, one adolescent and two adults). In other cases, the grave contains a larger number of less fragmentary bones belonging to the incomplete remains of several bodies, for example in the dolmenic necropolis of Beni Messous, or the tumuli of Gastel. Bazina XXII at Djebel Mazela has provided similar evidence, with a considerable number of remains that included skulls and long bones, as 64 65 66

Ben Younès 1986; 1988. These and the following figures in this paragraph are after Camps 1961. Djebel Mazela: Camps and Camps-Fabrer 1964, 75–79. El Ksour: Kallala et al. 2014, 29 and 41; see also Sanmartí et al. 2015.

https://doi.org/10.1017/9781108634311.008 Published online by Cambridge University Press

8 Numidian Burial Practices

well as 118 teeth, on the whole belonging to a minimum number of five individuals.67 Sometimes, though, a number of complete or nearly complete bodies were deposited, with skulls and long bones separately grouped. This is well attested in early Imperial hypogea at Sila, near Constantine, where skulls were carefully placed along the tomb’s wall or over pottery vessels.68 Additionally, these vessels contained other bones, especially small-sized ones, of several individuals, including children’s skulls and facial bones. Similar observations have been made in a series of fourth, possibly third, century BC rectangular pits at Tiddis, and in several bazinas of the same site.69 In the latter case, the articulated remains of hands, feet and cervical vertebrae were found in several vessels, which proves that they were deposited before the body was completely decomposed.

Individual and Collective Graves Another factor of variability is linked to the individual or collective character of the tombs. Certainly, most graves are collective, in sharp contrast with the evidence noticed in Fazzan, where individual tombs were the norm in protohistoric times.70 This is well attested in several bazinas (Djebel Mazela, Tiddis), in simple pits (Tiddis), in shaft graves (El Hkayma, Djebba) and in a number of dolmens (El Ksour-53, Henchir Mided, Beni Messous).71 On the contrary, most of the small inland dolmens excavated so far (at Djebel Mazela, plus El Ksour-647) accommodated the (very scanty and fragmentary) remains of only one body, or, as already said, were empty;72 some of the latter contained a pottery vessel with a stone inside, which is considered by Camps as a possible symbolic substitute for the remains of the deceased.73 The collective or individual character of tombs bears no relation with the treatment of the bodies. This is clearly indicated at the El Ksour necropolis, where the same treatment was given to the remains contained in Dolmen 53 (collective) and 647 (individual). At Dougga, as well, one of the dolmens excavated by Icard housed a single body, but another contained three, and another contained thirty.74 Finally, at Djebba, secondary inhumation is attested in three graves, one of which one (III) was individual, while the 67 69 71

72 74

Camps and Camps-Fabrer 1964. 68 Logeart 1936; Camps 1961: 492–93. Berthier 1956; Bussière 1998; Camps 1961, 494–95. 70 Mattingly 2003, 226. Bazinas: Camps and Camps-Fabrer 1964; Berthier 1956. Pits: Bussière 1998. Shaft graves: Ben Younès 1986; 1988; Krandel-Ben Younès 1993. Dolmens: Kallala et al. 2014, 41; Marras et al. 2009; Camps et al. 1991. Camps and Camps-Fabrer 1964; Kallala et al. 2014, 29. 73 Camps 1961, 490–91. Icard 1905.

https://doi.org/10.1017/9781108634311.008 Published online by Cambridge University Press

267

268

Joan Sanmartí et al.

remaining two (VI and IX) contained the bones of five and four individuals, respectively.75

Relationship between the Different Types of Monuments and the Treatment of the Body Secondary inhumation in dolmens is attested in 17 localities that are scattered over all the study area. It is also present in 11 tumuli (three in central Algeria, eight in eastern Algeria, one in Tunisia) and seven bazinas, most of them in Tunisia (two) and eastern Algeria (four). Its presence is also attested in a cave at Sila (eastern Algeria), in a choucha in the same region, and in the shaft graves necropolis of Djebba (Tunisian High Tell). Primary inhumation is attested in 18 dolmenic necropolises, of eastern Algeria (10) and Tunisia (eight); in four bazinas from the same areas (two cases in each), in three caves (one in each region), in 12 tumuli (three in central Algeria, five in eastern Algeria and four in Tunisia). It is also found in one choucha in eastern Algeria. In view of these data, it is clear that there is no correlation between the form of the treatment of the body and the type of monument in which they were deposited. However, it is possible that a more refined analysis, which would take into account the variants within the major forms of the treatment of the bodies, might indicate some significant association. For example, in the case of secondary inhumation, the deposition of just a few bone fragments, or even the complete absence of human remains, seems to be characteristic of some dolmenic necropolises of inland Algeria and the Tunisian High Tell (Djebel Mazela and El Ksour). On the contrary, the preservation of complete, if disarticulated, skeletons is seemingly typical of bazinas, pits and underground chambers of a limited area of eastern Algeria (Sila and Tiddis).

Cohabitation of the Dead and Living A final remark concerning the treatment of bodies derives from observations recently made at Althiburos. Fragments of human bones have been found in pre-Roman domestic areas of the site, in five different units of stratification. Three have been dated to the eighth century BC (remains of three adults), another one to the sixth-fifth century BC (one child) and the remaining one to the fourth century BC (one adult).76 Although we cannot 75 76

Krandel-Ben Younès 1993. Kallala et al. 2014, 71. Human remains studied by Thaïs Fadrique.

https://doi.org/10.1017/9781108634311.008 Published online by Cambridge University Press

8 Numidian Burial Practices

give many details about the precise context (due to the unavoidable fragmentation of the units of excavation in this deep stratified site), all the available evidence indicates that these layers are related to houses. Consequently, we may suppose that after excarnation some of the bones were kept in the dwellings; this may account, in part, for the small numbers in which they are found in some tombs. More or less similar practices are ethnographically attested, and Camps himself cites several examples.77 The presence of several adults rules out the possibility that these remains attest the practice of burying infants separately in settlement sites, which has been noticed in Fazzan.78 This practice has not been attested during the Imperial period, but a consistent number of Vandal layers of the same area also contain human bones. This may have some connection with the fact that Dolmen 53 of the El Ksour necropolis (about 3 km to the south of the city) was enlarged in the same period, and possibly reused for funerary purposes (the chamber has been plundered, so it is not possible be sure about this).79

Funerary Offerings The first aspect that needs to be stressed is that the (archaeologically attested) funerary goods in the Numidian area are generally poor, at least when compared with the lavish assemblages that are present in funerary contexts of different areas of Europe during the first millennium BC. This is a shared trait with the evidence from the western Maghrib and Fazzan prior to the classic Garamantian period.80 The Punic-type shaft grave necropolises of Djebba and El Hkayma, both within the Carthaginian territory, are important exceptions, which attest the influence of the Punic funerary practices in that area. As a rule, funerary offerings in the area under study are limited to animal bones, pottery vessels and, sometimes, small bronze or iron objects.

Animal Bones Camps’ synthesis on this issue contains a long list of animal species that have been found in tombs of the studied area.81 Among these, cow, sheep/ 77 80

81

Camps 1961, 538. 78 Mattingly 2003, 226. 79 Kallala et al. 2014, 39–50. Maghrib: see Bokbot, Chapter 10, and Papi, Chapter 9, in this volume. Fazzan: Mattingly 2001, 218. Camps 1961, 507–15.

https://doi.org/10.1017/9781108634311.008 Published online by Cambridge University Press

269

270

Joan Sanmartí et al.

goat, bird and horse are most frequent. A certain geographic pattern can be observed as regards the distribution of cow and horse, the former being particularly frequent in western Algeria, while the latter is typical of the dolmenic necropolises of eastern Algeria and the Tunisian High Tell. This has also been supported by recent work at the El Ksour necropolis, where Tomb 647 contained a horse bone, which represented the only funerary offering in the grave.82 Sheep/goats, however, are found all over the area studied in this chapter. Following Camps, birds are found in 13 necropolises, always in tombs with secondary inhumations, to which we must also add Dolmen 647 of the El Ksour necropolis. Among wild fauna, gazelles are relatively well attested, for example at Djebel Mazela, and in a recently excavated dolmen at Henchir Mided.83

Artefacts Tombs lacking any kind of artefacts are quite usual, especially in the dolmenic necropolises. This is well attested, for example, in several small dolmens at Djebel Mazela (a necropolis where funerary offerings are particularly scarce) as well as in the recently excavated El Ksour-647 dolmen, which is of the same type, and in the already mentioned large dolmen of Henchir Mided.84 In other cases, one or two vessels are found in small dolmens, a few more in large bazinas (but these contain several bodies, so that the ratio is more or less the same).85 The number of vessels is certainly a little bit larger in other kinds of tombs, particularly in pits at Tiddis and underground chambers at Sila;86 again, however, one must remember that these were collective graves that housed the remains of a large number of individuals. In addition, many of these vessels were intended to contain parts of the bodies, and consequently they cannot be considered as offerings. As already said, relatively important funerary assemblages are only found in the shaft-grave necropolises of Djebba and El Hkayma. Most of the vessels found in funerary contexts are locally produced. Following Camps, they may be classed in three main categories: votive miniatures (‘microcéramique’ in his terminology), ritual vessels and common utilitarian vessels. The ritual, and generally funerary, character of the first class is obvious. They are regularly found in tombs of many different types, including those in strongly Punicised areas, as El Hkayma and Djebba. Some of them, nevertheless, have been found in habitation 82 84 85

Kallala et al. 2014, 29–31. 83 Marras et al. 2009. Camps and Camps-Fabrer 1964, 79–81; Kallala et al. 2014, 29–31; Marras et al. 2009. Camps and Camps Fabrer 1964, 59. 86 Bussière 1998; Logeart 1936.

https://doi.org/10.1017/9781108634311.008 Published online by Cambridge University Press

8 Numidian Burial Practices

contexts at Althiburos, which might indicate either that they had other, non-funerary, meanings, or that funerary cult was also practised in the dwellings; a possibility that, as already said, is sustained by the finding of human bones in domestic contexts at the same site. The tradition of votive miniature vessels still persists today in the eastern Maghrib: they are found in marabouts, as well as in haouita and mzara sanctuaries, and even, though not frequently, on tombs. As Camps pointed out, these modern uses suggest that their function, if not strictly symbolic, might have been to burn perfumes. He also thought that the Libyans borrowed this practice from the Phoenicians.87 As for the vessel shapes that Camps interpreted as ‘ritual’, most of them have also been found in fair numbers in recent digs in the habitation site of Althiburos. One possible exception, nevertheless, is the so-called ‘vase coquetier’, a characteristic egg-shaped vessel, frequently decorated with geometric painted patterns that could have been inspired by ostrich eggs (which were frequently included in Punic tombs).88 However, these vessels appear to have been a specific trait of the Cirtean area, in particular the Gastel necropolis, but no settlement sites have been excavated in that region, which means that their specifically ritual and funerary use is yet to be proved. On the contrary, a ritual function may be reasonably claimed for a large tripod bowl with human heads decorating its feet that has been found in one of the pit graves at Tiddis, not far from Gastel.89 Other vessels found at Tiddis frequently bear elaborate painted decoration to which Camps also attributed a funerary symbolic meaning, mostly due to presence of motifs that have been interpreted as birds (which, as noted above, are also frequently found in graves) and women dancing, or aligned in a procession.90 But the shapes that bear this painted decoration have no peculiar traits that might indicate that these are specifically ritual vessels. Again, only the excavation of further habitation sites can give solid clues on this matter. To sum up, everything indicates that, except for miniatures, most of the vessels that have been found in graves were not essentially different from normal domestic pottery. The only clear difference is that most of them were tableware or, to a lesser extent, lamps (whose presence may witness a, rather limited, Punic influence), while cooking and storage wares were completely absent.

87 90

Camps 1961, 526. Camps 1961, 513.

88

Camps 1961, 287–92; Camps 1997.

https://doi.org/10.1017/9781108634311.008 Published online by Cambridge University Press

89

Bussière 1998.

271

272

Joan Sanmartí et al.

Let us also say that, contrary to other Mediterranean and contemporary autochthonous cultures (such as, the Etruscans, Tartessians, Iberians, Southern Celts and the Garamantians during the classical period), the Numidians seem to have been extremely reluctant to introduce foreign pottery into their tombs. An exception to this is at Henchir Mided where an Attic vicup (that is, a stemmed, double-handled wide cup, intended for wine consumption) dated to the second quarter of the fifth century BC has been found in a dolmen.91 However, the dig of Dolmen 53 of the El Ksour necropolis has produced almost exclusively local Numidian handmade pottery in the foundation levels of the tumulus that surrounded the funerary chamber; these are dated to the sixth-fifth centuries BC, a period when Punic imports are well attested at nearby Althiburos (albeit in small quantity).92 The same holds true for the twenty-four tombs excavated by Camps and Camps-Fabrer at Djebel Mazela, and for several bazinas, underground chambers and pits excavated at Tiddis and Gastel, in the Cirtean area.93 The only clear exceptions (leaving aside possible late re-uses) are found in third- to first-century BC necropolises in the areas controlled by Carthage, namely at Djebba and El Hkayma. Here, Punic ‘common’ pottery and transport amphoras, black-glazed tableware (both Punic and italic), unguentaria and lamps are quite common, along with a larger number of local handmade pottery that include the typical funerary miniatures and other vessels that are also frequently attested in the digs at Althiburos.94 The number and nature of items found in these necropolises, along with the type of tombs themselves, clearly attest strong Punic influence, which, nevertheless, did not extend to the treatment of the bodies (except for an isolated cremation at El Hkayma). As regards other items included in the funerary assemblages, weapons were extremely rare, as opposed to their relatively frequent presence in western Algeria. Other personal objects, such as bracelets, rings, fibulae, knives or beads were also present, but, on the whole, they were never very frequent.95 One particular trait that is shared with the evidence from Fazzan is the use of red ochre. Though not seemingly very frequent, this has been noticed in several sites, such as Djebba (only one instance),96 Smirat (Tunisia) and the Negrine area (Algeria).97 91 92

93 95

Ferjaoui 2010. Kallala et al. 2014, 49–50. A hardly datable Punic vase was found in another tomb of the same necropolis (number 42) along with local handmade pottery, while tomb 647 contained almost no pottery at all (Kallala et al. 2014, 29 and 32–35). Camps and Camps-Fabrer 1964. 94 Krandel-Ben Younès 1993; Ben Younès 1986; 1988. Camps 1961, 518–20; 527–29. 96 Krandel-Ben Younès 1993, 48. 97 Camps 1961, 524.

https://doi.org/10.1017/9781108634311.008 Published online by Cambridge University Press

8 Numidian Burial Practices

Funerary Banquets? Evidence for this kind of ceremony is extremely scarce. Nevertheless, the excavation of dolmen 53 at the El Ksour necropolis has afforded some data that may hint at their existence, at least in some particular monuments. These data come from the lower levels of the tumulus that surrounds the funerary chamber; these layers have produced a number of handmade Numidian pottery sherds that may be typologically dated to the seventh to fifth centuries BC, a chronology that is consistent with the 14C dating of the bones that were found inside the funerary chamber.98 These vessels were almost exclusively cups and a large jar, and many of them were covered with the red-slip that is typical of Numidian tableware; none were complete. None of the storage or cooking pottery shapes that have been recorded in the nearby habitat site of Althiburos were attested in this tumulus. It is also worth noticing that no bones (animal or human) have been found in these layers. There is no straightforward interpretation for this material, but it is not unreasonable to consider it the remains of vessels used in funerary ceremonies; if so, these would have involved only drinking. Somewhat different, but possibly related, evidence has been found in Dolmen IX at Djebel Mazela.99 Here the fragments of a large thick-walled vessel were found intentionally scattered over several square metres of the tumulus, and one fragment belonging to the same vessel was placed inside the funerary chamber. On the contrary, Dolmen 647 in the El Ksour necropolis did not yield any artefacts (either from the funerary chamber or from the tumulus). Given that Dolmen 53 at El Ksour is in several respects a remarkable monument – size, careful construction, structural complexity,100 number of interments (at least six) – we may conclude that there were social reasons that may explain the signs of possible funerary feasting here, behaviour that did not take place in other smaller and less carefully built dolmenic monuments.

Funerary Cults Convincing evidence for funerary cults has been observed at several sites. Among the most impressive cases, we may cite the large dolmen excavated at

98

99

Kallala et al. 2014, 49–50. Conventional radiocarbon age: 2480 ± 40 BP (Beta-283142), calBC 775–430. Camps and Camps-Fabrer 1964, 23–26. 100 Kallala et al. 2014 and also see below.

https://doi.org/10.1017/9781108634311.008 Published online by Cambridge University Press

273

274

Joan Sanmartí et al.

Makthar by Didier Pauphilet.101 This was a large (15 m x 7.5 m), complex megalithic structure, possibly dating to the fourth or third centuries BC, but still used during the early Roman Empire. It had four square funerary chambers, in which a large number of disarticulated bones were found. These chambers were preceded by six smaller rooms that contained anepigraphical stelae and a large number of pottery vessels. The latter were filled with animal bones, earth and ashes. They were most likely sacrifices linked to funerary cult. In other funerary monuments, the existence of a funerary cult is suggested by the presence of peculiar structural features. For example, a certain number of monuments, of different types, had a niche in the outer wall where offerings were supposedly placed, as suggested by the presence of burnt bird bones in one of them.102 These niches may have developed into much larger trefoiled inner halls that were not funerary chambers.103 They have been labelled as ‘chapels’. Some further observations on Monument 53 at the El Ksour necropolis are relevant here. Related to this dolmen, two stone alignments of unequal length have been found, pointing respectively north-west and east, and thus delimiting a V-shaped space open to the north-east. Since these structures do not seem to link Monument 53 with other tombs, the overall aspect of the dolmen strongly calls to mind the much older Saharan antenna tombs.104 Another similar case is attested at Sila, though in that case the ‘antennae’ are more regular, of equal length, and delimit a space open to the south, with a niche ‘chapel’ in the dolmen’s southern wall.105 The latter trait suggests that the antennae bounded a space devoted to funerary cult. If we leave aside the evidence from Makthar, data on worship acts involving sacrifices and offerings are extremely scanty, which is probably due to the fact that excavation has been generally limited to the funerary chambers. One striking instance, however, comes again from Dolmen 53 of the El Ksour necropolis.106 A ‘second life’ for the monument is attested, by the early third century AD, with deposition of at least two cooking pots that were placed on the low circular wall that delimits the building. These objects prove the existence of rituals that involved food consumption, which could be linked to new funerary depositions (this cannot be verified, since most of the funerary chamber has been plundered) or, rather, just to the cult to the ancestors. 101 104

Pauphilet 1953. 102 Camps 1961, 177–79, 545. Mattingly 2003, 201–03. 105 Camps 1961, 179.

https://doi.org/10.1017/9781108634311.008 Published online by Cambridge University Press

103 106

Camps 1961, 182–84. Kallala et al. 2014, 44–45.

8 Numidian Burial Practices

Final Remarks If we were to define the Numidian funerary practices with two words, these would certainly be ‘diverse’ and ‘complex’, which are evident in every aspect of the record, from the typology of the monuments, to their geographical distribution and to the observed ritual practices. This impression, however, is largely due to the lack of good chronological and contextual information. Many fresh data are obviously needed to improve our understanding of the chronological and the social, cultural and maybe ethnic reasons that may account for the wide range of variation that we have described in this chapter. Future research into some of these areas would help to overcome the existing limitations of our data. Complete excavation of funerary monuments is particularly important. In other words, investigation must not be limited to the sepulchral chambers, but ought to include the associated tumuli, if these exist. What is more, they should be extended beyond individual funerary monuments, to include large areas and several monuments. Such a strategy could provide crucial information about the chronological and/or social relationship between different burials. Such excavation may also be extremely helpful in order to reconstruct the processes related to the treatment of the bodies (especially how excarnation was performed, whether in open air areas prepared for exposure or, for example, in provisional pits) and ceremonies like funerary banqueting and worship. In other words, we need to improve our understanding of the funerary landscapes beyond what may be recorded and understood through surveys – however careful and technically refined these might be – and limited excavation. Complete excavation is also necessary not only to establish the foundation dating of monuments on firmer grounds (which is really needed), but also in order to understand the ‘new lives’, or re-use, experienced by some monuments. An important example of this is Monument 53 of the El Ksour necropolis, which was still used for a funerary cult in the third century AD and was later enlarged to form a large tumulus, in the fifth century AD. The persistence of autochthonous funerary monuments through Roman and Late Antique times constitutes a particularly interesting but still poorly studied issue. Such re-uses are attested at Makthar, where the dolmen excavated by Pauphilet (possibly built in the fourththird century BC) was still used in the late first century AD or even later.107 However, the hypogea from Sila, where specifically autochthonous rituals 107

Camps 1961, 191–94.

https://doi.org/10.1017/9781108634311.008 Published online by Cambridge University Press

275

276

Joan Sanmartí et al.

like excarnation are very well attested, were probably built in the late first or the second century AD.108 A still later date has been suggested for the Tafilalat chapel monuments and for the Gour Mausoleum (a large stepped bazina that was dated to the seventh century AD by Camps).109 The available data are still too scarce to establish the extent of these continuities and their causes, which could be linked in some cases to the phenomenon of the recreation of the past by discontent sectors of society in the Roman era.110 At the macro scale, one important need is to delimit the necropolises (including the very large ones, formed of thousands of monuments), to analyse their typological components and internal structure and to study their relationship with settlements, pathways, springs and other significant elements of the landscape. This may also be crucial in order to build up a map of hypothetical tribal or ethnic units, and to understand their relationship with the cities. Such information might in turn be essential to interpret the spatial distribution of the different funerary rites. No culture or social structure can be understood from the funerary remains alone. This means that research projects should include work on both the necropolises and the habitat sites. This is clearly proved by the work carried out at Althiburos, which has given clues not only for the dating of ceramic types that have also been found graves, but, as already said, for the understanding of the fate of the human remains after excarnation. Beyond the local complexities of Numidia, its funerary practices have much in common with those attested among the other autochthonous cultures of the Maghrib, while being quite different from those of the Phoenicians and other colonial and indigenous societies of the western Mediterranean area. This suggests the existence of a distinctive and strong common cultural background in western North Africa. Among the shared funerary customs, we may cite the almost universal practice of inhumation, the crouched or foetal position of the bodies, the absence, or limited number, of grave goods, the use of red ochre, as well as the fact that infants were regularly buried in the same tombs as adults. To this should be added, during the first millennium BC, the frequent existence of large sepulchral areas rather than discrete necropolises (which nevertheless are also attested). That being said, the similarities should not be pushed too far, 108 109

110

Camps 1961, 112–13. For a discussion of the Tafilalat chapel, see Bokbot, Chapter 10, this volume. For the Gour mausoleum, see Camps 1974. Kallala et al. 2014, 57–58; Mattingly 2011, 204–18; Fentress 2006.

https://doi.org/10.1017/9781108634311.008 Published online by Cambridge University Press

8 Numidian Burial Practices

since significant differences are also observed. As regards the monuments’ typology, a number of types are certainly common to the whole area, albeit with some variants, but the more elaborate seem to be peculiar to Numidia, particularly prior to the Classical Garamantian period. This is especially true for the large elaborate bazinas, for the dolmens and, at least to some extent, for the rock-cut chamber tombs. Given that these types – save for many small zenithal access dolmens – are intended for collective burials, their absence in Fazzan is quite logical, since individual burial is largely dominant in that region in protohistoric times. Finally, a possibly significant difference is also observed in the bodies’ funerary treatment, since excarnation seems to be particularly frequent in Numidia.

References Ben Younès, H. 1986. La nécropole punique d’El-Hkayma, Mai 1984. Revue des études Phéniciennes et Puniques et des Antiquités Libyques 2: 31–171. Ben Younès, H. 1987. Note sur les tumulus de Henchir el Mizouri (Sidi Alouane) au Sahel Tunisien. Revue des études Phéniciennes et Puniques et des Antiquités Libyques 3: 33–38. Ben Younès, H. 1988. La nécropole punique d’El Hkayma. Seconde campagne, septembre 1985. Revue des études Phéniciennes et Puniques et des Antiquités Libyques 4: 49–137. Berthier, A. 1956. Les bazinas de Tiddis. Libyca, Anthropologie, Préhistoire, Ethnographie 4: 147–53. Bussière, J. 1998. Quatre sépultures berbères protohistoriques de la nécropole orientale de Tiddis (Algérie). Antiquités Africaines 34.2: 31–43. Camps, G. 1961. Monuments et rites funéraires protohistoriques: aux origines de la Berbérie. Paris: Arts et Métiers Graphiques. Camps, G. 1973. Nouvelles observations sur l’architecture et l’âge du Medracen, mausolée royal de Numidie. Comptes-rendus des séances de l’Académie des inscriptions et belles-lettres 117.3: 470–517. Camps, G. 1974. Le Gour, mausolée berbère du VIIe siècle. Antiquités Africaines 8: 191–208. Camps, G. 1975. Recherches sur les plus anciennes inscriptions libyques de l’Afrique du Nord et du Sahara. Bulletin archéologique du Comité des travaux historique 10–11 [1974–1975]: 143–66. Camps, G. 1995. Les nécropoles mégalithiques de l’Afrique du Nord. In P. Trousset (ed.), L’Afrique du Nord antique et médiévale: monuments funéraires, institutions autochtones, Actes du VIe colloque international sur l’histoire et l’archéologie de l’Afrique du Nord (Pau, octobre 1993), Paris: Éditions du CTHS, 17–31.

https://doi.org/10.1017/9781108634311.008 Published online by Cambridge University Press

277

278

Joan Sanmartí et al.

Camps, G. 1996. Ellès. Encyclópédie Berbère 17: 2606–11. Camps, G. 1997. Le style de Gastel. Étude des céramiques d’une nécropole protohistorique d’Algérie. Antiquité Africaines 33: 39–48. Camps, G. 2001a. Incinération. Encyclópédie Berbère 24: 3710–14. Camps, G. 2001b. Inhumation. Encyclópédie Berbère 24: 3738–60. Camps, G. and Camps-Fabrer, H. 1964. La nécropole mégalithique du Djebel Mazela à Bou Nouara. Mémoires du Centre de recherches anthropologiques, préhistoriques et ethnographiques 3. Paris: Arts et métiers graphiques. Camps, G., Chamla, M. C. and Dastugue, J. 1991. Beni Messous. Encyclopédie Berbère 10: 1452–64. Christofle, M. 1951. Le Tombeau de la Chrétienne. Paris: Arts et Métiers Graphiques. Coarelli, F. and Thébert, Y. 1988. Architecture funéraire et pouvoir: réflexions sur l’hellénisme numide. Mélanges de l’Ecole française de Rome, Antiquité 100.2: 761–818. Fentress, E. 2006. Romanizing the Berbers. Past and Present 190: 3–32. Ferjaoui, A. 2010. Les relations entre Carthage et l’intérieur de l’Afrique, le cas de Zama Regia et sa région. In A. Ferjaoui (ed.), Carthage et les autochtones de son empire du temps de Zama. Hommage a Mhamed Hassine Fantar, Tunis: Institut National du Patrimoine, 341–52. Ghaki, M. 1997. ‘Le nouveau mégalithe’ de Makthar, rapport préliminaire. Revue des études Phéniciennes et Puniques et des Antiquités Libyques 10: 63–72. Icard, F. 1905. Note sur les dolmens de Dougga. Bulletin de la Société Archéologique de Sousse 256: 253–56. Kallala, N. and Sanmartí, J. (eds). 2011. Althiburos 1. La fouille dans l’aire du capitole et la nécropole méridionale. Collecció Documenta 18. Tarragona: Institut Català d’Arqueologia Clàssica. Kallala, N., Sanmartí, J., Jornet, R., Belarte, M. C., Canela, J., Chérif, S., Campillo, J., Montanero, D., Miniaoui, S., Bermúdez, X., Fadrique, T., Revilla, V., Ramon, J. and Ben Moussa, M. 2014. La nécropole mégalithique de la région d’Athiburos, dans le massif du Ksour (Gouvernorat du Kef, Tunisie). Fouille de trois monuments. Antiquités Africaines 50: 19–60. Khanoussi, M. 1993. La nécropole numide de Chemtou. In M. Harbai-Riahi (ed.), 30 ans au service du patrimoine: de la Carthage des phéniciens à la Carthage de Bourguiba, Tunis: Ministere des Affaires Culturelles, Institut National d’Archeologie et d’Art Tunis, 138. Krandel-Ben Younès, A. 1993. La nécropole rurale de Thigibba Bure (Djebba). Revue des études Phéniciennes et Puniques et des Antiquités Libyques 7–8 [1992–1993]: 179–286. Lassère, J. M. 1977. Ubique populus, peuplement et mouvements de population dans l’Afrique romaine de la chute de Carthage à la fin de la dynastie des Sévères (146 av. J.-C.–235 ap. J.-C.). Paris: CNRS.

https://doi.org/10.1017/9781108634311.008 Published online by Cambridge University Press

8 Numidian Burial Practices

Logeart, F. 1936. Grottes funéraires, hypogées et caveaux sous roche de Sila. Recueil des Notes et Mémoires de la Société archéologique de Constantine 63 [1935–1936]: 69–105. Longerstay, M. 1995. Les haouanet: état de la question. In P. Trousset (ed.), L’Afrique du Nord antique et médiévale: monuments funéraires, institutions autochtones. Actes du VIe colloque international sur l’histoire et l’archéologie de l’Afrique du Nord (Pau, octobre 1993), Paris: éditions du CTHS, 33–53. Marras, G., Doro, L., Floris, R., and Zedda, M., 2009. Il dolmen 102. Nota preliminare. In G. Tanda, M. Ghaki and R. Marras (eds), Storia dei paesaggi preistorici e protostorici nell’Alto Tell tunisino. Missioni 2002–2003, Tunisia: Ministère de la Culture et de la Sauvegarde du Patrimoine, 179–200. Martin, H. 1881. Excursion à la nécropole du Djebel Merah et de Ras el-Ain Bou Merzoug. Comptes rendus du Xe congrès de l’association française pour l’avancement des sciences, Algiers: 1138–42. Mattingly, D. J. (ed.). 2003. The Archaeology of Fazzan. Volume1: Synthesis. London: Society for Libyan Studies, Department of Antiquities. Mattingly D. J. 2011. Imperialism, Power and Identity. Experiencing the Roman Empire. Princeton: Princeton University Press. Mattingly, D. J. (ed.). 2013. The Archaeology of Fazzan. Volume 4: Survey and Excavations at Old Jarma (Ancient Garama) carried out by C. M. Daniels (1962–69) and the Fazzan Project (1997–2001). London: Society for Libyan Studies, Department of Antiquities. Pauphilet, D. 1953. Monument mégalithique à Mactar. Karthago 4: 51–82. Payen, E. (Comm.) 1863. Lettre sur les tombeaux circulaires de la Province de Constantine. Recueil des Notices et Mémoires de la Société archéologique de Constantine 8: 159–69. Rakob, F. 1979. Numidischer König Architektur in Nord Afrika. In H. G. Horn and C. B. Rüger (eds), Die Numider. Reiter und Könige nördlich der Sahara, Bonn: Rheinisches Landesmuseum, 119–71. Rakob, F. 1983. Architecture royale numide. In Architecture et société de l’archaïsme grec à la fin de la République romaine, Actes du Colloque international organisé par le Centre national de la recherche scientifique et l’Ecole française de Rome, Rome 2–4 décembre 1980, Rome: Ecole française, 325–48. Sanmartí, J., Kallala, N., Belarte, M. C., Ramon, J., Maraoui Telmini, B., and Miniaoui, S. 2012. Filling gaps in the Protohistory of the Easterm Maghreb: the Althiburos archaeological project (El Kef, Tunisia). Journal of African Archaeology 10.1: 21–44. Sanmartí, J., Kallala, N., Jornet, R., Belarte, M. C., Canela, J., Chérif, S., Campillo, J., Montanero, D., Bermúdez, X., Fadrique, T., Revilla, V., Ramon, and Ben Moussa, M. 2015. Roman dolmens? The Megalithic necropolises of Eastern Maghrib revisited. In. M. Díaz-Guardamino, L. García Sanjuán and D. Whitley (eds.), The Lives of Prehistoric Monuments. Iron Age, Roman and Medieval Europe. Oxford: Oxford University Press, 287–304.

https://doi.org/10.1017/9781108634311.008 Published online by Cambridge University Press

279

280

Joan Sanmartí et al.

Stone, D. L. 2007. Monuments on the margins: Interpreting the First Millennium BCE rock-cut tombs (Haouanet) of North Africa. In D. L. Stone and L. M. Stirling (eds), Mortuary Landscapes of North Africa. Toronto: University of Toronto Press, 43–75. Tanda, G., Ghaki, M., and Cicilloni, R. 2009. Storia dei paesaggi preistorici e protostorici nell’Alto Tell tunisino. Missioni 2002–2003. Tunisia: Ministère de la Culture et de la Sauvegarde du Patrimoine. Walsh, P. G. 1965. Massinissa. Journal of Roman Studies 55: 149–60.

https://doi.org/10.1017/9781108634311.008 Published online by Cambridge University Press

9

Revisiting First Millennium BC Graves in North-West Morocco emanuele papi

Introduction The historiography of ancient Morocco is a sequence of incursions and migrations. In the second half of the nineteenth century, the dolmens of Morocco were considered proof of the migration of ‘populations blondes’ to Africa from the Iberian Peninsula in the second half of the second millennium BC. Based on a comparison of megalithic tombs from Spain and the presence of light-haired indigenous people, this migratory theory was supported by two authoritative French scholars: the diplomatarchaeologist Charles Tissot and the neurologist-anthropologist Paul Brocca.1 Gabriel Camps also considered the megalithic dolmens found in three zones of Morocco (the Tangier peninsula, the Grand Atlas mountains south of Marrakech and the territory of the Béni-Snassen at the Algerian border) to be similar to the Iberian megalithic cist tombs of the second millennium BC and distinct from the Algerian and Tunisian dolmens.2 An alternative discourse has revolved around the importance of Phoenician contacts. Morocco was said to be inhabited instead by Phoenician colonists/ settlers from between the end of the second and the start of the first millennium BC. In 1943, in collaboration with the Vichy Republic, Jérôme Carcopino published Le Maroc antique, in which he argued that Phoenicians from Tyre discovered Morocco (the forerunner of America) and founded Lixus according to a planned system of colonisation. The motivating factors for the planned colonies were a desire for metals and the need to reduce demographic pressures on the motherland (‘ont vomi sur l’Afrique le trop plein de leur multitude’). Until the mid-1950s the prevalent theory in circulation was that in Morocco the Neolithic period persisted until the end of the first millennium BC when it was interrupted by the Phoenicians and the introduction of metals.3 The publication of a number of bronze objects, as well as 1

Tissot and Broca 1876.

2

Camps 1965.

3

Antoine 1952; Sáez Martín 1952.

https://doi.org/10.1017/9781108634311.009 Published online by Cambridge University Press

281

282

Emanuele Papi

Figure 9.1. The main centres of north-west Morocco between the eighth and third/second centuries BC.

rock carvings with representations of weapons south of Marrakech, resulted in the presumed anachronism of Morocco moving directly from stone to iron being replaced with the classic chronological sequence based on the use of metals in north-west Africa (‘Chalcolithic’ and ‘Bronze’).4 The Phoenician ‘colonisation’ hypothesis continued to be successfully proliferated. Every settlement from the first millennium BC was designated (and still continues to be designated) as ‘Phoenician’, from the so-called ‘Nécropoles phéniciennes de la region de Tanger’,5 to the settlements from the eighth to the sixth centuries BC (Fig. 9.1), where imported items were discovered.6 In the fifth century BC, the Phoenicians were replaced by the Carthaginians and their empire (commercial or territorial) in the western Mediterranean according to most historians and archaeologists.7 At the start 4 5

6

For bronze objects, see Camps 1960; Jodin 1956. For rock carvings, see Malhomme 1959. Ponisch 1967, but for Ponisch 1970, 84: ‘elles n’offrent en effet rien de commun avec celles d’époque phénicienne.’ El-Khayari 2004. 7 Contra Papi 2014.

https://doi.org/10.1017/9781108634311.009 Published online by Cambridge University Press

9 Revisiting First Millennium BC Graves in North-West Morocco

of the 1970s, the expansion of Carthage was tempered by the ‘Círculo del Estrecho’, a new theory that recognised the similarity of the territories around the Strait of Gibraltar and the specificity of the material culture, which was linked more to the Phoenician-Cypriot tradition than to the Carthaginian. Hegemony was thus passed from Punic Carthage to Punic Cadiz.8 Currently, archaeological data from the first-millennium BC communities of north-west Morocco are interpreted according to post-colonial parameters. This highlights the elements extraneous to the local traditions in order to integrate the communities of Morocco with the Phoenicians and Carthaginians of Andalusia in ‘colonial grounds’, with ‘ethnic and cultural porosity’,9 according to the processes of ‘hybridization, a word that entails the creation of new transcultural forms in colonial situation’.10 No genetic data from human remains of northern Morocco in the first millennium as yet support assumptions about the impact of population migration, although we do have some analyses for the previous and successive periods. The haplotypes of the chromosomes of goats (Capra hircus) in the Maghreb and the Iberian Peninsula indicate a common origin but it is impossible to determine the pathways of diffusion.11 The mitochondrial genetic data indicate population shifts from the Iberian Peninsula to Morocco across the Strait of Gibraltar from the tenth to the eighth millennium BC.12 Remains of Bos discovered at Atapuerca in northern Spain have been recognised as having African origins and demonstrate the existence of exchange in the second millennium BC in a south-north direction.13 A recent study of over a thousand individuals from the Iberian Peninsula detected over 10.6 per cent African genetic heritage, explained in the light of the more recent history of population movements from Morocco to Spain from the beginning of the eighth century AD onwards.14 My contribution examines the evidence for foreign, or colonial contact, based on archaeological data from two mid-first millennium BC funerary contexts from the Tangier peninsula in north-west Morocco, the necropolises of el Aïn Dahlia Kébira and Djebila to the south of Tangier (Fig. 9.2a) and from tumuli near Oujda (Fig. 9.2b) and from the Atlantic area, partially surveyed by G. Souville (Fig. 9.2c).15 I have excluded the tumulus of Volubilis from the end of the first century BC (Fig. 9.2c, no. 23), a commemorative citizen memorial lacking a burial, accompanied by Punic inscriptions that commemorate local chiefs, in which it is possible to identify a combination of Moorish, Punic, Hellenistic Greek and Roman traditions.16 I will examine 8 12 15

9 Papi 2016. Pappa 2009. 10 Aranegui et al. 2011 11 Pereira et al. 2009. Rhouda et al. 2009. 13 Anderung et al. 2005. 14 Adams et al. 2008. For necropolises, see Ponsich 1967; 1970. For tumuli, see Souville 1973. 16 Papi 2014.

https://doi.org/10.1017/9781108634311.009 Published online by Cambridge University Press

283

284

Emanuele Papi

Figure 9.2. a) Tangier region cemeteries; b) the tumuli surrounding Oujda; c) Atlantic region tumuli (distribution of the ‘chefferies’ based on Hamdoune 1993).

https://doi.org/10.1017/9781108634311.009 Published online by Cambridge University Press

9 Revisiting First Millennium BC Graves in North-West Morocco

the published data for these cemeteries, where we find neither anthropological characterisations nor archaeometric data, and will take into consideration the burials and the accompanying grave goods in an attempt to identify the characteristics of the tombs, the burial rituals, the origin of the primary materials and the objects. These data will then be incorporated with data from other northern Moroccan contexts, such as the certainly indigenous tumuli. Through this contribution, I will examine the particular characteristics of the north-west Moroccan communities and propose an alternative ‘African’ perspective to the ongoing discussion.17 The aim is to demonstrate how, in the cemeteries of the villages and in the tumuli surveyed, local communities and leaders of chefferies manifest indigenous characteristics and traditions that sometimes date far back in time. The traditional elements are homogeneous and deeply rooted, in contrast to the very few other elements that can be attributed to social and religious contexts outside the North African tradition. Some theories can be shown to be difficult to support, such as the dichotomy between foreign groups dedicated to craft and trade and local groups that exported raw materials, the diffusion of external techniques and food production to less advanced environments, and the assimilation of external cultural elements to create new identities.

Settlements Between the eighth and third centuries BC, 13 or 14 centres were founded in north-west Morocco within an area of approximately 10,000 km (Fig. 9.1). The oldest archaeological evidence appeared on the coast and dates to the eighth/seventh centuries BC, confirmed also by the presence of imported pottery, while other centres were added from the sixth to the third centuries BC. The presence of imported materials or foreign cult places (Herakles of Lixus or Astarte at Mogador) is insufficient to identify ‘colonies’ of foreign social groups (Mogador represents temporary and seasonal migration). Some Libyan place names, transliterated into Latin and Greek, indicate the connection between local ethnicities, chefferies, and settlements: Banasa/Baniurai, Sala/Salinsai-Salenses, Volubilis/Ouloubilianoi, Ouobrix/ Ouerberikai, Lixus/Lixitai (Figs 9.1 and 9.2).18 The sites cover territories of between less than 1 ha (Thamusida) and 10–15 ha (Lixus).

17

See above and Aranegui et al. 2011; Pappa 2009.

18

Hamdoune 1993.

https://doi.org/10.1017/9781108634311.009 Published online by Cambridge University Press

285

286

Emanuele Papi

We have little data on these new centres in terms of their structures and walls, stratigraphy and materials/objects, botanical and faunal remains, tombs and cemeteries. The settlements along the coast (Lixus), near riverbanks (Sala), river fords (Thamusida and Banasa) and in the vicinity of mines (Modagor and perhaps Tamuda) provide the most evidence for contact with foreign populations, suggested by imported ceramics, bronzes and foodstuffs and graffiti of foreign names (a hundred or so of these from Mogador), and they exhibit a higher level of cosmopolitanism and wealth. In addition to these main centres were farming villages and other centres of a more traditional character, such as the towns south of Tangier or in the mountainous regions, such as Tayadirt in the Middle Atlas or Oujda in the Rif (Fig. 9.2.a/b). To varying degrees, some communities had contact with foreign populations who wrote in Iberian (Banasa) and Punic (Mogador and Kouass), used materials produced on the Iberian Peninsula, the Punic world (few from Carthage), the Levant, Cyprus, Etruria and Egypt (vases and objects for banquets, wine amphorae), which publically signalled inequality of wealth. It is likely that these centres also served as markets and fairs of varying frequency and for the production of foodstuffs such as the salsamenta of the Atlantic coast, international trade as well as ceramics (Banasa, Kouass and Sidi Ali ben Ahmed/Thamusida). The centre for which we have the most information is Volubilis, founded in the third/second century BC, which was walled, had locations of cult and collective memory such as the tumulus, was the residence of leaders who used an alphabet derived from Phoenician cursive for public inscriptions, and had a monumental mausoleum such as the ones in Mauretania (Siga) and Numidia (Thugga).

Cemeteries Cemetery types in Atlantic Morocco are represented by cemeteries connected to villages, isolated tumuli, or small groups of tumuli. Fourteen cemeteries dating from the seventh to the fifth centuries BC have been identified in the region of Tangier connected to agricultural villages (Fig. 9.2a). The cemeteries of el Aïn Dahlia Kébira (84 graves) and Djebila1 (107 graves) have the most complete contexts and will be discussed here. In previous scholarship, these burials have been variously interpreted either as ‘Libyco-

https://doi.org/10.1017/9781108634311.009 Published online by Cambridge University Press

9 Revisiting First Millennium BC Graves in North-West Morocco

Phoenicians’,19 indigenous mixed with Phoenicians,20 indigenous,21 or a community in which ‘the distinct categories of “indigenous” and “Phoenician” need not always have meaningful identities’ with ‘intense cultural dialogue’, and sometimes of ‘Phoenician pedigree’.22 Unfortunately, the necropolis of Raqqada near Lixus has only been partially investigated, and therefore we cannot fully understand the differences between rural villages and one of the largest and richest centres of first-millennium BC Morocco.23

Burial Practices Only inhumation was practised. Most of the tombs have a uniform orientation (Fig. 9.3); at Aïn el Dahlia Kébira between 88° and 121° and at Djebila between 51° and 134°. At Aïn el Kébira Dahlia, where there is more data, the heads are oriented in two main groups: to the north-west (between 285° and 319°) and the south-east (between 201° and 257°). The bodies were deposited in the left lateral extended position (22 at Aïn Dahlia el Kebira and 10 at Djebila) or the right (eight at Aïn Dahlia el Kebira and five at Djebila), however, for the majority the side could not be determined. The precise age and sex of the deceased individuals are not known. The ratio of adults to children is approximately 4:1. At Djebila, burials were located over the whole area, but the children were predominantly concentrated in a zone on the summit (Fig. 9.4). Eight individuals are reported as having their arms folded towards the face, while the skulls of two children showed signs of postmortem trepanation. The bodies were covered in a red sediment (ochre?) which left traces of colour on the bones. The same practice was observed in the 87 tombs of the Rouaszi-Skhirat necropolis (around 200 km to the south, dating to the middle of the fourth to the third millennium BC), in the earlier excavated tombs in the region of Tangier at El Mers and El Mries, as well as at the necropolis of Marshan-Tangeri (where the 53 rock-cut tombs had an orientation of between 82° and 111°).24 Tombs were built from slabs of local limestone to form rectangular or trapezoidal cists using a technique of extraction, processing and disposal that was also used in the second millennium BC in the same context.25 The tombs A5 and 78 were built with walls of small blocks and the bodies 19 23 24

25

Ponisch 1967; 1970. 20 López Pardo 1990. 21 El Azifi 1995. 22 Pappa 2009. El-Khayari 2007. Rouaszi-Skhirat: Lacombe et al. 1990 with an anthropologically non-homogenous population. El Mers and El Mries: Jodin 1964. Marshan-Tangeri: Ponsich 1970, 230–36, Fig. 3. Jodin 1964.

https://doi.org/10.1017/9781108634311.009 Published online by Cambridge University Press

287

288

Emanuele Papi

Aïn Dahlia Kebira

Decubitus

N

N

70 62 60

50

Tombs

skull

40

30 22 20

t de

Direction of the skeletons

In

Direction of the tombs

te

rm

in a

ft

Ri gh

Le

0

bl e

8

10

Decubitus

Djebila N

N

120 98

100

Tombs

80

skull

60

40

20

10 5

le

t

ab

gh

in rm

Direction of the skeletons

In

de

te

Direction of the tombs

Ri

Le

ft

0

Ain Dahlia Kebira

Marshan

Djebila

90

90

80

80

83

Tombs

70

70 70

60

60

50

50

Tombs

N

40 30

40 30

29

22 20

20

10

10

0

Direction of the tombs

Adults Children

0 Adults Children

Numerical ratio of adults and children

Figure 9.3. Orientation of tombs and skeletal remains, position of bodies and percentage of adults/children of the cemeteries of Aïn Dahlia el Kebiria, Djebila and Marchan/ Tangier.

https://doi.org/10.1017/9781108634311.009 Published online by Cambridge University Press

9 Revisiting First Millennium BC Graves in North-West Morocco

Djebila

Number of burials per square 0 1–3 4–5 6–7 8–12

Number of children burials per square 0 1–2 3–4 5

Figure 9.4. Distribution of the tombs in the necropolis of Djebila.

were placed in cists of stone, but otherwise they do not differ from the others in terms of the grave goods or the position of the corpse. A13 was a burial chamber with an opening, surrounded and paved with slabs, however, the position of the deceased and the grave goods were also in accordance with the other burials. The tomb D5 was built with blocks in ‘grand appareil’, one of the oldest examples in Morocco, also used for the tombs of Moghoga and Ras Achakar (Fig. 9.2a). The attribution of the square masonry to a transfer of technology from a foreign context to local traditions is based on diffusionist theories and we cannot exclude, as we can for opus africanum, that it represents a technology that is attested simultaneously in different contexts in response to different construction needs.

Tumuli It is worth examining the entirely indigenous tumuli of Morocco to see how closely the burial practice resembles that of these supposedly ‘LibycoPhoenician cemeteries’. Souville listed 24 sites with tumuli in the region of

https://doi.org/10.1017/9781108634311.009 Published online by Cambridge University Press

289

290

Emanuele Papi

Sidi Allal El Bahraoui

Volubilis

Souk el Gour

Sidi Jari

Sidi Slimane

Lalla Mimouna?

Mzora

Sidi Bchir

Sidi Khelili

Bou Mimoun 20 m

Mausoleum of Augustus

Figure 9.5. Dimensions of monumental tumuli in the Atlantic region.

Larache, Meknes and Ouezzane (Fig. 9.2c).26 Their number is higher and here we have added five monumental tumuli of Bou Mimoun, Sidi Khlili, Lalla Mimouna, Sidi Allal el Bahraoui and Souk el-Gour.27 The data on the nine tumuli with diameter/length of between 30 and 100 m (Fig. 9.5) are summarised in Table 9.1. The bodies were buried in the centre or on one side of the tumulus, on the ground (Rommani, Zaërs), under a layer of raw clay (Sidi Allal elBahraoui), or in stone cists similar to those at Dahlia Aïn el Kebira and Djebila (Hasnoua, Bled Refit Khemis, Sidi Mohamed el Mhid, Lalla Ghano). The bodies were in lateral extended position, similar to those reported in the region of Tangier (Rommani, Hasnoua, Sidi Slimane). The tomb of Sidi Slimane has some exceptional characteristics: two individuals were placed in the burial chamber and two other individuals were placed in the corridor and in the front courtyard. Multiple burials in cists are also known in the region of Tangier, but for Sidi Slimane we cannot exclude the possibility of human sacrifice in the case of the two individuals in the corridor, considering that the mound G1/T3, the most important of Tayadirt, had two secondary burials where the bones showed signs of 26

Souville 1973.

27

Souville 1959 and Camps 1974.

https://doi.org/10.1017/9781108634311.009 Published online by Cambridge University Press

https://doi.org/10.1017/9781108634311.009 Published online by Cambridge University Press

Sidi Allal el Bahraoui

Bou Mimoun Not excavated

Sidi Bchir

Sidi Slimane

3

4

5

6

Body in right extended position, in pit under layer of clay (chape); amputated feet

Stone cist

Burial type

Handmade ceramics

Grave goods

Ø: 100 m H: 16 m Vol: 41,966 m³ Not excavated Ø: 60 m H: 8 m Vol: 7,536 m³ Ø: 47 m H: 6 m Vol: Body in the corridor in left 13 ivory and bone Tomb in the eastern half built objects, iron 3,468 m³ extended position; body of unfired brick and wood nails, amphora in courtyard in left with a ‘petit appareil’ façade: Mañá-Pascual extended position on corridor, courtyard, burial A4, Libyan a flagstone and covered chamber with ceiling of Thuya inscription with slabs; couple in the (Tetraclinis articulta), walledburial chamber in boxes/ up doors with clay cists, one in left extended and the other in left flexed position

Ø: 30 m H: 2.30 m Vol: 541 m³

Sidi Jari

2

Tumulus of clay, sand, stones and pebbles

Tumulus of stone and earth sur- Ø: 54/58 m H: 6 m Vol: 5,281 m³ rounded by a ring of squared blocks and by 167 monoliths and 2 larger monoliths to the W Not excavated Ø: 40 m

Mzora

1

Dimensions

Characteristics

No. Location

Souville 1973, 58 n. 10 Souville 1973, 130–33 n. 18; Arharbi 2009.

Souville 1973, 127.

Ponsich 1966b, 402 n. 22; Souville 1973, 39 n. 12. Souville 1959, 396.

Gozalbes Cravioto 2006 with previous bibliography

Bibliography

Table 9.1. Summary of data from the nine tumuli with diameter/length of between 30 and 100 m, region of Larache, Meknes and Ouezzane.

https://doi.org/10.1017/9781108634311.009 Published online by Cambridge University Press

Lalla Mimouna

9

Ø: 40 m

Not excavated 90 × 30 m Ø: 40 H: 5 m Vol: Stepped cylindrical tomb with c.2,000 m³ double concentric opus quadratum walls and courtyard

Sidi Khelili Gour

7 8

Dimensions

Characteristics

No. Location

Table 9.1. (Cont.) Burial type

‘céramique de Banasa’

Glass, ceramics

Grave goods

Souville 1973, 51 n. 1. Camps 1974 (cf. Fentress, Chapter 16, this volume for dating) Souville 1973

Bibliography

9 Revisiting First Millennium BC Graves in North-West Morocco

excarnation.28 In tumuli excavated around Oujda (25 out of 140 identified; Fig 9.2b), the bodies were laid on the ground individually. The position and orientation, however, cannot be determined.

Tomb Typologies Souville identified four tomb types based on construction technique and type of burial: 1. heaps of earth and stone; 2. coffin-shaped cist tomb; 3. ‘à chape’; and 4. built tomb.29 Following Camps’ classification, the Atlantic tumuli would all be ‘tertres’ of earth and stone of African tradition, the Mzora tumulus belong to the ‘grandes tertes à enceinte de monoliths’, and the Sidi Allal el Bahraoui tumuli to the ‘tertres à chape’; the Gour tumulus has a pair of concentric walls of opus quadratum.30 With the exception of the ellipsoidal tumulus of Sidi Khelli, the tumuli have a circular plan of between 30 and 100 m in diameter (Fig. 9.5) and the volume of earth for the reported dimensions is between 541 m³ and almost 42,000 m³. The tumulus of Mzora had the same diameter as the Medracen; the diameter of the Sidi Slimane tumulus was equal to the side of the Djedar C tumulus; and Sidi Khelili had a larger diameter than the ‘Tombeau de la Chrétienne’ (Fig. 9.6). Similar characteristics in building techniques and forms are found and there is no trace of the variety of shapes encountered in the region of Oujda, with its circular, rectangular, oval, square, semicircular and ‘à antennes’-shaped tumuli of piled or regularly arranged stones.31 We do not know whether the differences in forms and techniques found at Oujda are related to chronology or heterogeneous ethnic groups, or even to the status of the deceased (as is perhaps possible for the tumuli ‘à antennes’ of Saharan tradition). In contrast to Oujda and Tayadirt (Fig. 9.2), the Atlantic tumuli are neither found near settlements nor concentrated in cemeteries. The monumental tombs are distributed at an average distance of approximately 50 km from one another. Gour and Mzora were equipped with a square platform in front of the tomb, probably an altar for worshipping the deceased. The distribution of the ‘chefferies’32 and the locations of the monumental tumuli indicates a probable connection between the tombs of the tribal chiefs and the political-territorial entities of the ‘chefferies’, identifiable already in the first millennium BC. These territorial units can be shown from onomastic, historical, epigraphic and numismatic sources 28 32

Lambert 1967. 29 Souville 1959. Hamdoune 1993.

30

Camps 1961.

31

Voinot 1910.

https://doi.org/10.1017/9781108634311.009 Published online by Cambridge University Press

293

294

Emanuele Papi

Tombeau de la Chrétienne

Medracen Djedar C Sidi Slimane Souk el Gour

Sidi Allal el Bahraoui

Sidi Khelili

Mzora 0

10 m

Figure 9.6. Section of selected monumental tumuli in the Atlantic region compared to the ‘Tombeau de la Chrétienne’, the Medracen and Djedar C. (adapted from Camps 1961, Fig. 82)

to have continued for centuries during the Roman occupation. We do not know whether the placement of the tumuli of the tribal chiefs was determined by roads or by frontiers, but their distance from inhabited areas seems to rather indicate a territorial centrality for the surrounding communities.33

Chronology The bronze jewellery recovered at Rommani and Aïn el Aouda, the ceramics from Hasnouna and Sidi Mohamed el Mhid (next to the bodies), and the elephant tusks from Lalla Fatima al Douar Dehisset do not provide elements for dating. Five tumuli are datable with a certain level of approximation: an amphora from the third century BC was found in the tomb of Bled Riat el Khemis; a vessel ‘à chardon’ and fragments of ‘céramique de Banasa’ date the tumulus of Lalla Ghano and Lalla Momouna from between the fifth to third century BC, in the same way that the tumulus at Sidi Slimane is dated to the third century by the presence of a Mañá-Pascual A4 amphora. The tumulus at Mzora is dated to the fourth century BC. The chronology and interpretation of Gour is problematic: for Camps it should be dated to the seventh century AD on the basis of the 14C of certain carbon remains.34 However, there is some evidence that these were recovered from an Islamic pit or robber trench, and there is no doubt that the chronology, which contrasts with the type of technique employed, should be defined with more precision.

33 34

For a discussion of the influence of roads, see Bokbot, Chapter 10, this volume. Camps 1974. Also according to Fentress and Wilson 2016.

https://doi.org/10.1017/9781108634311.009 Published online by Cambridge University Press

9 Revisiting First Millennium BC Graves in North-West Morocco

Ain Dahlia el Kebira 5

0

10

15

20

25

30 A1 A2 A3 A4 A5 A6 A7 A8 A9 A10 A11 A12 A13 A14 A15 A16 A17 A18 A19 A20 A21 A22 A23 A24 A25 A26 A27 A28 A29 A30 A31 A32 A33 A34 A35 A36 A37 A38 A39 A40 A41 A42 A43 A44 A45 A46 A47 A48 A49 A50 A51 A52 A53 A54 A55 A56 A57 A58 A59 A60 A61 A62 A63 A64 A65 A66 A67 A68 A69 A70 A71 A72 A73 A74 A75 A76 A77 A78 A79 A80 A81 A82 A83 A84

WGG MJ CE OE SB IT GB IT - CE SB - CE MJ - CE IT - MJ SB - MJ GB - MJ SB - CE - MJ OE - IT - CE GB - MJ - CE SB - GB - IT SB - GB - CE SB - GB - OE IT - SB - GB - CE MJ - GB - OE - CE IT - SB - GB - OE MJ - IT - SB - GB - OE

MJ GB CE OE SB IT

0

5

10

15

20

25

30

35

40

45

50

55

60

65

70

Jewlery Ceramics Iron Tools Ostrich Eggs Without Grave Goods MJ = Metal Jewlery CE = Ceramics

OE = Ostrich Eggs SB = Shells and Bones

IT = Iron Tools GB = Glass Beads

Figure 9.7. Grave goods from the necropolis of Aïn Dahlia el Kebiria.

Grave Goods For the attestation and association of grave goods in the tombs of Aïn el Dahlia Kébira and Djebila, see Figures 9.7 and 9.8. Here too, there is abundant evidence for the African origins and tradition of these cemeteries.

https://doi.org/10.1017/9781108634311.009 Published online by Cambridge University Press

295

296

Emanuele Papi

0

Djebila

5

10

15

25

20

30

35

40 D1 D2 D3 D4 D5 D6 D7 D8 D9 D10 D11 D12 D13 D14 D15 D16 D17 D18 D19 D20 D21 D22 D23 D24 D25 D26 D27 D28 D29 D30 D31 D32 D33 D34 D35 D36 D37 D38 D39 D40 D41 D42 D43 D44 D45 D46 D47 D48 D49 D50 D51 D52 D53 D54 D55 D56 D57 D58 D59 D60 D61 D62 D63 D64 D65 D66 D67 D68 D69 D70 D71 D72 D73 D74 D75 D76 D77 D78 D79 D80 D81 D82 D83 D84 D85 D86 D87 D88 D89 D90 D91 D92 D93 D94 D95 D96 D97 D98 D99 D100 D101 D102 D103 D104 D105 D106 D107

WGG MJ CE OE SB IT GB IT - CE OE - CE SB - CE MJ - CE GB - CE IT - MJ SB - MJ GB - MJ OE - MJ OE - IT SB - IT SB - CE - MJ SB - GB - IT SB - GB - OE MJ - OE - CE MJ - OE - SB SB - GB - MJ SB - GB - MJ - CE SB - GB - MJ - OE SB - IT - MJ - OE - CE SB - IT - MJ - OE - GB SB - IT - MJ - CE - GB

0

5

10

15

20

25

30

35

40

45

50

55

60

65

70

MJ GB CE OE SB IT

Jewlery Ceramics Iron Tools Ostrich Eggs Without Grave Goods MJ = Metal Jewlery CE = Ceramics

OE = Ostrich Eggs SB = Shells and Bones

IT = Iron Tools GB = Glass Beads

Figure 9.8. Grave goods from the necropolis of Djebila.

Ceramics Pottery is present in approximately a third of the tombs (25 out of 84 at Aïn Dahlia el Kébira and 34 of 107 at Djebila), in association with jewellery, iron instruments and ostrich eggs. The vessels belong to two functional forms: 1. containers with globular body, neck, wide opening, with or without handles

https://doi.org/10.1017/9781108634311.009 Published online by Cambridge University Press

9 Revisiting First Millennium BC Graves in North-West Morocco

and 2. bowls. In some graves, a vase and bowl are paired, while in others only the vase or bowl appears. With the exception of a few wheel-turned examples, the majority of the vessels were handmade and were in all probability made locally, following technologies and forms already attested in the local megalithic cemeteries.35 The wheel-turned specimens decorated with stripes belong to the vessel type and technological repertoires of northern Morocco (‘céramique de Banasa’), the production of which is dated to before the fourth century BC.36 The ‘à chardon’ vases belong to a diffuse typological group in the centralwestern Mediterranean from the eighth to the first centuries BC, whose origin is attributed to local prehistoric repertoires or Phoenician-Punic or oriental containers. For now, the question remains unanswered.37 An ‘à chardon’ vase was found in the tumulus of Lalla Ghano in the Gharb.38

Glass Paste Beads Beads made from glass paste were used in limited numbers. They are predominantly blue in colour with a diameter of between 0.2 and 1.4 cm. The use of glass beads is also attested at Taydirt and in the Gour.39 It is not possible to know whether these were locally produced or imported (they are found at Carthage from the seventh century BC).40 At the end of the first millennium, the use of glass paste beads is attested in the cemeteries of Sala and in the second to third centuries AD.41 Eleven examples come from the military vicus of Thamusida.42 At Basra in the ninth to tenth centuries, there were six types of glass paste and the beads made at this site were exported as far as Awdaghust in the Kingdom of Ghana,43 illustrating the range of TransSaharan connections in the Islamic period.

Metals Figure 9.9 shows the location of gold, lead/silver, copper, iron and polycrystalline/chalcedony quartz deposits in the area between the Atlantic and the Mediterranean and between the Oued Moulouya and the Souss/Draa. 35 36

37 39 41 42

43

For the bowl see, for example, Ponsich 1970, Fig. 16. Kbiri Alaoui 2007, 116, hypothesises, on the basis of similarity, that the vase from tomb A30 was produced at Kouass. González-Alcalde 2009. 38 Rebuffat and Limane 2011, 75, M60. Lambert and Souville 1970. 40 Grose 1989. For Fazzan, see Cole et al. 2007. Boube 1977, Fig. 156 n. 7, Fig. 178 n. 7; 1999, 427. Akerraz et al. 2013, 308; for an example from Lixus see Clemente and Peraile 2001, 237, 245, Photo 3. Robertshaw et al. 2010.

https://doi.org/10.1017/9781108634311.009 Published online by Cambridge University Press

297

298

Emanuele Papi

Ceuta

Mediterranean Sea

Tangier

Fe Melilla

Pb

Fe

ukkos O.Lo

Pb

Atlantic Ocean

Fe

Ag

Pb Oujda

Pb Fe

Rabat

Meknes Bo u

Cu Fe

Re gr e g

SiO2

Fe Pb

Au Pb Ag Ag

SiO2

Cu

Pb

Pb Pb Cu Pb

Ag Cu

SiO2

Pb

Mogador

Cu

O.So

Cu Dr O.

SiO2

Cu

Cu Ag

Fe

SiO2

SiO2

Ag

Fe

Cu

Cu

Cu

Marrakech

Ag

Pb

Fe

Fe

Pb

Ag

ia er-Rb

Cu

Ag

Cu

SiO2 SiO2

SiO2

Oum

Fe

Fes Pb

Sn

Fe Fe

Au

O. Mo u lo uy a

O.Sebou

aa

Pb

Ag Pb

Cu

Pb Pb

us s

Cu

Ag

Cu

Cu Cu

0

200 km

Figure 9.9. Location of identified mines of north-west Morocco (adapted from Saadi 1975); Ag = silver, Au = gold, Cu = copper, Fe = iron, Pb = lead, SiO2 = chalcedony.

All of these minerals were used for the manufacture of the grave goods, but we only have archaeological data from Mogador, relating to iron production from the middle of the seventh to the middle of the sixth centuries BC. The excavations of Mogador, located on a peninsula 350 nautical miles from the Strait of Gibraltar, produced graffiti in Punic with both Semitic and non-Semitic names, which, in combination with the attestation of the cult of Astarte, allows the hypothesis that the site was visited seasonally by sailors and artisans, for the most part from the Iberian Peninsula. The imported materials include Cypriot vases, amphora of the Attic SOS type, as well as ceramics and amphora from Chios, similar to those manufactured on the Iberian Peninsula.44 The furnaces and slag found on the site indicate that in all probability the settlement was created for the 44

El Khayari 2004; on the recent surveys see Marzoli and El Khayari 2009 and 2010.

https://doi.org/10.1017/9781108634311.009 Published online by Cambridge University Press

9 Revisiting First Millennium BC Graves in North-West Morocco

reduction of iron extracted from the nearby (10 km) mines of Jebel El Hadid. Imported materials from the eighth and seventh centuries were also found at Sidi Driss on the Mediterranean coast and have been attributed to another centre for trade in the iron extracted from the nearby mine of Nakur, the exploitation of which is documented in the eighth and ninth centuries.45 In the two cemeteries in the Tangier region, iron was mainly used for scythes, knife blades or unidentifiable objects interpreted as weapons. Jewellery was present in most of the tombs with grave goods in association with vessels, ostrich eggs and tools. It is difficult to determine the gender of the deceased individuals on the basis of the jewellery sets. They are made of bronze, silver and gold and are combined with glass paste beads, perforated ostrich egg discs, shells and fish bones. Copper alloys were used in most of the jewellery in an indeterminable quantity. Jodin has already hypothesised the use of copper from Moroccan deposits.46 The shapes are attested in the Mediterranean, in many other sites of Morocco, including the Tafialet oasis, as well as in Mauretania.47 The attribution of the copper alloys to international trade rather than local production is based on the presence of some similar jewellery in Punic sites. At the beginning of the first millennium AD, lead was both imported and extracted from the argentiferous lead mines near Tamuda and on the Mediterranean coast.48 Silver was used in a much smaller quantity compared to bronze: a total of 163 g in the jewellery of Aïn Dahlia el Kebira and 221 g at Djebila in approximately 1/4 and 1/5 of the graves, respectively. Very few grams of gold were used for lobe and ‘nose’ pendants: 2 g at Aïn Dahlia el Kebira in 2 per cent of the graves and 3 g at Djebila in 3 per cent of the burials.

Shells The tombs yielded gastropod mollusc shells of the genus Cypraea, or cowrie shells, which were drilled and inserted into necklaces in different combinations: alone, with Cardium shells, with glass paste beads, bones and ivory, ostrich egg flakes, and silver pendants. The association with the skeletons indicates that the cowrie shells were intended for use by both adults and children, the sex of which are unknown. Cowrie shells were used 45 48

Kbiri Alaoui et al. 2004. 46 Jodin 1966. 47 Lambert and Souville 1970. An ingot at Volubilis came from Carthago Nova/Cartagena in Spain: Chatelain 1929. Ponisch 1966a.

https://doi.org/10.1017/9781108634311.009 Published online by Cambridge University Press

299

300

Emanuele Papi

in jewellery in 14 per cent of the burials at Aïn el Dahlia Kébira and 12 per cent at Djebila, half of which represented children. We do not know more detailed taxonomies of the cowrie shells and we cannot identify possible local sites of collection (in the southern Mediterranean Cypraea lurida, Erosaria spurca, Zonaria pyrum are common and more rarely Schilderia achatidea), or the imported species.49 The non-local use of cowrie shells coming from the Indian Ocean is known in the dolmens of Roknia and Bou Nouara in Algeria;50 cowrie shells from the Red Sea and the Indian Ocean were present at Carthage in the sixth and fifth centuries BC and at Merida in a tomb from the first century AD.51 Even the Cypraea moneta/ annulus used in the tenth and eleventh centuries as coins (wada’) for TransSaharan trade in West Africa are endemic in the Red Sea or the Indian Ocean.52 The earliest evidence for the presence of Cowrie moneta/annulus from the Indian Ocean in Atlantic European sites is documented at Zambujal north of Lisbon in the ‘Late Chalcolithic’.53 Some necklaces were also made using four perforated Cardium shells (Caerastoderma edule), which seem to be less popular in the jewellery than other bivalves.54 Cardium shells are common in the entire Mediterranean and the Atlantic coast of Africa. Some tombs have yielded non-identifiable shells, which were preferred in association with children.

Chalcedony Some agate beads streaked with white are attested. Chalcedony (agate and carnelian) was fairly widespread in luxury jewellery of Morocco in the middle of the first millennium BC: six beads in the main tumulus of Tayadirt, 12 in a necklace from Raqqada, others in a necklace from Kouass.55 In the Roman period, an incised carnelian ring and a carved slab were found at Thamusida in the valley of Sebou.56 It is also attested in the Sahara and the Tafilalt.57 It is possible that the stone was extracted from one of the many deposits of polycrystalline/chalcedony quartz found in Morocco (Fig. 9.9) from the region of Meknès to the Souss/Draa (also near Tayadirt), if not from the deposits of Sinai or continental Europe.58 49 51 52 53 55 56 58

Allan 1956. 50 Camps-Fabrer 1960, 96. Carthage: Reese 1991, 172, n. 88. Merida: Rodríguez Hidalgo et al. 2013. Lydon 2009, 75. For other evidence of cowrie in the Maghreb, see Camps-Fabrer 1960, 92–93. Reese 1991, 179, n. 138. 54 Camps-Fabrer 1960, 80. Tayadirt: Lambert 1967, 239–41. Raqqada: El-Khayari 2007. Kouass: Bridoux n.d. Akerraz et al. 2013, 365. 57 Lambert and Souville 1970. http://www.gemmo.eu/fr/agate.php.

https://doi.org/10.1017/9781108634311.009 Published online by Cambridge University Press

9 Revisiting First Millennium BC Graves in North-West Morocco

Fish Bones The necklace in the infant grave A30 was assembled with two vertebrae of Scombridae (more precise identification is not possible), a use that has few other attestations in the Maghreb.59 Five fish ‘écailles’ (for which more precise identification is not possible), perhaps decorated, were found in D58. No fishing tools were found in the tombs, only tools related to agriculture. Fishing activities, processing and trade of fish is attested in northern Atlantic Morocco from at least the fifth century BC by fillets of fish from Corinth in Mañá-Pascuál A4 amphoras produced around the Strait of Gibraltar.60

Worked Bone The body of D63 wore a finger ring made of bone.

Cereals Tombs A39 and D78 contained charred/burnt/carbonised wheat seeds, attesting to the deposition of agricultural products in the burials and cereal production in coastal villages as well as in the Tangier region in the middle of the first millennium BC. In northern Morocco, the oldest traces of agriculture and animal husbandry date back to the sixth millennium BC, documented by the charred remains of Triticum dicoccum at Khaf Taht el-Ghar, and by Triticum sp. (5209–5002 calBC), Hordeum vulgare and Triticum aestivum/durum (4988–4772 calBC and 4494–4350 calBC) from Ifri Oudadane.61 The grottos of El Khil near the settlement and necropolis of Djebila have yielded Triticum aestivum/durum and dicoccum seeds from contexts dating to the middle of the sixth millennium BC.62 Cultivation was integrated with sheep and goat husbandry.63 The theories of these economic practices spreading from Spain or of independent evolution remains an open question.64

59 61

62 63

64

Camps-Fabrer 1960, 100. 60 Maniatis et al. 1984; Munn 2003, 305. Ballouche and Marinval 2003: 5477–5068 calBC. For Ifri Oudadane, see Morales Mateos et al. 2013. Peña-Chocarro et al. 2011. Linstädter 2011; Linstädter et al. 2012 for Ifri Oudadane; Ouchaou and Amani 1997, for Kaf Taht el-Ghar. See discussion in Barich 2013.

https://doi.org/10.1017/9781108634311.009 Published online by Cambridge University Press

301

302

Emanuele Papi

Legumes Some tombs have yielded remains of legumes; only those from tomb D76 were possible to identify specifically as peas, otherwise only a generic identification is possible.65 Domestic legumes have a long history: the earliest evidence in Africa of the domestic lentil (Lens sp.) dates to the sixth millennium BC (5730–5561 calBC).66 Beans (Vicia faba) were cultivated or gathered from the sixth millennium in the Tangier region and appear in tomb D76 at Djebila and, complementary to agriculture, in the first–third century AD in the military vicus of Thamusida.67

Olive The reports of olive pits are not sufficient to identify specific species. The genus Olea has been identified on the Atlantic coast of Morocco from the fifth millennium BC.68 The charred remains of Olea europaea found at Lixus are the oldest remains attested thus far in Morocco (eighth–seventh century BC), although there is no data allowing for their attribution as either wild or cultivated.69 The attribution to the Phoenicians for the introduction of olive cultivation and olive oil production in Morocco remains uncertain.

Ivory The presence of elephant ivory is attested in two tumuli excavated in Gharb at Lalla Fatima/Dehisset (two tusks above the body) and Sidi Slimane (pyxis), and in the grave goods of tomb A30 in the Tangier region (ram’s head amulet). Between the fifth and fourth millennium BC, central Morocco was inhabited by elephants from the Savanna of the Loxodonta africana species, also attested at Lixus in the first millennium BC.70 In the fourth to third millennium, ivory jewellery was placed in the tombs at Rouazi-Skhirat;71 from the third millennium to after the first half of the second millennium, ivory from Morocco was exported to Spain, where, at 130 sites, 1,060 items have been found made from the tusks of 65 66 67 68 70

71

Ponsich 1970, 157: ‘activité agricole confirmée per les grains de blé, pois, févettes et olives.’ Morales Mateos et al. 2013. Tangier region: Peña-Chocarro 2011. Thamusida: Papi and Martorella 2007. Reille 1979; Ballouche and Damblon 1988. 69 Grau Almiero et al. 2010. Materials from Khef el Baroud between Rabat and Casablanca: Banerjee et al. 2011. Lixus: Grau Almero et al. 2010. Lacombe et al. 1990.

https://doi.org/10.1017/9781108634311.009 Published online by Cambridge University Press

9 Revisiting First Millennium BC Graves in North-West Morocco

walrus, mammoth, hippopotamus, sperm whale and three species of elephant, including the Loxodonta africana typical of Morocco.72 In Spain, the objects made from the tusks of Loxodonta africana were found mainly around the Strait of Gibraltar and have been attributed to barter between the populations of the two shores of the Strait; bell-shaped jars and worked metals found in Morocco would have been the counterpart of these exchanges. The ivory trade continued in the first half of the first millennium BC: 13 tusks were aboard the Phoenician wreck of the Bajo de la Campana (Murcia) that sank between the late-seventh and early-sixth centuries BC.73 Even in the fifth century AD, ivory arrived at the coast of Portugal. The raw material for these rare grave goods could have been found all over north-west Morocco and there is no evidence to suggest an exclusive trade with Mogador.74 According to Schuhmacher et al., ‘we still cannot exclude an even more distant Sub-Saharan origin of this African savannah elephant ivory.’75

Ostrich Egg Ostrich eggs were used as containers or in the form of perforated discs as necklace pendants. The same uses are attested in north Morocco starting in the sixth millennium BC.76 From the mid-first millennium BC (Fig. 9.10) the eggs have been found in settlements, and in the grave goods of tombs as containers with incised and painted decorations, cut in half as cups, or configured in discs with shells for necklace pendants.77 The distribution area of the eggs extends from the Mediterranean (Tangier, Melilla and Oujda) to internal sites of the Middle Atlas (Tayadirt) and the Atlantic coast at Mogador. In pre-Roman times, they are found in the settlements of Banasa, Volubilis and Lixus.78 At Thamusida, in the excavations of around 3 per cent of the settlement, approximately 150 egg fragments were found in strata dating from the first to the eleventh centuries AD. Morocco was a collection and distribution area of ostrich eggs (and perhaps also ostrich feathers), as is demonstrated 72 73 75 76

77

78

Schuhmacher et al. 2009. Mederos Martín and Ruiz Cabrero 2004; Sanmartín Ascaso 1986. 74 Pappa 2009. Schuhmacher et al. 2009. Bouzouggar and Collina-Girard 2005; Linstädter 2003, 49; for evidence from the necropolis of Rouazi-Skhirat see Lacombe et al. 1990; in general see Camps-Fabrer 1969. In settlements: Cintas 1954, 54. As decorations: Ponsich 1966b. As cups: Ponsich 1970, 20. For pendants: Lambert and Souville 1970, 64. Ponsich 1966b.

https://doi.org/10.1017/9781108634311.009 Published online by Cambridge University Press

303

304

Emanuele Papi

Mediterranean Sea

Tangier

Melilla

Atlantic Ocean

Lixus

Chafarinas islands

Ras Kebdana

Oujda

Banasa

Sala

Thamusida Volubilis

Tayadirt

Mogador

0

200 km

Figure 9.10. African habitat of the Struthio camelus, probable ancient habitat in Morocco, and distribution of ostrich eggshells in first millennium BC sites.

by the finds at Ras Kebdana (fifth–sixth centuries BC, in the lower valley of the Moulouya, Melilla), where the eggs were collected and then distributed towards the Iberian peninsula, the Balearic Islands and elsewhere in the Mediterranean.79 Currently the Struthio camelus (ostrich) habitat is between the Tropic of Cancer and the Equator, however, it is possible that in the first millennium BC ostrich was also common farther to the north, approximately up to the Grand Atlas as is indicated by rock carvings (there were reports in the 1970s of a colony of 300 individuals at Sidi Ifni).80 However, Struthio bones are absent at Lixus, Thamusida and in the Rif of northern Morocco. 79 80

Stambouli et al. 2004, 28,with evidence even on the Chafarinas Islands off the coast of the Rif. Rock carvings: Rodrigue 2006, 30. Camel colony: Ponsich 1970, n. 31.

https://doi.org/10.1017/9781108634311.009 Published online by Cambridge University Press

9 Revisiting First Millennium BC Graves in North-West Morocco

Conclusions The cemeteries of north-east Morocco follow a tradition attested in Atlantic Morocco from the fourth-third millennium and other contexts in the western Maghreb. Uniform characteristics in the inhumations are present in the position of the body, the orientation of the tombs, the presence of red pigment, the grave goods and the mixing of adult and children.81 It is difficult to differentiate gender or age on the basis of grave goods. Practices of cremation present in the Iron Age Iberian peninsula are absent, as are the supine deposition of bodies prevalent in Phoenician-Punic contexts. The burial tumuli of the chieftains are typical of north Atlantic Morocco, without architecture and decorations of Mediterranean origin such as in the mausolea of Numidia, while the Gour demonstrates Saharan elements. The form and techniques of the burials in the cemeteries are identical to the rectangular Moroccan cists of the second millennium BC, a model which is still used in some cases up to the middle of the first millennium. Many raw materials used to produce the grave goods were locally sourced or could be found within a radius of a few kilometres (cereals, legumes, olives, clay, shells, fish bones, elephant ivory); ostrich eggs and carnelian were probably imported from the pre-desert more than 500 km away, perhaps travelling along the Atlantic coast. There is no evidence for the first millennium BC to indicate Trans-Saharan goods/cargo; in the absence of taxonomic data, it is not possible to determine whether the cowries used as pendants were gathered on the Atlantic and Mediterranean shores or came from the Indian Ocean (the use of cowrie, however, still belongs to the local tradition). The metals used were available throughout north-west Africa; the exploitation of iron and copper is first attested in the first millennium BC, while we cannot identify the provenience of the gold and silver as coming from Africa or abroad. Trade with the Iberian Peninsula is attested from the third millennium, while from the first millennium, certain imported goods document the presence of Phoenician and Iberian traders in different centres of interior and coastal Morocco. Some jewellery pendants have a form that is probably Punic in origin and are common from the western Mediterranean to Saharan Africa, and were combined and integrated with other traditional pendants, and could suggest a difference in rank or gender. In this case, however, it is not easy to distinguish between imports and local production. Unlike the contemporary cemetery of Raqqada, near Lixus, we do not 81

See Fentress, Chapter 16, this volume.

https://doi.org/10.1017/9781108634311.009 Published online by Cambridge University Press

305

306

Emanuele Papi

find any symposial objects, local or imported. The liquid or solid contents of vessels are unknown and we should rule out wine at least for the absence of Vitis taxa and imported wine amphora. The African character of the cemeteries and tumuli is confirmed also in respect to the burial practices and grave goods of internal Morocco: the tombs of Oujda and the necropolis of the Middle Atlas. The attribution of the cemeteries of the region of Tangiers to supposed Libyco-Phoenician or even ‘colonial’ origins can thus be shown to be entirely constructed by those who wish to see all forms of culture in African contexts as introduced from elsewhere. Put so bluntly, we can obviously reject these hypotheses. However, the more insidious versions of creolisation and ‘third space’ are equally to be rejected here.82 Although it is clear that the local communities traded with Iberian and Western Punic merchants, their rituals and burial practices seem to have remained entirely local.

References Adams, S., Bosch, E., Balaresque, P. L., Ballereau, S. J., Lee, A. C., Arroyo, E., LópezParra, A. M., Aler, M., Gisbert Grifo, M. S., Brion, M., Carracedo, A., Lavihna, J., Martínez-Jarreta, B., Quintana-Murci, L., Picornell, A., Ramos, M., Skorecki, K., Behar, D. M., Calafell, F. and Jobling, M. A. 2008. The genetic legacy of religious diversity and intolerance: paternal lineages of Christians, Jews, and Muslims in the Iberian Peninsula. American Journal of Human Genetic 83: 725–36. Akerraz, A., Camporeale, S. and Papi, E. (eds). 2013. Sidi Ali ben Ahmed – Thamusida 3. Le matériel. Roma: Quasar. Allan, J. 1956. Cowry Shells of the World Seas. Melbourne: Australiana Society Publication. Anderung, C., Bouwman, A., Persson, P., Carretero, J. M., Ortega, A. I., Elburg, R., Smith, C., Arsuaga, J. L. Ellegren, H. and Götherstrom, A. 2005. Prehistoric contacts over the Straits of Gibraltar indicated by genetic analysis of Iberian Bronze Age cattle. National Academy of Sciences Proceedings 102: 8431–35. Antoine, M. 1952. Les grands lignes de la Préhistoire marocaine. Algiers: Les impressions Édita. Aranegui, C., López-Bertran, M. and Vives-Ferrándiz, J. 2011. The Strait and beyond: local communities in Phoenician Lixus (Larache, Morocco). In C. Sagona (ed.), Ceramics of the Phoenician-Punic World, Ancient Near Eastern Studies Supplement 36, Leuven, Paris, Walpole MA.: Peeters, 297–326. 82

Webster 2001; Malkin 2011.

https://doi.org/10.1017/9781108634311.009 Published online by Cambridge University Press

9 Revisiting First Millennium BC Graves in North-West Morocco

Arharbi, R. 2009. À propos de la chronologie du monument funéraire de Sidi Slimane: le tumulus de Koudia El Hamra. Bulletin d’archéologie marocaine 21: 246–49. Ballouche, A. and Damblon, F. 1988. Nouvelles données palynologiques sur la végétation holocène du Maroc. In C. Caratini (ed.), Palynologie, Écologie, Paléoécologie, Actes du Xe Symposium de l’Association de Palynologues de Langue Françaises, Paris: Institut Français de Pondichéry, 83–90. Ballouche, A. and Marinval, P. 2003. Données palynologiques et carpologiques sur la domestication des plantes et l’agriculture dans le Nèolithique ancien du Maroc septentrional (site de Kaf Taht el-Ghar). Révue d’Archéometrie 27: 49–54. Banerjee, A., Dindorf, W., Mikdad, A., Reischmann, T., and Schuhmacher, T. X. 2011. Die Elfenbeinfunde aus Kehf-el-Baroud (Ziaïda, Ben Slimane, Marokko) und die Frage des Nordafrikanischen Elefanten. Madrider Mitteilungen 52: 113–38. Barich, B. 2013. Northwest Libya from the early to the late Holocene: new data on environment and subsistence from Jebel Gharbi. Quaternary International 320: 15–27. Boube, J. 1977. Sala 3. Les nécropoles. Planches. Rabat: Musée des Antiquités. Boube, J. 1999. Les nécropoles de Sala. Paris: Editions Recherche sur les civilisations. Bouzouggar, A. and Collina-Girard, J. 2005. Les premiers résultats des datations des sites néolithiques au Cap de l’Eau (Maroc oriental). Nouvelles Archéologiques et Patrimoniales, Bulletin semestriel publié par les enseignants-chercheurs de l’Institut National des Sciences de l’Archéologie et du Patrimoine 6: 6–7. Bridoux, V. n.d. Activités archéologiques de l’École française de Rome, Kouass. http://www.efrome.it/la-recherche/programmes/detail-programme/detail/koua ss.html – accessed 23/05/2016. Clemente, I. C. and Peraile, I. I. 2001. Varia. Objetos diversos hallados en las excavaciones recientes. In A. C. Gascó (ed.), Lixus. Colonia fenicia y ciudad púnico-mauritana. Anotaciones sobre su ocupación medieval, Valencia: Saguntum Extra 4, 169–85. Camps, G. 1960. Les traces d’un Age du Bronze. Revue Africaine 104: 31–57. Camps, G. 1961. Aux origines de la Berbérie. Monuments et rites funéraires protohistoriques. Paris: Arts et métiers graphiques. Camps, G. 1965. Les dolmens marocaines. Libyca 13: 235–47. Camps, G. 1974. Le Gour, mausolée berbère du VIIe siècle. Antiquités africaines 8: 191–208. Camps-Fabrer, H. 1960. Parure des temps préhistoriques en Afrique du Nord. Libyca 8: 9–218. Camps Fabrer, H. 1969. La disparition de l’autruche en Afrique du Nord. Revue Africaine 106: 33–74. Chatelain, L. 1929. Note sur un lingot de plomb de Volubilis. Bulletin du Comité des Travaux Historiques 1928/1929: 416–417.

https://doi.org/10.1017/9781108634311.009 Published online by Cambridge University Press

307

308

Emanuele Papi

Cintas, P. 1954. Contribution à l’étude de l’expansion carthaginoise au Maroc. Publications de l’Institut des Hautes-Etudes Marocaines 56. Paris: Arts et métiers graphiques. Cole, F., Hoffmann, B., Parton, H., Sauer, E. and Mattingly, D. J. 2007. Fazzan project survey small finds report. In D. J. Mattingly (ed.), The Archaeology of Fazzan. Volume 2, Site Gazetteer, Pottery and other Survey Finds. London: The Society for Libyan Studies, Department of Antiquities, 463–99. El Azifi, M. R. 1995. Les nécropoles de la région de Tanger sont-elles phéniciennes? In M. H. Fantar and M. Ghaki (eds), Actes du III Congrès international des études phéniciennes et puniques, Tunis: Institut National du Patrimoine, 401–14. El-Khayari, A. 2004. Échanges entre le Maroc et la Méditerranée de l’époque phénicienne à l’époque tardo-républicaine. In A. Gallina Zevi and R. Turchetti (eds), Méditerranée occidentale antique: les échanges, Soveria Mannelli: Rubettino Editore, 149–68. El-Khayari, A. 2007. La présence phénicienne au Maroc. Les dossiers d’archéologie: 56–58. Fentress, E. W. B. and Wilson, A. I. 2016. The Saharan Berber diaspora and the southern frontiers of Byzantine North Africa. In S. T. Stevens and J. P. Conant (eds), North Africa under Byzantium and Early Islam, Dumbarton Oaks Byzantine Symposia and Colloquia 7, Washington, DC: Dumbarton Oaks Research Library and Collection, 41–63. González-Alcalde, J. 2009. Una aproximación cultural a los vasos caliciformes ibérico en cuevas-santuario y yacimientos de superficie. Quaderns de prehistòria i arqueologia de Castellò 27: 83–107. Gozalbes Cravioto, E. 2006. El monumento protohistórico de Mezora (Arcila, Marruecos). Archivo de Prehistoria Levantina 26: 323–48. Grau Almiero, E. and Pérez Jordá, G. Iborra Eres, M. P. 2010. La gestión de los recursos naturales. In C. Aranegui and H. Hassini (eds), Lixus 3, Valencia: Saguntum Extra 8: 61–63. Grose, D. F. 1989. Early Ancient Glass: Core-Formed, Rod-Formed, and Cast Vessels and Objects from the Late Bronze Age to the Early Roman Empire, 1600 BC to AD 50. New York: Hudson Hills Press. Hamdoune, C. 1993. Ptolémée et la localisation des tribus de Tingitane. Mélanges de l’Ecole Française de Rome. Antiquité 105: 241–89. Jodin, A. 1956. Les civilisations au sud de l’Espagne et l’néolithique marocaine. In Congrès Préhistorique de France, Poitiers-Angoulème, XVe session, Paris: Société préhistorique française, 564–78. Jodin, A. 1964. L’Âge du bronze au Maroc: la nécropole mégalithique d’El Mriès (Oued Bou-Khalf, Tanger). Bulletin d’Archéologie marocaine 5: 11–45. Jodin, A. 1966. Les gisements de cuivre du Maroc et l’archéologie des métaux. Bulletin d’Archéologie marocaine 6: 11–27. Kbiri Alaoui, M. 2007. Revisando Kuass (Asilah, Marruecos). Talleres cerámicos en un enclave fenicio, púnico y mauritano. Valencia: Saguntum Extra 7.

https://doi.org/10.1017/9781108634311.009 Published online by Cambridge University Press

9 Revisiting First Millennium BC Graves in North-West Morocco

Kbiri Alaoui, M., Siraj, A. and Vismara, C. 2004. Recherches archéologiques maroco-italiennes dans le Rif. Africa Romana 15: 567–604. Lacombe, J. P., Daugas, J. P. and Sbihi-Alaoui, F. Z. 1990. La nécropole néolithique de Rouazi-Skhirat (Maroc), présentation de l’étude des sépultures. Bulletins et Mémoires de la Société d’Anthropologie de Paris 2: 55–60. Lambert, N. 1967. Tayadirt, une nécropole protohistorique en haute Moulouya. Libyca 15: 215–60. Lambert, N. and Souville, G. 1970. Influences orientales dans la nécropole mégalitique de Tayadirt (Maroc). Antiquités Africaines 4: 63–74. Linstädter, J. 2003. Le site néolithique de l’abri de Hassi Ouenzga (Rif Oriental, Maroc). Beiträge zur allgemeinen und vergleichenden Archäologie 23: 85–138. Linstädter, J. 2011. The Epipalaeolithic Neolithic transition in the Eastern Rif Mountains and the Lower Moulouya valley, Morocco. In J. F. Gibaja Bao, A. F. Carvalho and A. F. Bicho (eds), The Last Hunter-Gatherers and the First Farming Communities in the South of the Iberian Peninsula and North of Morocco, Proceedings of the work shop (Faro 2009), Algarve: Universidad de Algarve Promotoria Monográfica 15, 92–100. Linstädter, J., Medved, I., Solich, M. and Weniger, G. C. 2012. Neolithisation process within the Alboran territory: models and possible African impact. Quaternary International 274: 219–32. López Pardo, F. 1990. Sobre la expansión fenicio-púnica en Marruecos, Algunas precisiones a la documentación arqueológica. Archivo Español de Arqueología 63: 7–41. Lydon, G. 2009. On Trans-Saharan Trails: Islamic Law, Trade Networks, and CrossCultural Exchange in Nineteenth-Century Western Africa. Cambridge: Cambridge University Press. Malhomme, J. 1959. Corpus des Gravures rupestres du Grand Atlas. Rabat: Service des antiquités du Maroc. Malkin, I. 2011. A Small Greek World: Networks in the Ancient Mediterranean. Oxford: Oxford University Press. Maniatis, Y., Jones, R. E., Whitbread, I. K., Kostikas, A., Simopoulos, A., Karakalos, C. and Williams, C. K. 1984. Punic amphorae found at Corinth, Greece: an investigation of their origin and technology. Journal of Field Archaeology 11: 205–22. Marzoli, D. and El-Khayari, A. 2009. Mogador (Essaouira, Marokko), Vorbericht über die Kampagnen 2006 und 2007. Madrieder Mitteilungen 50: 81–101. Marzoli, D. and El-Khayari, A. 2010. Vorbericht Mogador (Marokko) 2008. Madrieder Mitteilungen 51: 61–108. Mederos Martìn, A. and Ruiz Cabrero, L. A. 2004. El pecio fenicio del Bajo de la Campana (Murcia, España) y el comercio del Marfil norteafricano. Zephyrus 57: 263–81. Morales Mateos, J., Pérez-Jordá, G., Peña-Chocarro, L., Zapata, L., Ruíz-Alonso, L., López-Sáez, M. and Linstädter, J. A. 2013. The origins of agriculture in

https://doi.org/10.1017/9781108634311.009 Published online by Cambridge University Press

309

310

Emanuele Papi

north-west Africa: macro-botanical remains from Epipalaeolitic and Early Neolitic levels of Ifri Oudadane (Morocco). Journal of Ancient Society 40: 2659–69. Munn, M. L. Z. 2003. Corinthian trade with the Punic west in the Classical period. In C. K. Williams II and N. Bookidis (eds), Corinth, The Centenary, 1896–1996, Princeton: The American School of Classical Studies at Athens, 195–218. Ouchaou, B. and Amani, F. 1997. Etude préliminaire des grands mammifères du gisement de Kaf-taht-el Ghar (Tetouan, Maroc). Préhistoire Anthropologie Méditerranéennes 6: 53–60. Papi, E. 2014. Punic Mauretania? In J. Crawley Quinn and N. Vella (eds), The Punic Mediterranean: Identities and Identification from Phoenician Settlement to Roman Rule, British School at Rome Monographs, Cambridge: Cambridge University Press, 199–215. Papi, E. 2016. L’invention du Círculo de l’Estrecho. In M. Coltelloni-Trannoy, V. Bridoux and V. Brouquier-Reddé (eds), Le cercle du détroit dans l’antiquité. L’héritage de Miguel Tarradell, Karthago Supplement 29, Louvain: Peeters, 105–20. Papi, E. and Martorella, F. 2007. Il grano della Tingitana, In E. Papi (ed.), Supplying Rome and the Empire, Journal of Roman Archaeology Supplement 69, 85–96. Pappa, E. 2009. Reflections on the earliest Phoenician presence in north-west Africa. Talanta 40–41 [2008–2009]: 53–72. Peña-Chocarro, L., Pérez Jordà, G., Abel Schaad, D., Gibaja, J. F., López-Romero, E., López-Sáez, J. A., Moreno-García, M. and Amani F. 2011. La campaña de excavación 2011 en las cuevas de El Khil (Achakar, Tánger, Marruecos). Bienes Culturales. Revista del Instituto del Patrimonio Histórico Español 9: 546–61. Pereira, F., Queirós, S., Gusmão, L., Nijman, I. J., Cuppen, E. Lenstra, J. A., Davis, S. J. M., Nejmeddine, F. and Amorim, A. 2009. Tracing the history of goat pastoralism: new clues from mitochondrial and Y chromosome DNA in North Africa. Moleculal Biology and Evolution 26: 2765–73. Ponsich, M. 1966. Le trafic du plomb dans le détroit de Gibraltar. In R. Chevallier (ed.), Mélanges d’archéologie et d’histoire offerts à A. Piganiol. Paris: S.E.V.P.E. N., 1276–77. Ponsich, M. 1966b. Tanger. Un oeuf d’autruche décoré. Bulletin d’Archéologie Marocaine 6: 461–64. Ponsich, M. 1967. Nécropoles phéniciennes de la région de Tanger. Tangier: Editions marocaines et internationales. Ponsich, M. 1970. Recherches archéologiques à Tanger et dans sa région. Paris: C.N.R.S. Rebuffat, R. and Limane H. 2011. Carte archéologique du Maroc antique. Le bassin du Sebou 1. Au sud du Loukkos, Villes et site archéologiques du Maroc 2. Rabat: Royaume du Maroc, Ministère de la Culture, Institut National des Sciences de l’Archéologie et du Patrimoine.

https://doi.org/10.1017/9781108634311.009 Published online by Cambridge University Press

9 Revisiting First Millennium BC Graves in North-West Morocco

Reille, M. 1979. Analyse pollinique du lac de Sidi Bou Rhaba, littoral atlantique (Maroc). Ecologia Mediterranea 4: 61–65. Reese, D. S. 1991. The trade of Indo-pacific shells into the Mediterranean basin and Europe. Oxford Journal of Archaeology 10: 159–67. Rhouda, T., Martínez-Redondo, D., Gómez-Durán, A., Elmtili, N., Idaomar, M., Díez-Sánchez, C., Montoya, J., López-Pérez, M. J. and Ruiz-Pesini, E. 2009. Moroccan mitochondrial genetic background suggests prehistoric human migrations across the Gibraltar Strait. Mitochondrion 9: 402–07. Robertshaw, P., Benco, N., Wood, M., Dussubieux, L., Melchiorre, E. and Ettahiri, A. 2010. Chemical analysis of glass beads from Medieval Al-Basra (Morocco). Archaeometry 53: 55–79. Rodrigue, A. 2006. L’Homme et les fauves dans le Haut Atlas marocain. Anthropozoologica 41.2: 29–35. Rodríguez Hidalgo, A., Gibello Bravo, V., Menéndez Menéndez, A., Sanabria Murillo, D. and Sánchez Hidalgo, F. 2013. Un ejemplar de Cypraea pantherina en una tumba alto imperial de Augusta Emerita. Zephyrus 72: 183–93. Saadi, M. 1975. Les anciennes mines du Maroc (de la Préhistoire au XIXème siècle): province du nord (1:200,000). Rabat: Éditions du Service Géologique du Maroc. Sáez Martín, B. 1952. Sobre una supuesta Edad del Bronce en Africa Menor y Sahara. In Actes du Congrès panafricain de préhistoire. II session, 659–62. Schuhmacher, T. X., Cardoso, J. L. and Banerjee, A. 2009. Sourcing African ivory in Chalcolithic Portugal. Antiquity 83: 983–97. Sanmartín Ascaso, J. 1986. Inscripciones fenicio-púnicas del sureste hispánico. In G. Olmo Lete and M. E. Aubet (eds), Los Fenicios en la Péninsula Ibérica, Aula Orientalis 4, Barcelona: Editorial Ausa, 89–103. Souville, G. 1959. Principaux types de tumulus marocains. Bulletin de la Société préhistorique de France 56: 394–402. Souville, G. 1973. Atlas préhistorique du Maroc I. Le Maroc atlantique. Paris: CNRS. Stambouli, A., Malek, F., Bouzzouggar, A. and Otte, M. 2004. Datations radiométriques des sites néolithiques du Nord marocain entre Tanger et Cap de l’Eau. In M. Otte, A. Bouzzouggar and J. Kozlowski (eds), La Préhistoire de Tanger (Maroc), Liège: Université de Liège, 27–31. Tissot, J. C. and Broca, P. 1876. Sur les monuments mégalithiques et les populations blondes du Maroc. Revue d’anthropologie 3: 385–404. Voinot, L. 1910. Les tumuli d’Oudjda. Bulletin de Société de Géographie et Archéologie de Oran 30: 516-28. Webster, J. 2001. Creolizing the Roman provinces. American Journal of Archaeology 105: 209–25.

https://doi.org/10.1017/9781108634311.009 Published online by Cambridge University Press

311

https://doi.org/10.1017/9781108634311.009 Published online by Cambridge University Press

part iv

Looking West

Published online by Cambridge University Press

Published online by Cambridge University Press

10

Protohistoric and Pre-Islamic Funerary Archaeology in the Moroccan Pre-Sahara youssef bokbot

Introduction Despite more than 150 years of archaeological research in the Maghrib and the Sahara, it is still difficult to establish a universally accepted chronological framework for Moroccan protohistory. While it is generally agreed that its beginning corresponds with the appearance of metal objects around the start of the third millennium BC, its end is much more imprecise, vague and uncertain. The Mediterranean littoral and its hinterland first entered written history around the eighth century BC, at the date of the earliest evidence for a Phoenician presence. However, the other continental regions of Morocco were not part of this schema. These areas were at the margins of ancient knowledge. During the centuries of the ‘Phoenician presence’, only the urban centres changed their character; the countryside, on the other hand, retained a protohistoric lifestyle. For a long time, protohistoric archaeology in pre-Saharan Morocco was limited to the study of funerary monuments. This state of affairs is primarily due to the fact that these monuments are easily identifiable in the field as they are very distinct from Punic, Roman or Islamic tombs. Moreover, the excavators preferred to dig funerary monuments as they were well bounded and contained structures in which they were certain to find grave goods, sometimes including valuable objects in ceramic or metal. Researchers into the protohistory of Moroccan Sahara are confronted by countless limitations of the data and lacunae that are difficult to resolve.1 What are the reasons for the rarity, not to say the absence in some areas, of protohistoric settlements? Conversely, how can the diversity in funerary monuments and their concentration in well-defined areas be explained? Were these necropolises the work of sedentary people? Is the geographical proximity between tumulus cemeteries, rock art and areas of mining activities evidence for their probable contemporaneity? How can 1

Bokbot 2001; 2005.

https://doi.org/10.1017/9781108634311.010 Published online by Cambridge University Press

315

316

Youssef Bokbot

we explain the abundant depictions of metal objects engraved on rocks with the few authentic metallic objects? The funerary monuments of pre-Saharan Morocco take the form of heaps of stone or mounds of earth covering funerary chambers built above ground or dug under ground level. These chambers were sometimes fitted with quite complex architectural elements. A significant number of these monuments were either directly or indirectly associated with steles bearing Libyco-Berber inscriptions. There also exists a spatial and topographic recurrence between funerary monuments and rock art. Monuments could be isolated, or organised into small groups, but sometimes they also formed real necropolis areas where they can be counted in their hundreds or even thousands. These types of tombs are often known to local people, who refer to them using various Berber or Arabic (dialectal) appellations and often attribute them to foreign populations: akerkur/ikerkar (Arabic dialect: kerkur/kraker), rjem/larjam, imirš/imaršen, amariy/imariyn (‘mound, stone heap’), qbur lahlaliyin (‘tombs of the Hilalians’), mdinet lihud (‘Jewish city’), qbur ljuhala (‘tombs of the pagans’) or qbur lbartqiz (‘tombs of the Portuguese’).2 But despite there being more than 10,000 tumuli recorded, only a few dozen have been systematically excavated. The dating of these cemeteries is far from being precisely established. Their typological diversity may suggest a long chronology that could stretch from the Late Neolithic to the generalisation of funerary practices linked with the expansion of Islam after AD 1000. Protohistoric research in the pre-Saharan zone is clearly underdeveloped compared to the north of Morocco (Fig. 10.1). The first excavations were only conducted in 1938 in the Tafilalat. The necropolis of Erfoud is known to us through the work of Armand Ruhlmann.3 Research conducted about twenty years later led to the discovery in the necropolises of Taouz, Braber and Bouïa of tumuli with ‘funerary chapels’, a type of funerary monument very common in the Saharan areas of the Maghrib.4 Further west of the Tafilalat, Djamila Jacques-Meunié carried out survey work in the valley of the Wadi Draa.5 She discovered various groups of funerary monuments, the most important of which is located at Foum Larjam. More recently, between 2003 and 2005, a set of funerary monuments associated with a complex of rock engravings were discovered and excavated in the basin of the Wadi Noun.6 2 6

Bokbot 2015. 3 Rulmann 1939. Bokbot et al. 2011.

4

Bokbot 1991.

https://doi.org/10.1017/9781108634311.010 Published online by Cambridge University Press

5

Jacques-Meunié 1958.

10 Funerary Archaeology in the Moroccan Pre-Sahara

Bouia

317

Erfoud Braber Tafilalet Taouz

Wadi Draa Foum Larjam

Adrar Zerzem

0

100

200

300

400

500 km

N

Figure 10.1. Distribution map of cemeteries from the Moroccan pre-Sahara cited in the text (made by M. Sterry).

The study of these monuments shows that this region, although today inhospitable desert, was open to several currents of cultural influences, which contributed to a wealth of architectural forms, in relation to the religious and funerary beliefs.

Funerary Zones Erfoud The necropolis of Erfoud is situated on the left bank of the Wadi Ziz, 2 km south-east of the city of Erfoud. It extends over the slopes of the Jabal Erfoud and Ghabt Lahjar – ‘forest of stones’. Excavated in 1938 by Armand Ruhlmann, it is comprised of many small groups of no more than ten tumuli that are relatively distant from one another.7 The diameter of these tumuli varied between 4 and 8 m and the height between 1 and 2 m, built on a fairly uniform model (Fig. 10.2). The outside of the tombs were constructed from dry stone, generally in the shape of a truncated cone. These structures cover funerary chambers entirely dug below the level of the ground in the form of a cist, the base, sides and top of which were made 7

Rulmann 1939.

https://doi.org/10.1017/9781108634311.010 Published online by Cambridge University Press

318

Youssef Bokbot

Figure 10.2. Tumulus ‘à fosse’ from Erfoud (after Ruhlmann 1939, 74 re-drawn by M. Hawkes).

of large slabs. The grave goods found in the five excavated tumuli comprised a socket, fragments of iron, remains of a spear or javelin head and a few wooden fragments. Three of the tumuli contained collective or multiple burials while the other two were for individual burials. Bodies were most commonly laid supine.8 Based on the frequency of this type of inhumation rite in the tumuli with cists of the Erfoud necropolis, Stéphane Gsell suggested a recent dating for them, contemporary with the Roman occupation north of the Atlas.9 However, Ruhlmann’s excavations did not yield any archaeological materials that help in closely dating these monuments. During a survey mission in 1986, a small bronze or copper bell was discovered in the rubble of a looted tumulus.10 It is an uncommon object, not previously known from protohistoric funerary monuments in Morocco, possibly of Punic origin or inspiration. This bell resembles another one found in the ‘tell’ of Mogador.11 Similar objects were found in Libyco-Berber tombs in Tangier,12 as well as in Punic tombs in Carthage.13 According to Anziani, these small bells were used as ornaments rather than utilitarian objects, but disappeared in the fourth century BC.14 A general date for the cemetery in the first millennium BC can, thus, be postulated. However, it is worth referencing here a similar object found in one of the Garamantian cemeteries of the Wadi al-Ajal (TAG012), dated to the first half of the first millennium AD, a much later date compared to what has been suggested for the Maghribian bells.15 8 11 14

9 Rulmann 1939. Gsell 1927, Tome 6. 10 Bokbot 1991, 401–02. Ponsich 1970, 147. 12 Jodin 1966, 173–75. 13 Gauckler 1915, 1–2. Gauckler 1915, 1–2. 15 See Mattingly et al., Chapter 2, this volume.

https://doi.org/10.1017/9781108634311.010 Published online by Cambridge University Press

10 Funerary Archaeology in the Moroccan Pre-Sahara

Figure 10.3. Bouïa Necropolis, tumulus Type 1 (after Margat and Camus 1959, 350, re-drawn by M. Hawkes).

Bouïa The Jabal Bouïa is a small massif of primary outcrops emerging in the plain of the Tafilalat and bordering the Wadi Ghéris. An estimated 1,200 tumuli are scattered across the entire surface of the Jabal Bouïa. This necropolis was discovered by J. Margat and A. Camus in 1957.16 The three excavated tumuli each correspond to well-represented types in the necropolis. Unfortunately, the excavators were not archaeologists and in the published report only the raw results of their work were delivered, without further analysis or dating. The most common burial form was the Type 1 conical tumulus, with a pointed top, a rounded base and a diameter of between 4 and 8 m (Fig. 10.3). Also, relatively frequent were the larger Type 3 tumuli.17 They were circular in shape, with diameters between 10 and 16

Margat and Camus 1959.

17

Bokbot 1991, 403–11.

https://doi.org/10.1017/9781108634311.010 Published online by Cambridge University Press

319

320

Youssef Bokbot

Figure 10.4. Bouïa Necropolis, tumulus Type 3 (after Margat and Camus 1959, 354, re-drawn by M. Hawkes).

15 m. Orientated on the eastern side, there was an entrance leading into a ‘funerary chapel’ or set of antechambers (unconnected to the burial chamber, which lay in the centre of the tumulus). In front of this opening there were flat-topped, circular or square, structures that might have been used as offering or sacrifice tables (Fig. 10.4).18 The rarest form was the Type 2.19 This was a large tumulus with sides measuring between 10 and 15 m, but a square or rectangular shape and a flattened top. It also featured a ‘funerary chapel’, which was separated from the funerary chamber, and possible offering tables (Fig. 10.5). This type is similar to those found at Taouz and Braber (described below).

18

Bokbot 1991, 403–11.

19

Bokbot 1991, 403–11.

https://doi.org/10.1017/9781108634311.010 Published online by Cambridge University Press

10 Funerary Archaeology in the Moroccan Pre-Sahara

Figure 10.5. Bouïa Necropolis, tumulus Type 2 (after Margat and Camus 1959, 352, re-drawn by M. Hawkes).

Taouz The necropolis of Taouz spread across three hills located on the right bank of the Wadi Ziz, north of the village of Taouz at the south-eastern end of the Tafilalat, with a total of more than 100 tumuli concentrated at the bottom or mid-slope. Only two tumuli were entirely excavated by Jacques Meunié and Charles Allain.20 Five others were simply cleared on the eastern side in order to reveal their antechambers (Fig. 10.6). One of the excavated tumuli (T1 on Fig. 10.6) was circular with a diameter of 16 m and a height of 2.90 m.21 It was delimited by a dry stone wall with an opening on its eastern side onto a 5 m long corridor that ended in a niche. Along the axis of the corridor, two monolithic pillars had been placed to support the flagstones of the ceiling. The funerary chamber situated in the centre was located on the same axis as the corridor, but was clearly separated from it (Fig. 10.7). The second tumulus (T4) was rectangular in shape, 13 m long and 8 m wide.22 It was also surrounded by a dry stone wall and had a ‘funerary chapel’ on the east side with three openings 20

Meunié and Allain 1956.

21

Meunié and Allain 1956.

22

https://doi.org/10.1017/9781108634311.010 Published online by Cambridge University Press

Meunié and Allain 1956.

321

322

Youssef Bokbot

Figure 10.6. Necropolis of Taouz with location of tombs cited in the text (Image © 2016 DigitalGlobe, Google Earth).

Figure 10.7. Necropolis of Taouz, tumulus (T1) ‘à chapelle’ (photo Margat, Camps 1961, Pl. X.1).

https://doi.org/10.1017/9781108634311.010 Published online by Cambridge University Press

10 Funerary Archaeology in the Moroccan Pre-Sahara

Figure 10.8. Necropolis of Taouz, tumulus (T4) ‘à chapelle’ (after Meunié and Allain 1956, 74, re-drawn by M. Hawkes).

leading to six recesses (Fig. 10.8). The excavation of these recesses revealed fragments of bone and traces of hearths. The funerary chamber was rich in grave goods, primarily items of jewellery. No chronological attribution was suggested.

Braber The tumuli of Braber are located on the edge of the Hamada du Guir cliff, not far from the border-crossing between Algeria and Morocco of Hassi Braber. It has approximately 50 tumuli, only two of which were studied by C. Allain and J. Meunié.23 They were both tumuli with ‘funerary chapels’ (or antechambers), composed of corridors, perpendicular galleries and niches. Opposite the entrance of the funerary chapel, there were small rectangular structures that could again be interpreted as offering tables. An unusual element characteristic of Braber was a rectangular tumulus with three stepped levels and three openings in its south-east side. 23

Meunié and Allain 1956.

https://doi.org/10.1017/9781108634311.010 Published online by Cambridge University Press

323

324

Youssef Bokbot

Figure 10.9. Necropolis of Braber, stepped bazina with funerary chapel (Meunié and Allain 1956, 82, re-drawn by M. Hawkes).

The tomb measured 11 m wide, 13 m long and 3.50 m high.24 This tumulus can thus be linked to both the rectangular stepped type as well as to the tumulus with funerary chapel (Fig. 10.9).25

Foum Larjam In the Moroccan dialect, Foum Larjam means ‘the pass of the tumuli’. It is a vast necropolis extending over 20 km along the chain of Jabal Ben Selmane, 24

Meunié and Allain 1956.

25

Meunié and Allain 1956.

https://doi.org/10.1017/9781108634311.010 Published online by Cambridge University Press

10 Funerary Archaeology in the Moroccan Pre-Sahara

Figure 10.10. General map of the necropolis at Foum Larjam (yellow dots are tumuli; labelled are hilltop sites – Middle Draa Project, made by L. Rayne and M. Sterry).

situated in the area where the Wadi Draa bends and changes from a northwest to south-east course to an east-west direction (Fig. 10.10).26 The tumuli were conical with a pointed top and a corbelled funerary chamber. The excavators argued that a ‘window’ was constructed half-way up on the eastern side of the tumulus, giving the name to this distinctive and very localised type, the ‘tumulus à lucarne’, which seems restricted to the region of the Draa bend.27 A radiocarbon date published in the 1980s and two dates recently reported hint to a long use of the cemetery, from the early first millennium BC to mid-first millennium AD.28

Adrar Zerzem The Adrar Zerzem is located in the region of Guelmim, on the SaharoAtlantic slopes of the Anti-Atlas. It is a rocky ridge orientated east-west overlooking the right bank of the Wadi Eç-çayad. On this ridge there is a very interesting archaeological complex with various types of remains: 26 27

28

Jacques-Meunié 1958. Investigation at the cemetery has recently resumed as part of the ‘Middle Draa Project’, a joint venture between INSAP and University of Leicester under the direction of the present author and David Mattingly; see Mattingly et al. 2017 for report on the first field season. The cemetery also includes many mound cairns and some suspected examples of tumuli with funerary chapels. The date of 2360 ± 250 BP – 1076 calBC to calAD 133 (Gif-2912) is published in Delibiras et al. 1982. The new radiocarbon dates, published in Mattingly et al. 2017, are: 2649 ± 8 BP: 894–791 calBC (OxA-33338), for a mound cairn from LAR001; and 1589 ± 30 BP: calAD 406–542 (OxA33720), for an excavated corbel cairn from LAR005.

https://doi.org/10.1017/9781108634311.010 Published online by Cambridge University Press

325

326

Youssef Bokbot

Figure 10.11. Adrar Zerzem, overview of the necropolis (Image © 2016 CNES/Astrium, Google Earth).

rock engravings showing a great variety of themes, funerary tumuli associated with other circular dry-stone structures and an open-air settlement with abundant archaeological material on the surface.29 The engravings are located on the rocky outcrops overlooking the surrounding alluvial plain (Fig. 10.11). More than 300 engravings have been reported, mainly representing cattle and wild fauna, but also geometric patterns, schematic chariots and ‘libyco-berber’ riders.30 They cover a broad chronological range, from the Neolithic to protohistory. The necropolis comprises c.30 funerary monuments, of which only three have been excavated.31 These had almost circular outer walls with diameters varying between 7.40 and 7.80 m. The drystone walls were made of squared limestone blocks closely placed together. The walls survive to c.0.5–1 m in height; they served to contain a disordered accumulation of stones of various sizes. Originally however, they may have measured 1.70 m or more in height and therefore constituted a sort of cylindrical drum. Inside these structures, funerary chambers were dug as shafts into the ground. These were oval in shape with vertical sides and a flat base, oriented south-east to north-west and were marked out at ground level by a row of large stones. Inside the pits, the bodies were laid on a layer of small schist rocks arranged like a bed.

29

Bokbot et al. 2008.

30

Bokbot et al. 2008.

https://doi.org/10.1017/9781108634311.010 Published online by Cambridge University Press

31

Bokbot et al. 2008.

10 Funerary Archaeology in the Moroccan Pre-Sahara

Figure 10.12. Adrar Zerzem, tumulus ‘à antennes’: a) superstructure; b) inhumation.

Figure 10.13. Adrar Zerzem, tumulus ‘à niche’: a) superstructure; b) inhumation.

The excavated examples all contained female inhumations (Fig. 10.12).32 The grave goods of Tumulus 1 consisted of two iron bracelets, one for each forearm, ostrich eggshell beads and a copper cylinder very close to the neck, as well as a copper ring and three flint flakes.33 Traces of leather were also noted under the skeleton. It seems that the body, most probably wrapped in a leather shroud, was placed in a flexed position lying on its right side. On the south-east side of the monument were two antennas, 32

Bokbot et al. 2008.

33

Bokbot et al. 2008.

https://doi.org/10.1017/9781108634311.010 Published online by Cambridge University Press

327

328

Youssef Bokbot

Figure 10.14. Grave goods from tombs at Adrar Zerzem: a) pendant in bronze; b) bead in green stone; c) tubular bead in bronze; d) earring in bronze; e) ostrich eggshell beads; f) pierced marine shell (detail, not in scale).

each with a small circular structure at the outermost end. One was filled with sediment containing many pieces of charcoal and burnt animal bones, which suggests a use for those features as sacrifice altars or offering tables. Tumulus 2 had a funerary chapel consisting of a niche built directly in the facing of the outside wall on the eastern side (Fig. 10.13); two parallel lines of small flat stones, which can be considered as antennas, flanked the niche.34 The grave goods from this monument consisted of two copper earrings, beads made of a green stone, ostrich eggshell beads and elements of necklaces made of bronze and marine shells (Fig. 10.14). The samples taken from the funerary deposits of the two tumuli gave largely contemporaneous radiocarbon dates extending, in calendar years, between the second and the fourth centuries AD.35

New Types of Tumuli Tumuli with Cruciform Chapel The cemeteries from the region of Tafilalat/Guir featured a type of funerary monuments that symbolised the hybridisation of two cultures, one from 34 35

Bokbot et al. 2008. El Graoui et al. 2010; Bokbot et al. 2011; Tumulus 1 (Beta-259298): calAD 120–260 and 280–330 calAD; Tumulus 2 (Beta-259300): calAD 140–390. The latter showed signs of reuse, dated between the end of the eighth and the tenth century AD: (Beta-259299): calAD 780–980.

https://doi.org/10.1017/9781108634311.010 Published online by Cambridge University Press

10 Funerary Archaeology in the Moroccan Pre-Sahara

Figure 10.15. Tumuli with funerary chapels in form of a cross from North Africa: a–b) and d–f) from Taouz; c) and g) from Bouïa; h) from Besseraiani; i) from Fedj el-Koucha; j) from Jorf Torba; k) from Braber.

the south and the other from the north: the tumulus with funerary chapel in the form of a cross (Fig. 10.15). The chapel is possibly linked to the Saharan ritual of sleeping with the dead, while the cross shape was possibly brought in from northern Morocco and may derive from a Christian influence. This would suggest

https://doi.org/10.1017/9781108634311.010 Published online by Cambridge University Press

329

330

Youssef Bokbot

Figure 10.16. Painted panel from Jorf Torba (after Camps 1986, re-drawn by M. Hawkes).

a mid-first millennium AD date for this kind of tomb. The chapels of the tumuli of the Tafilalat and the Guir present almost exclusively this cross-shaped plan. The chapel of Tumulus 3 at Bouïa had three niches (in the shape of a Greek cross), whilst those of Tumulus 5 at Taouz and of Braber South have five niches in the shape of a Lorraine cross. The excavation of the chapel in Tumulus 3 revealed charcoal, maybe evidence of a ritual or funeral banquette. Painted or engraved stelae were found in the cruciform chapel of the monuments of Jorf Torba in the Wadi Guir. One of these shows six people, including two women, each one holding an object in the shape of a Latin cross (Fig. 10.16).36 Some of the other stelae have geometric patterns similar to those framing the paleochristian stelae of Volubilis and Altava, in western Algeria.37 What was the exact function of these chapels and why were ornate tablets and stelae deposited, and sacrifices conducted, there? It should first be noted that the development of these rectangular chambers or cruciform chapels was not driven simply by the desire to dedicate an area for sacrifice and perhaps communal meals. These chapels and chambers flanking the funerary structure or penetrating into its mass in fact had an additional function that justifies the way they developed. A text by Herodotus relating to the Nasamones seems to provide an explanation (and incidentally implies that it was a widespread and long-lasting tradition): ‘When they exercise divination they approach the monuments of their ancestors, and there, having said their prayers, compose themselves to 36

Camps 1984.

37

Bokbot 2003.

https://doi.org/10.1017/9781108634311.010 Published online by Cambridge University Press

10 Funerary Archaeology in the Moroccan Pre-Sahara

sleep. They regulate their subsequent conduct by such visions as they may then have.’38 Several converging elements enable us to suggest that the people who built the tumuli with chapels of the Tafilalat – Guir were Christians. These monuments could be the work of newly Christianised Amazighs, who adopted the symbol of their new religion while continuing to bury their dead using ancestral traditions and practising animist rituals, such as sleeping with the dead. They probably were nomadic or semi-nomadic groups established along the pathways of transhumance in the Atlantic Sahara and southern slopes of the Atlas. The tumuli with funerary chapel, in fact, are strictly localised in the pre-Saharan zone bordering the south of the Saharan Atlas and in the more hospitable regions of the desert.39

Tumuli with Windows (‘à Lucarne’) The very limited spread of these monuments, which are confined only in Jabal Selmane, in the south of Zagora between Tagounit and Mhamid,40 means they cannot be compared to any other North African and Saharan protohistoric burial types. As previously stated, available radiometric determinations from the cemetery give a broad chronological frame for these kind of tumuli, which is early first millennium BC to mid-first millennium AD (see above for details). Several hypotheses have been proposed to explain the location of the ’window’. It could be possible to envisage that it was built to carry out successive burials, but its dimensions are too small for an adult body to fit inside. The location of the ’window’, to the south-east, on a slope exposed to the sun may suggest another hypothesis: the sun rays passing through the aperture and reaching the inside of the burial could allow a rapid desiccation of the bodies, thus avoiding the problems linked to their decomposition. However, we think that the window has to be related with a specific cultural practice. Several clues found in Foum Larjam led Gabriel Camps to think there might be a parallel between these windows and the thirst-quenching devices of Jewish tradition found in the tombs of Ras Shamra (Ugarit) in northern Phoenicia.41 The presence of Jewish communities in Morocco is well known and it is probable that the Phoenician expansion on the Moroccan coast brought the first Jews into this region.42 This hypothesis deserves to be retained, even more so since Jewish traditions tend to suggest that the tumuli of 38 41

Herodotus, Histories, 4.172. 39 Camps 1965. 40 Bokbot, 1991, 450. Camps 1961, 517. 42 A. El Khiyari pers. comm.

https://doi.org/10.1017/9781108634311.010 Published online by Cambridge University Press

331

332

Youssef Bokbot

Foum Larjam correspond to the site of Tidri, the first centre of Jewish establishment in the Draa.43 The tumuli with windows of Foum Larjam were often accompanied by rectangular structures similar to offering tables. They may have been used to carry out sacrifice ceremonies as they could be linked to a ritual of purification of the soul and driving away of malevolent spirits. The excavation of these structures has revealed the presence of small hearths, ashes, pieces of charcoal and small burnt bones, probably of hares or birds.44 These elements may correspond to sacrifices or funerary feasts. Another possible explanation ritually linking the ‘window’ and the ancillary structures can be envisaged: on the rectangular structure, fires may have been lit to create smoke that would infiltrate inside the tomb. The aperture thus enabled both libations to be poured and smoke of the sacrifice to penetrate inside the funerary chamber to purify the soul of the deceased. This libation rite is foreign to the African and Paleo-Amazigh world. Such a ritual device could have originated in the libation holes of the tombs of the Semitic east. André Simoneau discovered two copper or bronze earrings with ovoid pendants in the clearing of a tumulus in Foum Larjam.45 These earrings can be related to those of Tayadirt in the upper Moulouya valley, and are Punic in inspiration. The tumulus ‘à lucarne’ of Foum Larjam provides evidence that allow us to consider the question of the adoption of some Semitic rituals by the Amazighs, alongside the first contacts with Phoenician sailors. The funerary rites show an eastern Semitic influence in particular in the libation ritual and the practice of animal offerings. This ritual is reported also in Garamantian tombs,46 and it is possible that there too it was introduced by Semitic populations.

Discussion and Concluding Remarks While all of the above-mentioned funerary monuments have individual characteristics, they also have common features that point towards a cultural or ethnic zone largely open to Saharan influences. The excavations of these tumuli, which the few available absolute dates suggest belong somewhere in a chronological range from the fifth century BC to the sixth century AD, have revealed new archaeological and anthropological elements able to enrich our knowledge of the proto43 46

Gattefossé 1939. 44 Bokbot 1991, 450. 45 Simoneau 1972, 16. See Mattingly et al., Chapter 2, this volume.

https://doi.org/10.1017/9781108634311.010 Published online by Cambridge University Press

10 Funerary Archaeology in the Moroccan Pre-Sahara

history of pre-Saharan Morocco and of the Maghrib in general. They also prove the perpetuation of prehistoric funerary traditions until late periods, an observation that needs to be considered in the debate on the chronology and true reach of the Islamisation of these territories, which was finalised only in the fourteenth century.47 It must be noted that very few of the funerary monuments of North Africa have undergone absolute dating. The absence of such dating, characteristic of the European archaeology literature has for a long time discouraged researchers in pre-Saharan funerary archaeology. The few radiocarbon analyses carried out on other Moroccan tumuli are: the Bazina of El Gour48 on the plateau of the Saïs-Meknès, dated seventh century AD;49 the necropolis of Malou Sidi Lahcen50 in the eastern Rif, dated to third–first centuries BC;51 the necropolis of Daya Chiker in the corridor of Taza, dated to the fifth–second centuries BC;52 and the necropolis of Adrar Zerzem the Wadi Noun.53 This limits the scale of any analysis of the available data and confers a preliminary status to any interpretation attempt. One must hope that systematic research on these types of monuments will be carried out in the future in order to fill in the gap in the current data. The tumulus necropolises of the Moroccan pre-Sahara are frequently located close to mineral deposits of lead, copper or silver, some of which were exploited in ancient times. This geographic proximity suggests that the economy of the human societies of the time were based on the exploitation of these metal deposits. The archaeological finds reveal an influence of Punic taste. This very deep penetration of eastern influences in the Moroccan hinterland most likely was related to isolated ethnic-religious groups, punctuating a trade route for copper, gold and other goods valued by the Carthaginians and later the Romans. Among the grave goods of the necropolises of Bouïa, Erfoud and Foum Larjam were carnelian beads, bronze earrings and a small bronze bell.54 All of these ornaments are of Punic inspiration. The penetration of this Punic influence in the far south of Morocco is to be linked to the very probable existence of a commercial route linking the Tafilalat and the Draa to the Mediterranean and the Atlantic. The spatial distribution of the tumuli within the necropolises was likely determined by criteria of social hierarchy. The discrepancies in the diversity and wealth of the grave goods between the different tumuli could be 47 50 52

Bokbot et al. 2011, 320. 48 Camps 1960. 49 Camps 1974. Mikdad and Eiwanger 2000. 51 Ben-Ncer 2003, 25–34. Abdelaziz Bouroumi, pers. comm. 53 Bokbot et al. 2011, 320.

https://doi.org/10.1017/9781108634311.010 Published online by Cambridge University Press

54

Bokbot 1991, 401–02.

333

334

Youssef Bokbot

explained by social classifications. The larger tombs presenting valuable grave goods might have belonged to tribal chiefs or princes. The question of the emergence of Moorish kingdoms needs to engage with these large tombs that we tend to think of as princely. Their importance needs to be taken into consideration as they imply a strong enough power to coerce dependants into doing considerable work, but also because they indicate a naturalisation of tribal chiefs in the process of transforming into kings. Consequently, the evidence seems to indicate that powerful confederations, if not kingdoms, were established in Morocco as early as the fourth century BC. It is not surprising that the origins of the emergence in Morocco of the monarchical structures remain surrounded by much mystery. Their power, which ended up spreading over huge territories, in fact derives from the power of tribal chiefs who managed to spread their influence wider and wider. The localisation of these funerary monuments along geographic features corresponding to natural passageways could suggest their relation to transhumant people. However, the majority of the large groups of tumuli are found near wadis and cultivable and irrigable plains, where oases have existed in historical periods. Everything seems to indicate that these tombs were the work of sedentary people practising agriculture. In fact, all of the tombs correspond to primary inhumations suggesting that these people were sedentary and living in the immediate vicinity. In none of these monuments did excavators identify secondary burials characteristic of communities of nomadic pastoralists. However, it is surprising to note the absence of ceramics in these pre-Saharan tombs. I would argue this absence is linked to the traditions and ritual beliefs of the populations who colonised these areas.55 In any case, the grave goods deposited in the preIslamic tombs of the pre-Saharan zone are limited to personal ornaments and weapons. In the larger necropolises (Bouïa, Taouz, Braber, Tazarine, Foum Larjam, Adrar Zerzem) there were some relatively elaborate tombs. Some of the tumuli have chapels, ‘windows’ and offering tables, evidence that sacrifices or ritual ceremonies were carried out more or less regularly. These tumuli seem to have contained collective burials; there are only rare examples where only a single individual was found in a tomb. The burials seem to have been successive and probably familial. These necropolises were accompanied by remains of settlements and were 55

Tombs in Fazzan, Libya were similarly aceramic until around the second century BC, see Mattingly et al., Chapter 2, this volume.

https://doi.org/10.1017/9781108634311.010 Published online by Cambridge University Press

10 Funerary Archaeology in the Moroccan Pre-Sahara

Figure 10.17. Jabal Bouïa, pre-Islamic necropolis and fortified settlement (after Margat and Camus 1959, 349, re-drawn by M. Hawkes).

located near water points that most likely permitted a regular and continued agriculture. The survey work that I have carried out in the Tafilalat region has enabled me to discover traces of settlements, enclosures and walls in the immediate surroundings of the tombs.56 These settlements have never been the focus of scientific excavations and it is therefore difficult to 56

Bokbot 1991, unpublished thesis.

https://doi.org/10.1017/9781108634311.010 Published online by Cambridge University Press

335

336

Youssef Bokbot

place them in a precise chronological framework. Some sites, such as Jabal Bouïa, might have been occupied in pre-Roman times. Others could be contemporary with the Roman colonisation or are pre-Islamic or even of later date. The constant presence of this type of settlement near the protohistoric funerary monuments suggests their contemporaneity. In the example of the Jabal Bouïa (Fig. 10.17), the settlement remains are located on the highest and better protected part of the massif.57 The settlement overlooks the entire pre-Islamic necropolis and the surrounding plain. It constitutes a fortified site of the ‘éperon barré’ or promontory fort type. The remains of buildings show a great variability of forms: rectangular, semi-circular, polygonal or ovoid. The lower parts of the structures were made of a double facing of orthostatically placed slabs to a height of over 1 m. The gap between the two rows of facing was filled by a mixture of dirt and gravel. The construction technique is commonly called ‘Berber masonry’.58 The fortified settlement of Jabal Bouïa presents many similarities with the deserted villages of the mountains dominating the city of Oujda reported by J. Marion, with those in the bend of the Wadi Draa, not far from the necropolis of Foum Larjam mentioned by J. Gattefossé as well as with those published by A. Simoneau along the Wadi Draa near the LibycoBerber rock-art site of Foum Chenna in Tinzouline.59 The architecture of the houses and enclosures of this type of ‘Berber’ settlement lacks clear planning. There are no straight lines, right angles or parallel walls. Surface archaeological surveying has rarely provided reliable chronological evidence. It is possible that not all of these sites were occupied in the same periods. But in the absence of excavations, it is impossible to detect changes in settlement and house morphology. However, there are some chronological markers that can be considered. The majority of these settlements were located well beyond the Roman limes. The survey reports highlight the absence of Roman ceramics, which it is important to highlight is in contrast to what is known from Fazzan.60 Additionally, Islamic remains are also very rare. None of these sites has previously been surveyed or excavated; however, according to the material collected on the surface, a chronological range from the fifth to the sixth century BC can be suggested.61

57 60 61

Margat and Camus 1959, 345–70. 58 Lawless 1972, 114–21. 59 Simoneau 1972, 29. See Mattingly et al., Chapter 2, and Gatto et al., Chapter 3, this volume. The recent work of the Middle Draa Project has added significantly to the numbers of such sites and has confirmed a pre-Islamic date for them, Mattingly et al. 2017.

https://doi.org/10.1017/9781108634311.010 Published online by Cambridge University Press

10 Funerary Archaeology in the Moroccan Pre-Sahara

Ruins of fortified settlements are often found in the immediate neighbourhood of pre-Islamic necropolises. In some cases, tumulus tombs can even be found within the inhabited space, as is the case at Jabal Bouïa. Stéphane Gsell gave an explanation for this coexistence: ‘il est vraisemblable qu’on a voulu établir ces demeures des morts auprès de l’asile des vivants, et nous avons ainsi une indication, assez vague, du temps où ceux-ci, (les ruines), faisaient usage.’62 The presence of tumuli within the inhabited area itself could be explained by the prominent status of the people buried in these monuments. It is highly likely that these people would have wanted to place their settlements under the symbolic protection of figures of high social rank or of venerated common ancestors. It is important to note that both metal and flint weapons were found in some funerary monuments of pre-Saharan Morocco, like in those of eastern Morocco.63 These weapons are to be related to the settlements of defensive type found near the necropolises. These weapons and this type of settlement could reflect two opposing lifestyles. In the pre-Saharan regions, nomads and sedentary people must have coexisted. The balance between these two ways of living may only have tipped in favour of nomadism with the great migrations of Arabic nomadic tribes in the thirteenth and fourteenth centuries. While the regions located to the north of the Atlas mountain range were very open to Mediterranean and Atlantic influences, eastern and southern Morocco on the other hand were turned towards the Sahara. The range of the Atlas seems to have constituted a barrier between the two cultural areas. However, there were various contacts following natural pathways. By studying closely the geography of Morocco, one can note that the Wadi Draa constitutes a great pathway from east to west. The continuity of the mountain ranges, from the Atlantic in the west to the Hamada du Guir in the east, makes communication in a north-south direction very difficult. The only pathway between the south and the north of Morocco is located at the eastern end of the High Atlas. Following the Wadi Daoura and its tributary, the Ziz, and the Wadi Saoura and its tributary, the Guir, one can reach the Moulouya valley. This route leads to the Mediterranean and the Atlantic. The study of the protohistoric funerary monuments of Morocco shows that important groups of monuments are located in the south of the Atlas 62

63

Gsell 1927, Tome 5, 237: ‘It is likely that they would want to establish these homes of the dead near the refuge of the living and we therefore have an indication, if quite vague, of the time when these [ruins], were in use.’ Bokbot 1991, 504.

https://doi.org/10.1017/9781108634311.010 Published online by Cambridge University Press

337

338

Youssef Bokbot

and in the east in the steppe or pre-Saharan zones. These necropolises follow major circulation routes; they are found along the Wadi Draa, the Wadi Ziz, the Wadi Guir and the Wadi Noun. This indicates that the valleys played an important role in facilitating contacts and cultural influences. During the protohistoric and pre-Islamic times, and still today, the valleys seem to have played the role of major circulation routes. Through the Wadi Ziz and the Wadi Guir and their respective Saharan expansions, the Wadi Daoura, the Wadi Saoura and the valley of the Wadi Moulouya, it is possible to reach the Mediterranean in the north and the Sahara in the south. It is through these natural topographic pathways that the Tafilalat and the Draa were in contact with both the civilisations of the Mediterranean world and of the Saharan world. The discovery of ornaments of Punic influence in the tombs of the Tafilalat at Erfoud, Bouïa and Taouz bears witness to this opening to Mediterranean trade. Hence, the Tafilalat and the Draa used to be a great crossroads where Mediterranean and Saharan influences met, sometimes leading to their fusion. The Tafilalat, the Draa and the Wadi Noun seem to have played the role of intermediaries between northern Morocco and the Sahara during protohistory. The medieval caravan towns of Sijilmassa and Noul Lamta inherited this strategic role as they found themselves on a very important communication node between the Maghrib and Sub-Saharan Africa, but also used as a hub and a relay point for caravans going to Egypt and to Sudan and as a considerable outpost through which all the goods traded towards the Mediterranean and Europe would pass.

References Ben-Ncer, A. 2003. Etude du tumulus n°1 de R’jem Souk (Province de Taza, Maroc). In M. Khanoussi (ed.), Afrique du Nord antique et médiévale. Protohistoire, cités de l’Afrique du nord, fouilles et prospections récentes, Actes du VIIIe colloque international sur l’histoire et l’archéologie de l’Afrique du Nord (Tabarka, 8–13 mai 2000), Tunis: Èdition de l’Institut National du Patrimoine, 25–34. Bokbot, Y. 1991. Habitats et monuments funéraires du Maroc protohistoriques. Unpublished PhD thesis, Université d’Aix-en-Provence, France. Bokbot, Y. 2001. Protohistoire du Maroc présaharien: bilan et perspectives. In Actes du Colloque International 1ères Journées Nationales d’Archéologie et de Patrimoine, Rabat: Èdition SMAP, 90–98.

https://doi.org/10.1017/9781108634311.010 Published online by Cambridge University Press

10 Funerary Archaeology in the Moroccan Pre-Sahara

Bokbot, Y. 2003. Tumulus protohistoriques du pré Sahara marocain: indices de minorités religieuses? In Actes du VIIIème Colloque International sur l’Histoire et l’Archéologie de l’Afrique du Nord, Tabarka, Tunisie, Tunis: Èdition de l’Institut National du Patrimoine, 35–45. Bokbot, Y. 2005. Hydrogéologie et protohistoire du Tafilalet, quelques aspects de la recherche archéologique en milieu présaharien. In L’eau: source de vie à travers les âges. Jardin des Hespérides, Bulletin semestriel de la Société Marocaine d’Archéologie et de Patrimoine 1, Novembre 2004 – Avril 2005, Rabat: Société Marocaine d’Archéologie et de Patrimoine, 42–44. Bokbot, Y. 2015. Protohistoire du Maroc. Encyclopédie Berbère 39: 6543–61. Bokbot, Y., Onrubia-Pintado, J., Rodriguez-Rodriguez, A., Rodriguez-Santana, C. G., Velasco-Vasquez, J. and Amarir, A. 2008. Le complexe funéraire et cultuel d’Adrar Zerzem (Anti-Atlas, Maroc): résultats préliminaires. In Etudes d’Antiquités Africaines. Lieux de cultes: aires votives, temples, églises, mosquées, Actes du IXème Colloque international sur l’histoire et l’archéologie de l’Afrique du Nord antique et médiévale, Tripoli, 2005, Paris: Èditions du CNRS, 23–33. Bokbot, Y., Onrubia-Pintado, J. and Salih, A. 2011. Néolithique et protohistoire dans le bassin de l’Oued Noun (Maroc présaharien), quelques données préliminaires. In Actes du 1er colloque de Préhistoire Maghrébine, Tamanrasset, 2007, Alger: CNRPAH, 305–22. Camps, G. 1960. Un mausolée marocain: la grande bazina de Souk el Gour. Bulletin d’Archéologie Marocaine 4: 47–92. Camps, G. 1961. Aux origines de la Berbérie: monuments et rites funéraires protohistoriques. Paris: Art et métiers graphiques. Camps, G. 1965. Les monuments funéraires à niche et à chapelle dans la protohistoire nord-africaine. In Congrès préhistorique de France, XVIème session, Monaco 1959, Paris: Société préhistorique française, 321–28. Camps, G. 1974. Le Gour, mausolée berbère du VIIème siècle. Antiquités Africaine 8: 191–208. Camps, G. 1984. Rex maurorum et romanorum, recherches sur les royaumes de la Maurétanie des VIème et VIIème siècles. Antiquités africaines 20: 183–218. Camps, G. 1986. Funerary monuments with attached chapels from the northern Sahara. African Archaeological Review 4: 151–64. Delibrias, G., Guillier, M. T. and Labeyrie, J. 1982. GIF Natural Radiocarbon Measurements IX. Radiocarbon 24.3: 291–343. Gattefossé, J. 1939. Juifs et chrétiens du Drâa avant l’Islam. Bulletin de la société de préhistoire du Maroc 3–4:39–66. Gauckler, P. 1915. Nécropoles puniques de Carthage. Paris: Èdition A. Picard. Gsell, S. 1927. Histoire ancienne de l’Afrique du Nord. Tomes 5, 6. Paris: Hachette. El Graoui, M., Bokbot, Y., Jungner, H. and Searight-Martinet, S. 2010. Datation radiocarbone sur des ossements mis au jour dans un tumulus à l’Adrar Zerzem, Oued Eç-çayad, région de Taghjijt (Sud marocain). Sahara 21: 77–80.

https://doi.org/10.1017/9781108634311.010 Published online by Cambridge University Press

339

340

Youssef Bokbot

Jacques-Meunié, D. 1958. La nécropole de Foum le-Rjam, tumuli du Maroc présaharien. Hespéris 45: 95–142. Jodin, A. 1966. Mogador: Comptoir phénicien du Maroc atlantique, Vol II. Rabat, Tanger: Èdition Marocaines et Internationales. Lawless, R. I. 1972. The lost Berber villages of eastern Morocco and western Algeria. Man 7: 114–21. Margat, J. and Camus, A. 1959. La nécropole de Bouïa au Tafilalet. Bulletin d’Archéologie Marocaine 3: 345–70. Mattingly, D. J., Bokbot, Y., Sterry, M., Cuénod, A., Fenwick, C., Gatto, M. C., Ray, N., Rayne, L., Janin, K., Lamb, A., Mugnai, N. and Nikolaus, J. 2017. Longterm history in a Moroccan oasis zone: The Middle Draa Project 2015. Journal of African Archaeology 15.2: 141–72. Meunié, J. and Allain, C. 1956. Quelques gravures et monuments funéraires de l’extrême sud-est marocain. Hespéris 43: 51–86. Mikdad, A. and Eiwanger, J. 2000. Recherches prehistoriques et protohistoriques dans le Rif oriental (Maroc). Rapport préliminaire. Beiträge zur Allgemeinen und Vergleichenden Archäologie 20: 109–67. Ponsich, M. 1970. Recherches archéologiques à Tanger et sa région. Paris: Èdition du CNRS. Ruhlmann, A. 1939. Les tumuli dans les recherches de préhistoire dans l’extrême sud marocain. Bulletin de la société de préhistoire du Maroc 6: 42–51. Simoneau, A. 1972. Nouvelles recherches sur les gravures rupestres du Haut Atlas et du Drâa. Bulletin de la société de préhistoire du Maroc 8: 15–36.

https://doi.org/10.1017/9781108634311.010 Published online by Cambridge University Press

11

Burial Practices in Western Sahara joanne clarke and nick brooks

Introduction Since 2002, the University of East Anglia’s Western Sahara Project has undertaken a series of field seasons in the POLISARIO-controlled areas or ‘Free Zone’ of Western Sahara (Fig. 11.1). This work has involved intensive survey and excavation in a 3 km by 4 km area north of the settlement of Tifariti, known as the TF1 study area (Fig. 11.2), and extensive survey throughout the Northern and Southern Sectors of the Free Zone.1 Fieldwork has focused on the recording of funerary monuments and other stone-built features, rock art, surface scatters of archaeological materials and palaeo-environmental indicators. Dating has been carried out on human remains from two burials in the TF1 study area and on charcoal from test excavations of surface scatters of chipped stone and pottery.2 In addition, a number of indicators of past humid conditions from throughout the Free Zone have been dated and are awaiting publication.3 Radiocarbon dating and typological comparisons of material culture and funerary monuments with known dated material from across the Sahara indicates human occupation of the region during the Late Acheulian, and again in the pastoral period, from the mid-fourth millennium BC to the first millennium AD. Furthermore, chipped stone scatters with high numbers of Ounan points suggest an early Neolithic utilisation of the region.4 The non-self-governing territory of Western Sahara borders Morocco in the north, Algeria in the north-east and Mauritania in the east and south. To the west, Western Sahara is bounded by the Atlantic Ocean. Western Sahara is subject to a territorial dispute between Morocco and the indigenous Sahrawi independence movement, known by its Spanish acronym of POLISARIO. The territory is effectively partitioned into a Moroccan1 3

Brooks et al. 2018b. McLaren et al. 2018.

2 4

See Clarke and Brooks 2018, for full publication of radiocarbon dates. Pirie 2018.

https://doi.org/10.1017/9781108634311.011 Published online by Cambridge University Press

341

342

Joanne Clarke and Nick Brooks

Figure 11.1. Map of Western Sahara showing places mentioned in the text.

controlled zone in the west and north, and a POLISARIO-controlled zone in the east and south. Holocene human occupation and funerary activity is widespread throughout the Free Zone, but is particularly concentrated along watercourses. The Northern Sector of the Free Zone is associated with a particularly dense drainage network, with the majority of major channels ultimately feeding into the Saguia al-Hamra, an east-west drainage system that flows into the Atlantic near the modern city of el Aaiun (Fig. 11.1), and which is still active after periods of significant rainfall. The Saguia alHamra is the only major east-west drainage system directly linking the Atlantic with the Saharan interior between the Wadi Draa in southern Morocco, and the Senegal River. The Saguia al-Hamra originates northeast of Tifariti and is fed by numerous drainage channels that flow from south to north through landscapes peppered with funerary and related monuments.

https://doi.org/10.1017/9781108634311.011 Published online by Cambridge University Press

11 Burial Practices in Western Sahara

Figure 11.2. Map of the TF1 Study Area showing monument distribution.

Watercourses have been important routes of trade and movement throughout Saharan history. During the Mid Holocene, watercourses such as the Wadi Draa, the Saguia al-Hamra and their tributaries, would have become more important for human populations as the climate became drier after the acceleration of Saharan desiccation in the fourth

https://doi.org/10.1017/9781108634311.011 Published online by Cambridge University Press

343

344

Joanne Clarke and Nick Brooks

millennium BC.5 Such features would have constituted refugia where water existed at or close to the surface, and plant and animal resources persisted after surrounding areas became desiccated.6 In this chapter, we argue that archaeological evidence from funerary architecture supports the supposition that Western Sahara was part of an extended interaction sphere that operated via the major south-north and east-west drainage systems that characterise the Saharan interior more generally. There are clear cultural links between Western Sahara and the southern and central Sahara and these are most clearly observed in the similarities that exist in the funerary architecture, discussed below. We also argue that a small but significant suite of cultural behaviours, recorded in both the Northern and Southern sectors, suggest links via the Atlantic with north-western Morocco and perhaps Iberia. These links may have been facilitated by an Atlantic maritime route that utilised the inland water courses, such as the Wadi Draa and the Saguia al-Hamra.

The Funerary Monuments of Western Sahara and Their Wider Saharan Context The Western Sahara Project has documented an extraordinary wealth of stone-built monuments, which throughout the Sahara are firmly associated with pastoral populations.7 These are interpreted as dating to the period c.3000 BC to AD 800 on the basis of similarities with dated funerary monuments from Libya and Niger.8 Radiocarbon dates on long bones from the individuals interred in two simple conical tumuli in the TF1 study area have returned dates between c.AD 480–820, but on stylistic grounds they are likely to fall in the first half of the first millennium AD.9 The interments within the tumuli and the general morphology of the funerary architecture are in keeping with wider Saharan funerary practices of the period. Although only two monuments have been radiocarbon dated and these are late in the overall relative sequence of funerary monuments,10 the density and variety of monuments suggest that the region around Tifariti was utilised for a considerable period of time.

5 7 8 9 10

6 Brooks 2010; Cremaschi 2002; di Lernia 2002; deMenocal et al. 2000. Brooks 2006; 2010. Chaix 1989; Clarke and Brooks 2018; di Lernia 2006; Ferhat et al. 1996; Paris 1997; 2000. di Lernia and Manzi 2002; Paris 1996; Paris and Saliège 2010. Clarke et al. 2018, 161, Table 6.1; Clarke and Brooks 2018. Y. Gauthier 2015; Paris and Saliège 2010.

https://doi.org/10.1017/9781108634311.011 Published online by Cambridge University Press

11 Burial Practices in Western Sahara

A total of 400 stone-built features (funerary and other ceremonial monuments) were recorded during systematic survey in the TF1 study area. A further 437 funerary and 337 ceremonial features have been recorded through extensive survey in the Northern and Southern Sectors of the Free Zone, and are likely to represent just a small fraction of the monuments present in these regions.11 In addition, Gauthier has supplemented these data through the mapping of monuments seen on high-resolution Google Earth imagery and Milburn has undertaken independent studies of specific types of funerary monuments in Western Sahara.12 The funerary monuments and other examples of ceremonial architecture have been grouped into 32 morphological types (Table 11.1). Many of these are identical or similar to monument types found throughout western and central Sahara, including Morocco, Mauritania, Algeria, Libya and northern Niger.13 Some, such as simple mounds or tumuli, platform tumuli, bazinas, corbeille monuments and V-type monuments, have a wide spatial and chronological distribution across the Sahara and confirm that Western Sahara sits firmly within a wider Saharan pastoral tradition.14 Mound-shaped tumuli are by far the most numerous funerary monuments in the TF1 area, and more widely in the Sahara (see Table 11.1). In their simplest form, they are mounded stones ranging from c.1 m to over 10 m in diameter and up to c.1.75 m in height (Fig. 11.3). Their plans are usually circular, or sometimes oval. Kerb stones are visible around the circumference of some tumuli, while others have false entrances, niches or offertories (‘annexes’). A subgroup are known as platform tumuli because they are situated on top of a circular or sub-circular pavement of stones usually no more than c.0.2 m high and in most cases outlined by a kerb (Fig. 11.4). In some cases, platforms have been formed by extending or accentuating a natural platform through the use of boulders as kerb stones to ‘complete’ the platform or perimeter (Fig. 11.5).15 Bazinas are ‘drum-shaped’ monuments (cairns or tombs) and are found in small numbers in both the Northern and Southern Sectors of Western Sahara, but only two have been recorded in the TF1 study area (see Table 11.1). They can also be equated with chouchet (or choucha in the singular). Their perimeter is usually constructed of dry stone courses, giving them a 11 12 13

14

Brooks et al. 2006; Brooks et al. 2009. Y. Gauthier and C. Gauthier, 2007; 2008; Y. Gauthier 2009; Milburn 1996; 2005. Almagro Basch 1946; Baistrochi 1987; Brooks et al. 2009; di Lernia et al. 2002; Mattingly 2003; Vernet 2007; see also Mattingly et al., Chapter 2, Bokbot, Chapter 10, and Sanmartí et al., Chapter 8, this volume. Y. Gauthier 2009; Sivilli 2002, 23, Fig. 3.2. 15 Brooks et al. 2018a, 45.

https://doi.org/10.1017/9781108634311.011 Published online by Cambridge University Press

345

https://doi.org/10.1017/9781108634311.011 Published online by Cambridge University Press

Crescent with tails

Chapel monument Compound monument Complex monument Crescent antenna Crescent, mounded Crescent, paved Crescent, regular

Bazina with mound Boulder burial Corbeille

Bazina, stepped

Axle-type monument Bazina

TYPE

Elongated monument tapering to point at each end of long axis. Monument with perimeter of layered dry-stone wall constructed with flattish slabs, round or square. Bazina whose construction consists of stepped layers decreasing in size from base. Bazina monument on which tumulus mound has been constructed. Burial inside hollow boulder sealed by rock wall. Circular monument defined by upward facing flat stones angled outwards from centre of monument. Monument with well-defined false-entrance Structure in which multiple monument types are combined Tumulus with line of orthostats at ‘front’ and enclosure at ‘back’. Tumulus from which extend paved arms of uniform width. Tumulus with short extensions shaped like a pastry croissant. Paved area from which extend paved extensions; tumulus near centre. Elongated tumulus from which emanate curved arms, reducing in height and tapering to a point from the central tumulus. Regular crescent whose arms are extended by linear stone arrangements.

Build stone features

DESCRIPTION

Table 11.1. List of monument types and their frequency.

1

1 2 13 14 12 2 7

2 1 8

5

5 14

N. Sector

0

0 0 0 0 1 0 18

0 0 1

0

2 14

S. Sector

FREQUENCY

0

2 1 15 3 3 8 3

10

0

0 2

TF1

https://doi.org/10.1017/9781108634311.011 Published online by Cambridge University Press

Stone emplacement

Stone disc/disc tumulus

Standing stones

Ridge monument

Other stone structure Stone platform

N-lith Orthostats, cairn with

Monument & towers

Goulet (Type 2) Half-circle/ half-ring

Funerary complex Goulet (Type 1)

Collection of stone structures physically linked or contained. Tumulus on pavement from which corridor emerges before curving back to form enclosures; tumulus at edge of monument, at narrow end. Similar to GL1 but with more circular shape and tumulus off-centre. Semi-circular arrangement of stones, often several layers deep and flush with ground, or series of such arrangements. Monument (tumulus or bazina) with small auxilliary towers, generally arranged in a line or arc. Flat stone balanced on a number (N) of upright stones. Mound/tumulus associated with line of orthostats but lacking size of ridge monument or characteristictics of complex monument. Structure outside of categories defined here. Circular or rectangular stone pavement, constructed from a single and more-or-less flush with ground. Monument whose principal apparent feature is a line of vertical flat stones, typically pointed to give tooth-like appearance. Individual upright stones or groups of such stones occurring in isolation or as dominant element of a stone feature (e.g. large standing stone with low crude tumulus). Does not include orthostats otherwise embedded in or associated with tumulus or other monument. Round stone structure built from more than one course of stones, with top surface elevated above ground level, but of low relief. Small pile of stones or cairn, too small to be described as a tumulus. 5

0

4

7

4

2

4

2 12

0 5

0

0 0

9 1

0

1 0

1 0

1

3 0

2 11

1

2

5

0

2 0

0 1

0

11

0 18

https://doi.org/10.1017/9781108634311.011 Published online by Cambridge University Press

(Linear) V-type monument

Tumulus, platform

Tumulus with arms

Tumulus

Stone ring

Stone outline

TYPE

Table 11.1. (Cont.)

Circular or other stone outline, with stones generally placed on ground or (less frequently) embedded in ground, and not necessarily close together. Circular arrangement of stones evenly and usually tightly spaced, often associated with goulets but also occurring in isolation. Monument constructed from piling stones into a cairn, varying from very low relief to over 1.5 m in height, and from around 1 m to over 10 m in diameter. Tumulus with low relief extensions forming arms or ‘wings’ (distinct from crescents). Tumulus constructed on an actual or ‘symbolic’ platform, the latter consisting of circular pavement or augmented bedrock around the tumulus. Small circular mound or platform from which emanate two linear features, often constructed from infilled parallel stone alignments.

Build stone features

DESCRIPTION

1

7

0

11

5

129

189

8

2

18

S. Sector

3

7

N. Sector

FREQUENCY

0

8

0

228

22

5

TF1

11 Burial Practices in Western Sahara

Figure 11.3. Tumulus, WS023 before excavation.

Figure 11.4. Schematic illustration of a platform tumulus.

wall-like appearance (Fig. 11.6). Bazinas are usually circular, but they can also be rectangular and even polygonal. They may be mounded or flat on top and they apparently have a rubble infill. Some bazinas are stepped, and a number of very large examples were identified during the extensive survey (Fig. 11.7). While most bazinas recorded in the study areas are less than 2 m in height, these large stepped bazinas are up to 5 m in height. Monuments of this type are well known in the Wadi al-Ajal, Libya.16 Corbeille monuments are relatively common in the Northern Sector and the TF1 study area but rare in the Southern Sector (see Table 11.1). They 16

Type 5, see Mattingly et al., Chapter 2, this volume; Brooks et al. 2018a, 46; Mattingly et al. 2008.

https://doi.org/10.1017/9781108634311.011 Published online by Cambridge University Press

349

350

Joanne Clarke and Nick Brooks

Figure 11.5. Low tumulus bounded by kerb stones.

Figure 11.6. An example of a bazina.

https://doi.org/10.1017/9781108634311.011 Published online by Cambridge University Press

11 Burial Practices in Western Sahara

Figure 11.7. A large stepped bazina in the Southern Sector of the Free Zone.

Figure 11.8. An example of a corbeille.

are comprised of a ring of upright stones around an interior, which may simply be the existing ground surface or a modified surface usually more or less flush with the ground. In most instances in the TF1 area, the stones forming the rings are small and sit low to the ground (Fig. 11.8). Corbeilles identified to date in the Free Zone differ from those of the Libyan Fazzan,

https://doi.org/10.1017/9781108634311.011 Published online by Cambridge University Press

351

352

Joanne Clarke and Nick Brooks

Figure 11.9. An example of a V-shaped monument.

which usually have a standing stone at their centre, are often associated with circular annexes, and frequently occur as components of more complex structures.17 In contrast, the examples from Western Sahara tend to occur in isolation and generally are not associated with standing stones or annexes, although they often have a pair of stones that are larger, more upright, and/or of a different composition, and mark a particular point around their circumference and define an orientation.18 Another class of stone feature, the V-shaped monument, spreads horizontally over the land surface and can be split into a number of different sub-classes. These include antenna monuments, with or without a central tumulus (Fig. 11.9), crescents, mounded crescents and other features with low relief, bilateral extensions. Crescents are the best represented form of V-shaped monuments in Western Sahara and are common in the Northern Sector but rare in the Southern Sector and the TF1 study area (see Table 11.1). They are characterised by curved extensions or arms and include ‘regular’ crescents and crescents with tails. The latter exhibit more linear extensions or arms, and include paved crescents and crescent antenna monuments. A typical characteristic of V-shaped monuments is 17

Y. Gauthier and C. Gauthier 2004.

18

Brooks et al. 2018a, 47.

https://doi.org/10.1017/9781108634311.011 Published online by Cambridge University Press

11 Burial Practices in Western Sahara

Figure 11.10. An example of a goulet.

that their extensions are in very low relief, comprising elongated pavements, often bounded by kerb stones. In contrast with the examples described above, which have a broad geographical range, other monument types have quite restricted ranges. To date, goulets (Fig. 11.10) have only been recorded in southern Morocco, in northern parts of Western Sahara and in northern Mauritania, with a single example recorded in the Southern Sector of the Free Zone (Fig. 11.11). Goulets located to the east of Bir Moghrein in Mauritania were first described by Gobin and Monod in the 1930s-1940s.19 These monuments can be very large (upwards of 500 m in length) especially when seen as complexes with multiple parts (for example, stone rings and stone concentrations). The most common form of goulet in the Free Zone, based on those recorded to date, consists of two parallel stone alignments in low relief (forming a ‘gulley’ or ‘corridor’), usually extending from the edge of an area densely paved with small stones (Fig. 11.10). After some distance, the stone alignments forming the corridor diverge and curve back towards a paved area. A small, low tumulus is often constructed on the paved area, at the very edge of the monument. However, not all goulets include such a 19

Gobin 1937; Monod 1948.

https://doi.org/10.1017/9781108634311.011 Published online by Cambridge University Press

353

354

Joanne Clarke and Nick Brooks

Figure 11.11. Distribution map of goulets (Image Landsat/Copernicus © Google Earth).

tumulus. Where a tumulus is absent, the corridor sometimes extends all the way through the paved area to the rear perimeter of the monument. Complex monuments (Fig. 11.12) have been recorded only in the northern areas of Western Sahara and in southern Morocco (see Table 11.1).20 Complex monuments are very distinctive. Unless they have been heavily disturbed, they are characterised by an alignment of orthostats (between c.5 m and c.12 m long), either straight or slightly curving, which serves as a ‘facade’ for a tumulus that usually faces north-east, east, or south. The orthostats extend beyond the limit of the tumulus and very often, up to three or so of the stones in the centre of the face will be noticeably higher than the rest. In some instances, a small mound of stones might be present up against the centre of the face and a false entrance can be made out. There is usually a rectangular, paved area of stones marked out and attached to the rear of these monuments and these can sometimes be subdivided by stone alignments. External annexes can also be found both to the front and the rear, even occasionally attached to the end of an

20

Searight 2004, although compare the In-Aghelachem complex, Wadi Tannzuft, Libya – di Lernia et al. 2002, 102–16.

https://doi.org/10.1017/9781108634311.011 Published online by Cambridge University Press

11 Burial Practices in Western Sahara

Figure 11.12. An example of a complex monument.

orthostat face. In many instances, there are oval configurations of small boulders in front of these monuments, often a few metres distant.21 Ridge monuments (Fig. 11.13) have been recorded only in the Southern Sector of the Free Zone, in Mauritania, and in the Anti-Atlas of Morocco.22 Typically, these features consist of a line of orthostats that can be straight, sinuous, or arced, and can run either in front of a tumulus, or over it. The orthostats are highest in the centre of the monument, reaching in some instances well over 1 m in height, and tapering down close to ground level at the extremities. A distribution map of dated funerary monuments across the Sahara, compiled by Sivilli,23 demonstrates that comparative materials for the monuments of Western Sahara are found across great distances throughout the Sahara – from south-central Algeria, south-west Libya, northern Niger, Nubia and even Ethiopia. Sivilli also illustrates that there is a great range of dates for different monument types. For example, she dates tumuli to 5600–1200 BP, platform monuments between 5000 and 1200 BP, bazinas to between 4400 and 1700 BP and ‘V’ shaped antenna monuments at 21 22

23

Brooks et al. 2018a, 50. Stone alignments are known elsewhere in the Sahara, see, for instance, di Lernia et al. 2002, 27– 31; Sterry and Mattingly 2011, 103–16. Sivilli 2002, 23, Fig. 3.2.

https://doi.org/10.1017/9781108634311.011 Published online by Cambridge University Press

355

356

Joanne Clarke and Nick Brooks

Figure 11.13. An example of a ridge monument.

5400, 3500 and 3000 BP. In addition, a large number of radiocarbon dates were acquired for a variety of monument types by Paris,24 but some of the key-hole monuments he excavated in Niger have been re-dated, using accelerator mass spectrometry, and their date range has changed from 5610 to 4515 BP to 4480 ± 30 to 4225 ± 30 BP.25 Further dates, which range from 4447 ± 46 BP to 4301 ± 46 BP, were obtained from the mineral fraction of bones (bioapatite) from monuments recently excavated in the Tassili.26 Although some of the original dates now appear to be imprecise by today’s standards, it is clear that many of the funerary monuments of the Sahara originated with Pastoral populations and therefore pre-date the heyday of the Garamantes. In summary, tumuli, bazinas and certain types of antenna monuments occur throughout Central and Western Sahara and many of these forms are found in association with the Libyco-Berber script, connecting the Free Zone of Western Sahara with regions further east.27 This link is reinforced by the presence of the Libyco-Berber script in the regions examined and as reported by literature from the colonial period.28 In contrast, goulets, complex monuments and ridge monuments are restricted to parts of

24 27

Paris 1996, 270–71, Table 42. 25 Paris and Saliège 2010. Y. Gauthier 2009. 28 Pellicer Catalán et al. 1974.

https://doi.org/10.1017/9781108634311.011 Published online by Cambridge University Press

26

Berkani et al. 2015.

11 Burial Practices in Western Sahara

Figure 11.14. WS001 standing stone monument with quartz scatter.

Morocco, Mauritania and Western Sahara and appear to belong to a western tradition of monument building. Evidence for connections with north-western and western Morocco is also demonstrated in the large and complex standing stone arrangements documented in the Northern Sector of the Free Zone. While standing stones occur elsewhere in the Sahara, the Northern Sector appears distinctive (in a Saharan context) in that it contains both groups of dozens of standing stones arranged so that they form particular geographical alignments (Fig. 11.14), or very large (c.2 m or more) solitary standing stones, often associated with small cairns or stone concentrations (Fig. 11.15). Standing stone arrangements are features of the Moroccan Atlantic region. Perhaps the best known is the Cromlech de Mzora, a stone circle measuring c.56 m in diameter.29 Other stone circles and single standing stones have been recorded throughout Atlantic Morocco.30 In some ways then, Western Sahara represents a mixture of east and west. On the one hand its pre-Islamic funerary architecture sits comfortably within the Saharan Pastoral tradition, with many of its monuments identical, or similar to, those found in Algeria and Libya, on the other hand, Western Sahara is distinctive for its apparent links with the Atlantic zone of 29

Bokbot 2008; Salisbury 2011.

30

Belmonte et al. 1999.

https://doi.org/10.1017/9781108634311.011 Published online by Cambridge University Press

357

358

Joanne Clarke and Nick Brooks

Figure 11.15. Single standing stone.

north-west Morocco. What that actually means in terms of cultural interaction and connectivity is not altogether certain but what seems to be clear is that this interconnectedness was complex and was undertaken over vast geographical distances, most probably via water courses that link the Atlantic coast with the Saharan interior.

The Funerary Monuments in a Possible Atlantic Context There are other aspects of Western Saharan funerary practices that we would argue bear striking adjacencies with an Atlantic (European?) cultural milieu; these are: • A particular emphasis on quartz as a material, for the construction of monuments, for the ‘carpeting’ of standing stone sites and funerary monuments in quartz chips, or simply as naturally occurring outcrops that become foci of cultural interest/behaviour. • The use of contrasting materials in the construction of funerary monuments. • The deliberate and specific ‘referencing’ of both landscape features and other monuments including standing stone sites.

https://doi.org/10.1017/9781108634311.011 Published online by Cambridge University Press

11 Burial Practices in Western Sahara

Use of Quartz in Funerary Contexts A very small percentage of monuments recorded in the Free Zone of Western Sahara are associated with scatters of quartz chips on and around the monuments (Table 11.2). In the TF1 area, eight out of a total of 400 monuments are associated with quartz scatters. This situation appears to be reflected throughout the areas examined by the extensive survey, in which quartz scatters in association with monuments were present but extremely rare. Nonetheless, the extensive survey revealed some striking examples of the use of quartz, including: • An east-oriented crescent monument constructed entirely of quartz and situated within a large quartz outcrop. • A small funerary complex with very dense concentrations of quartz pebbles inside and outside a well-defined perimeter, housing a large hollow boulder that had been filled with quartz pebbles. • Concentrations of quartz chippings inside and outside the enclosures of goulets. • Quartz chippings scattered on and around tumuli and other monuments in varying densities (Fig. 11.14). The monuments associated with confirmed, deliberate concentrations of quartz include six simple tumuli, one platform tumulus, five goulets, two crescent antennae monuments and one crescent, two standing stone sites, one bazina and a low stone circle. Many of these represent classic ‘Saharan’ forms (crescent and antennae monuments, that fall under the broader category of ‘V-type’ monuments) but also forms restricted to the north-west Sahara (goulets), and other forms that occur both within and outside North Africa (tumuli, platform tumuli and standing stone sites). The morphological diversity of these forms is striking and indicates that the scattering of quartz was not systematically associated with selection criteria based on monument type. Furthermore, the infrequency of the practice suggests that it was not common in Saharan funerary traditions. di Lernia and Manzi do not record any deliberate placing or use of quartz in the monuments of Wadi Tanzzuft in central Libya even though quartz was common in the area,31 nor does Baistrochi et al. record the deliberate use of quartz in the Tadrart Akakus.32 In contrast, Vernet records very similar types of funerary monuments to those found in Western Sahara, and the deliberate placement and use of quartz on a small number of these, in Atlantic Mauritania.33 Further east, Mattingly has 31

di Lernia and Manzi 2002, 151.

32

Baistrochi 1987.

33

Vernet 2007.

https://doi.org/10.1017/9781108634311.011 Published online by Cambridge University Press

359

360

Joanne Clarke and Nick Brooks

Table 11.2. Monuments in the Northern and Southern Sectors with quartz scatters. Feature

Context/description

Northern Sector IR1-11 Type-2 goulet, with concentrations of quartz chippings in interior of enclosures in front of tumulus, either side of corridor. IR1-12 Low-density of quartz chippings throughout enclosures of type-1 goulet, with higher density concentrations in certain areas, including in association with pictograms, e.g. inside triangular arrangement of small pebbles. Also inside stone ring to east of goulet and just north of corridor axis. MS1-4 Bright quartz chippings scattered on prominent conical tumulus constructed on slope of wadi bank, and on adjacent surface between tumulus and low stone cairn. TF0-23 Crescent antenna with stone alignment emanating from centre, joining crescent with nearby natural ridge. Quartz scattered between arms of crescent. TF0-38 Low-density scatter of medium quartz pebbles and chippings in association with circle of standing stones, mostly absent from interior of circle, and mostly concentrated on the east side of the circle, around opening in circle. TF1 WS001 Standing stones site, made up of at least 65 orthostats. The ground between the stones is littered with quartz fragments; they are also present around the bases of the individual uprights. TF1 WS043 This rectangular, bazina-like monument could possibly be interpreted as a ‘chapel tumulus’. There are numerous small quartz fragments on the ground nearby, especially to the west. TF1 WS074 Tumulus with a scatter of small quartz fragments immediately to the southeast. TF1 WS075 Tumulus to northeast of WS074. Small quartz fragments scattered over it. TF1 WS164 Low-lying circle of cobble-sized stones nearly flush with the ground. There is a slight concentration of stones within the ring and occasional quartz fragments are present. TF1 WS353 This complex monument comprises a cairn with low-lying wings extending north and south by c. 2.75 m each. There appears to be a ridge through the body of the cairn joining the wings, and to the west of this, on top of the cairn, there is a single orthostat. There is a partly paved rectangle attached to the western side of the monument. There are also concentrations of quartz fragments nearby.

https://doi.org/10.1017/9781108634311.011 Published online by Cambridge University Press

11 Burial Practices in Western Sahara

Table 11.2. (Cont.) Feature

Context/description

TF1 WS354

This goulet is unusually oriented to the west. Outside the west end of the goulet, just north of the central gully, are two low-lying stone rings (no more than 0.2 m high). The closer one, only c. 4.25 m from the goulet, is a simple circle of cobbles. Further to the west by 2 m is a larger stone ring. There is a small spread of quartz fragments to the northwest of the smaller stone ring. Crescent antenna monument. On the ground to the east are spreads of quartz fragments. Quartz chippings in high densities in association with stone concentrations (possible degraded pictograms) and inside stone rings in area beyond corridor and enclosures of type-1 goulet.

TF1 WS387 TF6-44

Southern Sector AZ3-1 Crescent monument constructed entirely from large quartz boulders and pebbles, with small circular quartz tumulus enclosed in arms, constructed on extensive outcrop of quartz slightly elevated above surrounding plains. GF2-6 Dense concentrations of quartz within enclosures of type-1 goulet (only goulet recorded in Southern Sector to date). Presumably from small natural quartz outcrop adjacent to concentration of stone features in this (detailed survey) area. GF3-1 Quartz pebbles scattered on and around tumulus. GF3-2 Quartz pebbles scattered on and around platform tumulus. LD0-3 Very dense concentration of yellowish quartz pebbles within funerary complex defined by stone arc, and adjacent to complex outside arc beyond narrow clear area. Quartz pebbles placed on circular stone outcrop and cairns. Large hollow boulder filled with quartz pebbles.

noted the rare use of quartz in the Wadi Barjuj, south of Jarma in central Libya, which may suggest that the practice was more geographically widespread than simply the Western Sahara.34 Evidence from Niger supports this, in a spatially restricted area of the Air Mountains, Paris recorded six monuments (c.3 per cent of a total of 190 monuments with built superstructures) that were characterised by quartz scatters.35 One was a crescent monument, two were simple tumuli, one was a circular platform disc with an outside wall of three layers of stones and one, a platform tumulus. The relative rarity of quartz scatters in Saharan funerary contexts, and the diverse range of monument types with which this low-frequency use of 34

D. Mattingly pers. comm.

35

Paris 1996, 206–62.

https://doi.org/10.1017/9781108634311.011 Published online by Cambridge University Press

361

362

Joanne Clarke and Nick Brooks

quartz is associated, raises the possibility that this was an imported practice that was incorporated into existing funerary traditions at certain times, in certain places, and by certain groups that were not representative of the wider Saharan or western Saharan cultural complex. One possibility is that the practice was the result of contact with Atlantic Europe.

The Use of Contrasting Colours and Materials in the Construction of Monuments In the region of Zag, in southern Morocco and to the north-east of Tifariti, Milburn recorded a low tumulus comprised of starkly contrasting colours and materials; he records that it was ‘composed of a border of whitish stones forming an enclosure, standing mostly on end, about 40 cm high; they enclose a roughly circular area. . .There is a flatpacked top made with the same whitish stone.’36 Milburn noted nothing similar to this monument anywhere else in the region. In the region of l’Azawagh, Paris recorded a crescent tumulus that used two different types of stone to achieve a difference in colour, texture and appearance; limestone slabs for the base and small quartz blocks to build up the tumulus.37 Sivilli notes that Paris considered these to be related to groups from outside Africa introducing new cultural elements, which were then assimilated into the north-eastern Niger traditions.38 The use of contrasting construction materials is evident at a number of locations. Typically, this takes the form of a tumulus being constructed principally from one material (usually dark fine-grained granite or basalt), with a contrasting material (typically lighter, coarse-grained granite) used to construct a ‘cap’ to the monument. Examples of such monuments have been recorded by the Western Sahara Project at Lajuad.39

Monument Locations and Alignments Deliberate orientation of monuments towards the rising or setting sun, or on a luni-solar alignment is well documented across the Sahara and appears to be a ‘Saharan’ behaviour.40 Gauthier’s work with high-resolution Google Earth imagery has demonstrated different orientation rules for different types of monuments through time and with location. For example, crescent 36 39

Milburn 1974, 105. 37 Paris 1996, 230. 38 Sivilli 2002, 22. Clarke and Brooks 2018, 202. 40 Y. Gauthier 2009.

https://doi.org/10.1017/9781108634311.011 Published online by Cambridge University Press

11 Burial Practices in Western Sahara

Figure 11.16. An example of the alignment of the corridor of a goulet referencing a flattopped ‘mesa’ on the horizon.

monuments in Western Sahara tend towards an easterly orientation; only 4.4 per cent of antenna-type monuments (of which crescent monuments are a sub-type) have a westerly component to their orientation.41 V-type monuments, crescents and key-hole monuments all show, to a greater or lesser extent, orientation rules, some of which are very specific (for example, the key-hole monuments of the Tassili-n-Azjar).42 In some instances, however, monuments are not simply orientated to a particular direction, they are deliberately positioned to reference other monuments, or prominent landscape features, in a way that is reminiscent of the monuments of ‘Atlantic Europe’. Many of the monuments (with and without quartz scatters) reference other monuments and/or natural features, such as prominent hills, lines of sight with the horizon and wadi confluences. This is most evident in the construction of tumuli on prominent rocky outcrops, or antennae monuments where the longer ‘arm’ extends upslope, or goulets where the central corridor references features on the horizon. The most striking example of this last characteristic is a very large goulet near Tifariti (WS006) whose corridor is aligned to west with a prominent flat-topped hill (Fig. 11.16). 41

Brooks et al. 2018a, 73; Y. Gauthier 2009, 319.

42

Gauthier 2009, 321.

https://doi.org/10.1017/9781108634311.011 Published online by Cambridge University Press

363

364

Joanne Clarke and Nick Brooks

The silhouettes of many monuments have been enhanced by the natural topography, but the monuments also merged with and became the topography in a way that echoes the monument building of Europe in the Neolithic and Bronze Age.43 Monuments were most numerous along wadi courses, which would have been used as routeways and for grazing animals, but unusual or significant topography and geology may have been preferentially selected for clusters of monuments. The TF1 study area is a natural basin, and the monuments were positioned on the ridges above the basin affording views both inward towards the centre of the wadi basin, but also outward towards the wider Saharan landscape.44 WS001, a standing stone site situated on a slight rise of compacted (fluvial) sand and gravel on the western flank of Wadi Tifariti, was probably an island of dry land whenever Wadi Tifariti flooded in the past. This site also features a dense scatter of quartz chips around the bases of upright stones and amongst the stones more generally (Fig. 11.14), and is in direct site line of a high percentage of the tumuli and other funerary monuments situated on the ridges above it (Fig. 11.17). This consideration of visibility is often combined with a choice of location at transitional points in the landscape. For example, monuments are concentrated at the point where the Wadi Ternit changes course from east-west to south-north, on both banks where they are visible from lower elevations (an observation also made by Bokbot in this volume for monument clusters in the Wadi Draa). In the Southern Sector, where drainage systems are far less numerous than in the north, monuments are concentrated along granite ridges and basalt dykes, sometimes merging with and modifying these features (for example Aij), and in other cases situated to one side of a ridge or dyke.

Discussion While the functions of many of the stone monuments recorded by the Western Sahara Project are unknown, comparisons with other parts of the Sahara indicate that at least some and probably most of the larger tumuli and their variants are likely to be human or animal burials or a combination of both.45 The functions of goulets, groups of standing stones, stone rings, stone alignments and stone concentrations are more uncertain. Even 43 44 45

Bradley 1998, 122. A pattern noted in other Saharan regions, see Belmonte et al. 1999, S24. Camps 1961; Lihoreau 1993; di Lernia and Manzi 2002; Mattingly et al. 2008; Y. Gauthier 2009.

https://doi.org/10.1017/9781108634311.011 Published online by Cambridge University Press

11 Burial Practices in Western Sahara

Figure 11.17. Viewshed from the standing stone monument WS001 showing site lines to surrounding monuments. Note that a significant number of monuments can be seen from WS001.

excavation of these monuments may fail to yield any significant insights if they are simply surface features. Monuments are found in a wide variety of geographic and topographic contexts, including on slopes, flat sand and gravel plains, the summits of hills, escarpments, around the bases of hills, and along wadis. However, the most common location for these features, particularly in the Southern Sector, is on top of, and immediately adjacent to, granite ridges. These provide an abundant supply of construction material and allow certain stone features (those that might be described as monuments) to be situated in slightly elevated locations where larger structures will be visible from a distance and, perhaps more importantly, from where key landscape features are visible. To date, the two tumuli that have been dated in the Northern Sector fall at the tail end of the Pastoral period, but dates on charcoal from test pits of lithic and pottery scatters fall at the very beginning of the Pastoral period. Thus, it can be reasonably confidently stated that the occupation of the Northern Sector of Western Sahara, in the region of Tifariti, can be

https://doi.org/10.1017/9781108634311.011 Published online by Cambridge University Press

365

366

Joanne Clarke and Nick Brooks

bracketed between c.3500 BC and c.AD 500. Clearly, more radiocarbon dates are required for a better picture of the utilisation of this important region during both the early and later Neolithic. A principal argument of this paper is that western and central Saharan peoples may have had knowledge of, or interaction with, ‘Atlantic Europe’. The extent of this relationship, if it existed, is unknown, but connections between the Maghrib and southern Europe via the Mediterranean, at least from the Early Bronze Age, are well documented. There is now a growing body of evidence to indicate seafaring along the Atlantic coast as well. Early Bronze Age (EBA) Bell Beaker burials have been recorded at Khemisset in Morocco and Bell Beaker pottery is known from a number of locations in western Morocco.46 These occurrences of cultural elements traditionally associated with the European EBA close to the Moroccan Atlantic coast support the premise that interaction was not simply via the Mediterranean Sea, but that an Atlantic coastal route was also used. Schumacher et al. argue for trade in ivory from the African savannah elephant to Portugal and south-western Atlantic Spain during the European Chalcolithic, and with south-eastern Spain in the EBA, and infer an Atlantic route for this trade, based on the above chronological differences and on comparisons between ivory artefacts from Iberia and modern-day Morocco.47 Recently, a complete, unworked African Savannah elephant tusk was found in an elite Chalcolithic burial at the site of Valencia de la Concepción, near Seville in southern Spain.48 Indeed, as long ago as 1955, Jodin argued that an Atlantic trade route between Iberia and north-west Africa existed by the Neolithic.49 How far south these connections reached is uncertain due to a dearth of evidence, but more and more circumstantial evidence is coming to light all the time. The concept of an ‘Atlantic Europe’ has existed in the archaeological literature for many years, and relates to a specific set of cultural and ideological traits that span millennia and that are evident in the regions ‘fronting’ the Atlantic Ocean from Scandinavia to the Iberian Peninsula.50 The nature and extent of what constitutes Atlantic Europe is still debated, but archaeologists agree that its principal architectural expressions can be documented during the Neolithic and Bronze Ages as megalithic stone monuments that referenced both the landscape and astronomical features in culturally specific ways.51 There was also a clear preference for 46 48 50 51

Bokbot 2005; 2007; Bokbot and Ben-Nçer 2008. 47 Schumacher et al. 2009, 993. García Sanjuán et al. 2013. 49 Jodin 1955, 356. MacWhite 1951; De Laet 1976; Bradley 1998; Cunliffe 2001. Bradley 1998, 125; Scarre 2002, 9.

https://doi.org/10.1017/9781108634311.011 Published online by Cambridge University Press

11 Burial Practices in Western Sahara

contrasting colours and textures in the different uses of stone for the construction of monuments, including the special selection of quartz.52 In Western Sahara, seasonal watercourses linking the Saharan interior with the Atlantic coast are very likely to have served as refugia, within the context of both the immediately adjacent areas and the wider western Saharan region as a whole. The current climate of Western Sahara is significantly more humid that that of the interior Sahara, with annual average rainfall in the early–mid-twentieth century estimated around 50 mm in some upland areas. Given the high inter-annual variability of rainfall in Western Sahara, as in all arid and semi-arid regions, rainfall in individual years may be significantly higher than this. The dense vegetation and small, residual active channels in many of the major wadis that feed the Saguia al-Hamra attest to the importance of rainfall for the region even today. Even where channels are filled with wind-blown sand, their courses are often visible from the presence of dense vegetation cover, indicating locally elevated groundwater tables or increased soil moisture. The modern Sahrawi have a concept of drought, and it is common practice among some refugee families in the Tindouf camps, who maintain animal herds in the Free Zone, to migrate temporarily to the Free Zone when good rainfall results in significant pasture. Human activity in the Free Zone today is focused around certain locations where fresh water is available near the surface, such as Tifariti, Bir Lahlou, Bir Lahmar and Zoug. These locations effectively serve as ‘refugia’ today, and may have played a more important role in the past, when water may have been more abundant. The importance of the Wadi Tifariti, at least in the vicinity of the modern settlement of Tifariti, is highlighted by the dating of two tumuli to the first half of the first millennium AD, indicating human utilisation of this area millennia after desiccation made much of the Sahara – and particularly the Saharan lowlands – effectively uninhabitable. Given their role as refugia in a drying Sahara, watercourses such as the Saguia al-Hamra and its tributaries (including the Wadi Tifariti), must be considered as potential corridors of cultural interaction and exchange. Movement along these corridors and contact between different cultural groups provides a potential mechanism for the transmission of ideas, practices and ideologies between the different regions of the Sahara and farther afield. In terms of Atlantic connections, the better watered regions of the Sahara desert would have been attractive destinations for people landing on the arid coast, particularly those regions linked by water courses.53 It is along these routeways that we might expect to find evidence 52

Scarre 2002, 11.

53

Gasse 2000; 2002; Lancaster et al. 2002.

https://doi.org/10.1017/9781108634311.011 Published online by Cambridge University Press

367

368

Joanne Clarke and Nick Brooks

of the percolation of material and ideological aspects of Atlantic European culture. It is therefore unlikely to be coincidence that evidence for connections between the Saharan interior and Atlantic Europe should be found along and in the vicinity of these drainage systems.

References Almagro Basch, M. 1946. Prehistoria del Norte de Africa y del Sahara Español. Barcelona: Consejo Superior de Investigaciones Cientìficas, Instituto de Estudios Africanos. Baistrochi, M. 1987. Pre-Islamic megalithic monuments of the Northern Tadrart Acacus. In B. E. Barich (ed.), Archaeology and Environment in the Libyan Sahara: the Excavations in the Tadrart Acacus 1978–1983, Oxford: Archaeopress, 87–89. Belmonte, J. J., Esteban, C., Cuesta, L., Perera Betancort, M. A. and Jiménez González, J. J. 1999. Pre-Islamic burial monuments in northern and Saharan Morocco. Archaeoastronomy 24: S21–40. Berkani, H., Zazzo, A. and Paris, F. 2015. Les tumulus à couloir et enclos de la Tassili du Fadnoun, Tassili Azger (Algérie): Premières datations par la méthode du radiocarbone. Journal of African Archaeology 13.1: 59–70. Bokbot, Y. 2005. La civilización del vaso campaniforme en Marreucos y la cuestión del sustrato Calcolítico precampaniforme. In M. A. Rojo-Guerra, R. GarridoPena and I. García-Martínez de Lagrán (eds), El Campaniforme en la Península Ibérica y su Contexto Europeo, Valladolid: Universidad de Valladolid, Junta de Castilla León, 137–49. Bokbot, Y. 2007. Fouilles archéologiques dans le pays Zemmour. http://bokbot .e-monsite.com/pages/communiques-de-presse/neolithique-et-protohistoire -des-plateaux-de-zemmour.html. Bokbot, Y. 2008. Le cromlech de Mzora, témoin du mégalithisme ou symbole de gigantisme de pouvoir? Le Jardin de Hespérides. Revue de la Société Marocaine d’Archaéologie et du Patrimone 4: 25–29. Bokbot, Y. and Ben-Nçer, A. 2008. Decouvertes campaniformes récentes dans le plateau de Zemmour (Maroc). In M. Baioni, V. Leonini, D. Lo Vetro, F. Martini, R. Poggiani Keller and L. Sarti (eds), Bell Beaker in Everyday Life, Studi di archeologia preistorica 6, Florence: Museo e Istituto fiorentino di preistoria, 327–30. Bradley, R. 1998. The Significance of Monuments: On the Shaping of Human Experience in Neolithic and Bronze Age Europe. London: Routledge. Brooks, N. 2006. Cultural responses to aridity in the Middle Holocene and increased social complexity. Quaternary International 151: 29–49.

https://doi.org/10.1017/9781108634311.011 Published online by Cambridge University Press

11 Burial Practices in Western Sahara

Brooks, N. 2010. Human responses to climatically-driven landscape change and resource scarcity: learning from the past and planning for the future. In I. P. Martini and W. Chesworth (eds), Landscapes and Societies: Selected Cases, London, New York: Springer, 43–66. Brooks, N., Clarke, J., Crisp, J., Crivallaro, F., Jousse, H., Markievicz, E., Nichol, M., Raffin, M., Robinson, R., Wasse, A. and Winton, V. 2006. Funerary sites in the ‘Free Zone’: report on the second and third seasons of fieldwork of the Western Sahara Project. Sahara 17: 73–93. Brooks, N., Clarke, J., Garfi, S. and Pirie, A. 2009. The archaeology of Western Sahara: results of environmental and archaeological reconnaissance. Antiquity 83: 918–34. Brooks, N., Garfi, S. and Gauthier, Y. 2018a. Typology of stone features. In Clarke and Brooks 2018, 34–55. Brooks, N., Clarke, J., Gauthier, Y. and Gaugnin, M. 2018b. The extensive survey. In Clarke and Brooks 2018, 56–105. Camps, G. 1961. Aux origines de la Berbèrie. Monuments et rites funéraires protohistoriques. Paris: Arts et Métiers Graphiques. Chaix J. 1989. Le monde animal à Kerma. Sahara 1: 77–84. Clarke, J., Winton, V. and Wasse, A. 2018. The excavations. In Clarke and Brooks 2018, 146–76. Clarke, J. and Brooks, N. (eds). 2018. The Archaeology of Western Sahara. Oxford: Oxbow Books. Cremaschi, M. 2002. Late Pleistocene and Holocene climatic changes in the central Sahara. The case study of the south-western Fezzan, Libya. In Hassan 2002, 65–81. Cunliffe, B. 2001. Facing the Ocean. The Atlantic and Its Peoples 8000 BC–AD 1500. Oxford: Oxford University Press. De Laet, S. (ed.). 1976. Acculturation and Continuity in Atlantic Europe. Bruges: De Temple. deMenocal, P., Ortiz, J., Guilderson, T. and Sarnthein, M. 2000. Coherent highand low-latitude climate variability during the Holocene warm period. Science 288.5474: 2198–02. di Lernia, S. 2006. Building monuments, creating identity: cattle cult as a social response to rapid environmental changes in the Holocene Sahara. Quaternary International 151: 50–62. di Lernia, S, 2002. Dry climatic events and cultural trajectories: adjusting Middle Holocene pastoral economy of the Libyan Sahara. In Hassan 2002, 225–50. di Lernia, S. and Manzi, G. 2002. Sand, Stones and Bones: The Archaeology of Death in the Wadi Tannezzuft Valley (5000–2000 BP). AZA 3. Firenze: All’Insegna del Giglio. di Lernia, S., Bertolani, G. B., Castelli, R., Merighi, F. and Palombini, A. 2002. A regional perspective: the surveys. In di Lernia and Manzi 2002, 25–68.

https://doi.org/10.1017/9781108634311.011 Published online by Cambridge University Press

369

370

Joanne Clarke and Nick Brooks

Ferhat N., Striedter, K. and Tauveron, M. 1996. Un cimetière de boeufs dans le Sahara central: la nécropole de Mankhor. In La Préhistoire de l’Afrique de l’Ouest, Saint Maure: Editions Sépia, 103–07. García Sanjuán, L., Luciañez Triviño, M., Schuhmacher, T. X., Wheatley, D. and Banerjee, A. 2013. Ivory craftsmanship, trade and social significance in the southern Iberian Copper Age: the evidence from the PP4-Montelirio sector of Valencia de la Concepción (Seville, Spain). European Journal of Archaeology 16.4: 610–35. Gasse, F. 2000. Hydrological changes in the African tropics since the Last Glacial Maximum. Quaternary Science Reviews 19: 189–211. Gasse, F. 2002. Diatom-inferred salinity and carbonate oxygen isotopes in Holocene waterbodies of the Western Sahara and Sahel (Africa). Quaternary Science Reviews 21: 737–67. Gauthier, Y. 2009. Orientation and distribution of various dry stone monuments of the Sahara. In J. A. Rubiño-Martín, J. A. Belmonte, F. Prada and A. Alberdi (eds), Cosmology Across Cultures, ASP Conference Series 409, San Francisco: Astronomical Society of the Pacific, 317–30. Gauthier, Y. 2015. Pre-Islamic dry-stone monuments of the Central and Western Sahara. In Clive L. N. Ruggles (ed.), Handbook of Archaeoastronomy and Ethnoastronomy, Heidelberg: Springer-Verlag GmbH, 1059–77. Gauthier, Y. and Gauthier, C. 2004. Un exemple de relation monuments – art rupestre: monuments en corbeilles et grands cercles de pierres du Messak (Libye). Les Cahiers de l’AARS 9: 45–62. Gauthier, Y. and Gauthier, C. 2007. Monuments funéraires sahariens et aires culturelles. Les Cahiers de l’AARS 11: 65–78. Gauthier, Y. and Gauthier, C. 2008. Art rupestre, monuments funéraires et aires culturelles: nouveaux documents concernant le Messak, le Sud-Est du Fezzân et l’Oued Djerat. Les Cahiers de l’AARS 12: 1–16. Gobin, C. 1937. Notes sur les vestiges des tombes du Zemmour. Bulletin du Comité des Etudes historiques et Scientifiques de l’Afrique Occidental Française 20: 142–46. Hassan, F. A. (ed.). 2002. Droughts, Food and Culture: Ecological Change and Food Security in Africa’s Later Prehistory. New York: Kluwer. Jodin, A. 1955. Les problèmes de la civilisation du vase campaniforme au Maroc. In G. Souville (ed.), Contacts et échanges entre la péninsule Ibérique et le Nord-Ouest de l’Afrique devant les temps préhistoriques et protohistorique, Paris: Compterendu de Séances de l’Académie des Inscriptions et Belles Lettres (44, 142.1), 353–60. Lancaster, N., Kocurek, G., Singvi, A., Pandey, V., Meynoux, M. M., Ghienne, J. F. and Lö, K. 2002. Late Pleistocene and Holocene dune activity and wind regimes in the Western Sahara Desert of Mauritania. Geology 30: 991–94. Lihoreau, H. 1993. Djorf Torba. Nécropole saharienne antéislamique. Paris: Karthala.

https://doi.org/10.1017/9781108634311.011 Published online by Cambridge University Press

11 Burial Practices in Western Sahara

MacWhite, E. 1951. Estudios sobre las relaciones atlánticas de la peninsula hispánica en la Edad del Bronce. Madrid: Seminario de Historia Primitiva del Hombre. Mattingly, D. J. (ed.). 2003. The Archaeology of Fazzan. Volume I: Synthesis. London: Society for Libyan Studies, Department of Antiquities. Mattingly, D. J., Dore, J. and Lahr, M. M., with contributions by Ahmed, M., Cole, F., Crisp, J., Rodriguez Gonzales, M., Hobson, M., Ismayer, M., Leitch, V., Moussa, F., Nikita, E., Reeds, I., Savage, T. and Sterry, M. 2008. DMP II: 2008 fieldwork on burials and identity in the Wadi al-Ajal. Libyan Studies 39: 223–62. McLaren, S. J., Brooks, N., White, H., van der Veen, M., Gouldwell, T. and Guagnin, M. 2018. The environmental survey. In Clarke and Brooks 2018, 10–33. Milburn, M. 1974. Some stone monuments of Spanish Sahara, Mauritania and the extreme south of Morocco. Journal de la Société des Africanistes 44.2: 99–111. Milburn, M. 1996. Two types of enigmatic stone structures in the north-western Sahara. Council for Independent Archaeology Newsletter 20: 9–10. Milburn, M. 2005. More enigmatic stone structures of the north-western Sahara. Independent Archaeology 52: 6–8. Monod, T. 1948. Sur quelques monuments lithiques du Sahara occidental. Actas y Memorias de la Societé Española de Antropologia, Etnografia y Prehistoria XXIII, cuadernos 1–4:12–35. Paris, F. 1996. Les Sépultures du Sahara Nigérien du Néolithique à l’Islamisation: 1) Coutumes Funéraires, Chronologie, Civilisations; 2) Corpus de sépultures fouillés. Paris: Institute Français de Recherche Scientifique. Paris, F. 1997. Burials and the peopling of the Adrar Bous region. In B. E. Barich and M. C. Gatto (eds), Dynamics of Populations, Movements and Responses to Climatic Change in Africa, Roma: Bonsignori, 49–61. Paris, F. 2000. African livestock remains from Saharan mortuary contexts. In R. Blench and K. MacDonald (eds), The Origins of African Livestock, London: University College London Press, 111–26. Paris, F. and Saliège, J. F. 2010. Chronologie des monuments funéraires sahariens: problèms, méthode et résultats. Les Nouvelles de l’archéologie 120–121:57–60. Pellicer Catalán, M. P., Martínez, P. A., Pérez, M. S. H. and Socas, D. M. 1974. Aportaciones al Estudio del Arte Rupestre del Sáhara Español (Zona Meridional). La Laguna: Universidad de la Laguna. Pirie, A. 2018. The chipped stone. In Clarke and Brooks 2018, 177–96. Salisbury, G. J. 2011. Locating the ‘missing’ Moroccan megaliths of Mzora. Time and Mind 4.3: 355–60. Scarre, C. 2002. Introduction. In C. Scarre (ed.), Monuments and Landscape in Atlantic Europe: Perception and Society during the Neolithic and Early Bronze Age, London: Routledge, 1–14. Schuhmacher, T., Cardoso, J. L. and Banerjee, A. 2009. Sourcing African ivory in Chalcolithic Portugal. Antiquity 83: 983–97.

https://doi.org/10.1017/9781108634311.011 Published online by Cambridge University Press

371

372

Joanne Clarke and Nick Brooks

Searight, S. 2004. Monuments en pierre pré- ou protohistoriques du Sud-marocain, Bulletin de Société des Etudes et de Recherches des Eyzies 53: 89–103. Sivilli, S. 2002. A historical background: mortuary archaeology in the Sahara between colonialism and modern research. In di Lernia and Manzi 2002, 17–24. Sterry, M. and Mattingly, D. J. 2011. DMP XIII: Reconnaissance survey of archaeological sites in the Murzuq area. Libyan Studies 42: 103–16. Vernet, R. 2007. Le Golfe d’Arguin de la préhistoire à l’histoire: littoral et plaines intérieures. Nouakchott: Parc National du Banc d’Arguin.

https://doi.org/10.1017/9781108634311.011 Published online by Cambridge University Press

part v

Looking South

Published online by Cambridge University Press

Published online by Cambridge University Press

12

Burial and Society at Kissi, Burkina Faso sonja magnavita

Introduction Situated in the hinterland of the eastern Niger Bend in north-east Burkina Faso (Fig. 12.1), in the so-called Gourma area (that is, the bush-land on the right bank of the Niger River), the archaeological site of Kissi consists of an extensive cluster of adjacent settlement areas, including several burial grounds (Fig. 12.2). Its occupation during almost the whole Iron Age (c.third century BC to twelfth century AD) provides the opportunity to follow certain developments that local society underwent over more than a millennium. Spreading over an area of more than 300 hectares, the archaeological site lies on the northern shore of the Mare de Kissi (see Fig. 12.2), a small rainwater-fed lake, similar to – though smaller than – several other lakes in this region (that is, Mare d’Oursi c.35 km to the west, Mare de Darkoy c.6 km to the north, or Mare de Markoye c.15 km to the east, to name but the largest). The Kissi inhabitants were sedentary millet farmers with small livestock. They used the dune ridge to the north of the lake for farming and pasture.1 Likewise, they were pottery and iron producers from the initial settlement phase in the last quarter of the first millennium BC onwards.2 Their dwellings were made out of local clay (banco) and they resided on the same spot over many generations. Large anthropogenic mounds (Fig. 12.3) accumulated due to settlement activities, though these tell-like structures are potentially not in themselves settlement mounds: no remains of distinctive features have been found within these mounds that would hint at the existence of decayed buildings (such as, floors, hearths, pits); instead, such structures were found close by, at the foot of and between mounds.3 The mounds thus rather seem to be giant rubbish heaps, with a high concentration of broken pottery, animal bones and stone debris within

1

Magnavita et al. 2002.

2

Magnavita 2009.

3

Magnavita 2009.

https://doi.org/10.1017/9781108634311.012 Published online by Cambridge University Press

375

376

Sonja Magnavita

Figure 12.1. The Eastern Niger Bend and the location of Kissi.

a sandy-ashy sediment.4 Settlement at Kissi reflects a stable, long-standing sedentary occupation of the same lakeshore with neither clear predecessors nor successors. The iron-producing community might have emerged from the Later Stone Age farmers in the area, but since there is an almost complete lack of first millennium BC archaeological sites in northern Burkina Faso, data comparison is virtually impossible. Thus, without clear evidence for a relation between the last semi-nomad agriculturalists at the end of the second millennium BC and those iron-producing and sedentary millet farmers at the end of the first millennium BC,5 it must be taken into consideration that the latter were immigrants. If they were indeed immigrants, their area of origin remains unknown, due to the apparent lack of archaeological data in the broader region for much of the first millennium BC. The same uncertainties also concern the 4

Compare with Vogelsang 2000.

5

Breunig and Neumann 2002; 2004; Magnavita et al. 2002.

https://doi.org/10.1017/9781108634311.012 Published online by Cambridge University Press

12 Burial and Society at Kissi, Burkina Faso

Figure 12.2. Topography of the archaeological site of Kissi.

Figure 12.3. A typical mound at Kissi, and another one near the car in the background (Photo by P. Breunig).

abandonment of the sites around the thirteenth century AD. Possibly, these people left the whole region and migrated to somewhere else – but where to is still unclear. What we do know, however, is that from the

https://doi.org/10.1017/9781108634311.012 Published online by Cambridge University Press

377

378

Sonja Magnavita

settlement’s abandonment up to the present day, only nomadic or seminomadic groups have settled again near the Mare de Kissi, such as the Fulani and Tuareg groups, or their former dependants or slaves, the Rimaibe, Bella and Kel Tamesgayt.6 In the nineteenth century, Kissi belonged to the Fulani Emirate of Liptako. In 1827, a battle between Fulani and Tuareg (Tuareg Oudalan and related groups) took place at Kissi and ended with the defeat of the Fulani. The Tuareg then gave the region a new name, Oudalan, which remains up to today the name of the province.7 We can suspect that prior to the massive arrival of the Fulani and other pastoralist groups in the mid-second millennium AD, the Gourma of north-east Burkina Faso was inhabited mainly by Voltaic millet farmers.8 Closer to the Niger River, the population was more heterogeneous, and included subsistence communities focused on fishing (the Sorko) and hunting (the Gaw) that converged with immigrants arriving from several directions.9 This loose affiliation and/or coexistence is often referred to as proto-Songhay,10 but further west, the millet farmers of the Gourma in north-east Burkina Faso were Gur- (or ‘proto-Gur’) speakers, according to what is known from oral traditions.11 While information from excavations in several of the settlement areas at Kissi basically matches expectations of a quintessentially Iron Age Sahelian rural society, the overall picture becomes considerably more complex when taking into account the funerary data currently available. Archaeological investigations of the Kissi cemeteries allow interesting insights into the structure of the society formerly living at that location. Apparently, we are dealing here with a hierarchical society, in which trade and trading networks played an important role. The fact that exotic items such as brass jewellery, swords, cowries and glass beads reached the site of Kissi before the Arab-Islamic conquest of North Africa, show that that community was not cut off from the outside world but was rather bound into a larger network of contacts and exchange.12 However, the trade factor is but one of many interesting socio-economic aspects at Kissi that becomes visible when burial practices and material culture are considered. Also concerning the probable migration into and out of the region, the Kissi funerary evidence could give important clues, especially when new data will eventually become available from adjacent regions in the future.

6 10 12

7 8 Magnavita et al. 2002. Barral 1967. Madiéga 1982. 11 Insoll 1996; 1997. Madiéga 1982. Magnavita et al. 2002; Magnavita 2009; see also Magnavita 2017.

https://doi.org/10.1017/9781108634311.012 Published online by Cambridge University Press

9

Mayor 2011.

12 Burial and Society at Kissi, Burkina Faso

Burial Practice The Burial Grounds The burial areas of Kissi are located towards the south and centre of the site cluster (Fig. 12.2). Thus, they appear to be in relative proximity to residential areas rather than being distinctly separate. Admittedly, the dating of the settlement deposits is not sharp enough to ascertain which areas were definitely contemporaneous with graves (Fig. 12.4). Nevertheless, the position of the burial grounds in the vicinity of inhabited areas, rather than completely separated from them, implies that the dead were held in high regard rather than segregated. The burial areas can be recognised by stone slabs positioned above or close to graves (Fig. 12.5); however, not all graves were marked by a stone, and only in rare cases was it possible to assign a specific stone to a specific burial.13 These ‘gravestones’ might have experienced multiple re-positionings in the course of time. Similar stele cemeteries can be found at numerous places in the whole region as, for example, in the case of the nearby villages of Oursi, Darkoye, Beiga, Tin Agadel and also farther abroad. In fact, the stele cemeteries of Koukiya and Gao (120 and 170 km away from Kissi, respectively), and further sites in the Niger Valley and Inland Niger Delta displaying dressed stones (for example, Tondidarou),14 give an idea of the possible extent of similar practices, across cultural and religious boundaries. The associated conceptions behind this practice were possibly not uniform, ranging from inscribed or decorated epitaphs with memory function through to demonstration of authority, or function as shrines or sacrificial stones.15 While the original concept behind the dressed stones at Kissi is lost, it is nevertheless interesting that stone slabs are still employed by the Kel Tamesgayt, a Tuareg group living close to the Mare de Kissi, who use the southernmost cemeteries at Kissi for their own funerary practices. However, there seems to be some differences: while the modern users are taking stones from older sites, including large lower grinding stones, the Iron Age people at Kissi never made use of grinding stones in the context of burials. While re-positioning of stone stele might have occurred, all of the dressed stones in the burial zones originally came from places outside of Kissi. A local rock outcrop at Kissi consists of dark rock, likely a type of Gabbro or Amphibolite. This rock type was not employed in burial contexts, although it was in use in the settlement zones. Stone slabs at Kissi were mainly made of slate, sandstones, and basalt.16 13 15

Magnavita 2014. 14 Mauny 1961, 129–34. Insoll 2013, 156–57, and references therein.

16

Magnavita 2014.

https://doi.org/10.1017/9781108634311.012 Published online by Cambridge University Press

379

380

Sonja Magnavita

Settlement areas Atmospheric data from Reimer et al. (2004);OxCal v3.10 Bronk Ramsey (2005); cub r:5 sd: 12 prob usp[chron] Ki 49

KIA-12488 2294 ± 36BP

Ki 22

UtC-6467 1882 ± 34BP

Ki 3B

UtC-9322 1250 ± 40BP

Ki 22B KI-4887 1240 ± 25BP Ki 40

KIA-10613 1097 ± 29BP

Ki 40

KI-4885 1030 ± 25BP

Ki 14B UtC-9324 996 ± 37BP

AD 15

00

C

al

AD al 00 10

50

C

al 0C

/C BC C

al

50

AD

AD al

BC al 0C

C 00 10

15

00

C

al

al

BC

BC

Ki 40S KL-4344 930 ± 35BP

Calibrated date Burials Atmospheric data from Reimer et al. (2004);OxCal v3.10 Bronk Ramsey (2005); lin r:10 sd: 12 prob usp[chron]

KIA-8548 1456 ± 38BP

Ki 14/Gr. 12

UtC-5221 1431 ± 60BP

Ki 3/Gr. 14

UtC-5671 1393 ± 33BP

Ki 14/Gr. 12

Erl-5220 1363 ± 69BP

Ki 13/-110

Erl-5140 1209 ± 62BP

Ki 13/Gr. 4

Erl-5137 1154 ± 52BP

Ki 13/-100

Erl-5139 1124 ± 50BP al A BC /C

al B C

50 0C

al 0C 10 0

C

Ki 3/Gr. 10

BC

UtC-5670 1495 ± 45BP

AD

Erl-5222 1562 ± 45BP

Ki 3/Gr. 10

C al

Ki 14C/Gr. 12

15 00

Erl-5223 1586 ± 64BP

AD

Utc-9325 1590 ± 60BP

C al

Ki 13/Gr. 11 Ki 14C/Gr. 12

10 00

Erl-5224 1611 ± 56BP

al AD

UtC-9323 1616 ± 36BP

0C

Ki 14/-110 Ki 14C/Gr. 18

50

Erl-5219 1715 ± 80BP

D

Erl-3315 1829 ± 105BP

Ki 14/-60

al

Ki 14C/Gr. 2

Calibrated date

Figure 12.4. Radiocarbon dates from Kissi, settlement areas and burials.

https://doi.org/10.1017/9781108634311.012 Published online by Cambridge University Press

12 Burial and Society at Kissi, Burkina Faso

(a)

381

(b)

Figure 12.5. a) Stone stele cemetery Kissi 4, used by a nearby-living Tuareg group; b) trench in Kissi 13, with two stone stele still standing. A skull was buried 10 cm below the stone in the upper left corner.

The nearest sources for such rock types can be found within a radius of c.20 to 70 km around Kissi. Based on the distribution of stone slabs, we have identified seven distinct burial grounds at Kissi. It is possible that further burial areas existed in which no stone stele were used, thus hampering their recognition as zones of interments. There is only one instance where a burial was found in a settlement area, neighbouring a cemetery (Fig. 12.7f).17 It is thus likely that further burials existed at the remote periphery or even outside of the visible burial grounds. Human bones or teeth unrelated to a burial context were, however, not encountered in the excavated settlement areas. Three of the seven cemeteries (Kissi 1, 2 and 4) are relatively close to the lakeshore and are the only ones still in use by the Tuareg fraction that are settled half a kilometre north-west of the lake. The other four cemeteries (Kissi 3, 7, 13 and 14) are located further north-east, towards the centre of the site cluster. Excavations took place in three of the latter cemeteries: one small trench was dug at Kissi 3 (2 × 3 m), one at Kissi 13 (2 × 3 m), and two trenches at Kissi 14 (4 × 4 m and 4 × 5 m, the latter labelled Kissi 14C).18 A larger trench (10 × 10 m) placed in an area at the periphery of Kissi 3 (labelled Kissi 3B) proved not to be part of the actual burial ground of Kissi 3.

The Inhumations All trenches revealed single and multiple inhumations placed directly in the soil that is in earthen pits. The only exception was one large storage vessel that served in Kissi 13 as an urn for a body (Fig. 12.6), close to which were scattered fragments of at least one further, smashed urn. Inhumation in horizontal funerary jars, sometimes found assembled to form impressive graveyards like at Gandéfabou or Gountouré Irbidi, has long been thought 17

Magnavita 2014.

18

Magnavita 2014.

https://doi.org/10.1017/9781108634311.012 Published online by Cambridge University Press

382

Sonja Magnavita

Figure 12.6. A horizontal jar burial in the north profile wall of Kissi 13, Grave 7, oriented west-east, with the mouth of the vessel towards the west. A stone stele is right above the vessel.

to be the main pre-Islamic burial practice in north-east Burkina Faso.19 The discovery of the earthen burials at Kissi indicates that this burial custom has a greater time depth than urn burials and possibly provides clues to the different groups performing different funerary practices within the same region.20 At least at Kissi, jar burials seem to be exceptional, though admittedly the excavations investigated only a small part of the burial grounds. Body orientation was determined for 49 inhumations and was distributed as follows: approximately 92 per cent were oriented roughly west-east (for example, Fig. 12.7; this also encompasses south-west/north-east and north-west/south-east burials); 6 per cent roughly east-west (east-west or north-east/south-west); and only 2 per cent north-south. No inhumation was oriented south-north. Posture of head and body of these 49 determinable oriented skeletons were as follows: approximately 39 per cent were looking towards the south or south-west (for example, Fig. 12.7a–c); 10 per cent looking to the north (for example, Fig. 12.7d, Grave 23); 4 per cent looking to the west or upwards (for example, Fig. 12.7e); 2 per cent looking downwards (for example, Fig. 12.8); but none to the east; about 41 per cent were undeterminable. Approximately 39 per cent were lying on the right body side; 26 per cent had a supine position; 19

Kiethéga et al. 1993.

20

See also Sidibé 1980; Millogo and Koté 2000.

https://doi.org/10.1017/9781108634311.012 Published online by Cambridge University Press

12 Burial and Society at Kissi, Burkina Faso

Figure 12.7. Examples of west-east oriented inhumations at Kissi: a) Kissi 3, Grave 1; b) Kissi 3, Grave 3; c) Kissi 14, Grave 14; d) Kissi 14C, Graves 22, 23, 19 and 25 (from left to right); e) Kissi 13, Grave 2; f) burial in settlement area Kissi 22B.

6 per cent were flexed; 4 per cent were lying on the left body side or in a prone position; about 20 per cent were undeterminable. Thus, the vast majority of the interred bodies were oriented in a roughly west-east direction, with the head to the west and facing in a southerly direction, the latter resulting from a position on the right side of the body (Fig. 12.7). Most of the skeletons show an elongated posture with only slightly

https://doi.org/10.1017/9781108634311.012 Published online by Cambridge University Press

383

384

Sonja Magnavita

Figure 12.8. Example of a west-east oriented skeleton in prone position: Kissi 14C, Grave 17.

flexed legs. When the position of the hands was identifiable, they were often put together over or close to the pelvis (for example, Fig. 12.7). Single graves, as well as double (Fig. 12.9) and multiple graves (for example, Fig. 12.10a) occurred, but in some cases it was not possible to determine whether two or more skeletons represent a multiple burial (for example, Fig. 12.7d) or if they are single burials close to one another (discrete grave pits were discernible only in the lowest parts of the excavation trenches, where the grey infill of the pits became clearly visible against the leached-out dune sand). Due to the sandy, probably acid sediment, the conservation of the bones was extremely poor, limiting physical anthropological information to a minimum. In most cases, only very general statements were possible. Most burials seem to have contained adults, but one individual (Fig. 12.7c) was buried at a younger age, likely between six and eight, although an age of up to 15 cannot be ruled out.21 The trench at Kissi 3 revealed 15 graves and at least 20 individuals. The trench at Kissi 13 revealed at least 16 individuals belonging to 13 graves. In Kissi 14 a minimum of 15 individuals in 14 graves were found, and in trench Kissi 14C a minimum of 45 individuals were assigned to 33 graves. The poor preservation of bones likely explains the partial lack of body parts, but in some cases, partial burials (for example, solely of the

21

M. Webster, pers. comm.

https://doi.org/10.1017/9781108634311.012 Published online by Cambridge University Press

12 Burial and Society at Kissi, Burkina Faso

Figure 12.9. Example of a burial with two individuals: Kissi 14C, Grave 14.

Figure 12.10. Skull burials at Kissi: a) Kissi 3, Grave 12 (the individual in the centre shows also parts of the upper body); b) Kissi 13, Grave 1; c) Kissi 3, H8/-160cm; d) Kissi 14C, Grave 21 (with parts of the upper body).

skull or upper body) can be assumed (Fig. 12.10). Several graves were found heavily disturbed or destroyed. This is most likely the result of recurrent use of the sites as a burial place, rather than the result of robbing – although the latter cannot be excluded in some cases. Either careless or deliberate actions might have taken place, and the idea of deliberate, ritual destruction of an older grave in the course of a new inhumation is certainly not too far-fetched.

Grave Goods The majority of the burials contained grave goods in the form of jewellery and/ or weaponry. The most common were bead collars, arrows and a special type of dagger. Some graves had no grave goods at all (at least no preserved ones), while a few other graves (three complete, one destroyed) were especially richly

https://doi.org/10.1017/9781108634311.012 Published online by Cambridge University Press

385

386

Sonja Magnavita

furnished. Among the graves that lacked any recognisable goods is the single grave in settlement trench Kissi 22B (Fig. 12.7f) and the grave of the young individual mentioned above (Fig. 12.7c). Since only a part of the burial grounds were excavated and because the excavated graves span several hundred years of activity, it is not reasonable to assign specific hierarchical relations between burials. But generally speaking, it is obvious that this was not a totally egalitarian society. Status was expressed by accumulating jewellery and weapons, and only a minority of richly adorned graves had numerous exotic objects: Grave 10 in cemetery Kissi 3, for example, had numerous iron rings and arrows, as well as two daggers, a sword, two brass anklets, four finger rings and several other rings, all made of copper alloy, a bead necklace consisting of 165 pieces, including at least 15 carnelians and 39 glass beads, as well as an unknown number of cowrie shells, six of which have been preserved. Also wool textiles were indicated within this grave. In contrast to this, the individual in Grave 3 in the same cemetery wore only a simple bead necklace consisting of 74 quartz and nine chalcedony beads (see Fig. 12.12a). It seems that exotic elements like cowries, glass and brass jewellery, or impressive weapons like swords, were not accessible to everyone, and those who had access displayed them conspicuously (at least in death) – such as with cowries on the forehead, or a multitude of rings and bracelets. The position of most of the objects in the graves indicates that they were worn by the deceased when they were placed into the grave.22 Although we do not know if this was the common garb or if a person was rather adorned for this special purpose, the way in which these objects were worn or displayed sheds a very rare light onto traditional dressing customs or how things were used. Both iron and copper alloys were worn in the form of jewellery. Rings were worn on fingers and ears, and otherwise attached to items. Bracelets were worn on lower and upper arms; wristlets could be worn on one or both arms and were made out of iron or copper alloys, whereas armlets were only worn on one arm and were exclusively made of iron. Anklets were worn on both ankles or on one side only (Fig. 12.11). They were less common than bracelets and were mostly made of copper alloy. Beads were worn as bead collars on the neck (Figs 12.12a–d, f; 12.13b), in one or double rows, most likely by men and women. In some graves, one of which was likely a woman’s grave, bead waist-belts or aprons were found, in addition to a bead necklace (Fig. 12.12c–d). Beads were also worn on the calf or ankles 22

Magnavita et al. 2002; Magnavita 2009.

https://doi.org/10.1017/9781108634311.012 Published online by Cambridge University Press

12 Burial and Society at Kissi, Burkina Faso

Figure 12.11. Examples for the position of anklets in graves: a) Kissi 3, Grave 10; b) Kissi 14C, Grave 7.

(Fig. 12.12e). Cowries were found in only two graves, on the forehead of the buried persons (Fig. 12.13). Perhaps they were sewn onto a kind of support, such as a form of headdress.23 The position and association of different types of weapons in the graves can also shed light on their usage. A large curved dagger was found in situ on the left upper arm of the richly adorned person in cemetery Kissi 3, Grave 10 (Fig. 12.14). In the same grave, a double-curved sword was found along the left leg. Quivers filled with arrows and sometimes further projectiles, like harpoons, were laid down at the feet of the deceased (for example, Fig. 12.11a). Often, a special type of iron dagger was found close to arrows or a quiver (for example, Fig. 12.11a); it had an oval ringhandle forged onto a leaf-shaped blade (Fig. 12.15), which was sometimes decorated. The closeness of such items to arrows/quivers is certainly no coincidence; in the ethnographic record, such objects appear as equipment of archers.24 No vessels were found as burial goods, nor can we confirm clearly associated food offerings within the graves. Of course, it is possible that perishable offerings have not been preserved. However, some organic materials were preserved because of the high amount of metallic artefacts in some of the graves, which led to mineralisation of organic fibre or created an antibacterial milieu.25 Textiles made of animal fibres (probably sheep wool or dromedary hair) were found in several graves, 23

Magnavita et al. 2002.

24

Lagercrantz 1937.

25

Magnavita 2008.

https://doi.org/10.1017/9781108634311.012 Published online by Cambridge University Press

387

388

Sonja Magnavita

Figure 12.12. Examples of beads position in graves: a) stone bead necklace in Kissi 3, Grave 3; b) ostrich eggshell bead necklace in Kissi 14C, Grave 25; c) stone bead necklace and ‘cache-sèxe’ made of stone and ostrich eggshell beads in Kissi 14C, Grave 23; d) stone and glass bead necklace and ‘cache-sèxe’ made of iron and stone beads in Kissi 14C, Grave 2; e) bone beads on ankle in Kissi 14C, Grave 23; f) stone and glass bead necklace, Kissi 3, Grave 10.

Figure 12.13. The position of cowries in graves: a) Kissi 3, Grave 5; b) Kissi 3, Grave 10 (note also the bead necklace).

https://doi.org/10.1017/9781108634311.012 Published online by Cambridge University Press

12 Burial and Society at Kissi, Burkina Faso

Figure 12.14. Curved arm-dagger on left upper arm: Kissi 3, Grave 10.

Figure 12.15. Examples of looped daggers: a) Kissi 13, Grave 9; b) Kissi 3, Grave 10 (restored); c) an archer, equipped with bow, quiver and looped dagger, depicted on a ‘Benin-beaker’ (after Lagercrantz 1937, 393, Fig. 3).

all made in weft-faced plain weave. The fragments are, however, too small to determine whether these textiles were part of clothing, or whether they were remnants of shrouds. But, the textiles have adhered to the outer side of the weapons whereas the lower, body-facing part sometimes exhibits traces of leather remains. It seems unlikely that weapons and jewellery were worn below wool garments (Fig. 12.16). Rather, I suspect that leather clothing was worn and that a shroud or, perhaps, a mantle, was covering the deceased. A mat made of vegetal fibres was sometimes preserved as the outermost layer of organic remains on metal artefacts. It appears that the deceased, possibly after

https://doi.org/10.1017/9781108634311.012 Published online by Cambridge University Press

389

390

Sonja Magnavita

Figure 12.16. Textiles on an iron ring: a) with leather on the other side, b) facing the skeleton: Kissi 3, Grave 10.

wrapping them in a shroud, were eventually wrapped in a mat before being buried. To sum up, the burial practice at Kissi followed some general conventions. Burial grounds were in relative proximity to living areas. The same area was used as a burial ground over several hundred years. Long and narrow burial pits were dug close to earlier burials and in so doing some of the latter were destroyed – deliberately or unintentionally. The deceased were dressed and adorned according to their economic means and social status, and eventually wrapped in textiles and mats. A roughly west-east orientation was preferred, as were straight or only slightly flexed postures on the right body side or in supine position. Rare exceptions have been noted, with totally deviating orientations, tightly flexed or prone positions, as well as jar burials. Stone slabs, typically made out of non-local rock, were used and possibly re-used. Vessels and terracotta figurines were not components of burials at Kissi, although figurines were part of the local material culture.26

Cultural Parallels between Kissi’s Former Inhabitants and Modern Groups During Kissi’s final settlement phase in the early second millennium AD, the eastern Niger Bend was under control of the Kingdom of Gao and the emerging Songhay State. Whether political control reached into the 26

Magnavita 2001.

https://doi.org/10.1017/9781108634311.012 Published online by Cambridge University Press

12 Burial and Society at Kissi, Burkina Faso

Gourma region and Kissi was among those who had to pay tribute to Gao, we simply do not know.27 But what we do see is that contacts with trade centres at the Niger River existed, and exotic goods obtained through these connections were eagerly incorporated into the material culture of the people living at Kissi. Since some of the rarest exotic objects were found in the most richly adorned graves, we can suspect that some of these special objects (cowries, swords, curved arm dagger) were used to express high status. Until Kissi’s abandonment around the thirteenth century AD, the area was not Islamised. What gave the impetus for finally migrating out of the region is not clear. According to the pollen profile of the Mare de Kissi, the lakeshore vegetation was severely impacted by animals from the early fourteenth century onwards;28 this being a strong indication that cattle pastoralists, perhaps the Fulani, were then dominating the area. The question of whether these pastoralists came across an already largely abandoned area upon their arrival, or if they were competing with the sedentary agriculturalists for water resources and pasture, ultimately causing the latter to leave, must, at least for the moment, remain unanswered. From the region’s archaeological data we can see, however, that the abandonment of Kissi was not an isolated event; what some have called the ‘settlement mound culture’ of north-east Burkina Faso ceased to exist all over the region sometime after the thirteenth century.29 Accumulating rubbish in relatively steep mounds is one of the most prominent characteristics of the sedentary millet farmers once living at Kissi and neighbouring lakes – at least for archaeologists. Such features were also recorded by ethnographers among several Gur-groups, including the Gourmantché, Kurumba, Kasena and Lobi, among others.30 Especially revealing is the remark that such a rubbish heap ‘is made up of the waste from the house . . . yet its size demonstrates the age, vitality or importance of the house’.31 Such a meaningful association between high refuse mounds and high-status houses/families/individuals has been reported in a similar way at, for example, Gourmantché settlements.32 Although it is less straightforward to suppose that we can trace back cultural relations over a thousand years or more, there are several 27 29 30

31

Magnavita et al. 2002. 28 Ballouche 2001; Kahlheber 2004, 229. von Cziernewicz 2002. Gourmantché: Schweeger-Hefel and Staude 1972; Kurumba: Dittmer 1961; Geis-Tronich 1991; Kasena: Abasi 1995; Fiedermutz-Laun 2005; Lobi: Schneider 1991. Abasi 1995, 468. 32 Dittmer 1961.

https://doi.org/10.1017/9781108634311.012 Published online by Cambridge University Press

391

392

Sonja Magnavita

Figure 12.17. Bead belts or ‘cache-sèxe’: a) drawing of how Kurumba women wore them (after Schweeger-Hefel and Staude 1972); b) Kissi 14C, Grave 23; c) Kissi 14C, Grave 2.

noteworthy parallels in material culture and practices between modern Gur-groups and the ancient inhabitants of Kissi. Those might reveal Kissi’s deep rootedness to the Gourma side of the eastern Niger Bend, but more objectively speaking, they account for an unexpected persistence of some traditional elements. There are remarkable similarities, for example, in the design and wearing of traditional garb. Bead aprons or ‘cache-sexes’ were found in situ in several of the oldest Kissi graves, worn in a similar manner by Kurumba women (Fig. 12.17).33 Armlets made of iron displaying hooks and loops were found in some Kissi graves from at least the fifth to seventh 33

Schweeger-Hefel and Staude 1972.

https://doi.org/10.1017/9781108634311.012 Published online by Cambridge University Press

12 Burial and Society at Kissi, Burkina Faso

Figure 12.18. Armlets with hooks and loops: Kissi 3, Grave 10.

Figure 12.19. Musical instruments (not to scale): a) Kissi 3, Grave 11; b) Kissi 3, Grave 10; similar instruments were used, among others, by c) Gulmance and d–e) the Kurumba of Lurum (after Schweeger-Hefel and Staude 1972, 443).

centuries AD. Similar bracelets were worn by Kurumba men as magical protection against injury (Fig. 12.18).34 Bells and rattles from Kissi graves are likewise very close to instruments still in use by the Kurumba (Fig. 12.19). Loop daggers are, together with arrows, the most common weapon type found in Kissi graves, and appear in the first half of the first millennium AD (Figs 12.15, 12.20). Similar implements are known from ethnographic instances throughout West Africa, including several Gurgroups,35 but their pronounced time depth is now for the first time 34

Schweeger-Hefel 1965, 28–29.

35

Lagercrantz 1937; 1950.

https://doi.org/10.1017/9781108634311.012 Published online by Cambridge University Press

393

394

Sonja Magnavita

Figure 12.20. Looped daggers (not to scale): a) Kissi 3, Grave 10 (side and front view); b) Tiv; c) Konkomba (after Lagercrantz 1937, 391, Fig. 2); d) Bich’hwa, India (after Lagercrantz 1950, 230, Fig. 64); e) Moba (Dapong); f) Bassari; g–h) Bariba; i) Igbira; j) Namdjii (after Lagercrantz 1937, 390, Fig. 1).

https://doi.org/10.1017/9781108634311.012 Published online by Cambridge University Press

12 Burial and Society at Kissi, Burkina Faso

documented by the Kissi artefacts. Coming back to the burial practices, west-east orientation of the dead is common among several Gur-groups, as is wrapping the deceased in mats. What characterises the society living at Kissi during the first and early-second millennium AD is thus their sedentary lifestyle with a subsistence based on millet, arboriculture and small ruminants. Furthermore, the building of mud architecture and the deliberate accumulation of trash nearby, as well as the persistence of traditional beliefs in opposition to Islamisation, are also significant. The persistence in traditional beliefs can be seen through the retention of customs such as burying in a west-east direction, adornment of the deceased with jewellery and weapons, and the custom to allow the burial of several individuals within one grave. However, foreign customs were adopted by some, such as the wearing of arm daggers or exotic jewellery, probably for the purpose of heightening or expressing their own social or economic status. The structure of the society was thus not egalitarian, and this fits well with the fact that kingdoms and empires were emerging in the broader region, at the Niger River and in its immediate hinterland. Tracking this society’s migration out of the region is problematic because settlements and funerary remains of the post-fourteenth century are archaeologically largely unexplored in the broader region, with the exception of the Dogonland. As mentioned above, modern Gur-groups now living in adjacent regions towards almost all directions of the compass, from southeast (Gulmance) to northwest (Dogon), share a number of cultural traits with the ancient inhabitants of Kissi. Exploring this issue further by integrating historic archaeology, oral traditions, linguistic and perhaps genetic data might be a line of research worth following in the future. Conversely, tracing from where the Kissi inhabitants once immigrated into the region at the end of the first millennium BC should be equally possible to explore. At present, the lack of archaeological sites of the earlier first millennium BC in the broader region of the eastern Niger bend is the main hindrance to clarify this point satisfactorily. What we can do, however, is to exclude a northern, broadlyspeaking ‘Saharan’ wave of immigration into the Gourma at this time, if the very different funerary customs and domestic economy are a measure that matters.

https://doi.org/10.1017/9781108634311.012 Published online by Cambridge University Press

395

396

Sonja Magnavita

References Abasi, A. K. 1995. Lua-Lia, the ‘Fresh Funeral’: founding a house for the deceased among the Kasena of north-east Ghana. Journal of the International African Institute 65.3: 448–75. Ballouche, A. 2001. Un diagramme pollinique de la mare de Kissi (Oudalan, Burkina Faso). In S. Kahlheber and K. Neumann (eds), Kulturentwicklung und Sprachgeschichte im Naturraum Westafrikanische Savanne, Berichte des Sonderforschungsbereichs 268, Vol. 17, Frankfurt: Goethe-Universität Frankfurt, 129–35. Barral, H. 1967. Les populations d’éleveurs et les problèmes pastoraux dans le nord-est de la Haute-Volta (Cercle de Dori – Subdivision de l’Oudalan). Cahiers ORSTOM, Sér. Sciences Humaines IV.1: 3–30. Breunig, P. and Neumann, K. 2002. Continuity or discontinuity? The 1st millennium BC-crisis in West-African prehistory. In Jennerstrasse 8 (ed.), Tides of the Desert – Gezeiten der Wüste. Contributions to Archaeology and Environmental History of Africa in Honour of Rudolph Kuper, Africa Praehistorica 14, Köln: Heinrich-Barth-Institut, 491–505. Breunig, P. and Neumann, K. 2004. Zwischen Wüste und Regenwald. Besiedlungsgeschichte der westafrikanischen Savanne im Holozän. In K. D. Albert, D. Löhr and K. Neumann (eds), Mensch und Natur in Westafrika. Ergebnisse aus dem Sonderforschungsbereich 268 ‘Kulturentwicklung und Sprachgeschichte im Naturraum Westafrikanische Savanne’, Deutsche Forschungsgemeinschaft, Weinheim: Wiley-VCH Verlag, 93–138. Dittmer, K. 1961. Die sakralen Häuptlinge der Gurunsi im Obervolta-Gebiet, Westafrika. Hamburg: Kommissionsverlag Cram-De Gruyter. Fiedermutz-Laun, A. 2005. The house, the hearth and the granary – symbols of fertility among the West African Kasena. The Medieval History Journal 8.1: 247–65. Geis-Tronich, G. 1991. Materielle Kultur der Gulmance in Burkina Faso. Studien zur Kulturkunde 98. Stuttgart: Steiner. Insoll, T. 1996. Islam, Archaeology and History, Gao Region (Mali), c.AD 900–1250. Oxford: Archaeopress. Insoll, T. 1997. Iron Age Gao: an archaeological contribution. Journal of African History 38: 1–30. Insoll, T. 2013. Lithic and other non-metal small finds. In T. Insoll, R. MacLean and B. Kankpeyeng (eds), Temporalizing Anthropology. Archaeology in the Talensi Tong Hills, Northern Ghana, Journal of African Archaeology Monograph Series 10, Frankfurt: Africa Magna, 149–70. Kahlheber, S. 2004. Perlhirse und Baobab: Archäobotanische Untersuchungen im Norden Burkina Fasos. Ph.D. thesis, Goethe-Universität, Frankfurt.

https://doi.org/10.1017/9781108634311.012 Published online by Cambridge University Press

12 Burial and Society at Kissi, Burkina Faso

Kiethéga, J. B., Sidibé, S. and Bedaux, R. M. A. 1993. Les pratiques funéraires. In J. Devisse (ed.), Vallées du Niger, Paris: Editions de la Réunion des Musées Nationaux, 425–40. Lagercrantz, S. 1937. Ringdolche, Armdolche und Schlagringe in Afrika. Zeitschrift für Ethnologie 69: 389–443. Lagercrantz, S. 1950. Contribution to the Ethnography of Africa. London: K. Paul. Madié ga, Y. G. 1982. Contribution à l’histoire précoloniale du Gulma (Haute-Volta). Wiesbaden F. Steiner. Magnavita, S. 2001. A ceramic figurine from Kissi (Burkina Faso). Sahara 13: 128–29. Magnavita, S. 2008. The oldest textiles from sub-Saharan West Africa: woollen facts from Kissi, Burkina Faso. Journal of African Archaeology 6.2: 243–57. Magnavita, S. 2009. Sahelian crossroads: some aspects on the Iron Age sites of Kissi, Burkina 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, Journal of African Archaeology Monograph Series 2, Frankfurt: Africa Magna, 79–104. Magnavita, S. 2014. 1500 Jahre am Mare de Kissi. Eine Fallstudie zur Besiedlungsgeschichte des Sahel von Burkina Faso. Frankfurt: Africa Magna. Magnavita, S. 2017. Track and trace: archaeometric approaches to the study of early Trans-Saharan trade. In D. J. Mattingly, V. Leitch, C. N. Duckworth, A. Cuénod, M. Sterry and F. Cole (eds), Trade in the Ancient Sahara and Beyond, Cambridge: Cambridge University Press, Society for Libyan Studies, 393–413. Magnavita, S., Hallier, M., Pelzer, C., Kahlheber, S. and Linseele, V. 2002. Nobles, Guerriers, Paysans. Une nécropole de l’Age de Fer et son emplacement dans l’Oudalan pré- et protohistorique. Beiträge zur Allgemeinen und Vergleichenden Archäologie 22: 21–64. Mauny, R. 1961. Tableau géographique de l’Ouest africain au Moyen Age d’après les sources écrites, la tradition et l’archéologie. Dakar: IFAN Institut Fondamental d’Afrique Noire. Mayor, A. 2011. Traditions céramiques dans la boucle du Niger: ethnoarchéologie et histoire du peuplement au temps des empires pré coloniaux. Journal of African Archaeology Monograph Series 7. Frankfurt: Africa Magna Verlag. Millogo, A. K. and Koté L. 2000. Recherches archéologiques à Gandéfabou. Berichte des Sonderforschungsbereichs 268.14: 353–65. Schneider, K. 1991. La grande maison de Bindoute Da: histoire d’une habitation Lobi au Burkina Faso. Stuttgart: F. Steiner. Schweeger-Hefel, A. 1965. Frühhistorische Bodenfunde im Raum von Mengao (Ober-Volta, Westafrika). Graz: Böhlau in Kommission,. Schweeger-Hefel, A. and Staude, W. 1972. Die Kurumba von Lurum. Monographie eines Volkes aus Obervolta (Westafrika). Wien: Schendl Verlag.

https://doi.org/10.1017/9781108634311.012 Published online by Cambridge University Press

397

398

Sonja Magnavita

Sidibé, S. 1980. Archéologie funéraire de l’ouest africain: sépultures et rites. Ph. D. thesis, Université Paris Sorbonne, Paris. Vogelsang, R. 2000. Archäologische Forschungen in der Sahel-Region Burkina Fasos – Ergebnisse der Grabungskampagnen 1994, 1995 und 1996. Beiträge zur Allgemeinen und Vergleichenden Archäologie 20: 173–203. von Cziernewicz, M. 2002. Studien zur Chronologie der Eisenzeit in der Sahel-Zone von Burkina Faso/Westafrika. Ph.D. thesis, Goethe-Universität, Frankfurt.

https://doi.org/10.1017/9781108634311.012 Published online by Cambridge University Press

13

Burial Practices, Settlement and Regional Connections around the Southern Lake Chad Basin, 1500 BC–AD 1500 scott maceachern

Introduction The Lake Chad Basin constitutes an important crossroads in Africa, in the middle of the sweep of savanna that stretches from the Atlantic to the Nile and articulating the Central Sahara with lands to the south. This positioning implicated the region in human responses to Mid-Late Holocene environmental changes, especially those involving decreases in rainfall regimes and the disappearance of Lake Mega-Chad. Archaeological and other evidence indicates that these processes involved periodic population exchanges and cultural interchanges between the southern Sahara and the Lake Chad Basin. The period from c.1800 BC onward saw a development of agro-pastoral systems and an expansion of permanent settlements south of Lake Chad, first on Gajiganna Culture sites and then more widely. This movement towards more permanent occupations, first close to the lakeshore and later on the plains and around the Mandara Mountains to the south, was accompanied by a great differentiation of settlement patterns, including the appearance of anthropic mound sites at some population centres. These mound sites seem to have served as cultural foci on the landscape. Given increased permanence of settlements, it is not surprising that burials became more important elements of the archaeological record, nor that burial practices diversified through time. The preponderance of burials within habitation sites may in part reflect a concentration of archaeological effort on such sites, but may also indicate their role in stabilising community claims to territory and history. By the late-first millennium BC and early-first millennium AD, some of these habitation sites were substantially large, with populations probably in the high hundreds or low thousands. By the second millennium AD, internally differentiated cemeteries existed adjacent to certain sites. The resemblances and differences between burial practices in different parts of the region may provide us with valuable data on cultural relations around Lake Chad. They may also inform us about external relationships,

https://doi.org/10.1017/9781108634311.013 Published online by Cambridge University Press

399

400

Scott MacEachern

especially through the incorporation of foreign objects as grave goods. Despite their significance, burial practices in the Lake Chad Basin have not been examined systematically, in part because the region is split between four different countries and a number of different archaeological research programmes. This chapter attempts to provide an overview of variability in burial practices in the context of a very varied literature, as well as an evaluation of possible cultural connections across the region and beyond.

The Archaeological and Historical Background Increased moisture availability in the early Holocene led to the dramatic expansion of Lake Mega-Chad, and a concomitant opening up of areas now desert or Sahel for occupation by foragers, pastoralists and eventually farmers.1 This would certainly have increased the potential for contacts across the ‘green Sahara’, but paradoxically would likely have decreased the attractiveness of much of the area around the modern Lake Chad itself.2 We have remarkably little evidence for human occupation of the southern Lake Chad Basin until after 2000 BC,3 probably to a significant degree because the marshlands and Sudanian woodlands of the Early-Mid Holocene would have been most conducive to occupation by relatively small foraging communities. In contrast, the Early-Mid Holocene record is far richer for the northern Lake Chad Basin, in areas that are now part of the Sahara.4 Roset’s description of the density of cultural materials close to Adrar Bous as ‘absolument étonnante’, on sites with ‘. . .les dimensions de très gros villages. . .’ is indicative of the intensity of occupation in the Early Holocene.5 The progressive desiccation of the Sahara and the Lake MegaChad regression after 5000 calBC,6 with successive fossil beach ridges marking the shrinking lake, accord well with archaeological data suggesting a southward shift in centres of human occupation in the southern Sahara and Sahel during subsequent millennia.7 Most unfortunately, in part because of political limitations and in part because of the difficulty of access, there is very little known of the prehistory of the areas immediately 1 2 3 5 7

Drake et al. 2011; Kropelin et al. 2008; Schuster et al. 2009. See for example Brooks et al. 2005; Cruciani et al. 2010; MacEachern 2012a. Breunig et al. 1996. 4 Clark and Gifford-Gonzalez 2008; Haour 2003b; Sereno et al. 2008. Roset 1983: 134. 6 Drake and Bristow 2006; Schuster et al. 2005. Haour 2003b; Vernet 2002.

https://doi.org/10.1017/9781108634311.013 Published online by Cambridge University Press

13 Burial Practices around the Southern Lake Chad Basin

north-west (Diffa Region in Niger) and north-east (Kanim Region in Chad) of Lake Chad. The first evidence for substantial settlement in the southern Lake Chad Basin dates to approximately 1800 BC, with the appearance of eventually well over 100 Gajiganna Culture sites on the plains south-west of the retreating lake (Fig. 13.1).8 There is some evidence for contemporary habitation and lithic production sites further to the south of the Gajiganna area, in Cameroon and Nigeria,9 but on a much smaller scale. Economic and artefactual data indicate cultural relationships between Gajiganna communities and contemporary populations in the Central Sahara to the north and north-west.10 The economic adaptation may parallel that of the slightly earlier Tenerian culture in Niger, and there are some general similarities in decorative styles between Gajiganna ceramics and roughly contemporary ‘Culture du Nord-Est’ ceramics north of Niamey in Niger.11 Few data on settlement types or economies for the latter communities exist, while the definition of the Tenerian seems to be quite variable.12 The initial stage of Gajiganna occupation (c.1800–1200 BC) involved shifting settlements and economies oriented towards pastoralism, hunting and fishing, along with the utilisation of abundant wild grasses. The appearance of domesticated Pennisetum millet after about 1200 BC, and its subsequent domination of the archaeobotanical record, was associated with shifts towards more agro-pastoral economies, more permanent settlement and the first stages in accumulation of the mound sites that would be an important element in the prehistory of the region through the next two and a half millennia. The first millennium BC played a pivotal role in the prehistory of the region, and not only because of the expansion and sedentarisation of the Gajiganna culture until about 800 BC. This period seems to have been one of demographic expansion and cultural dynamism south of Lake Chad. Permanent occupation by farmers and pastoralists expanded well beyond the Gajiganna heartland, on the clay plains south and south-east of Lake Chad (again with ceramics of Saharan/Sahelian affinities) and in the more varied environments along the southern edge of the Lake Chad Basin.13 Breunig and Neumann argue that cultural discontinuity in the Gajiganna area after about 800 BC is associated with regional aridity during the earlyfirst millennium BC, and with a transition between Neolithic Gajiganna 8 10 12 13

9 Breunig et al. 2001. MacEachern 2012b. Breunig and Neumann 2002b; Klee et al. 2004; Wendt 2007. 11 Vernet 1996, 172–233. Haour 2003b; Smith 2006. Connah 1981; Gronenborn 1998; Holl 1988b; MacEachern 2012b; Marliac et al. 2000; Wiesmüller 2001.

https://doi.org/10.1017/9781108634311.013 Published online by Cambridge University Press

401

Figure 13.1. Culture areas and sites south of Lake Chad. https://doi.org/10.1017/9781108634311.013 Published online by Cambridge University Press

13 Burial Practices around the Southern Lake Chad Basin

Culture and subsequent Iron Age occupations.14 However, it needs to be noted that there is much more evidence for cultural continuity in other areas south of the lake. The final expression of the Gajiganna Culture involved the appearance of large, nucleated habitation sites like Zilum, surrounded by ditches and possibly walls and dating to the middle of the first millennium BC.15 The socio-political dynamics associated with both much larger site sizes and probable defensive features at this early period have not yet been elucidated (see below), but we should note that large, often defended sites would be characteristic of Iron Age occupation of the region in subsequent centuries. The appearance of metallurgy around the Lake Chad Basin continues to be a topic of some discussion, given anomalously ancient dates of before 3000 BP for copper and ironworking from Niger.16 Available data indicate a gradual and somewhat patchy introduction of ironworking south of Lake Chad from approximately the middle of the first millennium BC onward.17 As Connah suggested more than 30 years ago, iron tools would have appreciably increased the efficiency of agriculture, especially in the heavy firki clay soils close to Lake Chad, but in fact throughout the area.18 The spread of ironworking is probably thus generally associated with continuing increases in settlement sizes and densities across the southern Lake Chad Basin during the first millennium AD, even though some communities did not adopt the technology until centuries after their neighbours. The gradual introduction of sorghum, and a shift from the herding of cattle to the keeping of ovicaprines in more sedentary communities, was probably part of the same process, as agriculture became a predominant element in the economy of those communities. Large Iron Age sites, frequently anthropic mounds and sometimes walled, are found throughout the southern Lake Chad Basin in Nigeria, Cameroon and Chad during the first and early-second millennia AD.19 Similar but mostly unstudied sites in south-eastern Niger probably belong to the same general tradition of settlement.20 There is a good deal of evidence for differentiation in ceramic suites across this large area, but the

14 15 17 18 19 20

Breunig and Neumann 2002a. Breunig et al. 2006; Magnavita et al. 2004; Magnavita et al. 2006. 16 Haour 2003b, 217–18. Gronenborn 1998, 232–33, 241; MacEachern 2012b; S. Magnavita et al. 2009. Connah 1981, 160. Connah 1981; 1984; Holl 2001; Lebeuf 1969; MacEachern 2012b; S. Magnavita et al. 2009. Haour 2003a; 2003b.

https://doi.org/10.1017/9781108634311.013 Published online by Cambridge University Press

403

404

Scott MacEachern

association between ceramic variation and ethno-linguistic identities remains unclear in many cases. It is very likely that the mid/late first and early second millennia AD is the period during which the broad features of the modern constellation of Kanuri, Hausa and other Chadic-speaking populations came into being across the southern Lake Chad Basin, at least in some nascent form. We can infer this in part from the historical record, as well as through material continuities between archaeologically known cases and recent communities.21 The region has been historically attested in Arabic texts as a source of slaves from the late-first millennium AD onward, with the trade attaining substantial numbers in the mid-second millennium AD.22 The intensified slave trade and associated processes of predatory state expansionism are probably the reason for the initial occupation of easily defensible (that is non-plains) environments over the last millennium or more, particularly in the Mandara Mountains and Lake Chad islands.23 There is from the twelfth century onward abundant evidence for diplomatic and cultural interchanges between the states of the southern Lake Chad Basin (especially Kanim and then Borno and Baghirmi) and states north of the Sahara, with Central Saharan communities playing an intermediary role in some of those connections.

Connections and Separations The investigation of prehistoric intra-regional cultural connections in the Lake Chad Basin is seriously handicapped by the fragmentary nature of data for the period in question, especially given the fact that archaeological research is undertaken in four different countries, with data variously analysed and interpreted in the context of Anglophone, Francophone and Germanophone research traditions. Available archaeological data indicates significant distinctions in ceramics between Gajiganna and descendant communities south-west of Lake Chad on the one hand, and late Neolithic and Iron Age populations occupying the firki clay plains south and south-east of the lake, on the other. However, the detailed relations between those populations and other groups living in the region remain quite unclear. Similarities and differences between Lake Chad Basin populations and Saharan groups are equally complicated but, somewhat 21 23

Lebeuf 1969. 22 Austen 1992; Lange 1988; 1989; Lovejoy 2011. See also Harich et al. 2010. Baroin 2005; MacEachern 2012c.

https://doi.org/10.1017/9781108634311.013 Published online by Cambridge University Press

13 Burial Practices around the Southern Lake Chad Basin

paradoxically, because of their geographical separation may be easier to analyse in an introductory overview. There is a degree of disjuncture when we look at data on the changing Holocene relationships between southern Lake Chad Basin communities and populations further north in the southern Sahara. As noted above, the region in question was originally settled by Gajiganna and other late Neolithic populations, with discrete cultural similarities to contemporary peoples of the (modern) southern Sahara. This is true not merely for the ceramic assemblages, which are representative of a broader Mid Holocene tradition of ceramic production stretching across the southern Sahara from the Wadi Howar area to the Niger, but also for some elements of the lithic materials.24 In addition, the semi-sedentary, pastoralist economic orientation of the initial Gajiganna settlers parallels that of communities further to the north a millennium or two earlier. The archaeological data are paralleled by linguistic and genetic data that, although disputed in details, broadly emphasise the origins of modern Chadic- and Nilo-Saharan populations around Lake Chad in a complex set of mid Holocene interactions in the southern ‘green Sahara’.25 Our concentration on cultural connections to the north is in part explained by a comparative ignorance of mid Holocene cultural relations between the southern Lake Chad Basin and points further east, south or west – but there is no real doubt that the initial settlers south of the lake were significantly related to contemporary Saharan populations. Between about 1000 BC and AD 1000, the situation was substantially different. During this period, which also encompasses the time period associated with the Garamantian culture, economic and settlement practices in the southern Lake Chad Basin diverged from earlier pastoralist norms, with a greater investment in sedentism, cereal agriculture and larger, more complex community structures. It is quite possible that specialised cattle pastoralism came into being as an economic adaptation into the first millennium AD, whether ethnically based (as in more recent times) or not, but we see few traces of it in the archaeological record. The palaeo-climatological record indicates progressively drier conditions in the southern Lake Chad Basin between about 800 BC and AD 300, and this expression of a general late Holocene aridification is likely implicated in this divergence in cultural norms. To best appreciate the dynamics of this divergence, we would need data from the areas north-east and north24 25

Breunig et al. 1996, 141. Blench 1999; Cerny et al. 2009; Cruciani et al. 2010; Ehret 2006; Tishkoff et al. 2009.

https://doi.org/10.1017/9781108634311.013 Published online by Cambridge University Press

405

406

Scott MacEachern

west of Lake Chad, where the economic commonalities and cultural relations of the earlier mid Holocene ‘green Sahara’ were presumably being ‘pulled apart’ in progressively drier environments during this time. As noted, however, the archaeological record is essentially non-existent for that area before the second millennium AD, after which time there have been a number of small-scale investigations of sites associated with the Kanim-Bornu polities.26 If cultural and economic connections were being maintained between the southern Lake Chad Basin and Saharan communities (including the Garamantian region) during the first millennium BC and first millennium AD, we might expect to find some material traces of such connections, most obviously in the form of trade goods – potentially in burial as well as other contexts. It is remarkable, however, just how little material evidence for such long-distance trade exists in the region during this period. There are scattered finds of amazonite and carnelian beads from southern Lake Chad Basin sites, including Zilum, Blé, Houlouf and in addition Mege (date somewhat uncertain), Doulo Igzawa I, Ghwa Kiva and Aissa Dugjé.27 These are most likely of Saharan provenance and the carnelian played a significant role in Garamantian international trade, at least from the point of view of Roman authors.28 It is notable that carnelian and amazonite beads are also found in contemporary Niger Bend sites and probably somewhat later at Igbo-Ukwu.29 In general, though, the archaeological data suggest that the southern Lake Chad Basin was not playing a central role in sub-continental exchange networks through this period. As we will see, the situation changed considerably in the mid/late first millennium AD. A wider variety of exotic artefact types, including cowries and glass and carnelian beads, began to appear on sites in the region, although such material is never really abundant.30 In addition, the remains of at least a dozen horses were recovered during excavations at Aissa Dugjé, the earliest dating to calAD 620–890 (1310±60 BP; Δ13 -16.2; TO-7515).31 In later times, horses would of course be heavily implicated in warfare and elite display across the Sudanic zone, and the discovery of horse remains (in contexts indicating that their possession conferred some status) suggests that the origins of these associations date to the late first millennium AD at least. We should not push these inferences too far, but it is worth remembering that the chief ‘product’ exported from the 26 27 28 30

Haour 2008; Haour and Gado 2009. Personal field notes. See also, Gronenborn 1998, 237–38; S. Magnavita et al. 2009, 54. Insoll et al. 2004; Wilson 2012. 29 Insoll and Shaw 1997; MacDonald 2011. S. Magnavita et al. 2009. 31 MacEachern et al. 2001; MacEachern 2012b.

https://doi.org/10.1017/9781108634311.013 Published online by Cambridge University Press

13 Burial Practices around the Southern Lake Chad Basin

southern Lake Chad Basin during the historical period was slaves, and that the first historical mention of such a trade from the broader area dates to only perhaps two centuries later than the earliest horse remains from Aissa Dugjé. Horses were also historically central to the process of slave-raiding in the southern Lake Chad Basin, and until breeding populations were established, they would probably have constituted an important trade good in and of themselves. This is also the period associated with the discovery of pottery originating in the Lake Chad Basin on the site of Marandet, about 800 km to the north-west near the Aïr Mountains in Niger,32 a remarkable testimony to the long-distance movement of a relatively fragile artefact class at this time. During the second millennium AD, evidence for long-distance trade is thus significantly more common in the region. During the Garamantian period between approximately 1000 BC and AD 700,33 the material evidence for contacts between the populated areas south of Lake Chad and Saharan communities is thus quite minimal. Some potential reasons for this will be evaluated below, but it is important to remember that relations between these two areas were always dynamic through time. This involved an initial period of settlement south of Lake Chad when Saharan cultural connections were quite important, through a long period of relative differentiation and separation contemporaneous with the Garamantes, through what seems a period of steadily increasing contact from the late first millennium AD. It is not as if there was a gulf fixed between these regions that was eventually bridged after perhaps AD 700, but rather a developing set of late Holocene cultural configurations that interacted more strongly at some times than at others. I turn now to the evidence of burials and grave goods around Lake Chad, separating my coverage into areas that correspond primarily to different research programmes, themselves correlated with different geographical/ cultural zones.

Burials in the Southern Lake Chad Basin Gajiganna Culture and Succeeding Sites South-West of Lake Chad A number of burials were encountered during the original excavations on Gajiganna habitation sites.34 Four burials from the NA90/5A site probably date from approximately 1250–1000 BC; all were buried in a flexed 32 34

S. Magnavita 2017. 33 Mattingly 2003, 348–49. Breunig et al. 1996, 135; Wendt 2007, 16, 22.

https://doi.org/10.1017/9781108634311.013 Published online by Cambridge University Press

407

408

Scott MacEachern

position, with arms flexed near the head, and with no common orientation. There was no evidence for grave pits or grave goods. Two were adult males, and two juvenile females. A very similar, and very poorly preserved, interment was discovered in excavation on the NA90/5D site, but dating to the seventh century AD – which testifies to the stability of this burial practise. Carlos Magnavita notes the existence of five burials at the Gilgila (Gajiganna Phase II: c.1300–900 BC), Zilum (‘Gajiganna Phase III’: c.500 BC) and Gilgila, Elkido Nord and Dorota (early Iron Age: c.AD 100–600) sites in the same area.35 These are very similar to the other Gajiganna burials: burials within habitation sites, crouched positions with flexed arms, a variety of orientations and no evidence for grave goods. At Zilum, it appears that pits used for craft industries, probably tanning, were later reused as secondary burials.36 Most of the burials appear to be on their right sides. As noted above, there are very rare finds of central Saharan carnelian and amazonite in these sites dating from the mid-first millennium BC onward, none of which are certainly associated with burial contexts.

Daima and Other Sites in Borno Graham Connah excavated a number of anthropic (‘tell’) mound sites in Borno in the 1960s, most notably Daima.37 In the course of that fieldwork, he recovered the remains of four individuals from Borno 38, nine from Kursakata and approximately 67 from the large excavations at Daima. The period covered by these excavations runs from the first half of the first millennium BC (Borno 38 and Kursakata) through to perhaps 600 BC to AD 1200 (Daima). In general, the pattern of burials on these sites through most of this period was similar to that of the Gajiganna sites noted above. Burials were located in habitation sites, with grave pits (when detectable) averaging about a metre in depth. At Kursakata, the nine interments were sufficiently clustered in depth, in a relatively small excavation unit, that led Connah to believe they may be from a cemetery context.38 The skeletons were strongly flexed, usually with hands close to the face, and most often on their right sides. Directional orientation of the body was quite variable until Daima III times (probably c.AD 700–1200, significantly later than the burials from the Gajiganna area further to the

35 38

Magnavita 2008, 24–25, 36–43, 54–60. Connah 1981, 95.

36

https://doi.org/10.1017/9781108634311.013 Published online by Cambridge University Press

Magnavita 2008, 125.

37

Connah 1981.

13 Burial Practices around the Southern Lake Chad Basin

west), when almost all skeletons where orientations could be identified were buried with their heads to the west. There is also little evidence for grave goods until the Daima III period. Connah noted the considerably greater diversity of the grave goods from Daima III contexts, and their increased evidence for contacts beyond the immediate region of the firki clay plains.39 Daima III grave goods include iron, bronze/brass and copper ornaments, and a great diversity of carnelian, stone, glass and ostrich eggshell beads. The change in burial customs can be perhaps best represented by Burial I in the Daima III levels, a subadult individual buried with 4,181 beads and cowrie shells, of which approximately 2,850 are glass or carnelian beads. It is notable that some of the jewellery found in Daima III levels, for example, the necklace and bracelet associated with Burial 27, are quite similar to Garamantian examples that are presumably some centuries older.40 This may, of course, be due simply to limits in the possibilities of stringing a common repertoire of bead types. There is artefactual evidence for a good deal of cultural continuity between the Daima II and Daima III periods, but evidently by Daima III times, the inhabitants of the site were more integrated into exchange networks that articulated with Saharan trading systems. It is interesting that the contemporary Yau site in the Komadugu Yobe valley close to the border with Niger, north-west of the Gajiganna sites and also excavated by Connah, yielded no burials and almost no evidence of exotic artefacts beyond a single cowrie shell.41 This serves to remind us that there were probably significant differences in the ways that particular settlements articulated with wider exchange networks at this time. Detlef Gronenborn investigated a number of plains sites near Daima, and discovered 18 burials in Ndufu (probably late first millennium BC, found over approximately 1.5 m vertically), a single burial at Mege (date uncertain, but possibly the late first millennium AD) and one burial at Ngala (eighth century AD).42 Gronenborn believes that the concentration of burials at Ndufu again indicates a cemetery, and that layers of ash and fire-reddened clay associated with the burials was part of a funeral ritual. In all cases, these individuals were buried in strongly flexed positions, with no standard orientation of the skeleton. No grave goods were found with the 18 individuals buried at Ndufu, but carnelian beads were found with the Mege burial. 39 41

Connah 1981, 173–78. Connah 1981, 201–12.

40 42

Connah 1981, 174–75; Mattingly et al. 2009, 127,Fig. 15a. Gronenborn 1998.

https://doi.org/10.1017/9781108634311.013 Published online by Cambridge University Press

409

410

Scott MacEachern

Figure 13.2. Sub-adult burial from the Doulo Kwovré site (PMW 631).

Mandara Burials In the course of excavations on sites around the Mandara Mountains during the mid-1990s, we recovered burials from a number of sites, all dating to the Iron Age. Two burials were recovered from units on the Doulo Kwovré habitation site (PMW 631), a well-preserved sub-adult (Fig. 13.2) and a partial adult skeleton. Both were dated to the first half of the second millennium AD, on the basis of ceramic associations. Both were buried in a flexed position on their right sides: the sub-adult’s arms were flexed, and that individual appears to have been clasping an egg in their hands. About 20 cm vertical depth of the burial pit was detectable in the case of the sub-adult. Nineteen human burials were uncovered in the course of excavations at the Aissa Dugjé (PMW 642) site, which was occupied between the middle of the first and the early second millennia AD.43 Many of these were very poorly preserved; in addition, our agreements with local Islamic community leaders involved getting their approval to excavate when burials were found, and we did not continue with excavation after the initial discovery of six of the burials. In all of these cases, interments were discovered within the multiple mounds that characterise the site. For the most part, these were adult burials, often placed under cairns made of stones and pots, and in a number of cases probable L-shaped burial pits were noted, similar to 43

Bourges et al. 1999; MacEachern et al. 2001; MacEachern 2012b.

https://doi.org/10.1017/9781108634311.013 Published online by Cambridge University Press

13 Burial Practices around the Southern Lake Chad Basin

those made in non-Islamic communities in the Mandara Mountains today. Infants and young children were frequently buried in large pots, generally similar to those used for beer and water storage in the area today. In all nonpot burials, skeletons were flexed, usually on their right sides, and where hand/arm positions were detectable, they were drawn up to the face. Some of the pot burials were so tightly contracted as to suggest that the bodies were bound before being placed in the vessels. Intact pots were associated with a number of burials, and a carnelian bead may have been associated with one burial dating to the late first millennium AD. A concentration of four burials in a limited area of Unit 1 excavations may indicate part of a cemetery, as was noted from other sites. One characteristic of this site, and of other mound sites excavated in this area during the 1980s and 1990s,44 was the presence of isolated human bones, bone fragments and teeth in excavation units, not associated with intact burials. This probably testifies to the degree of disturbance caused by human activity on these mound sites through time. There were rare finds of carnelian beads from these sites, dating from the first millennium BC onward, but none were definitively associated with burial contexts. We should note as well that another form of burial practice informs the question of regional connections at Aissa Dugjé – that of the horses noted above. Skeletal remains from a dozen domesticated equids (ponies or small horses) were found at the site (Fig. 13.3), and at least four of these were in burial pits; these include one very old animal with hip dysplasia, unable to be ridden, probably fed while tethered and ultimately buried in association with a dog. It is probable that these animals were being bred locally and had been so for some time, given that local breeds of small equids were also characteristic of the region through the historical period. But they certainly constitute evidence of some inter-regional contacts in the Lake Chad Basin before the late first millennium AD, either across the Sahara or via the Nile Valley. Practices similar to the Aissa Dugjé horse burials are known from a number of late precolonial communities along the eastern edge of the Mandara Mountains, and it is likely that in early times at least these horses were kept more for their role as status symbols than as tools of war and aggression. The Thuliva Kwacha (PMW 768) mound site in Nigeria, within 200 m of the Cameroonian border, was not excavated, but burials in pots were eroding out the banks of a seasonal stream adjacent to the site in August 1993 (Fig. 13.4). The bones were very scattered, but seemed to 44

David and MacEachern 1988.

https://doi.org/10.1017/9781108634311.013 Published online by Cambridge University Press

411

412

Scott MacEachern

Figure 13.3. Horse burial from Aissa Dugjé (PMW 642).

include a variety of adults and sub-adults. The age of this site is unknown, but it probably dates to the late Iron Age. To date, no burials have been recovered from sites in the Mandara Mountains.

‘Sao’ and Related Sites Survey and excavations since the late 1940s by Jean-Paul Lebeuf and colleagues in areas of Cameroon and Chad close to Lake Chad (and primarily associated with modern Kotoko people) have uncovered the

https://doi.org/10.1017/9781108634311.013 Published online by Cambridge University Press

13 Burial Practices around the Southern Lake Chad Basin

Figure 13.4. Pot burial exposed in stream bank at the Thuliva Kwacha site (PMW 768).

remains of more than 600 sites, which Lebeuf designated as ‘Sao’ – a terminology associated with the pre-Islamic, pre-Kanuri populations of the region.45 These prehistoric sites certainly display similarities in material culture to others around the southern Lake Chad Basin, and use of the term ‘Sao’ to designate these specific sites is thus to some degree associated with a particular research programme. Over more than 20 years of fieldwork, this research programme involved the production of a map of archaeological sites and large-scale excavations on a number of anthropic mound sites.46 Students and colleagues, especially Augustin Holl, continued that work.47 One problem with some of the excavations is a relative lack of chronological control, so that cultural sequences are sometimes difficult to discern.48 The walled mound site of Mdaga, situated north of N’Djaména, certainly originated in the Iron Age, but different parts of the site provide radiocarbon sequences from the second century BC until the eighteenth century AD, and there is some evidence for mixing of materials from different parts of the site. It appears that approximately 57 skeletons were recovered in the course of excavations on the Mdaga site, 45 48

Lange 1989. 46 Lebeuf 1962; 1969; Lebeuf et al. 1980. For example Lebeuf et al. 1980.

47

https://doi.org/10.1017/9781108634311.013 Published online by Cambridge University Press

See especially Holl 1988a; 2001.

413

414

Scott MacEachern

along with a cemetery of pot burials about 200 m south-west of the main mound.49 The burials within the habitation mound seem to be distributed fairly evenly through the stratigraphy; in contrast with the other areas examined to this point, the majority of individuals appear to have been buried in an extended position and often on their backs, although this varies somewhat across the site. The orientation of the burials is quite varied. There is little information available on grave goods associated with these mound burials, but the recovery of more than 250 carnelian beads from the excavations is of interest. The pot burials off the mound itself probably date from the later part of the occupation sequence, given their proximity of the ground surface, and link the site to others in the region (see below). At least three of the pot burials also contained carnelian and/or glass beads, while another contained three copper bracelets. Work near Lake Chad in Cameroon by Holl has uncovered evidence for a complex cultural sequence that probably begins in the late first millennium BC.50 About a dozen burials were recovered from habitation contexts on a number of sites, and dating to the late first and early second millennia AD. In general, these resemble burials at Mdaga and other ‘Sao’ sites: most were straightforward inhumations, buried in extended or slightly flexed positions and with varying orientations, and with few grave goods except for pottery and rare carnelian beads. There were also a number of child pot burials at the Mishikwa site, directly comparable to those at Aissa Dugjé and probably also dating from the late first millennium AD. On the mound site of Houlouf, however, a similar burial pattern over the period of approximately AD 500–1400 (that is, extended inhumations with few grave goods, scattered through the deposits) was replaced c.AD 1500–1600 by a cemetery containing 25 adult pot burials, apparently arranged in a social ranking system.51 Grave goods were again dominated by a variety of artefacts made of copper alloys, and carnelian beads, of which over 900 were found.

Durbi Takusheyi Although they sit well outside the Lake Chad Basin, the extraordinary Durbi Takusheyi elite burials in the kasar hausa, between Katsina and Daura in north-central Nigeria, are worthy of note in this context.52 The seven tumuli at this site, with a long and complex history of 49 52

Lebeuf et al. 1980, 91–96. Gronenborn et al. 2012.

50

Holl 1988a; 2001.

https://doi.org/10.1017/9781108634311.013 Published online by Cambridge University Press

51

Holl 1994.

13 Burial Practices around the Southern Lake Chad Basin

investigations, probably were built at various times between the thirteenth and early sixteenth centuries AD, during a period of florescence of developing Hausa elites. The very rich grave goods recovered from these tumuli, including various copper alloy and gold objects, cowries and glass and carnelian beads, testify to active participation in exchange networks and systems of elite display extending to the Near East at least. It is likely that most of these objects were either procured directly from Saharan sources (for example the carnelian beads) or brought into the region along Central Saharan trade networks.

Interpretations from Grave Contexts Burial Contexts There was a good degree of consistency in burial practices in the southern Lake Chad Basin in the period under study, and also in the ways in which those practices changed through time. First, known burials through the entire period are found within or in very close proximity to habitation sites, both the mound sites that are such prominent points upon the plains landscape and on flat (but perhaps ditched/walled) sites like Zilum and Elkido Nord. A comparatively large number of burials are encountered in even quite modest excavations in some cases. This is of course very much related to the fact that all of the archaeological research programmes in the region have involved survey for habitation sites, so that off-site burials have much more rarely been located; it is entirely possible – and indeed, given the relationship between community sizes and numbers of recovered burials, quite likely – that at least some people were buried away from habitation sites. That being said, placement of the dead in close and constant contact with the living was a common cultural element, one that invokes complex historical relationships and often justification for residence and control over particular places.53 This is a striking contrast with modern practice, not merely in the nowIslamised plains but in the communities of the Mandara Mountains that seem to be at least in part historically related to Iron Age southern plains sites. In Mandara communities, demarcated cemeteries play similar roles of historical positioning through their spatial and ideological placement at the centre of dispersed farming communities.54 Southern Lake Chad Basin burial locations also provide a significant contrast with the contexts of 53

Chapman 2008; Evans 2005.

54

MacEachern 2002.

https://doi.org/10.1017/9781108634311.013 Published online by Cambridge University Press

415

416

Scott MacEachern

Garamantian burials, which involved placement in formal cemeteries away from nucleated habitation sites and, it appears, a more complex typology of tomb types.55 This tendency to bury at least some dead individuals in close proximity to the active living probably also accounts for some of the characteristics of the burials discovered, perhaps especially in mound sites: the orientations of inhumations tended to be quite variable (probably because of the complexity of depositional circumstances); there is abundant evidence for prehistoric disturbance and damage to the remains; and fragmentary human skeletal material is often found throughout the soil matrix more generally. There is significant differentiation in burial postures in different parts of the region. Skeletons in north-eastern Nigerian and northern Cameroonian sites (and thus south-west and south of Lake Chad), tend to be buried in flexed/contracted postures, usually with hands in front of the head and usually but by no means always on their right sides. On the ‘Sao’ sites and other sites south-east of and closer to Lake Chad, extended burials are more common even in pre-Islamic periods, and seem to be a majority on sites like Mdaga. The other marked contrast between the two areas is in the prevalence of human burials in large pots or double pots. On relatively weak chronological evidence, this practise may be older and is certainly more widespread in the south-eastern ‘Sao’ sites, and indeed the large ceramic vessels used in these burials are known as ‘Sao pots’ through the southern Lake Chad Basin. They may have been used first for burials of children and later for adults; certainly by the time of the Houlouf cemetery in the midsecond millennium AD, they were being used for high-status male burials. Such burials tend to be rarer and later south-west of the lake. The region around the Mandara Mountains seems to be to some degree an area of overlap between these two regional sub-traditions. Skeletons found within mound sites in that area are usually flexed/contracted and mostly lying on their right sides, as in northern Nigeria. Uniquely, as far as I am aware, at Aissa Dugjé, we have evidence for L-shaped burial pits and cairns of pots and stones above the burials. However, there were also a number of burials of sub-adults in large single pots on the site through the occupation period, which links Aissa Dugjé to burial practices on the ‘Sao’ sites. The practice of using large ‘Sao pots’ for adult burials seems to have spread more widely through the southern Lake Chad Basin in the early second millennium AD, and as noted above they are found in Chad, Cameroon and Nigeria.56 55

Mattingly et al. 2009; Mattingly et al. 2010.

https://doi.org/10.1017/9781108634311.013 Published online by Cambridge University Press

56

Connah 1981, 48, 55–57, 239–41.

13 Burial Practices around the Southern Lake Chad Basin

Artefacts Artefact patterning associated with burials through time in these sites is particularly interesting. There is relatively little evidence for the inclusion of grave goods with burials in the region until the late first millennium AD: some pottery is occasionally found in association with burials, but given the intensity of human activity and disturbance on habitation sites, it is often difficult to tell whether the association is deliberate or simply due to taphonomic processes. As noted above, there are scattered, very infrequent finds of carnelian and amazonite beads on the southern Lake Chad Basin in different contexts of the first millennium BC to mid-first millennium AD. Given the likely value of these exotic artefacts and their association with burials in later periods, it is very likely that these earlier discoveries were also originally associated with burials and later disturbed, but that hypothesis cannot be proven. The presence of these exotic goods does indicate at least some indirect exchange relationships between the southern Lake Chad Basin and the Central Sahara in Garamantian times. A number of different factors, acting together, may account for the small amount of material evidence for such relations: • these connections were extremely tenuous and sporadic during this period; • southern Lake Chad Basin communities were at the very end of subcontinental exchange networks; and/or • the goods most frequently exchanged in trade by these communities were subsistence goods (for example salt and dried fish), as was the case during the historical period, so that exotic artefacts only rarely made it to the very end of the line of trading relations. Any and all of these factors might play a part in explaining the rarity of exotic items in this region at different times over this period. We do not know whether any trade in slaves existed south of Lake Chad before approximately the end of the first millennium AD, although the defensive features around a number of settlements dating to the previous 1500 years suggests that at least a local trade in captives might have existed. It is clear that the frequency of these exotic artefacts, and especially carnelian, increased very significantly from the late first millennium AD onwards, both associated with burials and in contexts not obviously so related. This presumably indicates more significant connections, and an increased integration of the southern Lake Chad Basin into subcontinental socio-economic systems, during this period. The rich burials

https://doi.org/10.1017/9781108634311.013 Published online by Cambridge University Press

417

418

Scott MacEachern

from the Daima III period at Daima, Mdaga and Houlouf imply the emergence of some degree of social rankings within regional communities, expressed in part by the accumulation of valuable and exotic items that would eventually be used as grave goods. In general, there is a strong association between the number of burials and the number of exotic beads found on sites during this period. This suggests that these exotic artefacts were not restricted to funerary contexts, but rather certain communities.57 Although there are few exotic grave goods such as carnelian at Aissa Dugjé, the presence of horses on that site (some deliberately buried) provides similar associations of emergent political power and far-flung sociopolitical connections. The origins of the various copper-alloy artefacts found in burials from this same time period and onwards are more obscure, since they could either be of West African origin or Saharan or TransSaharan imports: on balance, for this time and area, the latter seems more likely.58 At the same time, these artefacts do provide similar evidence for social ranking in these later communities.

Discussion It is fairly obvious from the evidence surveyed in this chapter that archaeological investigations of relationships between the Central Sahara and the Lake Chad Basin are just beginning. There is relatively little that would permit researchers to detect movement of people from this latter area into the Garamantian world during the mid/late Holocene, especially given the cultural disruptions that would go along with enslavement. Some characteristic genetic features might indicate specific regional ancestry, but in general neither the timing nor the geographical spread of such characteristics are known with great accuracy.59 Isotopic approaches may provide data on the origins and movement of individuals,60 but the utility of stable carbon isotope ratios is somewhat impaired by the discovery of sorghum and millet on Garamantian sites, while spatial and temporal variability in strontium and oxygen ratios around the Sahara are not well understood. Discovery of artefacts and practices correlating to particular SubSaharan cultural systems might provide further insights into contacts 57

58 60

A similar situation takes place in Fazzan several hundred years earlier with a concentration of Mediterranean imports in associated settlement and funerary contexts, see Mattingly et al., Chapter 2, this volume. Craddock and Hook 1995. 59 For example Podgorná et al. 2013. Power et al., Chapter 4, this volume.

https://doi.org/10.1017/9781108634311.013 Published online by Cambridge University Press

13 Burial Practices around the Southern Lake Chad Basin

between these regions, but the specificity of such practices must be kept in mind. Pot burials were distinctive of southern Lake Chad Basin sites, for example, but there is no reason to think that such burials would be permitted among enslaved people in Fazzan. The fascinating discovery of a female burial with a lip-plug in a Garamantian cemetery certainly hints at SubSaharan affinities, but provides little further information than that, given the wide distribution of this artefact type.61 Other artefact types and production techniques (for example, hand-made pottery or roulette decoration) may also indicate Sub-Saharan origins, but again provide rather little specific information on where those origins might be. A significant sample of distinctive decorative motifs might show the movement of pots themselves, while analyses of forming practices in the ‘chaîne opératoire’ of ceramic production could indicate the presence of Sub-Saharan potters on Garamantian sites – but to this point samples do not exist for these investigations. It is likely that, in the medium term at least, archaeometric analyses of specific artefact types found on either Saharan or Sub-Saharan sites will continue to provide the most useful information on material flows, whether these involve ceramics, stone or glass beads or other materials that can often be sourced with some precision.62 In a recent conference paper, Wynne-Jones and Haour provocatively examined the implications of ‘small-world’ systems for the study of exchange in Africanist archaeology,63 while Haour has also written about networks of trust in trade systems around the Central Sahara, especially as these would have been transformed with the coming of Islam.64 These issues must be related in this region, as they no doubt are in many other areas of the world. Shared, super-regional ideological systems that can bind together far-flung centres and actors into interacting networks, even ‘thin’ ones (as with Islam in the second millennium AD Sahara, or Christianity in mediaeval Western Europe), can be a potent force for facilitating economic and cultural interchange. Such ideological systems may have a considerable time-depth in the Sahara and Sahelian zones,65 although the particular case of literate, universalist world religions is probably quite different from anything that had gone before. We have much less information for areas of such networks further to the south in Africa. There seems to be a complex, overlapping florescence of site nucleation, ironworking and figurine production that stretches from at least the Walasa area, south of Lake Chad, towards the Jos Plateau from the mid-first millennium BC onward, and that may involve 61 63 65

Mattingly et al., Chapter 2, this volume. 62 S. Magnavita 2017. Wynne-Jones and Haour 2014. See also Sindbaek 2007. 64 Haour 2017. MacDonald 1998.

https://doi.org/10.1017/9781108634311.013 Published online by Cambridge University Press

419

420

Scott MacEachern

some such shared cultural/ideological elements.66 It may be that southern Lake Chad Basin communities were more broadly engaged with elements of that network, as sedentism and its consequences became more and more significant after about 500 BC. However, that cultural system is known only in its vaguest outlines to this point. For the time periods associated with Garamantian culture in Fazzan (that is between about 1000 BC and AD 700), there is little evidence for interchange between the southern Lake Chad Basin and Central Saharan communities. The presence of rare carnelian and amazonite beads on Chadian sites does indicate that such interchange sporadically took place, although its nature is difficult to evaluate. Initially, quite egalitarian societies south of Lake Chad, where (in contrast to the lands south of the Niger Bend) gold was unavailable, would probably have been of interest to the Garamantian state chiefly as a source of slaves, especially during the latter part of this period.67 It is extremely unlikely that any enslavement of southern Lake Chad Basin peoples during this period would have involved the Garamantes directly, and much more likely that slaving ‘subcontractors’ (peripheral groups engaged in taking slaves and selling them to commercial centres) would have been located in the archaeologically nearly-unknown areas directly north-east and north-west of Lake Chad, and thus intermediate between Fazzan and the lands south of the lake. In southern Saharan environments experiencing increasing aridification, the agricultural resources of increasingly sedentary communities further to the south would have been significant as well, so any slave raids could have been a part of broader interaction networks. In the Mandara Mountains during the historical period, seasonal slave raiding, often by ‘subcontractors’, coexisted with significant trading and exchange networks.68 It is possible that some similar relations might have existed in the southern Lake Chad Basin in earlier times. There has been a good deal of discussion about the settlement nucleation and enclosing ditches that are found at Zilum and related sites south-west of Lake Chad from c.500 BC onward.69 Before these sites were located, it was assumed that socio-political contexts had not generated large villages and defensive features for a thousand years after that time. It is probable that the processes leading to such large and complex sites are primarily endogenous, but one wonders if slave raiding from the western and north-

66 68

Breunig et al. 2008; MacEachern 2013. 67 Fentress 2011; Mattingly 2011. MacEachern 1993. 69 Magnavita et al. 2004; Magnavita et al. 2006.

https://doi.org/10.1017/9781108634311.013 Published online by Cambridge University Press

13 Burial Practices around the Southern Lake Chad Basin

western borders of Lake Chad may have played a role in their formation as well. This is certainly more likely than the immigration of masses of refugee Near Eastern nobility proposed by other researchers.70 It is quite likely that substantial contacts between Fazzan and the Niger Bend began earlier than they did to the southern Lake Chad Basin, West African gold likely providing a significant impetus to the former.71 The late first/early second millennia AD was the period in which the environs of Lake Chad begin to be mentioned in Arabic historical sources, originally involving the state of Kanim and rather quickly its trade in slaves. It is also during this period that evidence for material contacts between the southern Lake Chad Basin and communities in the Central Sahara became much more significant. It is likely that these processes, historical recognition, enslavement and trade, were linked at some level. They might, for example, reflect an increase in the strength and reach of exchange systems around Lake Chad, probably with slaves increasingly sought as one of the primary resources tradable for exotic, high-status artefacts. As noted, substantial nucleated settlements had existed south of Lake Chad from the middle of the first millennium BC, but archaeological and historical evidence for local socio-political differentiation, active participation in subcontinental exchange systems, and large-scale slave raiding does not appear for over a thousand years after that. Increased participation in these exchange networks predates any strong evidence for Islam in the regions around Lake Chad, at least; most of the societies now occupying the areas where these archaeological sites are found remained substantially non-Muslim until after AD 1500. It appears more generally that, around the Central Sahara, the expansion of exchange networks predated the influence of Islam by at least a century or two. As Haour says, ‘. . .[T]he fundamental difference between the pre-Islamic and the Islamic contexts was not one of nature, but of scale. . .’72 and the number of people involved in inter-regional exchange systems was probably very small indeed. At the same time, Haour is undoubtedly correct that participation in shared Islamic sociocultural systems acted to facilitate the expansion of these trading systems. One important element in the dynamics of trade and exchange around Lake Chad may be the role of the Garamantian polity itself, and particularly its demise. It is quite likely that the environmentally linked disappearance of the Garamantian state in the mid-first 70 71

Lange 2004; compare with Magnavita and Breunig 2008. MacDonald 2011; S. Magnavita 2009. 72 Haour 2017.

https://doi.org/10.1017/9781108634311.013 Published online by Cambridge University Press

421

422

Scott MacEachern

millennium AD opened the field to ambitious former slaving subcontractors and traders living in more amenable environments around Lake Chad to the south. These would eventually appear in history within just a few centuries as the rulers of Kanim – still raiding for slaves yet further to the south and trading into and across the Sahara to the north. The greatly increased frequency of Saharan and other exotic artefacts south of Lake Chad might then have resulted in part from the possibilities of more direct participation in sub-continental exchange networks after Garamantian control of those networks ended, as well as to increasing levels of social and political hierarchy in the area. If that were the case, we might see this change in the socio-political balance of power as the long-term continuation of the processes of aridification that led to the appearance of Gajiganna people south-west of Lake Chad more than two millennia earlier, in parallel with the relative depopulation of the southern Sahara, and thus gave rise to the complex constellation of peoples that exist in the region today.

References Austen, R. A. 1992. The Mediterranean Islamic slave trade out of Africa: attentative census. Slavery & Abolition 13: 214–48. Baroin, C. 2005. What do we know about the Buduma? A brief survey. In C. Baroin, G. Seidensticker-Brikay and K. Tijani (eds), Man and the Lake. Mauduguri: Centre for Trans-Saharan Studies,199–217. Blench, R. 1999. The westward wanderings of Cushitic pastoralists. In C. Baroin, and J. Boutrais (eds), L’homme et l’animale dans le Bassin du Lac Tchad, Paris: Institut de Recherche pour le Développement, 39–80. Bourges, C., MacEachern, S. and Reeves, M. 1999. Excavations at Aissa Hardé, 1995 and 1996. Nyame Akuma 51: 6–13. Breunig, P., Eichhorn, B., Kahlheber, S., Linseele, V., Magnavita, C., Neumann, K., Posselt, M. and Rupp, N. 2006. G(l)anz ohne Eisen: Große Siedlungen aus der Mitte des ersten Jahrtausends BC im Tschadbecken von Nordost-Nigeria. In H. P. Wotzka (ed.), Grundlegungen. Beiträge zur europäischen und afrikanischen Archäologie für Manfred K. H. Eggert, Tübingen: Narr Francke Attempto Verlag, 255–70. Breunig, P., Franke, G. and Nusse, M. 2008. Early sculptural traditions in West Africa: new evidence from the Chad Basin of north-eastern Nigeria. Antiquity 82: 423–37. Breunig, P., Garba, A. and Hambolu, M. 2001. From ceramics to culture. Studies in the Final Stone Age Gajiganna Complex of north-east Nigeria. Berichte des Sonderforschungsbereichs 268 14: 45–53.

https://doi.org/10.1017/9781108634311.013 Published online by Cambridge University Press

13 Burial Practices around the Southern Lake Chad Basin

Breunig, P. and Neumann, K. 2002a. Continuity or discontinuity? The 1st millennium BC-crisis in West African prehistory. In T. Lenssen-Erz, U. Tegtmeier and S. Kröpelin (eds), Tides of the Desert – Gezeiten der Wüste. Köln: Heinrich-Barth-Institut, 491–505. Breunig, P. and Neumann, K. 2002b. From hunters and gatherers to food producers: new archaeological and archaeobotanical evidence from the West African Sahel. In Hassan (ed.) 2002: 123–55. Breunig, P., Neumann, K. and Van Neer, W. 1996. New research on the Holocene settlement and environment of the Chad Basin of Nigeria. African Archaeological Review 13: 111–43. Brooks, N., Chiapello, I., di Lernia, S., Drake, N., Legrand, M., Moulin, C. and Prospero, J. 2005. The climate-environment-society nexus in the Sahara from prehistoric times to the present day. Journal of North African Studies 10: 253–92. Cerny, V., Fernandes, V., Costa, M., Hajek, M., Mulligan, C. and Pereira, L. 2009. Migration of Chadic speaking pastoralists within Africa based on population structure of Chad Basin and phylogeography of mitochondrial L3f haplogroup. BioMedCentral Evolutionary Biology 9: 63. Chapman, J. 2008. Meet the ancestors: settlement histories in the Neolithic. In D. W. Bailey, A. W. R. Whittle and D. Hofmann (eds), Living well Together: Settlement and Materiality in the Neolithic of South-East and Central Europe, Oxford: Oxbow Books, 68–80. Clark, J. D. and Gifford-Gonzalez, D. 2008. Adrar Bous: Archaeology of a Central Saharan Granitic Ring Complex in Niger. Tervuren: Musée Royale de l’Afrique Centrale. Connah, G. 1981. Three Thousand Years in Africa: Man and His Environment in the Lake Chad Region of Nigeria. Cambridge: Cambridge University Press. Connah, G. 1984. An archaeological exploration in southern Bornu. African Archaeological Review 2: 153–71. Craddock, P. T. and Hook, D. 1995. Copper to Africa – evidence for the international trade in metal to Africa. In D. Hook and D. R. M. Gaimster (eds), Trade and Discovery: the Scientific Study of Artefacts from post-Medieval Europe and Beyond, London: British Museum Press, 181–93. Cruciani, F., Trombetta, B., Sellitto, D., Massaia, A., Destro-Bisol, G., Watson, E., Beraud Colomb, E., Dugoujon, J. M., Moral, P. and Scozzari, R. 2010. Human Y chromosome haplogroup R-V88: a paternal genetic record of early mid Holocene trans-Saharan connections and the spread of Chadic languages. European Journal of Human genetics 18: 800–07. David, N. and MacEachern, S. 1988. The Mandara Archaeological Project: preliminary results of the 1984 season. In D. Barreteau and H. Tourneux (eds), Le milieu et les hommes: recherches comparitives et historiques dans le bassin du Lac Tchad, Paris: Editions de l’ORSTOM, 51–80.

https://doi.org/10.1017/9781108634311.013 Published online by Cambridge University Press

423

424

Scott MacEachern

Dowler, A. and Galvin E. R. (eds) 2011. Money, Trade and Trade Routes in preIslamic North Africa. London: British Museum Press. Drake, N., Blench, R. M., Armitage, S. J., Bristow, C. S. and White, K. H. 2011. Ancient watercourses and biogeography of the Sahara explain the peopling of the desert. Proceedings of the National Academy of Sciences 108: 458–62. Drake, N. and Bristow, C. 2006. Shorelines in the Sahara: geomorphological evidence for an enhanced monsoon from palaeolake Megachad. The Holocene 16: 901–11. Ehret, C. 2006. The Nilo-Saharan background of Chadic. In P. Newman and L. M. Hyman (eds), West African Linguistics: Studies in Honor of Russell G. Schuh, Columbus: Ohio University Press, 56–66. Evans, J. 2005. Memory and ordination: environmental archaeology in tells. In D. W. Bailey, A. W. R. Whittle and V. Cummings (eds), Unsettling the Neolithic, Oxford: Oxbow Books, 112–25. Fentress, E. 2011. Slavers on chariots. In Dowler and Galvin 2011, 65–71. Gronenborn, D. 1998. Archaeological and ethnohistorical investigations along the southern fringes of Lake Chad, 1993–1996. African Archaeological Review 15: 225–59. Gronenborn, D., Adderley, P., Ameje, J., Banerjee, A., Fenn, T., Liesegang, G., Haase, C. P., Usman, Y. A. and Patscher, S. 2012. Durbi Takusheyi: a high-status burial site in the western Central Bilād al-Sūdān. Azania 47: 256–71. Haour, A. 2003a. Ethnoarchaeology in the Zinder Region, Republic of Niger: the Site of Kufan Kanawa. Oxford: Archaeopress. Haour, A. 2003b. One hundred years of archaeology in Niger. Journal of World Prehistory 17: 181–234. Haour, A. 2008. The pottery sequence from Garumele (Niger) – a former KanemBorno capital? Journal of African Archaeology 6: 3–20. Haour, A. 2017. What made Islamic trade distinctive – as compared to pre-Islamic trade? In D. J. Mattingly, V. Leitch, C. N. Duckworth, A. Cuénod, M. Sterry and F. Cole (eds), Trade in the Ancient Sahara and Beyond, Trans-Saharan Archaeology Vol. 1, Cambridge: Cambridge University Press and Society for Libyan Studies, 82–100. Haour, A. and Gado, B. 2009. Garumele, ville médiévale du Kanem-Borno? Une contribution archéologique. Journal of African History 50: 355–75. Harich, N., Costa, M. D., Fernandes, V., Kandil, M., Pereira, J. B., Silva, N. M. and Pereira, L. 2010. The trans-Saharan slave trade – clues from interpolation analyses and high-resolution characterization of mitochondrial DNA lineages. BioMed Central Evolutionary Biology 10: 138. Hassan, F. (ed.). 2002. Droughts, Food and Culture. New York: Springer. Holl, A. 1988a. Houlouf I: archéologie des sociétés protohistoriques du NordCameroun. Oxford: Archaeopress. Holl, A. 1988b. Transition du Néolithique a l’Age du Fer dans la plain peritchadienne: le cas de Mdaga. In D. Barreteau and H. Tourneux (eds), Le milieu et les

https://doi.org/10.1017/9781108634311.013 Published online by Cambridge University Press

13 Burial Practices around the Southern Lake Chad Basin

hommes: recherches comparitives et historiques dans le bassin du Lac Tchad, Paris: Editions de l’ORSTOM, 81–109. Holl, A. 1994. The cemetery of Houlouf in northern Cameroon (AD 1500–1600): fragments of a past social system. African Archaeological Review 12: 133–70. Holl, A. 2001. The Land of Houlouf: Genesis of a Chadic Polity 1900 BC–AD 1800. Ann Arbor: Museum of Anthropology, University of Michigan. Insoll, T., Polya, D., Bhan, K., Irving, D. and Jarvis, K. 2004. Towards an understanding of the carnelian bead trade from Western India to sub-Saharan Africa: the application of UV-LA-ICP-MS to carnelian from Gujarat, India and West Africa. Journal of Archaeological Science 31: 1161–73. Insoll, T. and Shaw, T. 1997. Gao and Igbo-Ukwu: beads, interregional trade and beyond. African Archaeological Review14: 9–24. Klee, M., Zach, B. and Stika, H. P. 2004. Four thousand years of plant exploitation in the Lake Chad Basin (Nigeria), part III: plant impressions in potsherds from the Final Stone Age Gajiganna Culture. Vegetation History and Archaeobotany 13: 131–42. Kropelin, S., Verschuren, D., Lézine, A. M., Eggermont, H., Cocquyt, C., Francus, P., Cazet, J. P., Fagot, M., Rumes, B., Russell, J. M., Darius, F., Conley, D. J., Schuster, M., Von Suchodoletz, H. and Engstrom, D. R. 2008. Climate-driven ecosystem succession in the Sahara: the past 6000 years. Science 320: 765–68. Lange, D. 1988. The Chad region as a crossroads. In M. el Fasi (ed.), General history of Africa, III. Africa from the seventh to the eleventh century, Paris and London: Unesco and Heinemann Educational Books Ltd, 436–60. Lange, D. 1989. Préliminaires pour une histoire des Sao. Journal of African History 30: 189–210. Lange, D. 2004. Ancient Kingdoms of West Africa: Africa-centred and CanaaniteIsraelite Perspectives. Dettelbach: J. H. Röll. Lebeuf, A. M. D. 1969. Les principautés Kotoko: essai sur le caractère sacré de l’autorité. Paris: Editions du CRNS. Lebeuf, A. M. D., Lebeuf, J. P., Treinen-Claustre, F. and Courtin, J. 1980. Le gisement sao de Mdaga (Tchad): Fouilles 1960–1968. Paris: Société d’Ethnographie. Lebeuf, J. P. 1962. Archéologie tchadienne: les Sao du Cameroun et du Tchad. Paris: Hermann. Lebeuf, J. P. 1969. Carte archéologique des abords du Lac Tchad (Cameroun, Nigeria, Tchad). Paris: Editions du CNRS. Lovejoy, P. E. 2011. Transformations in Slavery: A History of Slavery in Africa. Cambridge, New York: Cambridge University Press. 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, London: Leicester University Press, 71–103.

https://doi.org/10.1017/9781108634311.013 Published online by Cambridge University Press

425

426

Scott MacEachern

MacDonald, K. C. 2011. A view from the south: sub-Saharan evidence for contacts between North Africa, Mauritania and the Niger, 1000 BC–AD 700. In Dowler and Galvin 2011: 72–82. MacEachern, S. 1993. Selling the iron for their shackles: Wandala-montagnard interactions in northern Cameroon. Journal of African History 34: 247–70. MacEachern, S. 2002. Beyond the belly of the house: space and power in the Mandara Mountains. Journal of Social Archaeology 2: 179–219. MacEachern, S. 2012a. The Holocene history of the southern Lake Chad Basin: archaeological, linguistic and genetic evidence. African Archaeological Review 29.2–3: 1–19. MacEachern, S. 2012b. The prehistory and early history of the northern Mandara Mountains and surrounding plains. In N. David (ed.), Metals in Mandara Mountains’ Society and Culture, Trenton, NJ: Africa World Press, 29–67. MacEachern, S. 2012c. Rethinking the Mandara political landscape: enslavement, climate and an entry into history in the second millennium AD. In A. Ogundiran and J. C. Monroe (eds), Power and Landscape in Atlantic West Africa: Archaeological Perspectives, Cambridge: Cambridge University Press, 309–38. MacEachern, S. 2013. Time on the timeless continent: history and archaeological chronologies in the southern Lake Chad Basin. In T. R. Pauketat and J. Robb (eds), Big histories, human lives: tackling problems of scale in archaeology. Santa Fe: School for Advanced Research Press, 123–44. MacEachern, S., Bourges, C. and Reeves, M. 2001. Early horse remains from northern Cameroon. Antiquity 75: 62–67. Magnavita, C. 2008. Studien zur endsteinzeitlichen und früheisenzeitlichen Besiedlung im südwestlichen Tschadbecken (1300 BC–700 AD). Frankfurt am Main: Goethe-Universität. Magnavita, C. Breunig, P. 2008. Facts and speculations: a reply to D. Lange’s ‘The emergence of social complexity in the southern Chad Basin towards 500 BC: archaeological and other evidence.’ Borno Museum Society Newsletter 72/75: 71–83. Magnavita, C., Breunig, P., Ameje, J. and Posselt, M. 2006. Zilum: a mid-first millennium BC fortified settlement near Lake Chad. Journal of African Archaeology 4: 153–69. Magnavita, C., Breunig, P., Ishaya, D. and Adebayo, O. 2009. Iron Age beginnings at the southwestern margins of Lake Chad. In S. Magnavita et al. 2009: 27–58. Magnavita, C., Kahlheber, S. and Eichhorn, B. 2004. The rise of organisational complexity in mid-first millennium BC Chad Basin. Antiquity 78.301: http:// antiquity.ac.uk/projgall/magnavita/index.html. Magnavita, S. 2009. Sahelian crossroads: some aspects on the Iron Age sites of Kissi, West Africa. In S. Magnavita et al. (eds.) 2009, 79–104. Magnavita, S., Koté, L., Breunig, P. and Idé, O. (eds.). 2009. Crossroads/Carrefour Sahel. Cultural Developments and Technological Innovations in 1st

https://doi.org/10.1017/9781108634311.013 Published online by Cambridge University Press

13 Burial Practices around the Southern Lake Chad Basin

Millennium BC/AD West Africa. Frankfurt: Journal of African Archaeology Monograph Series. Magnavita, S. 2017. Track and trace: archaeometric approaches to the study of early Trans-Saharan trade. In D. J. Mattingly, V. Leitch, C. N. Duckworth, A. Cuénod, M. Sterry and F. Cole (eds), Trade in the Ancient Sahara and Beyond, Trans-Saharan Archaeology Vol. 1, Cambridge: Cambridge University Press, Society for Libyan Studies, 393–413. Marliac, A., Langlois, O. and Delneuf, M. 2000. Archéologie de la région MandaraDiamaré. In C. Seignobos and O. Iyebi-Mandjek (eds), Atlas de la Province Extrême-Nord, Cameroun, Paris: Editions de l’Institut de Recherche pour le Développement, 71–76. Mattingly, D. J. (ed.). 2003. The Archaeology of Fazzan: Synthesis. London: Society for Libyan Studies and Department of Antiquities. Mattingly, D. J. 2011. The Garamantes of Fazzan: an early Libyan state with transSaharan connections, In Dowler and Galvin 2011: 49–60. Mattingly, D. J., Abduli, H., Aburgheba, H., Ahmed, M., Ali Ahmed Ismaia, M., Baker, S., Cole, F., Fenwick, C., Rodriguez, M. G., Hobson, M., Khalaf, N., Lahr, M., Leitch, V., Moussa, F., Nikita, E., Parker, D., Radini, A., Ray, N., Savage, T., Sterry, M. and Schörle, K. 2010. DMP IX: summary report on the fourth season of excavations of the Burials and Identity team. Libyan Studies 41: 89–104. Mattingly, D. J., Lahr, M. and Wilson, A. 2009. DMP V: investigations in 2009 of cemeteries and related sites on the west side of the Taqallit Promontory. Libyan Studies 40: 95–131. Podgorná, E., Soares, P., Pereira, L. and Černý, V. 2013. The genetic impact of the Lake Chad Basin population in North Africa as documented by Mitochondrial diversity and internal variation of the L3e5 haplogroup. Annals of Human Genetics 77: 513–23. Roset, J. P. 1983. Nouvelles données sur le problème de la Néolithisation du Sahara méridional: Air et Ténéré, au Niger. Cahiers ORSTOM, Serie Géologie 13: 119–42. Schuster, M., Duringer, P., Ghienne, J. F., Roquin, C., Sepulchre, P., Moussa, A., Lebatard, A. E., Mackaye, H. T., Likius, A., Vignaud, P. and Brunet, M. 2009. Chad Basin: paleoenvironments of the Sahara since the Late Miocene. Comptes Rendus Geosciences 341: 603–11. Schuster, M., Roquin, C., Duringer, P., Brunet, M., Caugy, M., Fontugne, M., Taïsso Mackaye, H., Vignaud, P. and Ghienne, J. F. 2005. Holocene Lake Mega-Chad palaeoshorelines from space. Quaternary Science Reviews 24: 1821–27. Sereno, P. C., Garcea, E. A. A., Jousse, H., Stojanowski, C. M., Saliège, J. F., Maga, A., Ide, O. A., Knudson, K. J., Mercuri, A. M., Stafford, T. W. Jr., Kaye, T. G., Giraudi, C., N’siala, I. M., Cocca, E., Moots, H. M., Dutheil, D. B. and Stivers, J. P. 2008. Lakeside cemeteries in the Sahara: 5000 years of Holocene

https://doi.org/10.1017/9781108634311.013 Published online by Cambridge University Press

427

428

Scott MacEachern

population and environmental change. PLoS ONE 3.8: e2995. doi:10.1371/journal.pone.0002995 Sindbaek, S. M. 2007. The small world of the Vikings: networks in Early Medieval communication and exchange. Norwegian Archaeological Review 40: 59–74. Smith, A. B. 2006. The Tenerian. In Clark and Gifford-Gonzales 2006, 201–43. Tishkoff, S. A., Reed, F. A., Friedlaender, F. R., Ehret, C., Ranciaro, A., Froment, A., Hirbo, J. B., Awomoyi, A. A., Bodo, J. M. and Doumbo, O. 2009. The genetic structure and history of Africans and African Americans. Science 324: 1035–44. Vernet, R. 1996. Le Sud-Ouest du Niger de la préhistoire au début de l’histoire. Niamey, Paris: IRSH, Editions Sépia. Vernet, R. 2002. Climate during the late Holocene in the Sahara and the Sahel: evolution and consequences on human settlement. In Hassan 2002: 47–64. Wendt, K. P. 2007. Gajiganna: Analysis of Stratigraphies and Pottery of a Final Stone Age Culture of Northeast Nigeria. Frankfurt: Africa Magna Verlag. Wiesmüller, B. 2001. Die Entwicklung der Keramik von 3000 BP bis zur Gegenwart in den Tonebenen südlich des Tschadsees. Frankfurt am Main: J.W.-GoetheUniversität. Wilson, A. 2012. Saharan trade in the Roman period: short-, medium- and long-distance trade networks. Azania 47: 409–49. Wynne-Jones, S. and Haour, A. 2014. Precolonial African trade networks and the ‘Small World’ paradox. Annual Meeting of the Society for American Archaeology. Austin, TX.

https://doi.org/10.1017/9781108634311.013 Published online by Cambridge University Press

part vi

Linguistic Aspects of Migration and Identity

Published online by Cambridge University Press

Published online by Cambridge University Press

14

The Linguistic Prehistory of the Sahara roger blench

Introduction The Sahara today is fairly uniform linguistically; Hassaniya and other Arabic dialects, Berber and Saharan1 (Teda and Zaghawa) cover most of the terrain from west to east. However, this state of affairs is relatively recent and reflects both the current aridity of the Sahara (which discourages ethnolinguistic diversity) and political and religious movements over recent millennia which have also tended to induce uniformity. But it would be a mistake to read this back into the past; the ‘green Sahara’ of the early Holocene and the variety of material culture uncovered by archaeologists points to both the presence of many more languages and cultures, representing phyla different from those spoken in the present. The expansion of languages such as Hassaniya and Tuareg are demonstrably recent and a prior history is there to be uncovered. The indications of a former linguistic geography are all fragmentary; inscriptions in Libyco-Berber and Punic/Latin scripts, terms without etymologies in present-day languages and the characteristics of residual populations which may reflect former distinct ethnolinguistic groups whose languages have been assimilated. Written records, iconography and archaeology can all provide general indications of this more diverse past. The ‘Tehenu’ appear in the Egyptian Fifth Dynasty sources (3200 BC) as livestock keepers of the Western Desert.2 Egyptian records speak of the incursions of the I-S-B-T-U, usually identified with the Asbytes of Herodotus, nomads from the deserts west of the Nile who raided settlements in the Eighteenth Dynasty (1543–1292 BC). Moreno García presents a re-evaluation of this account suggesting it was only at the point where 1

2

The term ‘Saharan’ contains a potential ambiguity in linguistic discussions. It refers both to the desert as an adjective and is applied to the subgroup of the Nilo-Saharan phylum which includes Kanuri, Teda-Daza and Beria. The chapter will try and make clear at each point which usage is intended. Vernet and Onrubia-Pintado 1994, 56.

https://doi.org/10.1017/9781108634311.014 Published online by Cambridge University Press

431

432

Roger Blench

there was competition for resources that the mobile herders were presented in a negative light.3 Herodotus (c.500 BC), in Book 4 of his Histories in particular, gives a long account of the geography and characteristics of the Libyan tribes, which grows more exotic with distance westward from the known terrain of Egypt and the Nile. Nonetheless, the oases of Siwa, Awjila and Garama, the capital of the Garamantes in Fazzan, can easily be recognised in his account. Beyond that, Herodotus claimed that a ten days’ march west of the Garamantes lived the Atarantes ‘who have no names of their own’. It has been speculated that the name Atarantes derives from Berber adrar (plural idaren) ‘mountain.’ The challenge is to sift out the colourful and fantastical elements and develop an interpretation which is congruent with both synchronic linguistics and archaeology. The object of this chapter4 is to draw together these fragments to present a new model for the linguistic prehistory of the Sahara and to explore how this could be correlated with the developing regional archaeology. It begins by summarising the present-day linguistic situation and the existing evidence for inscriptions. The second half of the chapter explores the significance of the ‘green Sahara’ concept for ethnolinguistic diversity and suggests that knowledge of the residual foraging groups in the Sahara provide some clues to former subsistence strategies. I discuss the importance of substrate lexicon, in particular for animals, in uncovering Berber-forager interaction in the southern Sahara. The final section synthesises the data presented and suggests avenues for further research.

The Present-Day Situation General At present, the dominant languages of the Sahara from west to east are Hassaniya Arabic, Tuareg Berber, Teda-Daza and the related Beria [=Zaghawa], Arabic again, with Beja on the Red Sea coast. Isolated pockets of Sub-Saharan languages exist within this, most notably islands of 3 4

Moreno García 2014. This chapter represents the fruition of years of discussion with scholars from a variety of different disciplines. Particular thanks to Vaclav Blažek, Nick Drake, Maarten Kossmann and Lameen Souag for unpublished data and critiques. The chapter was ably summarised at the Leicester workshop ‘Burials, Migration and Identity’ by Tim Insoll and has benefited from comment both at the workshop and subsequently.

https://doi.org/10.1017/9781108634311.014 Published online by Cambridge University Press

14 The Linguistic Prehistory of the Sahara

Table 14.1. Arabic dialects of the Sahara. Location Hassaniya Moroccan Algerian Uled Suliman Shuwa Chadian Bedawi

Mauretania and adjacent areas Morocco Algeria Libya, Niger, Nigeria Nigeria, Cameroun, Chad Chad, Sudan Egypt

Songhay, such as Tabelbala and Kanuri.5 These islands are usually thought to be relatively recent; the Kanuri for example, is almost certainly a relic of the Trans-Saharan slave trade, dating from the medieval period. NonTuareg Berber languages such as Zenaga, Tetserret, Awjila and the extinct al-Fuqha, form pockets within Arabic and other Berber.6 Figure 14.1 shows the overall current linguistic situation.7

Arabic The Arabic spoken in the Sahara is divided into a number of dialect groups as shown in Table 14.1. Hassaniya Arabic (‫)ﺣﺴﺎﻧﻴﺔ‬, the language of the Moors, is a conservative Arabic dialect originally spoken by the Beni Hassān, who are thought to have expanded from North Africa from the fifteenth century onwards.8 It is likely that it assimilated a largely Berber population, of which Zenaga is the last remaining island.9 Although the Hassaniya must have interacted with other in situ populations, there is little evidence in unidentified loan-words for the genetic affiliation of these populations. Shuwa Arabic is the language of the westernmost pastoral nomads who arrived from Sudan in the Kingdom of Kanim in the thirteenth century.10 Their incursions, which seem to have been very violent, are the subject of a letter of complaint from the Shehu of Kanim, preserved in the Cairo Archives.11 The Shuwa must have intermarried extensively 5 6 7

8

Tabelbala: Souag 2010; Kanuri: Fuchs 1961. Zenaga: Tayne-Cheikh 2008; 2010; Tetserret: Lux 2013. The map represents the situation as it was c.AD 2000. However, the political upheavals in the past decade may well have changed the linguistic geography of the Sahara yet again, with the Songhay in particular fleeing many settlements. 9 Taine-Cheikh 1979. Al-Chennafi and Norris 1981. 10 Zeltner 1970. 11 Walz 1978.

https://doi.org/10.1017/9781108634311.014 Published online by Cambridge University Press

433

434

Roger Blench

Figure 14.1. Map of the Sahara showing the overall current linguistic situation.

with indigenous populations as today they cannot be distinguished in terms of physical appearance. It is likely that the Arabic they speak is continuous with standard Chadian Arabic and with the Arabic of the Baggara cattle nomads of Sudan.12 The Uled Suliman is a clan of camel herders who were originally found in southern Libya. However, at some point in the twentieth century they began to leave Libya for Niger and then to migrate to north-east Nigeria, where they are now well established.

Berber Mainland Languages The Berber languages constitute a major branch of the Afroasiatic language phylum and are spoken both by settled and nomadic populations along the North African coast and far down into the Sahara, presently reaching the borders of Nigeria. Today, Berber languages are confined to a series of islands surrounded by Arabic except where they touch Sub-Saharan African languages (Fig. 14.2). This is a considerable geographical range, but it has been regularly argued that Berber culture and by implication, people, reached as far as the Nile confluence.13 Bechhaus-Gerst claimed to detect loans from Berber into Nubian and Behrens adduced cultural evidence from rock paintings, etc. 12

De Pommerol 1997.

13

For instance, Behrens 1989.

https://doi.org/10.1017/9781108634311.014 Published online by Cambridge University Press

14 The Linguistic Prehistory of the Sahara

Figure 14.2. Map of the Sahara showing distribution of Berber languages.

Such a stretch is not inconceivable geographically, but the evidence for this remains weak, both linguistically and archaeologically.14 Nonetheless, Berber must at one point in the past have been the dominant population throughout much of North Africa and the Sahara.15 Although the Tuareg are presently the most widespread group, found across much of Algeria, Niger and southern Libya,16 their expansion is probably relatively recent as they may have entered the south-central Sahara as late as the sixth century AD.17 The Berber remain a highly mobile group, the Tuareg in particular. For example, they have been forming new communities in the coastal cities of West Africa and are adept at maintaining a strong media presence. The Zenaga in south-west Mauretania were a significant group when first described, but are now down to some 300 speakers.18 North of Agadez in Niger live the Tetserret, whose language shows correspondences with Zenaga and who are now encapsulated by the Tuareg.19 Other islands of Berber speakers occur within the Arabic-speaking zone further east, most notably at Awjila (‫)ﺃﻭﺟﻠﺔ‬, formerly at al-Fogaha and Siwa.20 Furthermore, it is often claimed that Berbers reached the Canaries at an unspecified date in the past, leading to the formation of the Guanche, the now-vanished aboriginal population.21 14 15 16 18 19 20 21

Bechhaus-Gerst 1989; cf. The negative evaluation in Kossmann 2013. Basset 1952; Camps 1980; Willms 1980; Ameur 1990; Brett and Fentress 1996; Blench 2001. Bernus 1981. 17 Camps 1974. Faidherbe 1877; Masqueray 1879; Basset 1933; Dubié 1940; Nicolas 1953; Taine-Cheikh 2008. Attayoub 2001; Lux 2013. Awjila: Paradisi 1960; Putten 2013; al-Fogaha: Paradisi 1963; Siwa: Laoust 1932. Wölfel 1965.

https://doi.org/10.1017/9781108634311.014 Published online by Cambridge University Press

435

436

Roger Blench

Despite an abundance of information, there are a series of major unanswered questions about the affiliations, origins and date of diversification of the Berber languages.22 Berber is Afroasiatic, and its nearest relative is likely to be Semitic. Yet when deep-level Arabic borrowings are weeded out, it retains only a very small corpus of established Afroasiatic roots, pointing to a ‘long tail’, a split from Afroasiatic at quite some time-depth. When and where this took place is highly uncertain. Similarly, the dates of the primary expansion of Berber are problematic; its extremely low internal diversity points either to a recent epoch or to an episode of language levelling. Evidence from Neo-Punic and Latin borrowings suggests a date for Proto-Berber of AD 100–200. Under no circumstances can Berber be identified with the Capsian (c.11000–9000 BC) or even the first stages of the Neolithic in North Africa (c.6000 BC onwards), both of which are far too early to be reconciled with Berber internal diversity. If this is indeed so, what process is in consilience with the archaeological record? The only way to account for the distinctiveness of Berber is to suppose that the speakers of the proto-language must have been resident somewhere for a long period, diverging from Afroasiatic but not diversifying internally.23 At a point in the more recent past, a social or economic change must have transformed their society, stimulating a major expansion. I have argued that this was pastoralism, on the basis that a quite detailed lexicon of livestock-keeping can be reconstructed for Proto-Berber.24 This ought to correspond to the expansion of pastoralism across the Central Sahara, which occurred around 4000–3000 BC.25 The earliest dates for cattle in Africa are debated because it is difficult to be sure that skeletons represent domesticated species. Wild cattle existed in north-east Africa, and by the time of Nabta Playa, that is c.8000 BC, they may have been managed by humans.26 Di Lernia has radiocarbon dated a large number of cattle burials in the Massak in southern Libya, and they give a fairly consistent suite of dates pointing to the introduction of livestock c.6000 BC.27 Bones of small ruminants also occur in these burials, together with occasional other species such as equids (presumably wild ass). These dates are difficult to reconcile with the lack of diversity within Berber and there are three possible explanations: 22 23

24

Chaker 2002; Galand 1971. Alternatively, they could have diversified, but the branches that developed at that period were then assimilated by other languages. Blench 2001. 25 Brooks et al. 2009. 26 Gautier 1984; 1987; 2001. 27 di Lernia 2006.

https://doi.org/10.1017/9781108634311.014 Published online by Cambridge University Press

14 The Linguistic Prehistory of the Sahara

1. either the early wave of pastoral expansion was a quite different group of people who have vanished without trace; 2. or Berber has diversified extremely slowly compared with other world language families; 3. or Berber underwent a major episode of language levelling around AD 100–200, eliminating prior diversity. Postulating mystery populations is bad science, and the most likely explanation is that the expansion of pastoralism across the Sahara was indeed a primary migration of Berbers into the desert. The closeness of Berber lects is only explicable if we postulate a significant episode of language levelling, which judging by the distribution of Latin loan-words could have been as late as AD 200. The similarities between even the far-flung branches of Berber, Zenaga and Siwa, date from this period. This co-occurs with the establishment of the Roman limes in North Africa, suggesting that the concentration of population this induced stimulated the spread of a prestige dialect. It seems likely that residual populations remained after this event, but a third wave of expansion and assimilation occurred with the Tuareg dispersal, from perhaps the sixth century onwards, which in turn eliminated other more archaic and diverse Berber lects, such as the relatives of Tetserret. Finally, the Hassaniya expansion from the fifteenth century further pushed back the south-western Berber lects and precipitated the long decline of Zenaga and Azer, as well as probably displacing the original language of the Nemadi and the Imraguen. The idea that Berber underwent a language-levelling episode to account for its lack of internal diversity merits further exploration. Kossmann28 points out that there are well-assimilated Latin loans associated with agricultural terminology which are part of Proto-Berber, as well as numerous Arabic borrowings that are found virtually across its range.29 Latin loans must surely date to the period of the Roman conquest of North Africa, for example, after the end of the Third Punic War, 146 BC onwards. This points to a series of levelling episodes within Berber, facilitated by the mobile culture of the nomads, but equally important to farmers, to judge by the importation of the plough and various species of fruit tree. There was thus a strong pressure to keep communication systems open, which may in turn be associated with the importance of the camel from the first century AD onwards.30 Two other chapters in this volume, Ehret (Chapter 15) and Fentress (Chapter 16) approach the age and diversification of Berber very 28

Kossmann 2013

29

Kossmann 1999; 2013.

30

Brogan 1954; Shaw 1979.

https://doi.org/10.1017/9781108634311.014 Published online by Cambridge University Press

437

438

Roger Blench

differently. Ehret’s model of Berber is based on cognation counts, thus a variant of lexicostatistics, and glottochronology, that is, calibrating age with lexical divergence. Fentress largely accepts this chronology and then proposes an interpretation of the archaeology which is consistent with them. The dates proposed by Ehret are significantly older than those in the present chapter. While I believe, as do these authors, that Berber must indeed have split from the Afroasiatic phylum very long ago, reconstruction based on synchronic evidence points to a much more recent date. Neither Ehret nor Fentress engage with the historical linguistics of Berber, such as the painstaking work of Kossmann on Latin and Punic borrowings. Unless there is a convincing account of these aspects of Berber, it is hard to accept these early dates. This is not to discount the archaeological evidence marshalled by Fentress, which may correlate with the pre-Berbers (whose language disappeared with the levelling episode proposed here).

The Peopling of the Canaries When the Spanish first reached the Canary Islands in the sixteenth century, they found the inhabitants were a people called the Guanche, with four dialects spread across seven main islands. So entrancing were their dances that these were carried back to Europe and entered the repertoire of classical music, hence the ‘canaries’ in the harpsichord suites of J. S. Bach. Unfortunately, the fate of the language was less iconic, as by the end of the eighteenth century it had disappeared, with the speakers killed, dying from disease or being assimilated. Archaeology currently points to an initial settlement from the North African mainland around 1000 BC.31 There are many perplexing aspects of the culture of the Canaries. The most notable is the presence of mummified bodies, which use the specific techniques associated with Ancient Egypt. Similarly, there are small earth constructions that look very like attempts to replicate pyramids. Whether this implies the ancestors of the Guanche were carried to the Canaries by Egyptian ships remains a point for speculation. The records of Guanche are only those recorded by travellers and amateur enthusiasts. Unfortunately, little of the vocabulary is basic (there are, for example, almost no records of body parts), and it is impossible to identify the affiliation of Guanche unambiguously. The classic synthesis, by Wölfel, noted many similarities with Berber and a much smaller number 31

Milburn 1988; González and Tejera Gaspar 1990.

https://doi.org/10.1017/9781108634311.014 Published online by Cambridge University Press

14 The Linguistic Prehistory of the Sahara

with Basque.32 A persistent history of theories relating Basque to African languages, particularly Berber, goes as far back as Gabelenz, Wölfel and Mukarovsky.33 Most researchers who have looked at the fragmentary records of Guanche have identified it as a branch of Berber.34 Moreover, the very short inscriptions on rock in the Canaries which are in the old North African Numidian script further confirm the Berber link. We cannot say for certain that Guanche was originally a Berber language, but that there was a major influx of Berbers, which introduced the pastoral culture of North Africa and transformed the languages. Roman contact is now also demonstrated and it has been argued that the later inscriptions show familiarity with the principles of Latin writing.

Saharan (Kanembu, Teda-Daza and Beria) The only virtually undisputed languages of early residents of the desert that are still spoken are the Saharan languages. Saharan consists of a group of four languages, spoken in the region between Lake Chad and the border of Sudan. Saharan was first outlined as a subgroup of NiloSaharan by Lukas and consists of Kanuri, Teda, Beria (=Zaghawa) and the now extinct Sagato (=Berti).35 The Saharan languages subgroup is shown in Figure 14.3.

Kanuri-Kanembu Kanuri-Kanembu is spoken principally around Lake Chad today, but islands of Kanuri occur further north in the Sahara, most notably in the oasis of Fachi.36 Such islands are thought to date from the Trans-Saharan slave trade, although there is no clear evidence for this and these Kanuri outliers have not been studied linguistically. The Kanembu are found principally in communities east and north of Lake Chad, while Kanuri is situated within Nigeria. The Turkish physician, Evliya Çelebi (‫)ﺍﻭﻟﻴﺎﭼﻠﺒﻲ‬, 1611–1682, collected information about Kanim and the arrival of the yearly slave caravans in the late seventeenth century.37 Çelebi recorded two languages, Bornavi and Maiburni, both essentially Kanuri, from his Cairo informants.

32 35 37

Wölfel 1965. 33 Gabelenz 1894; Wölfel 1955; Mukarovsky 1964; 1969. 34 Galand 1988. Lukas 1952; for Sagato, see Petráček 1987. 36 Fuchs 1983. Çelebi 1995/2007. Habraszewski 1967 summarises all the information that can be gathered from Çelebi’s account of Borno.

https://doi.org/10.1017/9781108634311.014 Published online by Cambridge University Press

439

440

Roger Blench

Figure 14.3. Saharan languages subgroup.

Teda-Daza The Teda-Daza or Tubu inhabit the Tibesti mountains of Northern Chad and adjacent regions of Niger.38 There is every reason to consider these are long-term residents of the region. It is often supposed that these were the Ethiopian Troglodytes referred to by Herodotus, ‘who speak a language different from all other peoples, which resembles the cry of the bat’ and who were chased by Garamantian slavers in four-horse chariots, a technology independently confirmed by rock paintings.39

Beria (Zaghawa) The Beria (Zaghawa) live east of the Teda-Daza and may also have been present for a long time as they are mentioned in the Arab geographers.40 Al-Ya’qubi, in a description written around AD 890, refers to the Zaghawa in Kanim. References to the Zaghawa (=Beria) have a very long pedigree in the sources. The first modern account of the Beria is in MacMichael, written from the Sudan side of the border, while Chalmel provides the first extended account from the Francophone side.41 Since then, the anthropologists Joseph and Marie-Jose Tubiana have actively documented Beria culture.42 The Ounia, a residual population around Lake Ounianga in north-east Chad, speak a dialect of Beria.43

Songhay The Songhay languages are spoken along the Niger between Timbuktu and Gao, stretching into the Sahara of Niger and south and east into Benin Republic and Nigeria (Fig. 14.4). Songhay speakers are also known from 38 39 41 42

Cline 1950; Chapelle 1957; Kronenberg 1958; Baroin 1985. Herodotus, Histories 20; Lhote 1985. 40 Le Rouvreur 1989. MacMichael 1912; Chalmel 1931. For example, Tubiana 1964, 1985; Tubiana and Tubiana 1977.

https://doi.org/10.1017/9781108634311.014 Published online by Cambridge University Press

43

Fuchs 1961.

14 The Linguistic Prehistory of the Sahara

Figure 14.4. Map with distribution of Songhay languages.

Sudan, remnants of the pilgrimage to Mecca.44 An isolated population, the Kwarandzyey of Tabalbala (‫)ﺗﺒﻠﺒﺎﻟﺔ‬, live in a small community on the Moroccan-Algerian borderland.45 Songhay is often treated in earlier literature as if it was a single language, but it is now recognised as a complex cluster.46 Fortunately, much new, high-quality data has become available on Songhay lects and it is now possible to provide a brief account of the evolution of the language family.47 Songhay is undoubtedly Nilo-Saharan, as it shares a significant number of basic lexemes with remote geographical languages which are neither Afroasiatic nor Niger-Congo in origin. But it is distant from other NiloSaharan branches and suggests that either pre-proto-Songhay was spoken in an isolated community with little differentiation or that the relatives of proto-Songhay were subsequently assimilated by other languages. Songhay 44 46

47

Abu Manga 1995. 45 Souag 2010. The first study of the varieties of Songhay and its internal relationships is Nicolai 1981. This was a valuable beginning, but unfortunately Nicholai has followed up his initial work with a series of increasingly ill-founded claims which have had the effect of confounding the situation – for instance, the assertion in Nicholai 1990 that Songhay is a Berber creole has not gained any adherents. Nicholai 2003 is a voluminous tome which promises an in-depth account of these processes, but only serves to further confuse the linguistic issues. For example Heath 1998a; 1998b; Souag 2012.

https://doi.org/10.1017/9781108634311.014 Published online by Cambridge University Press

441

442

Roger Blench

appears to have come under strong Mande lexical and grammatical influence (perhaps specifically from an ancestor of Soninke) at an early stage in its evolution.48 At the same time proto-Songhay was diverging, it must also have been in contact with Berber influence, to judge by a small number of Berber borrowings in early or proto-Songhay.49 Songhay is traditionally divided into two primary subgroups, Northern and Southern, but a revised classification is now proposed by Souag.50 According to this hypothesis, the first split within the family was between Eastern Songhay, probably spoken around Gao, and North-western Songhay, somewhere further north; it was followed by a more prominent split between Western and Northern Songhay. Eastern Songhay is close to being a dialect continuum, although a handful of extra-riverine varieties at Hombori and Kikara in Mali and Djougou and Kandi in Benin show greater divergence. Western Songhay, spoken around Timbuktu and at Jenné, remained in fairly close contact with Eastern, a situation intensified not just by the ease of riverine trade but also by the Songhay Empire’s conquest of Timbuktu; as a result, words attested only in Eastern and Western Songhay can be securely reconstructed only for proto-Eastern. The divergence of the Northern Songhay split with the rest has been far more complete, thanks to its speakers’ dispersal in oases and desert areas dominated by speakers of Berber and Arabic.51 Figure 14.5 shows the internal structure of the Songhay languages using this new model. Two varieties (Tadaksahak and Tagdal) are spoken by nomadic groups; since agricultural vocabulary is reconstructible for proto-Northern Songhay, these are likely to result from a later change of lifestyle or language shift. The dispersal of Kwarandzyey as far as Tabalbala, 1000 km north of the rest of the family, has not yet been fully explained, but appears to be linked to the Trans-Saharan trade. Tabalbala was an important halt on a trade route linking Morocco and the Sahel. Souag points out that Songhay terms related to agriculture, such as ‘hoe’ (kumu), ‘sow seed’ (dzʊm) and ‘irrigated garden’ (ləmbu) are retained in Kwarandzyey.52 This suggests that their ancestors were brought there as horticulturalists, whether as slaves or through some other type of contract relationship.53 As with Berber, the puzzle of Songhay is that it is remote from other branches of its parent phylum, Nilo-Saharan, yet all its lects are very close to one another, implying a relatively recent dispersal. Souag observes that 48 52

Creissels 1981. Souag 2012.

49 53

Souag 2010. 50 Souag 2012. See also Souag 2015.

https://doi.org/10.1017/9781108634311.014 Published online by Cambridge University Press

51

Wolff and Alidou 2001.

14 The Linguistic Prehistory of the Sahara

Table 14.2. Proto-Songhay terms associated with urbanism (adapted from Souag 2012, 201). Gloss

Songhay

villages or town house room kitchen key clay

*ko(y)ra *hugu *gar *fuutay *karkabu *laabu

*indicates regular construction.

Figure 14.5. Internal structure of the Songhay languages.

lexemes concerned with both livestock and agriculture reconstruct in proto-Songhay, as well as words reflecting urbanism, such as ‘kitchen’ and ‘key’ (Table 14.2).54 This implies that it was the development of urbanism which kick-started the expansion of Songhay as we see it today and it is reasonable to associate it with the incipient urbanism occurring in the archaeological record from 54

Souag 2012.

https://doi.org/10.1017/9781108634311.014 Published online by Cambridge University Press

443

444

Roger Blench

c.200 BC. Park describes the ceramic phases which mark a transition to urbanism in the Timbuktu area.55 There is considerable evidence for a shift from aquatic subsistence to cereal agriculture and livestock. Importantly, none of the significant terms reconstructed for either architecture or livestock in proto-Songhay are Berber or Arabic loans. In subsistence terms, the Songhay apparently split into those who developed the urban motif and spread east and south as traders and those who adopted cattle pastoralism from their Berber neighbours and moved into the arid zones around Agadez. They became the ‘Wangarawa’ who traded gold and other products with the Hausa in the early Middle Ages, marked by a significant number of borrowings from Songhay into Hausa.56 The surveys reported in the Niger River Valley on the border of Benin and Niger indicate proto-urban settlement formation from around AD 400.57 Given this whole region is still Dendi-speaking, there is little doubt that these sites mark the first expansion of the Songhay along the Niger River.

Niger-Congo Azayr or Azer, once spoken in Wadan and Tichitt, was a form of Soninke, a Mande language, partly restructured via Berber. The status of the language is hard to gauge. Monteil reports that it was disappearing and TaineCheikh says that it was nearly extinct.58 Cleaveland, in his historical study of Walata, indicates that Azer place names still are used in the ‘upper’ town, although the language itself has gone.59

Inscriptional Languages Libyco-Berber Apart from Latin in North Africa and recent Arabic, the principal script found in the Sahara is the Berber script.60 Its modern form, Tifinagh (as in Neo-Tifinagh), is the written form of Tuareg and is still in use. Indeed, Tifinagh has undergone something of a renaissance as a consequence of its use in literacy materials in Mali. Older forms of the script, known as Libyco-Berber, occur spread across from the Central Sahara to Western Morocco, with fragmentary texts in the Canaries. The first inscriptions 55 58

Park 2010. 56 Skinner 1996. 57 Haour et al. 2011. Monteil 1939; Taine-Cheikh 1979. 59 Cleaveland 2002.

https://doi.org/10.1017/9781108634311.014 Published online by Cambridge University Press

60

O’Connor 1996.

14 The Linguistic Prehistory of the Sahara

Figure 14.6. Bilingual Latin/Berber inscription from Kef beni Fredj, Algeria (after Gsell 1933).

occur in the third century BC and continue through to the third century AD, but almost all texts are disappointingly short, hence the limited contribution of epigraphy to Berber history.61 The eastern variant was used in what is now Constantine, the Aurès region and Tunisia. Since 22 letters out of 24 can be transliterated and there are several Numidian bilingual inscriptions in Libyan and Punic, it is the best-deciphered variant.62 Figure 14.6 shows a bilingual Latin/Berber inscription from Kef beni Fredj (Algeria), which gives an idea of how these inscriptions can be transliterated.63 The western variant that was used along the Mediterranean coast from Kabylie to the Canary Islands has 61 63

Le Quellec 2011. 62 Zyhlarz 1932. CIL VIII, 5209: AE 2005; 1692: ILAlg 1, 137; Gsell 1933.

https://doi.org/10.1017/9781108634311.014 Published online by Cambridge University Press

445

446

Roger Blench

13 extra letters, which has made decipherment more uncertain.64 It is entirely possible that the underlying language is different from the Eastern inscriptions. One of the most problematic aspects is the language and inscriptions attributed to the Garamantes. The people generally known as Garamantes in classical records were placed by Pliny a 12 days’ journey from the Awjila, in the interior of Libya.65 They occupied the most habitable region of the Sahara: the Wadis al-Ajal and ash-Shati and the oases from Murzuq to Zuwila. At a later date, lbn Khaldun stated that Germanah (Jarma) was first settled by the Lauta or Luwwatah tribe, who also inhabited the coastal regions of Tripolitania.66 Sites in the vicinity of Jarma, the Garamantian capital of what is now known as Fazzan, have abundant inscriptions (Fig. 14.7).67 They are found cut or painted on dark grey amphorae, in the tombs of Garamantian cemeteries, such as those of Saniat bin Huwaydi.68 A recent project under the auspices of the British Library has digitised most of the known inscriptions and these are described in Biagetti et al.69 Although the inscriptions are in Berber characters, only some are decipherable. Various reasons for this have been suggested; either the messages were deliberately coded, so that only specific readers could understand them. Alternatively, they may have had a ‘ludic’ nature. The most exciting possibility is that they were in a non-Berber language, perhaps NiloSaharan or something unknown. Transliterated inscriptions for specialists to analyse are a high priority.

Punic Punic is an extinct Semitic language spoken in the overseas Phoenician Empire in North Africa, which included Carthage and some Mediterranean islands.70 The Phoenicians originated in what is now Lebanon and created a sea-borne empire. Carthage was founded c.800 BC and destroyed in 146 BC, but the language continued to be spoken until around the sixth century AD. It is known from inscriptions (most of them religious formulae) and personal name evidence (Fig. 14.8).71 The play Poenulus by Plautus contains a few lines in spoken Punic which has 64 65 66 67 68 71

Février 1965. On the ancient source references to the Garamantes, see Mattingly 2003, 76–90. Pliny Naturalis Historie 5.26. Daniels 1975; Mattingly 2003, 317–24; 2010, 346–50; 2013, 517–18. Mattingly 2010, 232–36. 69 Biagetti et al. 2012; 2015. 70 Segert 1976. Jongeling and Kerr 2005.

https://doi.org/10.1017/9781108634311.014 Published online by Cambridge University Press

14 The Linguistic Prehistory of the Sahara

Figure 14.7. Garamantian inscriptions in Berber characters from the Wadi al-Ajal (after Mattingly 2010, Fig. 6.8).

provided key evidence for its transliteration, because vowels are represented.72 A series of late trilingual funerary texts found in the Christian catacombs of Sirte, Libya are in Classical Greek, Latin and Punic. Al-Bakri (c.AD 1014–1094), an Andalusian Muslim geographer, describes an unknown language spoken in Sirte in the tenth century, so it is conceivable Punic survived as a spoken language into the medieval era. Part of the interest of Punic is that there are identifiable borrowings in Proto-Berber, which implies that the culture of Carthage preceded the break-up of Berber.73 In addition, the nature of the loans provides useful 72

Sznycer 1967.

73

Vycichl 1952.

https://doi.org/10.1017/9781108634311.014 Published online by Cambridge University Press

447

448

Roger Blench

Figure 14.8. Punic inscription from Carthage (Carthage Museum).

information on the interaction of the two cultures. Due in part to the destruction of Carthage’s libraries following the Third Punic War (149– 146 BC), records of the language are mainly in later neo-Punic.74 Examples of Punic loans into Berber are reviewed in Malášková and Blažek and Kossmann and are shown in Table 14.3.75 Kossmann notes the absence of Punic loans in Zenaga, the Berber language of southwest Mauretania, which may reflect the lifestyle of the desert nomads (most Punic loans are nouns associated with settled life), but possibly showing the migration of the Zenaga prior to the period of significant interaction.76

Former Ethnolinguistic Diversity The ‘Green Sahara’ The current aridity of the Sahara is not a good guide to its past status, as in the early Holocene (c.11,000 BC) it was a lush humid area with a chain of lakes and waterways connecting West to North Africa.77 This would have created both a ‘corridor’ for savannah and aquatic species to diffuse and a major expansion of resource opportunities for both Sub-Saharan and North African foragers. The waterways probably both drew populations from further east with specialisations such as crocodile and hippo hunting and attracted Maghribian foragers armed with bows and arrows 74 77

Kerr 2010. 75 Kossmann 2013, 58; Malášková and Blažek 2012. Drake et al. 2011.

https://doi.org/10.1017/9781108634311.014 Published online by Cambridge University Press

76

Kossmann 2013, 59.

14 The Linguistic Prehistory of the Sahara

Table 14.3. Punic borrowings into Proto-Berber (adapted from Malášková and Blažek 2012). Gloss

Neo-Punic

Phoenician š.q.d.m q.š.’ Phoenician z.t b.ʃ.l q.n n.ħ.š.t g.d.r Phoenician q.r.t n.r lamp, candelabrum To move, remove g.l.y To read q.r.’

Almond Cucumber Olive Onion Reed Bronze Fortified camp City Oil-lamp

Proto-Berber *ā-sāyīd *ā-y[i]ssīm *ā-zātīm *ā-b[i]zālīm ˙ *ā-yānīm ?*ā-niHās *ā-gādīr *yarat *ē-niHir əgəl *aɣriH

Berber gloss if different melon

copper ? To be polite

Tamasheq ‘to go away’

southwards as large mammals began to migrate north along these networks of rivers and lakes. Linguistic evidence suggests that the aquatic specialists and wielders of bone harpoons could have been Nilo-Saharan speaking. This period is too early for the spread of cattle-keeping and indeed there is no linguistic evidence for the reconstruction of livestock terms to this level in Nilo-Saharan.78 Brooks et al. have plotted the spread of cattle keeping across the Sahara and the pattern is clear.79 The oldest sites are in the Nile Valley (8000?-6000 BC) and the youngest sites in Mauretania (3000–2000 BC).80 di Lernia reports foragers corralling Barbary sheep between 7500 and 6500 BC at some Libyan sites.81 Populations must have moved in from north, south and east and they are likely to have been both linguistically and culturally diverse, something particularly suggested by the Gobero site in Niger.82 The linguistic affiliation of the North African forager populations who came south is difficult to establish as they probably represented a language phylum or phyla now vanished. The evidence for this is an expansion of microlithic ‘Ounanian’ points in west-central Sahara from c.9000 BC contributing to the ‘Epipalaeolithic’ of Northern Mali and Southern Algeria.83 These populations are called ‘Paleoberber’ in the literature, but there is no evidence they spoke a language in any way connected with modern Berber. 78

79 83

Although there is a strong association between cattle and the subsequent expansion of a major subgroup of Nilo-Saharan, the Eastern Sudnaic languages (Dimmendaal 2007). Brooks et al. 2009, 919. 80 di Lernia 2013. 81 di Lernia 2001. 82 Garcea 2013. Vernet et al. 2007.

https://doi.org/10.1017/9781108634311.014 Published online by Cambridge University Press

449

450

Roger Blench

Figure 14.9. Residual populations among modern Niger-Congo.

As they spread southwards they transferred the bow and arrow to a population which eventually became modern Niger-Congo. Although these foragers were broadly assimilated, there appear to be relic populations from this era; across the region there are remaining groups of foragers, often socially excluded, who maintain an archaic lifestyle. None of these retain a distinct language, but there is anecdotal evidence for unusual vocabulary relating to subsistence. Unfortunately, none of this is presented to modern linguistic standards. The next section summarises what is known about these populations and Figure 14.9 shows their locations.

Residual Populations Nemadi The Nemadi [=Ikoukou] are a small group of hunters who migrate between eastern Mauretania and adjacent parts of Mali.84 They speak a dialect of Hassaniya Arabic, or in some cases perhaps Azer. Arnaud reported that around Tichitt the Nemadi speak ‘Azeïr’ and the name Nemadi is an Azer expression meaning ‘master of dogs’.85 Brosset says that they speak Hassaniya, and that ‘their special vocabulary does not consist of vocables different from Hassaniyya, but of technical terms which need has forced them to create, which are forged from Arabic, Zenaga and maybe Azer’.86 According to Gerteiny, they speak ‘their own dialect, probably a mixture of Azêr, Zenaga and Hassaniyya, called Ikôku by the Moors’.87 It is more than 84

85

Anon n.d., Fondacci 1945; Fortier 2004; Gabus 1952; Laforgue 1917; Hermans 2013; TaineCheikh 2013. Arnaud 1906. 86 Brosset 1932. 87 Gerteiny 1967.

https://doi.org/10.1017/9781108634311.014 Published online by Cambridge University Press

14 The Linguistic Prehistory of the Sahara

possible that Nemadi preserves archaic terminology relating to hunting and other subsistence strategies.

Imraguen The Imraguen language is spoken by some 500 people in the Banc d’Arguin National Park on the Atlantic coast of Mauritania.88 The Imraguen are an isolated and endogamous fishing population whose origins and antiquity are unknown. The name Imraguen is Berber for ‘fishermen’. Ethnographic accounts suggest they have a knowledge of the sea and fishing techniques quite unlike fishing populations further north.89 According to Gerteiny, it is ‘a strange version of Hassaniyya restructured on an Azêr base’.90 Fortier says the Imraguen speak the ‘same’ language as the Nemadi, that is Hassaniya Arabic.91

Dawada The Dawada (Duwwud, Dawwada) are a distinctive, dark-skinned ethnic group of Fazzan, southern Libya, first reported by Hugh Clapperton, who visited them in the 1820s.92 They formerly lived around the Gabr’Awn oasis where they harvested brine shrimp in the salty lakes, drying and selling them to caravans.93 Their name, Dawada, means ‘worm folk’ in Arabic. An endogamous group, they were dispersed by the Qaddafi regime in the 1980s. No information is available on the likelihood that their language contains unetymologisable lexemes.

Ounia The Ounia people live around Lake Ounianga in north-eastern Chad. They were visited by Fuchs and there are scattered reports of them since including a field visit by the Tubianas.94 They are said to speak a variety of Beria (Zaghawa), although this is far from certain. There are persistent rumours that they speak a ‘secret’ language, but no data or confirmation are forthcoming. 88 89

90 93

Trotignon 1981. Lotte 1937; Anthonioz 1968; Maigret and Abdallahi 1976; Trotignon 1981; Maigret 1985; Topper 1993; Worms and Ould Eida 2002; De Noray 2006. Gerteiny 1967. 91 Fortier 2004. 92 Clapperton 1829. Limouzin 1951; Bellair 1951; Pauphilet 1953; Wellard 1964. 94 Fuchs 1961.

https://doi.org/10.1017/9781108634311.014 Published online by Cambridge University Press

451

452

Roger Blench

Haddad A population which has remained mostly unknown until recently is the Haddad, foragers who live among the Teda-speaking populations of the desert zone of Chad.95 In 1963, the Haddad were still largely foragers, deploying both nets and bows and arrows, and having a culture largely based around hunting techniques. Le Quellec and Civrac point out that net hunting is widely attested in Saharan rock art, as well as elsewhere in SubSaharan Africa.96 The Haddad were not a unitary group, with varied subgroups focused on specific techniques in different regions. Nonetheless, even in the 1960s, declining numbers of game meant that more were switching to livestock herding or craft. It is unclear whether any Haddad today still maintain a foraging lifestyle, but insecurity in Chad over decades and the spread of high-power weapons does not support a very optimistic prediction. Haddad is also a name for ‘blacksmith’ and there is a strong relationship between the caste of blacksmiths who live among the Teda and the foragers. A key question, however, is whether the Haddad are genuine survivors of a Late Stone Age–foraging tradition, or are a casted group which has adapted to foraging. Nicolaisen considers this at some length, but ultimately the question is not easily answered, because Haddad traditions are internally contradictory. It may be that their origins are diverse and that they have been brought together as a cultural category, something like the Mikea foragers of Madagascar.97

Substrate Words in Sahara Languages Berber Evidence The date of the Berber presence in the Sahara is debated, but in terms of the overall peopling of the region, it is relatively recent. Both the Zenaga and the Tuareg migrants must have encountered forager populations already in situ. One of these was undoubtedly Songhay, because of the evidence for intensive interaction between Songhay lects and Berber and the existence of ‘mixed’ speech-forms. South Berber borrows the word for ‘hippo’ from Songhay, for example. However, it seems there were also likely to have been other very different languages, because Tuareg has apparently adopted lexical items for significant flora and fauna from 95

96

Nicolaisen 2010. The monograph reflects the situation of the Haddad in 1963 although the text in question was written up by Ida Nicolaisen, based on her husband’s field notes. Le Quellec and Civrac 2010; Lindblom 1925. 97 Tucker 2003.

https://doi.org/10.1017/9781108634311.014 Published online by Cambridge University Press

14 The Linguistic Prehistory of the Sahara

Table 14.4. Comparative mammal lexicon in South Berber (Data for Zenaga, Tayne-Cheikh 2008 and Sudlow 2009; for Tuareg, Heath 2006; for Tetserret, Lux 2013). Species

Scientific

lion Leo africanus hare Lepus sp. antelope [?reedbuck] elephant Elephas maximus gazelle Gazella spp. striped hyena Hyaena hyaena

Tuareg

Zenaga

Tetserret

ăhărr t-e-mæ̏ rwæl-t t-ȅ-nher-t élu

waˀr tärämbuL änaˀr iyih

ar tmarwult ænarˤ

t-ašə̏ ŋkəṭṭ šæbójæn

äžänkuđ ärđ̣ äy

aʒonkəḍ tafagant

unknown sources. Table 14.4 shows a few selected examples of mammal names in South Berber, that is, Tuareg, Zenaga and Tetserret. Some of these, such as ‘lion’ and ‘hare’ are clearly cognate across all three languages. However, other mammal names are unrelated and of unclear origin. At least some of these may have been adopted from pre-existing forager languages. Terms for most species are cognate but the hyena and the elephant seem to have acquired unrelated names.

Lexical Evidence in Nilo-Saharan If the expansion westwards of Nilo-Saharan is associated with the Holocene ‘green Sahara’, then there should also be a link with aquatic subsistence terms. In Drake et al., it is argued that this is in turn also reflected in the distribution of bone harpoons, used to hunt large river fauna.98 Two species which are notable in Saharan rock art and which would have been key prey species for foragers are the crocodile and the hippo. Useful confirmation of this is that both of these can be reconstructed in Nilo-Saharan languages. Tables 14.5 and 14.6 marshal the lexical evidence for this. This assumes the root had an initial k-, deleted in East Sudanic and that Sara-Bongo-Bagirmi (SBB) languages underwent metathesis. This is just a preliminary insight into a large body of evidence. Work is in hand compiling a raft of water-related lexemes, including ‘river’ and ‘lake’. It is unclear just how old the core languages of Nilo-Saharan are in their 98

Drake et al. 2011.

https://doi.org/10.1017/9781108634311.014 Published online by Cambridge University Press

453

454

Roger Blench

Table 14.5. A Nilo-Saharan root for ‘crocodile’ (SBB = SaraBongo-Bagirmi). Family

Subgroup

Language

Attestation

Eastern Sudanic Eastern Sudanic Central Sudanic Central Sudanic Saharan Songhay

Ama Nubian SBB SBB West South

Afitti

arəm elim màrà *màr[à] kárám kààrày

Gula Mere *SBB Manga Zarma

Table 14.6. A Saharan root for ‘hippo’. Family

Subgroup

Language

Attestation

Gumuz Maban CS Songhay Songhay Songhay

Runga Sara North North South

Kokit Aiki Nar Tadaksahak Koyra Chiini Zarma

baŋa bùngùr àbà ágamba baŋa bàŋá

homeland area in the Ethio-Sudan borderland. But there is increasingly little doubt that the relatively sudden abundance of aquatic resources stimulated a rapid expansion westwards of fisher-foragers, with specialised subsistence strategies. The subsequent expansion of first Niger-Congo and then Afroasiatic speakers fragmented Nilo-Saharan, giving it the appearance of a chain of isolated branches. Nonetheless, its history within the framework of the Sahara is there to be reconstructed.

Synthesis and the Agenda for Further Research The present-day linguistic situation of the Sahara is a highly misleading guide to its past. The pre-Holocene language situation is probably unrecoverable, but once the humid period began, the so-called ‘green Sahara’, the desert must have been rich with languages, Niger-Congo and Nilo-Saharan in the south-centre, but unidentified forager languages in the north-centre. Residual forager populations which still survive give some indication of the possible interlocking subsistence specialties at this period.

https://doi.org/10.1017/9781108634311.014 Published online by Cambridge University Press

14 The Linguistic Prehistory of the Sahara

Prior to the expansion of Berber and then Arabic, unknown but distinct languages would have been spoken in both the Sahara and along the North African coast. It is generally assumed the listing of Saharan peoples in Herodotus and their varied customs reflects something of the ethnic diversity in the mid-first millennium BC. For convenience, these languages can be referred to as ‘Old North African’ (ONA) with no presuppositions as to their genetic affiliation(s). It is possible they were related to the former languages of the Iberian Peninsula, such as Tartessian. Archaeologically, these must be identified with the Capsian and its predecessors, although the languages spoken in the first period of the Neolithic in the Maghrib would also have been ONA. But the completeness with which Berber eliminated ONA means little can be said about it. The Berber roots which are not of Afroasiatic origin may reflect these languages, or simply the long period of differentiation from the mainstream of the Afroasiatic lexicon. The problematic inscriptions typical of the western Libyco-Berber area undoubtedly represent a distinct language, possibly some sort of creole between Berber and ONA. It is this language which was carried to the Canaries, hence the partial Berber character of Guanche. Surveys of burial grounds in ‘Numidia’ have shown remarkable diversity in architecture and burial practice and it is usually assumed this reflects ethnolinguistic diversity, although the disappearance of individual languages following the Berber dispersal makes it now difficult to tie these to particular populations.99 However, speakers of Old North African also spread far across the Sahara, attracted by the expansion of large game animals following the Holocene humid episodes. Its speakers carried the bow and arrow south, where they encountered both Nilo-Saharan speakers and another ethnic group that was later to develop into Niger-Congo. However, they were still in situ when the Tuareg expanded south, as their names for large animals were borrowed by the Tuareg, who might have been either unfamiliar or barely familiar with animals such as lions and elephants. By this period, the Sahara was sufficiently arid for species such as the hippo to be long-gone. Hence, it was only when the Tuareg reached the Niger and came into contact with the Songhay that they would have needed a name for this animal. Likely candidates for the last speakers of ONA would have been the Dawada and possibly the Nemadi, although Nilo-Saharan is another 99

Camps 1961; Sanmartí et al., Chapter 8, this volume.

https://doi.org/10.1017/9781108634311.014 Published online by Cambridge University Press

455

456

Roger Blench

possibility for their language. At any rate, the language clearly became extinct as a living entity within the early phases of the Tuareg and Hassaniya expansion. Nonetheless, it seems likely that further lexical analysis and toponymy should be able to recover more of the linguistic prehistory of the Sahara than is presumed at present. The major puzzle lies in understanding the role of Berber in assimilating these diverse populations. Berber is remote from other Afroasiatic languages, suggesting a period of diversification in isolation and then a primary wave of expansion. It seems credible to identify this with the first wave of pastoralism in the Sahara (c.6000 BC), but this is in marked contradiction to the low internal diversity of Berber. A possible explanation is that Berber underwent a major episode of language levelling at the time of the establishment of the Roman limes (AD 100–200) and perhaps an additional episode following the spread of Islam in North Africa. Two areas remain open for significant further work in respect of this model. The vast corpus of Saharan rock art remains poorly dated and has not often been analysed for possible economic interpretations. Illustrations of livestock and management practices need to be further mapped and compared with the archaeological and linguistic record, as do important changes, such as the introduction of the camel. The other area concerns inscriptions; these are far from fully catalogued, dated and convincingly translated. Old Libyan inscriptions predate the postulated episodes of language levelling and those which include problematic characters may provide clues to language diversity, linking both to the records of Herodotus and the diversity implicit in mortuary sites. We are unlikely to recover significant new languages, although more detailed elicitation of plant and animal names would be welcome. Lexical analysis shows that layers of archaic lexicon exist within Berber, and these may provide evidence for the rich mosaic of Saharan linguistic prehistory.

References Abu-Manga, A. A. 1995. The Songhai speech communities in the Sudan with special reference to the Songhai speakers of the Blue Nile. In R. Nicolai and F. Rottland (eds), Proceedings of the Fifth Nilo-Saharan Linguistics Colloquium, Köln: Rüdiger Köppe, 13–27. Al-Chennafi, M. and Norris, H. T. 1981. How the Hassaniyya vernacular of Mauritania supplanted Zenaga. The Maghreb Review 76.5–6: 77–78.

https://doi.org/10.1017/9781108634311.014 Published online by Cambridge University Press

14 The Linguistic Prehistory of the Sahara

Ameur, M. 1990. A propos de la classification des dialectes berbères. Etudes et Documents Berbères 7: 15–27. Anon. n.d. Chants de Femmes Nemadi. Label: Buda Musique, Ref: CD 926212. Anthonioz, R. 1968. Les Imraguen, pêcheurs nomades de Mauritanie. Bulletin de l’Institut fondamental d’Afrique noire 29, série B 3–4.2: 695–738. Arnaud, R. 1906. Chasseurs et pêcheurs du Tagant et du Hodh. La Géographie 13: 148–49. Attayoub, A. K. 2001. La Tətsərret des Ayttawari Seslem: identification sociolinguistique d’un parler berbère non-documenté chez les Touaregs de l’Azawagh (Niger). Mémoire de maîtrise. Paris: Institut national des langues et civilisations orientales. Baroin, C. 1985. Anarchie et cohésion chez les Téda. Cambridge: Cambridge University Press. Basset, A. 1933. Note sur les parlers zenaga. Bulletin du comité d’études historiques et scientifiques de l’Afrique Occidentale française 1933: 319–32. Basset, A. 1952. La langue berbère. London: Oxford University Press. Bechhaus-Gerst, M. 1989. Nubier und Kuschiten im Niltal. Sprach- und Kulturkontakte im ‘no man’s land’. Köln: Afrikanische Arbeitspapiere. Behrens, P. 1989. Langues et migrations des premiers pasteurs du Sahara: la formation de la branche berbère. In Libya Antiqua. Reports and Papers of the Symposium organized by UNESO in Paris, 16–18 January 1984, Paris: UNESCO, 31–53. Bellair, P. 1951. La ramla des Daouada (Fezzân). Travaux de l’institut de recherches sahariennes 7: 69–86. Bernus, E. 1981. Touaregs Nigériens. Unité culturelle et diversité régionale d’un peuple pasteur. Mémoires ORSTOM n. 94. Paris: Éditions de l’office de la recherche scientifique et techniques outre-mer. Biagetti, S., Kaci, A. A. and di Lernia, S. 2015. The ‘written landscape’ of the central Sahara: recording and digitising the Tifinagh inscriptions in the Tadrart Acacus Mountains. In M. Kominko (ed.), From Dust to Digital: Ten Years of the Endangered Archives Programme, United Kingdom: Open Book Publishers, 1–30. Biagetti, S., Kaci, A. A., Mori, L. and di Lernia, S. 2012. Writing the desert: the ‘Tifinagh’ rock inscriptions of the Tadrart Acacus (southwestern Libya). Azania: Archaeological Research in Africa 47.2: 153–74. Blench, R. M. 2001. Types of language spread and their archaeological correlates: the example of Berber. Origini 23: 169–90. Brett, M. and Fentress, E. 1996. The Berbers. Oxford: Blackwell. Brogan, O. 1954. The camel in Roman Tripolitania. Papers of the British School at Rome 22: 126–31. Brooks, N., Clarke, J., Garfi, S., and Pirie, A. 2009. The archaeology of Western Sahara: results of environmental and archaeological reconnaissance. Antiquity 83.322: 918–34.

https://doi.org/10.1017/9781108634311.014 Published online by Cambridge University Press

457

458

Roger Blench

Brosset, (Capt.) D. 1932. Les Némadi. Bulletin de l’Afrique de l’Ouest Française, Renseignements coloniaux: 337–46. Camps, G. 1961. Monuments et rites funéraires protohistoriques. Paris: Arts et métiers graphiques. Camps, G. 1974. L’âge du tombeau de Tin Hinan, ancêtre des Touareg du Hoggar. Zephyrus 25: 497–516. Camps, G. 1980. Berbères. Aux marges de l’histoire. Toulouse: Hespérides. Çelebi, E. 1995/2007. Evliya Çelebi Seyahatnâmesi, Topkapi Sarayi Bagdat 304 Yazamasinin Transkripsyonu –Dizini. 10 vols. O. S. Gökyay and Y. Dağh eds. Istanbul: Yapi Kredi Yayinlari. Chaker, S. 2002. L’écriture libyco-berbère. État des lieux et perspectives. Colloque annuel de la SHESL, Lyon-ENS, samedi 2 février 2002. Chalmel, (Capt.) 1931. Notice sur les Bideyat. Bulletin de la société de recherches congolaises 15: 33–91. Chapelle, J. 1957. Nomades noirs du Sahara. Paris: Plon. Clapperton, H. 1829. Journal of a Second Expedition into the Interior of Africa. London: John Murray. Cleaveland, T. 2002. Becoming Walāta. A History of Saharan Social Formation and Transformation. Portsmouth, NH: Heinemann. Cline, W. 1950. The Teda of Tibesti, Borku and Kawar in the Eastern Sahara. Menasha, WI: George Banta Publishing. Creissels, D. 1981. De la possibilité de rapprochements entre le Songhay et les langues Niger-Congo (en particulier Mandé). In T. Schadeberg and M. L. Bender (eds), Nilo-Saharan: Proceedings of the First Nilo-Saharan Linguistics Colloquium, Leiden, September 8 –10, Dordrecht: Foris Publications, 185–99. Daniels, C. M. 1975. An ancient people of the Libyan Sahara. In J. and T. Bynon (eds), Hamito-Semitica, The Hague: Mouston, 249–65. De Noray, M. L. 2006. Le livre des Imraguen: pêcheurs du banc d’Arguin en Mauritanie. Paris: Buchet-Chastel. De Pommerol, P. J. 1997. L’arabe tchadien: Émergence d’une langue véhiculaire. Paris: Karthala Editions. di Lernia, S. 2001. Dismantling dung: delayed use of food resources among Early Holocene foragers of the Libyan Sahara. Journal of Anthropological Archaeology 20: 408–41. di Lernia, S. 2006. Building monuments, creating identity: cattle cult as a social response to rapid environmental changes in the Holocene Sahara. Quaternary International 151: 50–62. di Lernia, S. 2013. The emergence and spread of herding in Northern Africa. In P. Mitchell and P. Lane (eds), Oxford Handbook of African Archaeology, Oxford: Oxford University Press, 527–40. Dimmendaal, G. J. 2007. Eastern Sudanic and the Wadi Howar and Wadi El Milk diaspora. Sprache und Geschichte in Afrika 18:37–67.

https://doi.org/10.1017/9781108634311.014 Published online by Cambridge University Press

14 The Linguistic Prehistory of the Sahara

Drake, N. A., Blench, R. M., Armitage, S. J., Bristow, C. S. and White, K. H. 2011. Ancient watercourses and biogeography of the Sahara explain the peopling of the desert. Proceedings of the National Academy of Sciences of the United States of America 108.2: 458–62. Dubié, P. 1940. L’îlot berbérophone de Mauritanie. Bulletin de l’Institut fondamental d’Afrique noire 2: 315–25. Faidherbe, L. L. 1877. Le Zenaga des tribus Sénégalaises. Paris: Leroux. Fevrier, J. G. 1965. La constitution municipale de Dougga à l’époque numide. Mélanges de Carthage 1964–65:85–91. Fondacci, P. H. (Capt.). 1945. Les Némadis. Paris: Mémoire du CHEAM n. 1009. Fortier, C. 2004. Au miroir de l’autre: Chasseurs (Némadi) et pêcheurs (Imraguen) dans un monde de pasteurs nomades (Mauritanie). Figures Sahariennes 7. http:// recherche.univ-montp3.fr/cerce/r7/c.f.htm, accessed 2/6/2016. Fuchs, P. 1961. Die Völker der Südost-Sahara: Tibesti, Borku, Ennedi. Vienna: W. Braumüller. Fuchs, P. 1983. Das Brot der Wüste. Sozio-Okonomie der Sahara-Kanuri von Fachi. Studien zur Kulturkunde 67. Wiesbaden: Franz Steiner. Gabelenz, G. 1894. Die Verwandschaft des Baskischen mit den Berbersprachen Nordafrikas. Braunschweig: A. C. Graf von der Schulenburg. Gabus, J. 1952. Contribution à l’étude des Némadi, chasseurs archaïques du Djouf. Bulletin de la Société Suisse d’Anthropologie et d’Ethnologie 52: 49–83. Galand, L. 1971. Unité et diversité du vocabulaire berbère. In A. Giuffre (ed.), Atti della settimana magrebina. Cagliari 22–25 Maggio 1969, Milan: Università di Cagliari, Istituto di Studi Africani, 17–20. Galand, L. 1988. Berberisch: der Schlüssel zum Altkanarischen? Almogaren 18– 19:7–16. Garcea, E. A. A. (ed.). 2013. Gobero: the No-Return Frontier. Archaeology and Landscape at the Saharo-Sahelian Borderland. Frankfurt: Africa Magna Verlag. Gauthier, A. 2001. The Early to Late Neolithic archeofaunas from Nabta and Bir Kiseiba. In F. Wendorf and R. Schild (eds), Holocene Settlement of the Egyptian Sahara Volume 1: The Archaeology of Nabta Playa, New York: Springer, 609–35. Gautier, A. 1984. Archaeozoology of the Bir Kiseiba Region. In F. Wendorf and R. Schild (eds), Cattle-Keepers of the Eastern Sahara, Dallas: Southern Methodist University Press, 49–72. Gautier, A. 1987. Prehistoric men and cattle in North Africa: a dearth of data and a surfeit of models. In A. Close (ed.), Prehistory of Arid North Africa. Essays in Honor of Fred Wendorf, Dallas: Southern Methodist University Press, 163–87. Gerteiny, A. G. 1967. Mauritania. New York: Frederic A. Praeger. González Anton, R. and Tejera Gaspar, A. 1990. Los aborígenes canarios: Gran Canaria y Tenerife. Oviedo: Ediciones Istmo. Gsell, S. 1933. La Tripolitaine et le Sahara au IIIème siècle. Mémoires de l’Académie des Inscriptions 43: 149–66.

https://doi.org/10.1017/9781108634311.014 Published online by Cambridge University Press

459

460

Roger Blench Habraszewski, T. 1967. Kanuri – language and people – in the ‘Travel book’ (Siyahetname) of Evliya Çelebi. Africana Bulletin 6: 59–66. Haour, A., Banni Guene, O., Gosselain, O., Livingstone Smith, A. and N’Dah, D. 2011. Survey along the Niger River valley at the Benin-Niger border. Nyame Akuma 76: 23–32. Heath, J. 1998a. Dictionnaire Songhay –Anglais- Français. Tome I, Koyra Chinni. Paris: L’Harmattan. Heath, J. 1998b. Dictionnaire Songhay –Anglais- Français. Tome II, Djenné Chinni. Paris: L’Harmattan. Heath, J. 2006. Dictionnaire touareg du Mali. Paris: Karthala. Hermans, J. M. 2013. Les Nemadis, chasseurs-cueilleurs du désert mauritanien. Paris: Edilivre. Jongeling, K. and Kerr, R. M. 2005. Late Punic Epigraphy: an Introduction to the Study of Neo-Punic and Latino-Punic Inscriptions. Tübingen: Mohr Siebeck. Kerr, R. M. 2010. Latino-Punic Epigraphy. A Descriptive Study of the Inscriptions. Tubingen: Mohr Siebeck. Kossmann, M. 1999. Essai sur la phonologie du proto-berbère. Köln: Rüdiger Köppe. Kossmann, M. 2013. The Arabic Influence on Northern Berber. Leiden: Brill. Kronenberg, A. 1958. Die Teda von Tibesti. Wiener Beiträge zur Kulturgeschichte und Linguistik XII. Vienna: Ferdinand Berger. Laforgue, P. 1917. Une fraction non musulmane en Mauritanie saharienne, les Némadi. Annuaire et mémoires du comité d’études historiques et scientifiques de l’Afrique occidentale française: 685–92. Laoust, E. 1932. Siwa. I. Son parler. Paris: Leroux. Le Quellec, J. L. 2011. Rock art, scripts and proto-scripts in Africa: the LibycoBerber example. In A. Delmas and N. Penn (eds), Written Cultures in a Colonial Context: Africa and the Americas, 1500–1900, Leiden: Brill, 3–29. Le Quellec, J. L. and Civrac, M. A. 2010. La chasse au filet sur les peintures rupestres du Sahara central et dans l’antiquité. Les cahiers des amis de l’art rupestre saharien 14: 255–62. Le Rouvreur, A. (ed.). 1989. Sahéliens et sahariens du Tchad. Paris: L’Harmattan. Lhote, H. 1985. Stations inédites de gravures rupestres au Hoggar. Dossiers et recherches sur l’Afrique 2: 153–61. Limouzin, (Capt.). 1951. La marche chez les Daouada. Bulletin de liaison saharienne, 5: 30–31. Lindblom, G. 1925. Jakt-och fångstmetoder bland Afrikanska folk: (with a retrospect in English). Stockholm: V. Pettersons. Lotte, L. 1937. Coutumes des Imraguen (côtes de Mauritanie, AOF). Journal de la société des africanistes 7.1: 41–51. Lukas, J. 1952. Umrisse einer ostsaharanischen Sprachgruppe. Afrika und Ubersee 36: 3–7.

https://doi.org/10.1017/9781108634311.014 Published online by Cambridge University Press

14 The Linguistic Prehistory of the Sahara

Lux, C. 2013. Le tetserrét, langue berbère du Niger: description phonétique, phonologique et morphologique, dans une perspective comparative. Berber Studies 38. Koln: Rudiger Koppe. MacMichael, H. A. 1912. Notes on the Zaghawa and the people of Gebel Midob, Anglo-Egyptian Sudan. The Journal of the Royal Anthropological Institute of Great Britain and Ireland 42: 288–344. Maigret, J. 1985. Les Imraguen, pêcheurs des côtes de la Mauritanie: une technique ancestrale. Vémes rencontres internationale d’archéologie et d’histoire d’Antibes, 24, 25, 26 octobre 1984, Juan-les-Pins: Association pour la promotion et la diffusion des connaissances arché ologiques, 205–14. Maigret, J. and Abdallahi, A. O. 1976. La pêche des Imraguen sur le Banc d’Arguin et au Cap Timiris (Mauritanie): techniques et méthodes de pêche. Notes Africaines 149: 1–8. Malášková, Z. and Blažek, V. 2012. Phoenician/Punic loans in Berber languages and their role in the chronology of Berber. Unpublished paper presented at the conference Rethinking Africa’s Transcontinental Continuities in Pre- and Protohistory. African Studies Centre, Leiden University, Leiden, 12–13 April 2012. Masqueray, E. 1879. Comparaison d’un vocabulaire des Zenaga avec les vocabulaires correspondants des dialectes Chawia et des Beni Mzab. Archives des missions scientifiques et littéraires 3.5: 473–533. Mattingly, D. J. (ed.). 2003. The Archaeology of Fazzān, Volume 1. Synthesis. London: Society for Libyan Studies, Department of Antiquities. Mattingly, D. J. (ed.). 2010. The Archaeology of Fazzān, Volume 3. Excavations of C. M. Daniels. London: Society for Libyan Studies, Department of Antiquities. Mattingly, D. J. (ed.). 2013. The Archaeology of Fazzān, Volume 4. Excavations at Old Jarma. London: Society for Libyan Studies, Department of Antiquities. Milburn, M. 1988. Guanche: prehistoric inhabitants of Tenerife and possible African affinities. Africa: Rivista trimestrale di studi e documentazione dell’Istituto italiano per l’Africa e l’Oriente 43.3: 475–81. Monteil, C. 1939. La langue Azer. Publications du comité d’études historiques et scientifiques de l’Afrique Occidentale française série B.5: 213–41. Moreno García, J. C. 2014. Invaders or just herders? Libyans in Egypt in the third and second millennia BCE. World Archaeology 46.4: 610–23. Mukarovsky, H. G. 1964. Baskisch und Berberisch. Wiener Zeitschrift für die Kunde des Morgenlandes 59–60:52–94. Mukarovsky, H. G. 1969. Baskisch und Berberische Entsprechung. Wiener Zeitschrift für die Kunde des Morgenlandes 62: 32–51. Nicholas, F. 1953. La langue berbère de Mauritanie. Mémoire de l’IFAN 33. Dakar: IFAN. Nicolaï, R. 1981. Les dialectes du songhay: contribution à l’étude des changements linguistiques. Paris: SELAF. Nicolaï, R. 2003. La force des choses, ou, l’épreuve ‘nilo-saharienne’: questions sur les reconstructions archéologiques et l’évolution des langues. Koln: Rudiger Koppe.

https://doi.org/10.1017/9781108634311.014 Published online by Cambridge University Press

461

462

Roger Blench

Nicolaï, R. 1990. Parentés linguistiques (à propos du songhay). Paris: Editions du Centre National de la Recherche Scientifique. Nicolaisen, I. 2010. Elusive Hunters: the Haddad of Kanem and the Bahr el Ghazal. Aarhus: Aarhus Universitetsforlag. O’Connor, M. 1996. The Berber scripts. In W. Bright and P. Daniels (eds), The World’s Writing Systems, New York: Oxford University Press, 112–16. Paradisi, U. 1960. Il berbero du Augila (materiale lessicale). Rivisita degli Studi Orientali 35: 157–77. Paradisi, U. 1963. Il linguaggio berbero di El-Fógǎha (Fezzân) (testi e materiale lessicale). Annali ns 13: 93–126. Park, D. P. 2010. Prehistoric Timbuktu and its hinterland. Antiquity 84: 1076–88. Pauphilet, D. 1953. Les Daouada, In P. Bellair, E. G. Gobert, P. Jodot and D. Pauphilet (eds), Mission Au Fezzan, 1949. Publications scientifiques I, Tunis: Institut de hautes études de Tunis, 115–32. Petráček, K. 1987. Berti or Sagato-a (Saharan) Vocabulary. Afrika und Übersee 70: 163–93. Putten, M. V. 2013. A Grammar of Awjila Berber (Libya): Based on Umberto Paradisi’s Material, Unpublished PhD thesis, Leiden University, Leiden. Segert, S. 1976. A Grammar of Phoenician and Punic. Munich: Beck. Shaw, B. D. 1979. The camel in Roman North Africa and the Sahara: history, biology and human economy. Bulletin de l’Institut fondamental d’Afrique noire série B 41.4: 663–721. Skinner, N. 1996. Hausa Comparative Dictionary. Köln: Rüdiger Köppe. Souag, M. L. 2010. Grammatical Contact in the Sahara: Arabic, Berber, and Songhay in Tabelbala and Siwa. Unpublished PhD thesis, University of London, UK. Souag, M. L. 2012. The subclassification of Songhay and its historical implications. Journal of African Languages and Linguistics 33.2: 181–213. Souag, M. L. 2015. Explaining Korandje: language contact, plantations, and the trans-Saharan trade. Journal of Pidgin and Creole Languages 30.2: 189–224. Sudlow, D. 2009. Dictionary of the Tamasheq of North-East Burkina Faso. Köln: Rudiger Köppe. Sznycer, Maurice 1967. Les passages puniques en transcription latine dans le Poenulus de Plaute. Paris: Librairie C. Klincksieck. Taine-Cheikh, C. 1979. Aperçus sur la situation sociolinguistique en Mauritanie. In Introduction à la Mauritanie. Centre de recherches et d’études sur les sociétés méditerranéennes, Centre d’études d’Afrique noire, Paris: Editions du CNRS, 169–73. Taine-Cheikh, C. 2008. Dictionnaire Zénaga-français (berbère de Maurétanie). Köln: Rudiger Köppe. Taine-Cheikh, C. 2010. Dictionnaire français-Zénaga (berbère de Maurétanie). Köln: Rudiger Köppe.

https://doi.org/10.1017/9781108634311.014 Published online by Cambridge University Press

14 The Linguistic Prehistory of the Sahara

Taine-Cheikh, C. 2013. Des ethnies chimériques aux langues fantômes: l’exemple des Imraguen et Nemâdi de Mauritanie. In C. de Féral (ed.), In and out of Africa: Languages in Question. In Honour of Robert Nicolaï. Vol. 1. Language Contact and Epistemological Issues, Leuven: Peeters, 137–64. Topper, U. 1993. Einige Notizen zur berberischen Fischerbevölkerung am Atlantik. Almogaren 23: 151–69. Trotignon, E. 1981. Données socio-économiques sur le village de pêcheurs Imraguen d’Iwik (Mauritanie). Bulletin de l’Institut fondamental d’Afrique noire série B 43.3–4: 319–60. Tubiana, M. J. 1964. Survivances pré-islamiques en pays zaghawa. Paris: Université de Paris. Tubiana, M. J. 1985. Des troupeaux et de femmes: mariage et transferts de biens chez les Beri (Zaghawa et Bideyat) du Tchad et du Soudan. Paris: L’Harmattan. Tubiana, M. J. and Tubiana, J. 1977. The Zaghawa from an Ecological Perspective. Rotterdam: Balkema. Tucker, B. 2003. Mikea origins: relicts or refugees? In Z. Crossland, G. Sodikoff and W. Griffin (eds), Lova/Inheritance – Past and Present in Madagascar. Discussions in Anthropology 14, Ann Arbor: Department of Anthropology, University of Michigan, 193–215. Vernet, R. and Onrubia-Pintado, J. 1994. La place des ancêtres de berbères dans le Sahara Néolithique. In Milieux, hommes et techniques du Sahara préhistorique, Paris: Harmattan, 53–67. Vernet, R., Otte, M., Tarrou, L., Gallin, A. and Géoris-Creuseveau, J. 2007. Fouille de la butte de FA 10 (Banc d’Arguin) et son apport à la connaissance de la culture épipaléolithique de Foum Arguin, nord-ouest du Sahara. Journal of African Archaeology 5.1: 17–45. Vycichl, W. 1952. Punischer Sprachinfluβ in Berberischen. Journal of Near Eastern Studies 11: 198–204. Walz, T. 1978. Trade Between Egypt and Bilād as-Sūdān 1700–1820. Cairo: IFAO. Wellard, J. 1964. The Great Sahara. London: Hutchinson and Co. Willms, A. 1980. Die dialektale Differenzierung des Berberishen. Berlin: Reimer. Wölfel, D. J. 1955. Eurafrikanische Wortschichten als Kulturschichten. Acta Salmanticensia. Filosofía y letras, 9.1. Salamanca: Universidad de Salamanca. Wölfel, D. J. 1965. Monumenta linguae Canariae. Austria: Akademische Druck. Wolff, E. and Alidou, H. 2001. On the non-linear ancestry of Tasawaq (Niger), or: how ‘mixed’ can a language be? In D. Nurse (ed.), Historical Language Contact in Africa, Köln: Rüdiger Köppe, 523–74. Worms, J. and Ould Eida, A. M. 2002. Les savoirs traditionnels des Imraguen, activités liées à la pêche. Nouakchott: Parc National du Banc d’Arguin. Zeltner, J. C. 1970. Histoire des Arabes sur les rives du Lac Tchad. Annales de l’Université d’Abidjan, ethnosociologie 2.2: 112–79. Zyhlarz, E. 1932. Die Sprache Numidiens. Zeitschrift für Eingeborenen-Sprachen 7: 275–80.

https://doi.org/10.1017/9781108634311.014 Published online by Cambridge University Press

463

15

Berber Peoples in the Sahara and North Africa Linguistic Historical Proposals christopher ehret

Introduction Over the period since the early Holocene, peoples belonging to either of two cultural and linguistic complexes have come to prevail across most of the vast Sahara region, from the Mediterranean in the north to the Sahel zone in the south. Through the southern Sahara, from the areas just north of the great bend of the Niger River at the west to the Nile River in the east, these populations most often spoke languages of the Nilo-Saharan language family. In North Africa, the northern half of the Sahara, and in the lands east of the Nile, societies speaking diverse languages of the Afroasiatic language family usually predominated.1 The primary focus here will be on what linguistic evidence can tell about the history of peoples of the Berber branch of Afroasiatic, and thus on historical developments in North Africa and the northern, central and Western Sahara, particularly in the period from around 2000 BC down to Late Antiquity. Understanding the deeper-time historical context out of which the Berber languages emerged requires, however, giving attention as well to the early history of the peoples who spoke Chadic languages, the nearest-related branch to Berber within the Afroasiatic family, and on the early historical relations between the Chadic and Nilo-Saharan speaking peoples of the Sahara and Sahel.

Berber Peoples and Language Histories The Berber languages form a relatively tightly related group within the very diverse, very deep-time Afroasiatic family. The time depth of the family as a whole exceeds 15,000 years.2 In contrast, the divergence between the most distant Berber languages is in the order of 4,000–3,500 years ago – that is to 464

1

Ehret 2011a.

2

Ehret 2011a; Ehret et al. 2004.

https://doi.org/10.1017/9781108634311.015 Published online by Cambridge University Press

15 Berbers in the Sahara and North Africa: Linguistic Historical Proposals

say, the ancestral, Proto-Berber language was spoken most probably in the span 2000–1500 BC – and the subgroups within the Berber group are still more recent in origin. The closest relative of Berber within the family is the Chadic branch.3 The traditional sorts of historical linguistic indicators of relationship strongly support this conclusion.4 For example, in a computational study currently being undertaken by Deven Vyas (Stony Brook), Andrew Kitchen of the University of Iowa and this writer, out of our 90-meaning list of core vocabulary we identify 14 instances, a very high proportion, in which Chadic and Berber languages share unique lexical innovations that replaced the older Afroasiatic root words for those meanings – for ‘bite’, ‘breast’, ‘cold’, ‘drink’, ‘ear’, ‘arm/hand’, ‘hear’, ‘know’, ‘root’, ‘say’, ‘stand’, ‘who?’ and, most telling of all, the pronouns ‘I’ and ‘you (singular)’. The Chadic branch has a considerably greater internal time depth than Berber. The proto-Chadic language and society, according to our findings, date to a median range of around 7,000 years ago, or in other words to sometime in the period around 5000 BC. Our estimates place the divergence of the separate Berber and Chadic lines of linguistic descent out of their common ancestor, proto-Chado-Berber, another 2,000–3,000 years earlier, in the broad range of the eighth or ninth millennium BC. Over such a long period, from as early possibly as 7000 BC down to 2000 BC, it is quite improbable that stable boundaries somehow persisted among the various daughter speech communities of protoChado-Berber. Shifts in language distributions and new population movements surely took place across different portions of North Africa at different times over that period. Exploring those possibilities depends, however, on the testimony of archaeology. The linguistic record of this long intervening span is no longer accessible for a very straightforward reason: the successive expansions of the various daughter languages of Proto-Berber since 2000 BC eventually drove all of its former sister languages out of existence, removing them as possible linguistic historical resources. The archaeological record may well reveal periods of cultural spread within North Africa in that long intervening period, but the peoples involved would have not been Berber per se, but would instead have spoken languages belonging to sister lines of linguistic descent from proto-Chado-Berber or possibly languages that split off even earlier from the line of Afroasiatic linguistic descent leading down to proto-Chado-Berber. 3

Newman 1980.

4

Bynon 1984.

https://doi.org/10.1017/9781108634311.015 Published online by Cambridge University Press

465

466

Christopher Ehret

So what about the Berber language group proper? Over the past 3,500 years, various Berber-speaking peoples participated in several periods of quite wide expansions. The several Berber family tree diagrams depict the key stages in this history (Figs 15.1, 15.3, 15.4, 15.6 and 15.7).

Arguing History from the Stratigraphies of Linguistic Relationship The patterns of cognate loss over the long term among a group of equidistantly related languages tend to a normal distribution. The larger the number of equidistantly related languages in the group, the more the distribution will approximate a bell curve. In most real-world cases, such groups consist of only a few languages. When languages of a group are equidistantly related to one another, their figures of mutual cognation will be consistent with a normal distribution; that is, they will distribute over a very narrow range, with a majority of the figures clustering around the middle of the range. The normative procedure in interpreting a matrix of this kind is to move from the examination of sub-matrices of closer relationship outward to the successively deeper levels of relationship with the language group. The Central Amazigh subgroup, because it contains the most languages, provides the largest distribution of cognation counts, making it a useful starting point in the analysis. Figure 15.1 displays the portion of the matrix relating to Central Amazigh. Wargla 69 Mzab 59

66

Shawiya

61 63 58 65

66 63 59 64

63 65 60 64

Beni-Snous 69 Rif 66 69 65 68

Sanhaja 71 Tamazight

Figure 15.1. Matrix showing relationship between Central Amazigh languages.

Two consistent distribution patterns stand out in the cognation counts for Central Amazigh. The higher range encompasses a West-Central Amazigh subgroup, which includes Beni-Snous, Rif, Sanhaja and Moroccan Tamazight. The cognation figures among these four distribute around a median of 68/69 per cent:

https://doi.org/10.1017/9781108634311.015 Published online by Cambridge University Press

15 Berbers in the Sahara and North Africa: Linguistic Historical Proposals

65

66

69 69

68

71

A lower, but equally tight and internally consistent range, with a median of around 63/64 per cent, characterises the next deepest level of relationship, of the West-Central Amazigh subgroup with Mzab, Wargla and Shawiya:

58

59

60

61

63 63 63

64 64

65 65

66 66

The Wargla-Mzab figure of 69 per cent stands outside this range, supporting the proposal that Wargla and Mzab constitute a closer related subgroup within Central Amazigh, on a similar time depth of initial differentiation as that of the West-Central Amazigh subgroup, with their cognation with each other at 69 per cent. The tree representation of this body of evidence (Fig. 15.2) depicts three equidistant subgroups of Central Amazigh (a further language, Shenwa, is included, although not cited in the previous cognation distribution figures): proto-Central Amazigh (median: 63/64%)

proto-Oasis (median: 69%)

Wargla

Mzab

proto-West-Central Amazing (median: 68/69%)

Shawiya

Shenwa Beni-Snous

Rif

Sanhaji Tamazight

Figure 15.2. Tree showing putative branches of Central Amazigh.

The linguistic geography of the branches of this tree implies a particular history of population spread. The languages of two of the subgroups, Oasis and Shawiya, are spoken in a relatively compact stretch of territories in eastern Algeria, consisting of the two oasis settlements in the far northern Sahara and, for Shawiya, the Aures mountain areas beginning 300 km north of the oases. In contrast, the languages of the third subgroup, West-Central Amazigh, extend across 1,000 km from central northern Algeria, where

https://doi.org/10.1017/9781108634311.015 Published online by Cambridge University Press

467

468

Christopher Ehret

Shenwa is spoken, to Tamazight of the central Atlas range in Morocco. A longestablished principle of argument from language distribution is that the area of greatest diversity within a language group is the area in which the group has longest been spoken. The areas of lesser differentiation are the areas to which the groups subsequently spread.5 These considerations locate the speakers of proto-Central Amazigh language somewhere in the region encompassing the Aures at the north and the Mzab and Wargla oases at the south. What kind of dating can we propose for the history of this group? As already noted, work on a more sweeping investigation of the chronology of the Afroasiatic language family, of which the Berber languages comprise a branch, is in progress but not yet completed. For now, there is an older kind of evidence that we can draw on for dating. The findings in a variety of cases in other parts of Africa suggest a possible timing of when the protoCentral Amazigh society first spread out and diverged.6 Elsewhere in the continent, a median percentage of cognation in the low to middle 60 per cent range in core vocabulary indicates that a span of time somewhere in the range of 1,500–1,600 years has elapsed since that divergence began. Taken together, these considerations favour the conclusion that the protoCentral Amazigh language was spoken by a society residing in the later first half of the first millennium AD in what is today eastern Algeria, perhaps in the northernmost parts of the Sahara or on the southern margins of the Atlas range and certainly no farther north than the Aures Mountains. By around the fifth century AD, the linguistic evidence indicates, people from this society set out on a quite rapid expansion that spread their presence and their language as far west as central Morocco within no more than a century or two. The reader will note that inferences drawn from linguistic evidence about the dates or locations of previous historical developments can never be as precise as the findings of archaeology. But what they do quite well is to identify for the investigator the most probable span of time and the most probable span of regions in which to look for the material record of those societies or the developments identified from the linguistic testimony.

Extending the Story Back in Time Having considered the evidence relating to the cohering of a more recent major subgroup of Berber, we can now extend our analysis, step by step, 5

Sapir 1916 is the classic introduction to the relevant methods.

https://doi.org/10.1017/9781108634311.015 Published online by Cambridge University Press

6

Ehret 2000; 2011a.

15 Berbers in the Sahara and North Africa: Linguistic Historical Proposals

farther back in time by examining the evidence for the relationships of the rest of the Berber languages to the Central Amazigh subgroup, beginning with the languages with the highest ranges of cognation with Central Amazigh. We look first at the evidence relating to two languages, Shilh in western Morocco and Matmata at the other extreme of the Central Amazigh distribution, in southern Tunisia. Comparing the Shilh and Central Amazigh figures immediately brings to light a further factor that often affects matrices of cognation counts. That factor is a history of either long or intensive contacts among speakers of closely related languages with, as a consequence, extensive word borrowing among those languages. Because of the relative closeness of the relationships of the Berber languages, it is often difficult to distinguish the words that one language has borrowed from a sister language from those that each language has separately retained from their common protolanguage. One sort of numerical indicator is nevertheless a dead giveaway that this kind of contact history took place, even if not all the particular borrowings can be individually identified – namely, cognation figures that fall distinctly outside the normal distribution formed by the remainder of the figures. In the Shilh case, the cognate counts of Shilh with the easterly Central Amazigh languages, with which the speakers of Shilh had not historically been in contact, present a consistent, tight distribution: 52

54

55

56

In contrast, the figures of Shilh with nearby and adjacent western languages of Central Amazigh – Rif, Sanhaja and Central Atlas Tamazight – are offset far higher, at 60, 68 and 66 per cent respectively, from this tight range, as shown in Figure 15.3. The patterning of the Shilh figures overall is consistent with the conclusion that the primary direction of borrowing was from early Shilh. The size of the offset indicates that on the order of 10 per cent or more of basic vocabulary in the western languages of Central Amazigh came into those languages by borrowing from early Shilh (the over-high percentages are marked in boldface in Figure 15.3). This kind of word borrowing is a defining feature of what has been called the ‘intensive’ category, in which loan-words penetrate into even the most basic and borrowing-resistant part of the vocabulary.7 ‘Intensive word borrowing’ can take place in either of two quite different historical situations. On the one hand, this scale of word borrowing can pass from the 7

Ehret 2011a.

https://doi.org/10.1017/9781108634311.015 Published online by Cambridge University Press

469

470

Christopher Ehret

Wargla 69 Mzab 59

66

Shawiya

61 63 58 65

66 63 59 64

63 65 60 64

54

52

55

Beni-Snous 69 Rif 66 69 65 68 56

60

Sanhaja 71 Tamazight 68

66

Shilh

Figure 15.3. Matrix for Shilh and western languages of Central Amazigh.

language of a dominant majority society into the language of a persisting subordinate, minority community. The notable African instances of this kind of linguistic effect are all small communities of hunter-gatherers living among food producers. That solution clearly does not apply in this instance. The alternative explanation for this kind of borrowing is a particular kind of conquest history, in which a minority intrusive group, in a relatively short period of time, conquers and establishes its rule over a much larger existing population. Thereafter, gradually, the conquered population shifts over to the language of the conquerors. In this kind of situation, the conquered population tends to carry over into their new language a great many of the words of their former language, including even words from the most basic portions of the vocabulary. The anomalously high cognate counts of Shilh with the westernmost languages of the Central Amazigh group thus have a probable dual historical implication. First of all, the high-skewed counts in Rif, Sanhaja and Tamazight indicate that, before the arrival of West-Central Amazigh speakers, the populations of all of Morocco and extending into far western Algeria spoke an early form of Shilh. Second, the people who brought the early West-Central Amazigh ancestral language into these regions most probably arrived as conquerors, rather than as a large, gradually intruding population. The Matmata figures with the Central Amazigh languages (shown in Fig. 15.4) present, in contrast, an outwardly similar pattern in some respects; but the overall distribution of its cognation figures is different, and the historical implications are different as well. With geographically distant members of Central Amazigh, the Matmata percentages, like those of Shilh, fall mostly in the middle 50 per cents, with three at 55 per cent, although with one reaching 59 per cent. Except for that one 59 per cent figure, all the higher figures for Matmata are with languages – Shawiya, Mzab and Wargla at

https://doi.org/10.1017/9781108634311.015 Published online by Cambridge University Press

15 Berbers in the Sahara and North Africa: Linguistic Historical Proposals

respectively 62, 62 and 58 per cent – that would in earlier centuries have been neighbours of the Matmata or members of the same medieval trading networks. Unlike Shilh, however, Matmata has no extraordinarily high-skewed figures, and even including the higher counts, the figures of Matmata with Central Amazigh fall in a tight range, 55–62 per cent, with a median of 58 per cent (Fig. 15.4): Matmata 58 62

Wargla 69 Mzab

62

59

66

59 55 55 55

61 63 58 65

66 63 59 64

Shawiya 63 65 60 64

Beni-Snous 69 Rif 66 69 65 68

Sanhaja 71 Tamazight

Figure 15.4. Matrix relating to links between Matmata and Central Amazigh languages.

This pattern typically indicates that the language in question is an outlying member of the overall group, related at a slightly greater remove to the rest. It supports the postulation of a Wide-Central Amazigh group, in which Matmata would form one primary branch coordinate with Central Amazigh as the second branch (Fig. 15.5). proto-Wide Central Amazigh (median: 58%)

proto-Central Amazigh (median: 63/64%)

proto-Oasis (median: 69%)

Matmata

Wargla

Mzab

proto-West-Central Amazigh (median: 68/69%)

Shawiya Shenwa Beni-Snous Rif

Sanhaja Tamazight

Figure 15.5. Tree illustrating relationship of Matmata with Central Amazigh languages.

A corollary inference in a case like this – where the highest percentages of cognate sharing are with nearby languages, and most of the low figures are with languages far away – is that inter-language borrowing since the breakup of the common proto-language has slightly inflated the figures for

https://doi.org/10.1017/9781108634311.015 Published online by Cambridge University Press

471

472

Christopher Ehret

nearby languages. Thus, the 55 per cent figures of Matmata with its far distant sister languages probably give a better sense of the actual time depth than does the median figure, 58 per cent. This consideration would be in keeping with a dating of the consolidation of this society to around the beginning of the first millennium AD. The societal divergencies depicted by this overall tree, in conjunction with the desert-fringe location today of the Matmata branch in Tunisia, support the conclusion that the ancestral Wide Central Amazigh society took shape among Amazigh communities living on the desert margins of southern Roman Africa and adjacent parts of modern-day eastern Algeria. The proposed location of the proto-Wide-Central Amazigh society, on the fringe or just beyond the fringe of the Roman world, would be in keeping, as well, with proposals that the subsequent proto-Central Amazigh spread around the fourth to fifth century emerged out of the eastern part of this same zone. On a tangential but germane note, scholars have previously applied the group name Zenati to many of the languages identified here as forming the Wide-Central Amazigh group.8 There are two issues with this name. The lesser issue is the lack of full congruence between the proposed memberships of Zenati and Wide-Central Amazigh. As recently defined by Kossman,9 Zenati does mostly coincide with Wide-Central Amazigh. But it also includes a couple of languages that the evidence presented here would exclude, and it leaves out one language, Central Atlas Tamazight, that the cognation evidence as well as the evidence of lexical innovation strongly support placing in the West Central Amazigh subgroup. It also encompasses communities, notably the Sanhaja of Srair, who do indeed speak a Central Amazigh language today, but who are not Zenata in their historical affiliations. The more important problem from an archaeological and historical point of view, as the instance of Sanhaja of Srair reveals, is that Zenati is an historical anachronism. Three major Berber tribal groupings dominated the Maghrib between around AD 1000 and 1500 – the Masmuda, roughly 8

9

Kossman (1999) cites a number of grammatical innovations supporting the Zenati grouping. Not all of them occur in all the languages that have been classified here as Wide-Central Amazigh, and some of the shared features overlap into neighbouring languages, such as Nefusa, the cognation percentages of which indicate strong word borrowing interactions with, in particular, speakers of the Oasis languages of Central Amazigh. Grammatical borrowing is less common than word borrowing, but it does occur, especially between closely related languages and especially in the earlier periods of their divergence from a common ancestry. This kind of borrowing from Central Amazigh languages is thus certainly possible in the case of Nafusa. Kossman 1999.

https://doi.org/10.1017/9781108634311.015 Published online by Cambridge University Press

15 Berbers in the Sahara and North Africa: Linguistic Historical Proposals

473

corresponding with the medieval Shilh; the Sanhaja, primarily composed of Zenaga-related groups in the western Sahara and western Maghrib; and the Zenata, whose affiliates inhabited most of the remaining areas from eastern and north-eastern Morocco to Tunisia and from whose name the term Zenati comes. The claim by particular tribes to a Zenata affiliation may indeed be rooted in attenuated historical links back to the period of the early Central Amazigh linguistic expansion, of the fourth to sixth centuries AD, but if so, that possibility needs to be separately established. To avoid the risk of reading more recent history into the deeper past, it is important to apply historically neutral names to the linguistically reconstructed societies of earlier times.

Extending the Story Still Farther Back in Time How do Amazigh languages farther to the east and south connect with the history we have considered so far? Most of the languages spoken east of Algeria and southward through the Sahara appear to stand off from WideCentral Amazigh at about the same distance of relationship as does Shilh. Including the cognation figures for those languages, but leaving aside for the moment Zenaga, Siwa and Kabyle, yield an expanded matrix (Fig. 15.6). Tuareg 79 Ghat 51

56

Ghadamis

56

56

61

Nafusa

47

46

57

57

56 56

55 53

55 55

61 63

58 62

51

52

51

52

62

59

66

Shawiya

49 49 52 52

52 47 49 51

51 49 48 52

51 53 50 56

59 55 55 55

61 63 58 65

66 63 59 64

63 65 60 64

47

46

48

55

46

54

52

55

Matmata Wargla 69 Mzab

Beni-Snous 69 Rif 66 69 64 68 56

60

Sanhaja 71 Tamazight 68

66

Shilh

Figure 15.6. Matrix tracing relationship between Wide-Central Amazigh and various Saharan groups.

Nafusa, like Shilh, has a dual set of cognation figures with respect to the Wide Central Amazigh branch. Three individual Nafusa cognation

https://doi.org/10.1017/9781108634311.015 Published online by Cambridge University Press

474

Christopher Ehret

percentages, with Wargla and Mzab of Central Amazigh and with nearby Ghadamis fall in the low 60s. But the remainder form a tight distribution covering roughly the same range as those of Shilh with Central Amazigh and having a median figure of 52, close to the 54 of Shilh with Central Amazigh: 50

51

52

53

56

The figures for Ghadamis spread across roughly the same range as those of Shilh with Wide Central Amazigh, at 46–55, with almost the same median, 51.

48

51 51 51

49

55 55

52

The Tuareg dialects sampled here, Ghat and Tuareg of Hoggar, have figures with Nafusa, Shilh and Wide Central Amazigh, which again cover a similar range:

46

49 49 49

47 47

51 51 51

52 52 52 52

53

56 56 56 55 56

Factoring in geographical distance shows that this distribution falls, however, into two groupings of figures. The communities that lay along or close to the Tunisian and western Libyan routes of the Trans-Saharan trade routes of the past 1,300 years, and close to each other – Ghat and other Tuareg, Ghadamis, Nafusa, Wargla and Mzab – speak languages whose percentages of cognation with one another mostly cluster at the high end of this distribution. These percentages are shown in italics in Figure 15.6:

51

53

55 55 55

56 56 56 56

Even higher figures, noted in boldface, connect Nafusa with the languages of the geographically closest desert oasis trading communities – Ghadamis with Nafusa at 61 per cent and, separately, Nafusa with Wargla and Mzab at 61 and 63 per cent respectively. Here we again encounter a common phenomenon of linguistic history – a point already touched on in the discussion of Matmata: cognation counts among related speech communities tend to be progressively higher the nearer geographically and the more intertwined

https://doi.org/10.1017/9781108634311.015 Published online by Cambridge University Press

15 Berbers in the Sahara and North Africa: Linguistic Historical Proposals

historically the communities have been over the long term. What takes place with this kind of history is that, as the daughter languages diverge out of their common protolanguage – and as they continue overall to gradually change and diverge from one another over the centuries – their geographical and historical contacts sporadically encourage a counter-effect, namely the diffusion from time to time of new words or new grammatical usages from one evolving sister language to another. As a result, geographically closer languages in a cluster of evolving sister languages tend to have higher percentages of cognate words and more grammatical sharings with one another than they do with their otherwise equally distantly related sister languages evolving farther away. An additional factor in the case of Nafusa, the lands of which lay adjacent to the outlets of the Saharan trade, is likely to have been long-term close relations with the merchant traders who spoke the Oasis languages. If one sets aside the figures from the languages that have long been in historical contact, the distribution of the remaining figures for Tuareg, Ghadamis, Nafusa and Shilh with Central Amazigh strongly approximate a bell curve, with its median at 51 per cent:

46

47 47

48 48

51 51 51 51 51

49 49 49 49

52 52 52 52 52

53

55

56

If we then divide this distribution into two, one for the Ghadamis and Nafusa and one for the Tuareg counts – again leaving aside the figures for the pairs of languages long in historical contact with each other – this picture still holds. In the Ghadamis case, the median lies again at 51 per cent. In the Tuareg case, the median is between 49 and 51: Ghadamis and Nafusa versus Central Amazigh

48

49

50

51 51 51

52 52

53

56

Tuareg versus Wide Central Amazigh

47

49 49 49

51 51

https://doi.org/10.1017/9781108634311.015 Published online by Cambridge University Press

52 52 52 52

475

476

Christopher Ehret

The overall implication of these distributions is that Tuareg, Ghadamis, Nafusa, Shilh and Wide-Central Amazigh constitute five co-distant, sister subgroups of a wider branch to which we can give the geographically descriptive name, East-West Amazigh.

The Question of Siwa Interpreting the human history reflected in these figures entails undertaking one more task, preliminary to formulating a tree of these relationships. That task in turn leads to the discovery that a sixth branch, consisting of the Siwa language, is best also included in East-West Amazigh. Zenaga 31

Siwa

34 34

46 48

Tuareg 79 Ghat

34

44

51

56

Ghadamis

38

47

56

56

61

Nafusa

38

51

47

46

57

57

41 38

49 52

56 56

55 53

55 55

61 63

58 62

Wargla 69 Mzab

41

50

51

52

51

52

62

59

66

Shawiya

38 36 41 42

51 47 49 47

49 49 52 52

52 50 49 51

51 54 48 52

51 53 50 56

59 55 55 55

61 63 58 65

66 63 59 64

63 65 60 64

44

41

47

46

48

55

46

54

52

55

56

60

68

66

Shilh

38

38

41

42

42

46

40

51

47

51

45

53

56

55

48

Matmata

Beni-Snous 69 Rif 66 69 Sanhaja 65 68 71 Tamazight

Kabyle

Figure 15.7 Full Berber matrix of cognate counts.

First, as the matrix in Figure 15.6 reveals, the recurrent median of around 51 per cent for the inter-subgroup cognation figures operates for subgroups that are contiguous or would have been contiguous before the intrusion of Arabic into many areas. But for the subgroups at the far ends of the distribution, the cognation counts drop markedly lower – to 46/47 per cent for Shilh versus Tuareg and to 48 per cent for Shilh versus Ghadamis. As we have already seen, this pattern is the typical outcome of an initial stage of history in

https://doi.org/10.1017/9781108634311.015 Published online by Cambridge University Press

15 Berbers in the Sahara and North Africa: Linguistic Historical Proposals

which the speakers of the protolanguage spread out relatively rapidly and evenly over an extended set of new territories. The protolanguage in consequence begins to diverge into a chain of dialects extending across those areas. Because of contacts, this continues among the adjacent emerging dialects along the chain; the sister languages of later eras that evolve out of these dialects will tend to show successively lower percentages of cognation with one another the farther apart they are in the chain. The corollary interpretive principle in such a case is that the lowest percentages in the chain – those least inflated by contacts in the early dialect chaining period – rather than the median cognation, will thus more nearly represent the actual time depth of relationship within the group. The lowest percentage range – of Tuareg, Ghadamis and Matmata at 46–48 per cent with the distant Shilh language – would comport with a median dating of the proto-East-West Amazigh society to the late first half of the first millennium BC. This particular consideration has a further corollary implication for how we must treat the evidence from the Siwa language, spoken far to the east in Egypt. In the full Berber matrix of cognate counts (Fig. 15.7), the Siwa figures with the five previously proposed sub-branches of East-West Amazigh approximate a bell curve, with six out of the thirteen counts clustering at 46–49 around a median of about 47/48 per cent:

41

44

46 46

47 47 47

48

49 49

50

51

52

The median of 47/48 per cent for Siwa’s relationship to each of Tuareg, Ghadamis, Nafusa, Shilh and Wide Central Amazigh coincides with the Shilh low figures of 46–48 per cent with the geographical distant Tuareg, Ghadamis, and Matmata languages. The relationship of Siwa to each of the previously identified branches, in other words, lies at the same time depth as the relationships of those branches to one another. Siwa is therefore a sixth subgroup of the East-West Amazigh branch. Figure 15.8 portrays the overall tree of East-West Amazigh relationships (the Awjilah language has been very tentatively added to the tree as a seventh subgroup, but this proposal remains to be properly established). The linguistic geography of the subgroups strongly favours a central place of origin for the East-West Amazigh branch. Four of the six (or seven) subgroups are today spoken in Tunisia and adjacent parts of Libya and Algeria. Tuareg expansion into the Sahara, beginning about a millennium ago, emanated out of the areas of Libya and Algeria

https://doi.org/10.1017/9781108634311.015 Published online by Cambridge University Press

477

478

Christopher Ehret

1000 BC proto-East-West Amazigh (median: 47/48%) 600 BC? 500 BC

proto-Wide-Central Amazigh (median: 58%)

0 AD 100?

proto-Central Amazigh (median: 63/64%)

AD 400? AD 500

proto-West-Central Amazigh (median: 68/69%)

proto-Oasis (median: 69%)

AD 600?

proto-Tuareg (median: 79%)

AD 1000 AD 1100

ilh Sh

ig az m Ta

nh

aj

h

a

if R

no -S ni

Sa

us

a w en Sh

Be

b

a iy

za

aw Sh

a gl

at

ar W

m at M

M

a

a N

da ha

af

m

us

is

t ha G G

Aw jila Ah Tu agg ar ar eg

Si

w

a

AD 1500

Figure 15.8. The overall tree of East-West Amazigh relationships.

immediately south of Tunisia, probably from around where the Ghat dialect is spoken today. The speakers of Nafusa and Ghadamis languages reside, respectively, immediately east and south of Tunisia, while the proto-Wide-Central Amazigh language, as already argued, most probably originated in southern Tunisia and adjacent desert-fringe areas of far eastern Algeria. Just two branches, Shilh far to the west and Siwa far to the east, have spread out more distantly. The linguistic geographical indications are very strong for placing the proto-East-West Amazigh society in Tunisia and Libya.

More Distant Relationships: Kabyle and Zenaga Left aside thus far have been the relationships of the East-West Amazigh branch to Kabyle, spoken in northern Algeria, and to Zenaga, spoken today by remnant populations in Mauretania. The Kabyle cognation figures present a pattern similar to those of Shilh and Siwa with the East-West Amazigh branch. Its figures with languages located in distant areas – with Zenaga, Siwa, Tuareg, Ghadamis, Nafusa and Matmata – distribute compactly, with a median of just 40/41 per cent, markedly lower than the low internal range of 46–48 of the East-West Amazigh branch (Fig. 15.7):

https://doi.org/10.1017/9781108634311.015 Published online by Cambridge University Press

15 Berbers in the Sahara and North Africa: Linguistic Historical Proposals

38 38

40

41 41

42

46

On the other hand, the Kabyle lands lie in portions of northern Algeria where, for centuries before the expansion of Arabic, the surrounding languages would all have belonged to the West-Central Amazigh subgroup. The impact of encounters with these peoples is clear in Kabyle’s anomalously high counts in the mid-50s with Rif, Sanhaja and Tamazight. In this case, the source of the contact influences in Kabyle were WestCentral Amazigh languages. Word borrowing from the languages of one subgroup into the language of a different subgroup has a particular giveaway side effect. Many of the borrowed words occur of course not just in the particular donor language of the words, but also in other languages of the subgroup to which the donor language belongs. Because of this fact, the apparent cognate counts of the borrowing language with languages closely related to the donor language can be expected also to rise, but not by as much as they will with the language or languages from which the borrowings directly come. Kabyle’s significantly over-high figures of 56, 55 and 53 per cent are with Tamazight and with the nearby Rif and Sanhaja languages. With other, geographically more distant Central Amazigh languages – in keeping with the normative patterns of this kind of borrowing – the Kabyle counts are somewhat raised above the median, at 45 and 47 and, in two cases, reach 51 per cent, but are not skewed nearly as high as the Kabyle counts with Rif, Sanhaja and Tamazight. As in the case of East-West Amazigh, the actual time depth of Kabyle’s relationship to the rest of Berber is better indicated by its cognation with geographically distant languages at 38–46 per cent, with a median of 40/ 41 per cent. The figures for Zenaga present a parallel analytical challenge. They run lower overall than the Kabyle counts with East-West Amazigh languages, distributing from the low 30s into the low 40s (Fig. 15.7).

31

34 34

36

38 38 38

41 41 41 42

44

Geography features at both extremes of this distribution: Zenaga’s highest counts, at 44 and 42 per cent, are respectively with the language geographically closest to it, Shilh, and with the language second closest, Tamazight of the Moroccan Atlas. Zenaga speakers must once have been a significant population element in Morocco, and the importance of the Sanhaja tribal groups in the medieval record of the Western Sahara and parts of Morocco suggests that this presence remained a significant factor

https://doi.org/10.1017/9781108634311.015 Published online by Cambridge University Press

479

480

Christopher Ehret

into the period between AD 1000 and 1500. The low Zenaga figures are with geographically distant languages, such as Tuareg and Ghadamis at 34 per cent. The lowest figure of all, 31 per cent, is of Zenaga with the language farthest away of all, Siwa, 4,500 km to the north-east, in Egypt. As already set out for the East-West Amazigh branch and Kabyle, and for the same reasons, Zenaga’s low cognate figures with geographically distant languages, at 31 and 34 per cent, probably offer a more accurate sense of the actual time depth of the divergence of Zenaga from the rest of Berber.

Summing Up: Linguistic Proposals on Berber History The evidence presented to this point identifies three deep sub-branches of the Berber branch of the Afroasiatic family, Zenaga, Kabyle and East-West Amazigh. Zenaga, with the lowest cognation figures of all, reaching 31–34 per cent with geographically distant languages of the branch, constitutes one primary sub-branch. The sister primary sub-branch, Amazigh, divides in turn into two sub-sub-branches, East-West Amazigh and Kabyle, with the lowcognation range between geographically distant languages at 38–46 per cent, with a median of 40/41. Figure 15.9 presents the overall classification. This linguistic descent tree portrays a history in which Proto-Berber diverged first into ancient ancestral Zenaga and proto-Amazigh. A range of cognate counts for Zenaga mostly in the 30s, reaching the low 30 per cents with geographically more distant languages, suggests that this initial divergence began in the second millennium BC, probably before mid-millennium and perhaps early in first half of the millennium. This first divergence produced the communities ancestral to the Zenaga, whose histories, as far as is known, played out entirely in the western Berber areas. Some centuries later, probably around 1000 BC, a second divergence took place, with proto-Amazigh diverging into the ancient ancestral form of Kabyle on the one hand and the proto-East-West Amazigh language on the other. The second period of divergence gave rise to both western and eastern populations – the ancestral communities of Kabyle speakers probably in the eastern or central Mahgreb, and the proto-East-West Amazigh, who spread all across the areas of North Africa to the east of the Maghrib. The presence of ‘Libyan’ mercenaries in Egypt and the rise of the Libyan 22nd Dynasty, from around 950 to 730 BC, probably mark the eastern frontiers of this era of expansion. The proto-East-West Amazigh communities of around 1000 BC may for several centuries have maintained the sense of belonging to one

https://doi.org/10.1017/9781108634311.015 Published online by Cambridge University Press

15 Berbers in the Sahara and North Africa: Linguistic Historical Proposals

481

Proto-Berber (median: 38%) 1500 BC?

1500 BC

1000 BC?

1000 BC

Proto-Amazigh (median: 41%)

Proto-East-West Amazigh (median: 47/48%) 600 BC? 500 BC

Proto-Wide-Central Amazigh (median: 58%)

0 AD 100?

Proto-Central Amazigh (median: 63/64%) Proto-Oasis (median: 69%)

AD 400? AD 500

Proto-West-Central Amazigh (median: 68/69%)

AD 600?

Proto-Tuareg (median: 79%)

AD 1000 AD 1100

ga

le

na Ze

ilh

by

Sh

ig az m Ta

Ka

h

if

a

R

aj nh

w

no -S ni

Be

Sa

us

a

a

en Sh

Sh

aw

iy

b

a gl

za M

ar

at m at M

af N

W

us

a

a

is

t

m

ha G

da ha G

jila Ah Tu agg ar ar eg

Aw

Si

w

a

AD 1500

Figure 15.9. The overall classification of Berber languages.

society and therefore developed little in the way of dialect differentiation. Why might the linguistic divergence within proto-East-West then begin a few centuries later? One possible factor may be commercial growth and the establishment of trading towns by foreigners at the coast from the ninth and eighth centuries BC onward. If each different area in this way gained its own outlets and separate ability to exploit the trade, the new relations may well have undermined earlier senses of belonging to a common, wider non-commercial society. The breakup of a sense of belonging to a common proto-East-West society therefore might have been the consequence not of new population movements around the eighth and seventh centuries BC, but of the development of disparate and competing local interests. Finally, around the fourth or fifth century AD, a further major period of divergence appears in the linguistic record, with the expansion of Central Amazigh speakers westward as far as central Morocco. The origin area of these movements cannot be precisely pinned down from the language evidence alone. Broadly speaking, however, the linguistic developments suggest that the source area of this new era of expansion lay in the northern Sahara of far eastern Algeria or possibly in adjacent mountain regions. The language-contact evidence, as discussed previously, shows that the

https://doi.org/10.1017/9781108634311.015 Published online by Cambridge University Press

482

Christopher Ehret

expanding Central Amazigh speakers of fifteen centuries ago very often established themselves in areas of the Maghrib where early forms of Shilh had previously been spoken.

Early Berber and Amazigh Economies An even more important capacity of the linguistic toolkit is that its methods open windows for us into the cultures and livelihoods of ancient peoples. People have words in their languages for things and activities only if they know of those things. If we can reconstruct the presence in a protolanguage of a word for a particular cultural item or activity, then we know that the item or activity, at the least, was part of the cultural knowledge of the people who spoke the protolanguage. If we can reconstruct a cluster of items relating to a particular feature of culture, then we have good reason to think that the feature in question was a salient element in the cultural experiences of those people. The reconstruction of the features of material culture and economy at different stages in the history of a group of related languages in turn greatly enhances the prospects of correlating archaeological with linguistic findings. The reconstructed lexicons of the economic pursuits of the Proto-Berber society provide an especially instructive example of what is required if one is to infer the practices of food production from lexical evidence. First, a caveat – the possession by a protolanguage of a generic word for a particular animal that humans have domesticated is not by itself enough to make the case. The presence in Proto-Berber of a generic term for ‘cow’, for example, is non-determinative of domestication because northern Africa was part of the wild range of cattle. In contrast, however, because the wild range of goats did not include northern Africa, the presence of a term for ‘goat’ in Proto-Berber would indicate that the Proto-Berber speakers at least knew of domesticated goats. But again, the generic term alone does not require that the Proto-Berber themselves raised goats – only that they were in close contacts with people who did raise the animals. So how do we know from reconstructed ancient lexicon that people are raising livestock? One kind of determinative evidence is their possession of breeding terminology – of terms for different life stages and breeding states of particular domestic animals. A second kind of determinative evidence consists of verbs or nouns specifically referencing animal tending, such as verbs for herding animals or for driving them to water or for milking, or nouns for livestock enclosures and for livestock herders. An additional

https://doi.org/10.1017/9781108634311.015 Published online by Cambridge University Press

15 Berbers in the Sahara and North Africa: Linguistic Historical Proposals

indicator sometimes encountered, when domestic stock has high social salience, is the presence of colour terms specifically applying to animals. In the case of the Berber languages a large body of evidence of all these kinds is available. We can already reconstruct verbs and nouns for the activities of keeping domestic animals, such as words for ‘to herd’ and ‘to milk’ along with a large noun lexicon of breeding terms. Table 15.1 identifies the extensive corpus of terms reconstructed thus far to each of the first three stages of Berber history, the Proto-Berber, proto-Amazigh and proto-East-West Amazigh periods. From the lexicon in Table 15.1, we can see that the Proto-Berber raised cattle, sheep and goats, and they milked their animals as well. At least three terms for equids reconstruct to the Proto-Berber language. Since the donkey had long been a domestic animal farther east in Saharan Africa, the Proto-Berber may well already have kept them also.10 The wild range of donkey, though, used to include parts of northern Africa, so we cannot rule out the possibility that the Proto-Berber word for ‘donkey’ might have referred to the wild counterpart. A second reconstructed root for an equid specifically denoted the horse. The lexical indications of horse raising – the presence of separate terms for mare and for foal and colt – trace to the following proto-Amazigh period and hence to the second half of the second millennium BC. Here the evidence may fit well with what we know from other historical sources. This dating would be in keeping with the arrival of horses in Egypt around the beginning of the Second Intermediate Period, with that knowledge then spreading westward to the Proto-Berber soon after, but with the actual breeding of horses taking hold only in proto-Amazigh era, sometime later in the second millennium BC. An additional major animal in the economies of more pastoral Berber societies in recent eras has been the camel. A common generic term for the camel occurs across the languages of all three deep branches of the group – as, for example, alɣwem in central Moroccan Tamazight, alɣem in Kabyle and äyi’m in Zenaga. The overall linguistic evidence, however, does not support tracing the raising of this animal back to the Proto-Berber or even to the proto-Amazigh or proto-East-West Amazigh periods. Breeding terms, the determinative evidence of the raising of any particular animal, reconstruct in the case of cows, goats and sheep back to Proto-Berber, and for the horse, back to proto-Amazigh (Table 15.1). But breeding terminologies for camels appear to have come into being only much later, with different 10

Ehret 2011b.

https://doi.org/10.1017/9781108634311.015 Published online by Cambridge University Press

483

484

Christopher Ehret

Table 15.1. Proto-Berber root words showing they raised cattle, sheep and goats and milked their animals. Proto-Berber

184.

*isu *aHeɣu *ilawki *äzger *amäka

722. 193. 226.

131. 713. 368. 579. 439.

*tixsi *tiHeli *iHel*tiHatten *ikrer *abejoj *ibeker *izmer *ižagay *iɣuH*iɣaịḍ *agurH *azay *yyis *agmar *eks *ameksa *zzeg ˙˙

‘cow, head of cattle’ ‘bull calf; heifer’ ‘calf’ ‘bull’ (Baamarani ‘bull’; Zenaga ‘petit boeuf’; Tamashek ‘large calf’) ‘young male domestic animal’ (Tuareg ‘(young?) bull’; Zenaga ‘young goat/ram/bovine/camel’) ‘small stock, sheep, goat’ ‘ewe’ ‘sheep’ ‘ewe’ ‘ram’ ‘wether’ (Zanaga ‘castrated camel’) ‘young ram’ ‘lamb’ ‘(young?) billy-goat’ ‘nanny-goat; ewe’ ‘kid’ ‘castrated animal’ ‘donkey’ ‘horse’ ‘unidentified equid’ (Zenaga ‘equid (gen.)’) ‘to herd’ ‘herder’ ‘to milk’

Proto-Amazigh

368. 207. 172. 617.

*iberkew *Huwal*akrew *aHaulil *agmar *aHug *aHeras *ik’efay

‘calf not yet weaned’ (Tuareg; Kabyle) ‘small livestock (gen.)’ (Siwi ‘ewe’; Kabyle ‘goat’) ‘lamb’ ‘donkey’ (Tuareg ‘young donkey’; Kabyle ‘wild donkey’) ‘mare; horse’ (see Proto-Berber entry for this root) ‘foal, colt’ ‘red/brown’ (animal colour) ‘fresh milk; fresh (of milk)’

https://doi.org/10.1017/9781108634311.015 Published online by Cambridge University Press

15 Berbers in the Sahara and North Africa: Linguistic Historical Proposals

Table 15.1. (Cont.) Proto-Berber Proto-East-West-Amazigh

708. 180.

*-funas*aziḍ ˙ *aHenbay

‘cow’ ‘donkey’ ‘having a black head’ (of an animal)

Reconstruction notes: 1. The numbered roots in this table are from Kossmann (1999); the roots without numbers are the author’s reconstructions, using Kossmann’s criteria. 2. The consonant represented by H was a labial consonant of uncertain pronunciation.

Berber languages and language subgroups, such as Zenaga and Tuareg, each developing their own separate terminologies. The implications of this evidence are twofold: 1. the generic term for camels does not go back to Proto-Berber, but spread later with the extension of the knowledge of the animal; and 2. camel raising was not present until after, and probably well after, the proto-East-West-Amazigh period of the middle or early first millennium BC. An introduction of the camel to North Africa late in that millennium or at the beginning of the first millennium AD would fit this evidence quite well. The Proto-Berber clearly also made use of grains in their diet, as their reconstructed root words in Table 15.2 for cooked grain, bread and bread dough. But were they cultivators of those crops? To consider this question, as for livestock raising we must identify just what it is that constitutes determinative evidence. The most direct evidence is the presence in the protolanguage of verbs specifically denoting the processes of cultivation and of nouns specifically referring to cultivated land. As was the case for introduced animals, the presence of the generic term for a particular crop shows knowledge of the crop among the speakers of the protolanguage, but does not by itself show that they themselves cultivated it. The Proto-Berber society’s word for the date (an introduced, cultivated crop) shows them to have, at the least, been in contact with cultivators and to have known of dates. The word itself testifies that dates had been earlier introduced from the east: it was a loanword in Proto-Berber, coming ultimately from Middle Egyptian. At the next stage in the Berber linguistic history, proto-Amazigh, the available evidence does include a term specifically for cultivated ground, as well as additional terms re-emphasising the significance of grains in their

https://doi.org/10.1017/9781108634311.015 Published online by Cambridge University Press

485

486

Christopher Ehret

Table 15.2. Proto-Berber root words for cooked grain, bread and bread dough. Proto-Berber 196. 209.

*tiHeyni *aHun

413.

*taglal-

‘date’: loan-word from ancient Egyptian (Middle Egyptian bni) ‘cooked grain’ (in different languages this term names different grains and manners of cooking them) ‘bread’

Proto-Amazigh 411. 588. 171. 266. 578.

*tigemmi *takidirt *aHern *ibzan *arikt

‘cultivated land’ ‘ear of grain’ ‘flour’ (cf. also Mzab wirin < *weHern) ‘semolina’ ‘bread dough’

Regional agricultural loan-words of Roman origin in Maghrib languages 395. 416. 361. 362.

*targa *iger *tayuga *ayug

‘irrigation channel’ ‘(ploughed) field’ ‘pair (of oxen)’ ‘ox (especially for ploughing)’

An additional agricultural word of regional spread only in Maghrib 569.

*krez

‘to plough’

Reconstruction note: The reconstructions presented here are from Kossman (1999). The consonant represented as H was a labial consonant of uncertain pronunciation.

diet. A verb lexicon of cultivating activities has yet to be reconstructed, however, to either era. Taking this evidence at face value, one could argue that the Proto-Berber and their proto-Amazigh descendants of the second millennium BC were above all livestock raisers, although acquainted with agricultural activities and with agricultural foods. We could, in this light, understand these peoples as initially having been pastoralists living around and alongside more sedentary cultivating communities and trading for food with them. One factor most probably skews our understandings, however – that our evidence from one entire primary branch of Berber, Zenaga, comes from the

https://doi.org/10.1017/9781108634311.015 Published online by Cambridge University Press

15 Berbers in the Sahara and North Africa: Linguistic Historical Proposals

last remaining language of that group, which has been spoken for probably the past 2,000 years far to the south, in tropical desert steppe environments where only pastoral pursuits are possible and where the crops of North Africa and the northern half of the Sahara mostly do not successfully grow. To reconstruct a word back to a proto-language, we need to locate its modernday reflexes in at least one language from each of at least two of the primary branches of the language group. Berber has two primary branches, Amazigh with its numerous languages versus the single-language branch, Zenaga. To attribute a word to Proto-Berber we must therefore identify it both in the one existing Zenaga language and in at least one Amazigh language. If the Zenaga did once upon a time possess additional words for cultivation cognate with those found in Amazigh languages, the movement 2,000 or more years ago of the Zenaga southward into the tropical desert and desert steppe environments of the western Sahara would have rendered many of those words obsolete. If Amazigh languages retained a particular Proto-Berber root word, but Zenaga lost it, we would no longer be able to reconstruct that word with certainly back to Proto-Berber, even though it had actually been in that language. If the Proto-Berber were indeed, at least to some extent, cultivators of crops, as seems probable, they seem very likely not to have used the plough in their farming. Multiple Latin loan-words related to ploughing and one loan-word for irrigation (Table 15.2) – each with a general regional distribution across the Central Amazigh, Shilh and Kabyle languages of the Maghrib – affirm the history of agricultural intensification that we know from other sources to have taken place in the Roman period.

Early History of the Chadic Branch Chadic, which with Berber forms the Chado-Berber branch of Afroasiatic, has a far larger number of languages than Berber and a much greater time depth, and a majority of the Chadic languages have been less closely studied and are less well known than the typical Berber language. Nevertheless, the overall number of Chadic languages for which some lexical data exists provides a wide sweep of evidence for historical reconstruction and allow one to paint a broad picture of the long-term history of Chadic societies. The distributions of the present-day Chadic languages imply a settlement history in which, following on the breakup of the protoChadic society – at a date sometime, as proposed here, around 5000 BC –

https://doi.org/10.1017/9781108634311.015 Published online by Cambridge University Press

487

488

Christopher Ehret

the early daughter communities of proto-Chadic fairly rapidly expanded all across the southern Lake Chad Basin. Scholars have previously divided Chadic into either three primary branches or four primary branches.11 The geographical distribution of the branches, whether three or four, in subsequent eras supports a scenario in which the initial period of settlement established the early speakers of Chadic languages across a belt of lands, from as far west as probably the modern western Hausa-speaking areas in northern Nigeria to as far east as the areas around the Guerra mountains of the present-day country of Chad. These expanding daughter communities, it can be proposed, moved into regions previously inhabited by peoples who spoke languages of the Western Sahelian branch of the Nilo-Saharan family. The presence of Western Sahelian loan-words in proto-Chadic, including even a few core vocabulary items, provides the primary support for this conclusion.12 The geography of the extant Western Sahelian languages backs up this historical inference as well. The modern-day Western Sahelian languages fall into two subgroups that are distant in relationship to each other and also geographically distant from each other.13 One Western Sahelian subgroup, Songay, consists of a single language, although with a number of distinct dialects. Its speakers inhabit the areas of the country of Niger immediately ‘west’ of the westernmost modern Chadic language, Hausa, and also the areas farther west along the great bend of the Niger River in Mali. The speakers of the second subgroup of Western Sahelian languages, Maban, reside at the other geographical extreme, in eastern Chad, just ‘east’ of the easternmost modern-day Chadic societies. The presence of Chadic languages in all the intervening lands between Songay and the Maban languages, in conjunction with evidence of the early Western Sahelian loan-words in Chadic, makes it highly probably that Western-Sahelianspeaking peoples previously, perhaps in the earlier sixth millennium BC, occupied all the areas from Wadai in the east to the bend of the Niger in the west. The spread of the descendant communities of the proto-Chadic society across the southern Lake Chad Basin, it can be argued, broke up a former continuity of Western Sahelian peoples across those regions and over time incorporated those peoples into the newly emerging Chadic societies. The proto-Chadic linguistic evidence allows an additional interesting proposal. The incoming Chadic speakers seem likely to have been the 11

12

Three branches: Jungraithmayr and Shimizu 1981; Jungraithmayr and Ibriszimow 1994; four branches: Newman 1977. Ehret 2006. 13 Ehret 2001, Chapter 4.

https://doi.org/10.1017/9781108634311.015 Published online by Cambridge University Press

15 Berbers in the Sahara and North Africa: Linguistic Historical Proposals

introducers of the knowledge of cattle to the peoples of still another language family, Niger-Congo, who have long inhabited the savanna and rainforest belts farther south, between the Lake Chad Basin and the Atlantic coast. This is at least the implication of the very wide diffusion of one of the two proto-Chadic generic root words for ‘cow’, *nakw-, in a modified shape *naka or *naNka to a great many of the Niger-Congo peoples of West Africa. But where did the immediate forebears of the proto-Chadic people, the pre-proto-Chadic society, reside before the spread of their descendant societies across the southern Lake Chad Basin? Two pieces of evidence would best make sense if the proto-Chadic speakers had moved south into the Chad Basin from areas in the Central Sahara west of Tibesti. The words for two domestic animals, sheep and donkeys, in the Saharan branch of the Nilo-Saharan family are loan-words from early Chadic. The proto-Saharan language was spoken in the eastern parts of Central Sahara, in Tibesti and around the northern sides of Lake Mega-Chad, during the same broad era as the initial Chadic expansion into the southern parts of the basin.14 A plausible location for the pre-proto-Chadic society that would account for these contacts and allow for the subsequent expansion of their protoChadic descendants into the areas south of Lake Mega-Chad would be the areas west of the proto-Saharans. Ténéré and the Akakus might be promising areas to look for this population. Partly because the Chadic branch is of significantly greater time depth than Berber, and partly because the lexicons of herding and cultivation have been inadequately collected for most of the Chadic languages, the selection of root words semantically determinative of food production that we can currently reconstruct for proto-Chadic are fewer than for ProtoBerber. The data nonetheless are sufficient to make the case that the protoChadic people of around 7,000 years ago were herders of livestock and, in all probability, cultivators of crops. All the reconstructed proto-Chadic names for domestic animals – cow, sheep, goat and donkey – are for animals that are non-native to the countries in which the proto-Chadic people would have lived. That fact is enough to show that the proto-Chadic acquaintance in each case was most probably with the domestic version of the animal in question. So far the only breeding terms for which we have extensive collections of data are for the males of domestic animals. But linguists have recorded these terms all across the family, and this work allows us to reconstruct a complex of 14

Ehret 2011a ; Saxon 1980.

https://doi.org/10.1017/9781108634311.015 Published online by Cambridge University Press

489

490

Christopher Ehret

proto-Chadic root words for male domestic stock (Table 15.3). That evidence is enough on its own to certify to the keeping of domestic cows, sheep and goats by the proto-Chadic society. But a much fuller set of terminology could have been reconstructed, with terms for the young of the animals and probably for castrated livestock, had more of the collectors of Chadic paid attention to the other life and reproductive stages of domestic animals. At least three nouns for cultivated fields trace back to the proto-Chadic language, certifying to the practice of some kind of cultivation by the proto-Chadic communities. In those fields, it appears, the proto-Chadic speakers tended a variety of tropical African food plants, including three grain species, yams and, most strikingly, a considerable variety of melons and calabashes (Table 15.4). The proliferation of terms for melons and calabashes has a potential further implication. A proliferation of crop names in this manner is commonly an indicator that an accompanying proliferation of crop varieties has been taking place. In turn, that kind of proliferation is typically the outcome of the domestication of those crops. The lexical evidence suggests, in other words, that proto-Chadic cultivators of c.7,000 years ago were engaged not just in pre-domestication cultivation, but were already domesticating at least the cucurbits. The probable broad contemporaneity of the proto-Chadic society with the Mid-Holocene phase of drier climate offers a possible enabling context for the proto-Chadic population expansion from the Central Sahara into the southern Lake Chad Basin. Scholars have tended often to see drier conditions in the Sahara as pushing people to migrate southward. But the more important factor in the Mid-Holocene might have been opportunity. Drier climatic conditions would have opened up new lands to the expansion of Chadic peoples southward. A drying of climate would have provided new settlement opportunities because it would have led to the retreat of bush and woodland in the southern Lake Chad Basin and to the expansion of grassland areas well suited both to the grazing of livestock and the raising of the steppe and dry savanna crops of the proto-Chadic society, such as melons and calabashes.

What to Look For? The overall longue durée historical picture that emerges from the considerations of the Berber and Chadic linguistic records offers a variety of suggestions of what we might look for as our archaeological information

https://doi.org/10.1017/9781108634311.015 Published online by Cambridge University Press

15 Berbers in the Sahara and North Africa: Linguistic Historical Proposals

491

Table 15.3. Proto-Chadic root words for male domestic stock. Proto-Chadic cow

*ɬa *nakw-

bull

*mə(n)ts-

male domestic animal goat billy-goat

*təŋgal*akw*bəgr-

billy-goat sheep ram

*bənsər*təm(ək) *əgəm

ram/billy-goat

*mbak-

donkey

*kwaara

proto-Chadic: occurs widely in all sub-branches WCh: Kir; CCh: Njanye, Mwulyen, Bacama; Masa: Lame, Peve, Zime WCh: Goemai mIs, Gera mizi; CCh: Wamdiu mıcu, Higi Nkafa mca, Higi Baza mce, Kapsiki mca, Higi Futu mcu; Masa: Lame mənʃi CCh: Fali Jilbu tıŋgeli, Fali Mucella, Gude taŋgara ‘billygoat’; ECh: Mubi teŋgel ‘bull’ proto-Chadic: occurs widely in all sub-branches WCh: Karekare bugure, Pidhimdi pokirti; Wandai bəgala, Geji bagəla ‘ram’; CCh: Kilba, Hildi bura, Fali Kiria bʊraa, Mwulyen bwaara, Bacama bɔgʊrrɛy, Gudu bwarə (also BERBER: Proto-Berber *ibeker ‘young ram’) WCh: Hausa; Ngizim; CCh: Glavda, Dghwede, Gava proto-Chadic: occurs widely in all sub-branches WCh: Angas-Sura *ŋgam; Goemai, Karekare, Bole, Ngamo, Dera, Ngizim, Bade; CCh: Hwona, Bura, Cibak, Margi, Ngwaxi, Kilba, ram/billy- Hildi, Wamdiu, Margi, various Higi; Masa: Lame, Masa, Peve WCh: Dwot mbakn ‘he-goat’; Kir mbakləm ‘he-goat’, mbakparım ‘ram’; Wangdai bəgwunʒi, Geji bagonmiʒi ‘he-goat’; Dwot mbakn ‘he-goat’; CCh: Fali Jilbu/Mucella bəga, Fali Bwagira bəgən ‘sheep’; Mwulyen mbaaga, Bacama mbaga ‘ram’; Glavda mbakulaka, Gava mbakuləka ‘ram’ proto-Chadic: occurs widely in all sub-branches

Reconstruction note: The phonological reconstructions of the Chadic roots in this table and in Table 15.4 follow the rules of vowel and consonant correspondence, as laid out in Ehret (1995). These rules combine the findings of Newman (1977) on Chadic vowels with those of Jungraithmayr and Shimizu (1981) and Jungraithmayr and Ibrizsimow (1994) on the consonants. An important source for herding and farming terms in many of the little-known Chadic languages is Kraft (1981).

grows. In the early Holocene, communities speaking languages of the Chado-Berber branch of the Afroasiatic language family became important historical actors across probably considerable parts of the northern half of the Sahara. These peoples would initially have spread from the east across the Sahara, taking advantage of the amelioration of climate in the early Holocene to introduce the older Afroasiatic mixed hunting and wild-graincollecting economy to these regions.15 15

Ehret 2011a.

https://doi.org/10.1017/9781108634311.015 Published online by Cambridge University Press

492

Christopher Ehret

Table 15.4. Proto-Chadic speakers tended a variety of tropical African food plants. Proto-Chadic cultivated plot *wiy-

sorghum

*mar*wal*wəxiya

*dəwrPennisetum

*məɗ-

fonio (?)

*gər-

yam

*dawy-

melon

*(m)bawd-

melon melon calabash

*ɓok’o *kab*ɗay(w)-

calabash

*kwəɗ-

calabash

*kwal-

WCh: Tangale yeyə, Kir wiye, Boghom wiye; CCh: Njanye wuyə, Mwulyen wiyo; ECh: Mubi wiya WCh: Angas, Cip, Sura, Goemai, Bole; ECh: Dangla WCh: Guruntum wal; CCh: Fali Jilbu (w)ole WCh: Wangai, Kir, Geji, Buli, Dwot, Palci; CCh: Ngwaxi, Kilba, Hildi, Wamdiu, Margi, various Higi, Fali Gili, Wandala, Glavda, Dghwede, Gudu; Masa: Lame ‘sorghum’, Peve ‘Pennisetum’ WCh: Hausa; Sura gyewuro, Wangdai dɘwro, Guruntum gyuuri, Seya gyoro, Dwot dəwro ‘Pennisetum’; CCh: Masa dawn ‘sorghum’ WCh: Karekare, Gera, Gole, Pero, Tangale, Ngizim; CCh: W.Margi, Ngwaxi, Njanye, Gava WCh: Hausa gero; CCh: Higi Futu gərwa, Mwulyen gwoyo, Gudu gəəwaa ‘sorghum’, Mafa gagar ‘Pennisetum’ WCh: Hausa, Gera, Bole, Seya, Bade (dauya); Masa: Masa, Banana, Lame; ECh: Kera ədaw WCh: Boghom mbode; Masa: Masa buda, Banana puda; ECh: Mokilko boode, Bidiya booda CCh: Njanye, Mwulyen; Masa: Lame, Peve, Zime WCh: Hausa kabeewaa; Bade kaabiyan; CCh: Bacama kpa ‘calabash’ WCh *ɗay: Angas ɗe, Sura ɗa, Cip ɗa, Goemai ɗayi; Jimi ɗee; ECh *ɗay-w-: Bidiya ɗawa; Migama ɗewwa ‘melon’ WCh: Gera kwaɗa; Seya kwət; CCh: Dghwede, Nakatsa kwata; Masa: Banana gwətə WCh: Tangale kwali; Seya kwalat; CCh: Gava kwəla

The proto-Chado-Berber society, suggested here to date to around 8000–7000 BC, cannot be said from the available lexical evidence to have yet been raisers of domestic animals or cultivators, and that conclusion comports with what we know of the northern Saharan and North African archaeology of that period. Proto-Chadic and Proto-Berber, the two identified early descendant languages of proto-Chado-Berber, each developed its own almost entirely separate lexicon of food production. A single breeding term for a male goat or sheep does occur in both proto-Chadic and ProtoBerber – proto-Chadic *bəgr- ‘billy-goat’ and Proto-Berber *ibeker ‘young ram’. But, as a solitary item, it may well be an early borrowed term from one language group into the other, perhaps spread in conjunction with the initial diffusion of livestock across the northern Sahara.

https://doi.org/10.1017/9781108634311.015 Published online by Cambridge University Press

15 Berbers in the Sahara and North Africa: Linguistic Historical Proposals

The lexical reconstructions reveal, on the other hand, that the protoChadic communities of around 5000 BC clearly were food producers in economy (Tables 15.3 and 15.4). That conclusion again, certainly in the case of livestock raising, is in keeping with the archaeological evidence in the Sahara. The reconstructed lexicon indicative of cultivation by the proto-Chadic society predicts, moreover, that archaeological investigations should eventually uncover, as well, the presence of at least predomestication cultivation in the Central Sahara by or before 5000 BC, and probably also the early domestication of some crops, notably cucurbits. A likely region in which to look for archaeological markers of the original proto-Chado-Berber society of around the eighth millennium BC would be the central and central-northern parts of the Sahara, from Ténéré and the Akakus to Fazzan. The proto-Chadic language most probably evolved out of the Chado-Berber communities who lived in the southern parts of this expanse. Where the linguistic forebears of the Proto-Berber might have resided in those times is a more difficult question to answer. Their lands probably lay in more northerly areas, possibly in Libya. From the early second millennium BC onward, however, the picture for Berber history begins to take shape. The complex internal relations among the Berber languages and the extensive available linguistic evidence allow much more detailed proposals on the subsequent locations, histories and expansions, especially of the various Amazigh peoples and their languages. The possible archaeological manifestations of these histories are among the topics explored by Elizabeth Fentress in the next chapter.

References Bynon, J. 1984. Berber and Chadic: the lexical evidence. In J. Bynon (ed.), Current Progress in Afro-Asiatic Linguistics: Papers of the Third International HamitoSemitic Congress, London, 1978, Amsterdam, Philadelphia: J. Benjamins, 241–90. Ehret, C. 1995. Reconstructing Proto-Afroasiatic (Proto-Afrasian): Vowels, Tone, Consonants, and Vocabulary. Berkeley, Los Angeles: University of California Press. Ehret, C. 2000. Testing the expectations of glottochronology against the correlations of language and archaeology in Africa. In C. Renfrew, A. McMahon, and L. Trask (eds), Time Depth in Historical Linguistics, Vol. 2, Cambridge: McDonald Institute for Archaeological Research, 373–99. Ehret, C. 2001. A Historical-Comparative Reconstruction of Nilo-Saharan. Köln: Rüdiger Köppe Verlag.

https://doi.org/10.1017/9781108634311.015 Published online by Cambridge University Press

493

494

Christopher Ehret

Ehret, C. 2006. The Nilo-Saharan background of Chadic. In P. Newman and L. Hyman (eds), West African Linguistics: Studies in Honor of Russell G. Schuh, Studies in African Linguistics, Suppl. 11, Columbus: Ohio State University, 56–66. Ehret, C. 2011a. History and the Testimony of Language. Berkeley, Los Angeles, London: University of California Press. Ehret, C. 2011b. A linguistic history of cultivation and herding in North-eastern Africa. In A. G. Fahmy, S. Kahlheber and A. C. D’Andrea (eds), Windows on the African Past: Current Approaches to African Archaeobotany, Vol. 3, Frankfurt am Main: Africa Magna Verlag, 196–208. Ehret, C., Keita, S. O. Y. and Newman, P. 2004. The origins of Afroasiatic. Science 306.5702: 1680–81. Jungraithmayr, H. and Shimizu, K. 1981. Chadic Lexical Roots. Berlin: D. Reimer. Jungraithmayr, H. and Ibriszimow, D. 1994. Chadic Lexical Roots. Berlin: D. Reimer. Kraft, C. 1981. Chadic Wordlists. 3 vols. Berlin: D. Reimer. Kossmann, M. 1999. Essai sur la phonologie du proto-berbère. Köln: Rüdiger Köppe Verlag. Newman, P. 1977. Chadic classification and reconstructions. Afroasiatic Linguistics 5.1: 1–42. Newman, P. 1980. The Classification of Chadic within Afroasiatic. Leiden: Peers. Saxon, D. E. 1980. The History of the Shari River Basin, c.500 BC–AD 1000, Unpublished PhD thesis, UCLA, USA. Sapir, E. 1916. Time Perspective in Aboriginal American Culture. Canadian Geological Series.

https://doi.org/10.1017/9781108634311.015 Published online by Cambridge University Press

16

The Archaeological and Genetic Correlates of Amazigh Linguistics elizabeth fentress

Introduction This chapter is written in response to that of Christopher Ehret in this volume, aiming to show how the linguistic ‘family tree’ he sketches for the Tamazight, or Berber, language family might relate to the history of Amazigh peoples.1 This is, of course, a tall order. Any more than an impressionistic correlation between historical linguistics and archaeological evidence is going to fall straight into a series of perfectly well-known pitfalls: that language change does not necessarily imply massive invasions has been constantly repeated, not least by Colin Renfrew, while the various processes by which new languages or their variants may be diffused have been thoroughly theorised, a subset of the diffusion versus acculturation debate. Renfrew, for example, makes a short list of reasons for linguistic replacement that includes: elite dominance, possibly migrationary, which is of course related to the ‘invasion’ hypotheses, the demographic effects of new technology (this relates to his view that Neolithic expansion resulted in the diffusion of the Indo-European language family), the effects of system collapse, the consequences of the formation of a lingua franca and contact-induced language change, this latter reflecting disease models: both of the final ones fit nicely into the modern diffusion of the use of English.2 Now, North Africa has four very straightforward whole or partial historical language changes, all of which relate to foreign conquest or immigration. The first sees the introduction of Punic around 800 BC, first on the coast but progressively deep into the interior, the second the introduction of Latin from the first century BC, the third the introduction of Arabic, beginning with the initial attacks and 1

2

I am very grateful to Josephine Quinn for her careful reading and comments on this paper, and to the various participants in the conference for their corrections and additions. Maria Gatto saved me from various errors. Maarten Kossmann has given me valuable insights and bibliography on linguistic questions, particularly in sharing his chapter ‘Berber subclassification’ for The Oxford Handbook of African Languages Forthcoming. Renfrew 2002, 65.

https://doi.org/10.1017/9781108634311.016 Published online by Cambridge University Press

495

496

Elizabeth Fentress

Islamisation of the late seventh century AD but penetrating only gradually and incompletely into rural areas over time, and finally the introduction of French with the first French colonial conquests in 1848 and the subsequent use of the language for law, government and instruction. All of these introductions are perfectly well known and have blindingly obvious archaeological correlates, not to mention textual ones. The languages in question were sustained by abundant texts, which maintained a normative effect on the languages, without, however, keeping them static, as the dialectical variation between Tunisian, Algerian and Moroccan Arabic show. What Christopher Ehret has engaged in is a far more difficult process, teasing out the various strands of Tamazight, the Berber languages. Here there is no textual support (beyond a few tombstones and rock-cut inscriptions, and the names on Egyptian documents) and precious little historical or archaeological evidence on which to pin the branching structures he observes.3 These analyses are based on the ratio of similarity between different languages, using a 92-word list as a basis. It is not the first such attempt: indeed, Blažek presented a roughly similar tree in 2010, based on the same evidence.4 But it is unique in proposing a chronological framework for the various new introductions. Now, this is based on an argument that language change, while not constant and based on individually random changes, over long time periods tends to follow normal distributions. The differences between two languages would then have a measurable value that can be roughly quantified in chronological terms.5 While there is some agreement that the earliest measurable separation of the Amazigh languages lies around 2000 BC, there is none for the rest of the shifts. What I want to do here is to write a (pre)historical novel in which his proposal of the dates and form of the branching of the Berber languages (see Fig. 15.9 and Fig. 16.1 for their distribution), as well as the still incomplete information from Berber linguistics, is used to suggest a recurrent pattern of movement from the east and south by mobile Berber groups, fitting into other archaeological and historical information. It follows a recently published article written with Andrew Wilson in which we examined one of the last of these regroupings, the Zenatic, which united the languages of the southern Tell with that of the Western Saharan oases.6 I will briefly summarise the reasoning in that paper further along, but it seems sensible here to start at the beginning of the observed branches from the parent trunk. 3 6

Ehret, Chapter 15, this volume. 4 Blažek 2010. 5 Ehret 2010, 106–7. Fentress and Wilson 2016. The term was coined by the Berber linguist Edmond Destaing (2002 – reprint of 1894 paper).

https://doi.org/10.1017/9781108634311.016 Published online by Cambridge University Press

16 The Archaeological and Genetic Correlates of Amazigh Linguistics

Figure 16.1. Modern distribution of Amazigh languages.

The basic theory would hold that dramatic changes within languages would be triggered by expansions of population, creating incursions into previously occupied territories, and the substitution of one branch or language for another.7 Less dramatic changes are obviously happening all the time, as groups interact with one another, or move away from one another, creating separate and evolutionary changes in what was once a common language, developing slowly diverging daughter languages.8 What we are looking at here are less the various processes involved in branching off, but instead those moments when over a large area it appears that a single dialect has effectively eliminated its contemporaries, becoming the sole ‘parent’ of the subsequent dialects in those area. Now, the archaeological correlates of this sort of thing are themselves extremely impressionistic, and it can be understood in what follows that I am attempting to make a best fit between story and data, rather than proposing any certainties. Indeed, my divergences from Ehret’s interpretations, in what was originally intended to be an entirely complementary paper, are a clear indication that the evidence is so slight, even with this new component, that reasonable people can give very different reconstructions of the history of language diffusion. Genetic data, which, like the linguistic data, derives from modern populations, is used in a similar way, as geneticists attempt to tease out evidence for population movements from observed patterns. Less advanced, it is still worth considering its evidence. 7

Renfrew 2002, 65.

8

Ehret 2010; Chapter 15, this volume.

https://doi.org/10.1017/9781108634311.016 Published online by Cambridge University Press

497

498

Elizabeth Fentress

These differences are clearest in the early stages of this process: indeed, here the story can only be imagined, particularly in my case because I am not in any way a prehistorian, and there is both little agreement among practitioners and a generalised failure to calibrate radiocarbon dates, or, indeed, to cite them in full.

Proto-Berber: 2000–1500 BC (Fig. 16.2) Ehret’s characterisation of the initial stages of the Afroasiatic languages in or near Ethiopia around 13000 BC is highly convincing, although not undisputed by those who cling to a Middle Eastern origin for the family.9 The proto-Afroasiatic language is very old indeed, going back to stages where grain was gathered, but not cultivated and perhaps not even ground.10 Nor are there words in common for herding and milking, which suggests that animal domestication did not appear until the full separation of the language groups. Thus, Afroasiatic as a whole did not spread with Neolithic technologies, but was already divided at their inception. The early speakers of Chado-Berber languages would have expanded westwards across the Sahara between 9000 and 7000 BC, branching into the Chadic languages, settling in the southern and central Sahara and the Chad basin, and the Proto-Berber languages spoken in the northern Sahara and the Maghrib. The exact date of this separation is in no way clear, but it seems not implausible that it coincides with the first evidence of animal domestication in the Nile Valley and the Central Sahara, around 6000 BC,11 or, perhaps even more likely, with the drying of the Sahara a millennium later, which would have separated the two groups into northern and southern branches.12 This is the moment when dairying techniques are first attested in the Libyan Sahara.13 There is, however, no certainty that the hunter-gatherers of the Capsian in the Maghrib spoke a language related to Proto-Berber: they may have, but they also may have spoken an entirely different language (Blench’s ‘Old North African’)14 for which we have no evidence whatsoever. Genetic data, based on samples from modern populations, seems to support this idea. A recent article asserts that the genetic structure of the Y chromosome observed in North Africa, has a distinct east-west cline, 9 11

12

Blench 2013, 54. 10 Ehret 2010. di Lernia 2006, 2013; Guilaine 2011, 81; Blench, Chapter 14, this volume. However, Ehret would place the division somewhat earlier. Broodbank 2013, 204. 13 Dunne et al. 2013. 14 Blench, Chapter 14, this volume.

https://doi.org/10.1017/9781108634311.016 Published online by Cambridge University Press

16 The Archaeological and Genetic Correlates of Amazigh Linguistics

Figure 16.2. The diffusion of Zenaga, c.2000–1500 BC.

meaning specific components become less frequent in the more westerly populations.15 This could have been engendered by early food-producing societies originating in East Africa and relying principally on pastoralism, which would have favoured genetic drift from the east because of their greater mobility.16 They would have expanded across North Africa into an environment containing few humans. This ties only approximately with the archaeological evidence for domestication. While there is evidence for an intensified management of wild aurochs and taming of Barbary sheep in the Nile Valley and the Sahara during the Early Holocene,17 there is no doubt that domesticated cattle, sheep, goats and dogs were introduced from the Middle East by the end of the seventh millennium BC.18 In North Africa, however, there is evidence for considerable holdover from the earlier, Capsian period. At inland Neolithic sites like the Grotte Cappelletti in the northern Aurès and the Dahmous el Ahmar, near Tebessa, the flint tradition remained very conservative, and snails were still an important source of food in the fourth millennium BC.19 Haua Fteah in Cyrenaica, too, shows goats being introduced at a fairly advanced

15 18 19

Arredi et al. 2004. 16 Arredi et al. 2004. 17 di Lernia 2013. Mitchell 2005, sidebar in Chapter 2; Barich 2010, 207; Barker 2013; di Lernia 2013. Barich 2010, 211; Roubet 2011, 13–26. Roubet gives the earliest dates for the two sites as 3965–3575 and 3640–3260 calBC.

https://doi.org/10.1017/9781108634311.016 Published online by Cambridge University Press

499

500

Elizabeth Fentress

level of domestication while the flint tradition remained Capsian.20 In all these cases, there is no evidence for plant domestication: indeed, at Haua Fteah, there is no sign of domesticated grain until the seventh century BC.21 This leaves a very long time, during which we imagine pastoral Proto-Berber speakers interacting at some level with the preexisting Mechtoid populations who lived primarily in cave sites, and whose genes a recent study has identified in the Tunisian oasis town of Chenini.22 In Morocco, the Cardial, a pastoral Neolithic with sheep and goats, was in evidence since the mid-sixth millennium, sharing characteristics with that of the Iberian peninsula, from which it may have arrived and, again, concentrated in cave sites.23 The very slow progression of this Neolithic inland from the coast is suggested by a long period during the fifth millennium BC during which there is evidence for pottery but not domestication.24 Thus what we might suggest was the ‘pre-Berber’ Neolithic in Africa contains both Iberian and earlier, Mechtoid elements, and remained distinctly conservative.25 In the sixth millennium, in the Egyptian Western Desert oases of Farafra and Dakhlah, there is a dense concentration of pastoral sites, again without a trace of cultivated grain, though there is clear evidence for sheep and goats.26 A ceremonial centre at Nabta Playa, in the southern Egyptian Western Desert, goes back to the beginning of the seventh millennium and it was in use until at least the end of the fourth millennium BC.27 What is significant here is the first instance of the tumuli which subsequently became important markers for Saharan populations, spreading into the Northern Sahel in Niger and deep into the Sahara.28 Although the fairly rapid drying-out of the Sahara from

20 23

24

25

26 28

Douka et al. 2013. 21 Lucarini et al. 2015. 22 Fadhlaoui-Zid et al. 2013. The site of Hassi Ouenzga in the Rif and others in the region of Tangiers and Oran, date from the fifth millennium BC, although their excavators tend to associate them with a Mediterranean Neolithic emanating from Spain rather than a south-eastern source. However, a necklace of ostrich shell pearls certainly seems to indicate some Saharan contact (Linstädter 2003, 111). Other sites include Achakar and Dar es Soltan. Hutterer et al. 2014, 85. Note, however, the Cardial Moroccan sites, of the second half of the sixth millennium BC where cereal cultivation, sheep and goats, all near-eastern origin, are found. These seem to be related to the Spanish littoral. Guilaine 2011, 83; Ballouche and Marinval 2003. As Broodbank says, ‘the romantic notion that the Capsian groups of the Mediterranean zone were Proto-Berbers is sadly unlikely’: 2013, 205. See also Blench, Chapter 14, this volume, with the argument that they spoke an otherwise unattested language he calls ‘Old North African’. Barich 2016. 27 Bobrowski et al. 2014. Barich 2010, 216. For the dating of the various types of funerary monument, Paris and Saliege 2011.

https://doi.org/10.1017/9781108634311.016 Published online by Cambridge University Press

16 The Archaeological and Genetic Correlates of Amazigh Linguistics

5000 BC onwards made life there more difficult, there are large numbers of sites with very little cultural material scattered through the Western Desert from Egypt to northern Sudan; these seem to be short-term camps.29 They demonstrate that at that date there were still people who were prepared for desert travel, probably using the donkey, domesticated in Africa, as a pack animal.30 The fourth millennium, like the fifth, seems to have been characterised by movements towards and away from the Nile and into the Sahara, both in Nubia, where the A-Group sites show a continuity from agricultural sites near the river to more pastoral groups in the Western Desert. Barley and emmer wheat were cultivated, and the pastoral economy included cattle, sheep and goats as well as dogs.31 Most burials were shaft graves very dissimilar to those of the Sahara, but tumuli with concentric rings are also found, sometimes as isolated burials in the desert. With the arid climatic crisis at the end of the fourth millennium BC, which corresponds to the beginning of dynastic Egypt, sites in Nubia are less frequent, smaller and short-term, suggesting higher mobility out into the deserts around the Nile. At the same time, in the Sahel and the Ahaggar, we find ever-more elaborate tumuli.32 The desert was thus densely occupied, though we have no way of knowing what language the highly mobile occupants were speaking: it may have been Proto-Berber, but it also may not, and Proto-Berber may have been only one of a number of languages left over from the Green Sahara.33 However, Ehret’s evidence for the antiquity of Zenaga suggests that, perhaps by 2000 BC, Proto-Berber speakers had moved westwards, eventually reaching the Atlantic coast of Morocco: they are hardly likely to have moved there along the coast of North Africa, where the Gulf of Syrtis provides a significant barrier to movement,34 and where, as we have seen, Neolithic sites stubbornly conserve their Capsian traditions. It seems not impossible that they were moving in from the Northern Sahara, along the base of the Atlas. A major group of sites which might initiate this period is that studied in the Wadi Noun by Bokbot, Onrubia-Pintado and Salih, although the two tumuli nearby dated by radiocarbon are substantially 29 31

32

33

Riemer 2004. 30 For the domestication of the donkey, Mitchell 2017; Rossel et al. 2008. Gatto 2006. The A-Group sites date mainly to the second half of the fourth millennium BC, while the most recent to 3100–2900 BC. Paris and Saliège (2011, 278) give numerous radiocarbon dates for megaliths in the Central Sahara, ranging for the earliest type from 6000 to 4200 BP (c.4900–2800 calBC). The latest type, the ‘tumulus à cratères’, dates from 3700 to 1100 BP (c.2100 calBC–calAD 900). Blench, Chapter 14, this volume. 34 For travel by land along this coast, Quinn 2011.

https://doi.org/10.1017/9781108634311.016 Published online by Cambridge University Press

501

502

Elizabeth Fentress

later.35 These are relatively small open sites, characterised by pottery, microliths and ostrich-shell beads. Their distribution, running down a long east-west valley which runs into the Atlantic in the Anti-Atlas, emphasises the southern aspect of these sites, which seem to relate more to the Sahara than to the Tell. In these groups, we can perhaps see the first wave of the new arrivals from the Sahara and eventually the Nile region, who may have been speaking a language related to that identified as the earliest Berber language in the Maghrib, Zenaga. This is spoken now in south-western Mauretania and Senegal, by a pastoral people specialising in sheep and cows. Their presence today so far south may support the idea that they moved to the Atlantic via the desert, spreading up into – or down from – Morocco and exploiting marine as well as pastoral resources. The persistence of the early language in the south would be due to the concentration of subsequent groups of Berber speakers in the Central Sahara and the Tell, as well as its isolation by the waves of Arabic language.36 But the modern position of Zenaga does not, of course, necessarily mean that the first Zenaga speakers lived in southwestern Mauretania: indeed, they could have been part of the first wave of movement of pastoral Berber communities moving west from the Nile Valley, as further discussed. This first diffusion of the Berber languages is the most difficult and hypothetical, and based mainly on Ehret’s proposal that the first western spread of groups speaking Proto-Berber was no earlier than 2000 BC. However, if, as Ehret contends, the split from the Chadic languages occurred with the Neolithic or even earlier, that leaves several millennia unaccounted for, in which languages may have split and new daughter languages developed – as we only have the one relict Zenaga language in Mauretania it could represent a later splitting off from the main Berber stem, which could have been spoken in the Maghrib since the fourth or fifth millennium but for which we have no other evidence. Indeed, any number of reconstructions would fit its appearance there, as long as they are only slightly earlier than the next coherent group of languages, which Ehret refers to as Proto-Amazigh, and which established themselves around 1000–800 BC. He quite convincingly sees these as related to the Libyans that appear in Egyptian history and material culture in the Twenty-first and Twenty-second Dynasties, between 1200 and 800 BC.37 Where did these emerge from? 35

36

Bokbot et al. 2011: the excavated tumuli date between the second and the fourth centuries AD, see Bokbot, Chapter 10, this volume, for further details. Blench, Chapter 14, this volume. 37 Morkot 2016.

https://doi.org/10.1017/9781108634311.016 Published online by Cambridge University Press

16 The Archaeological and Genetic Correlates of Amazigh Linguistics

Proto-Amazigh: 1000 BC (Fig. 16.3) I would suggest that, like the preceding Berber languages, proto-Amazigh emerged near the Nile, according to a pattern which we can see repeating in the archaeology of previous millennia. The largely pastoral, mobile Amazigh speakers would have found themselves in close contact with the intensely agricultural populations of the Fayum and the Nile, as well as with the highly developed Nubian populations to the south of them.38 Their relationship with the settled populations of the Nile Valley may have been relatively complementary. The more mobile Libyan population comes into much sharper focus at the beginning of the first millennium BC,39 by which time they seem to have adopted horses, increasing their mobility, their range and their ability to form aggressive, larger groupings.40 Furthermore, we might expect equid-based mobility to produce nomadic élites, in control of resources and exchange, and engaged in frequent raiding of both pastoral and sedentary populations. Over the latter, riders would have had a decisive advantage. The stele of Pasenhor, dated to 731 BC, mentions the group of the Meshwesh descending from ‘Buyuwawa the Libyan’.41 Libyans are known as early as the Eighteenth Dynasty, when they supplied cattle to Amenhotep III (1386–1349 BC). They are represented on paintings and papyri as nomadic pastoralists armed with hatchets, bows and arrows, spears and daggers. They were tattooed, or perhaps ritually scarred, as are the Garamantes in later periods.42 A substantial period of strife with the Egyptians, particularly under Ramesses III, reveals their linguistic as well as cultural differences from their neighbours on the Nile, who settled them in ‘strongholds of the Victorious King, [where] they hear the language of the [Egyptian] people, serving the King, he makes their language disappear’.43 Given Libyan mobility, this intense and prolonged contact with the Egyptians, to whom they were also supplying trade goods, will have stimulated groups of them to occupy new areas. Of these, we know in 38

39

40

41 43

For the linguistic evidence that the Proto-Amazigh speakers were largely pastoral, Ehret, Chapter 15, in this volume. Libyan nomadic pastoralists are reported from the Egyptian Western Delta from the third millennium BC, Moreno Garcia 2014. For the impact of the horse on pastoral societies, Anthony et al. 1986. For the linguistic evidence for horses in Berber society, Ehret, Chapter 15, this volume. Ritner 2009, 18. 42 Leahy 1985. The Great Harris Papyrus, British Museum EA 9999 (http://www.britishmuseum.org/research /collection_online/collection_object_details.aspx?objectId=114379&partId=1, with reference wherein).

https://doi.org/10.1017/9781108634311.016 Published online by Cambridge University Press

503

504

Elizabeth Fentress

Figure 16.3. The diffusion of Proto-Amazigh, c.1000 BC, into Fazzan and beyond.

particular of Fazzan, where the work of Charles Daniels shows that the defended promontory settlement of Zinkekra, dating from around 1000 BC, was settled by a group using the full Near-Eastern package of wheat, barley, dates, figs, sheep and goats.44 Animals were kept on the site, and multiple mortars cut into the living rock. The proliferation of these sites down the escarpment suggests that a considerable number of people arrived in the course of time. The occupation of Fazzan implies that the oasis at Siwa was already settled, although the earliest archaeological evidence there, a cemetery, seems to be as late as the Twenty-sixth Dynasty (685–525 BC). Indeed, the fact that its language dates to the moment of this second expansion may well provide evidence for the early date of its settlement. The dates of the other escarpment villages of the Wadi al-Ajal during the first half of the first millennium BC confirm the idea of a significant thickening of the occupation of Fazzan during that time, with the arrival of new groups.45 Further into the desert to the south, in the Tadrart Akakus the relationship of these ‘caballine’ sites, characterised by engravings of horses, to the previous pastoral ones has been well discussed and understood.46 But little work has yet been done on their extension westward, where Herodotus tells us of tribes called the ‘Atarantes’ (from *Kel Tadrart, the people of the 44

Pelling 2005; Mattingly 2010, 525.

45

Liverani 2006, 371.

https://doi.org/10.1017/9781108634311.016 Published online by Cambridge University Press

46

Liverani 2000; 2006.

16 The Archaeological and Genetic Correlates of Amazigh Linguistics

mountains?) or the Atlantes (from *Kel Ataren, ‘the people of the West’?).47 Further west, in the Saharan Atlas there are also large amounts of engravings of horses and chariots.48 Another proxy for such spread during the second half of the first millennium might be the tumuli, particularly those ‘à caisson’ described by Gabriel Camps. These stretch from southern Tunisia into the valley of the Draa, south of the Anti-Atlas, with its greatest concentration at the cemetery of Aïn Sefra, south-west of Laghouat at the southern tip of the Saharan Atlas.49 They are poorly dated: At Aïn Sefra, bronze bracelets and ostrich-shell beads were found, as well as offerings of sheep and birds, but no iron, which might suggest a date in the early first millennium BC. The engravings of the lower Draa, which show chariots and metal armaments and are characterised by Salih and Hackendorf as ‘lybico-berbères’,50 clearly form part of this same complex, and have been dated (by a piece of decorated ostrich eggshell) to 1232–793 calBC.51 I would propose to associate the groups that created these tumuli with the same proto-Amazigh speakers that settled Fazzan, moving first west along the northern Sahara and then northwards into the Aurès, and the Tell, where defended promontory sites such as Ichoukkane on the piedmont of the northern Aurès, with their cemeteries of tumuli, become increasingly common – indeed, Ichoukkane might reveal itself as a northern equivalent of Zinkekra.52 This is a coherent cultural complex, involving, at least in Fazzan, proto-urban communities, large tumuli for the élites and extensive contacts with the desert, shown by the continuous arrival of the ostrich shells used in the tombs. The earliest settlements would relate to this new, far more complex society, characterised, like the Libyans in Egypt, as having ‘kings’ at the time of the settlement of Carthage, at least in the literary record.53 The spread of the tumuli into the mountains of the Hodna and further north, so clear on the Atlas Archéologique de l’Algerie, though completely undated, seems to show horse-using pastoral communities of this date and later reaching to the coast, as do the stele showing horses of the Kabyle.54 47

48 49 51

52 53

54

Liverani 2004: note however the relationship between Atlas and Atlantes. See also Gatto et al., Chapter 3, this volume, and Blench, Chapter 14, this volume. Camps and Hachid 1982; Lhote 1984; Hachid 1992; Amara 2003. Camps 1961, 69; Bokbot et al. 2011. 50 Salih and Hackendorf 2002. Grébénart 1975 cited by Bokbot et al. 2011, 2790 ± 105 BP, 1232–793 calBC (no lab information given). Gsell 1901, 1.16–18; Morizot 1997, 42–44. Vergil, Aeneid 4.197. Note that Iarbas is said to be the son of a Garamantian nymph, raped by Jupiter Ammon. Gsell 1902/1911.

https://doi.org/10.1017/9781108634311.016 Published online by Cambridge University Press

505

506

Elizabeth Fentress

The spread of this cultural complex, distinguished by engravings of chariots, by villages in the Sahara and the southern Tell and tumuli in a broad band across the desert and pre-desert, must be associated with a gradual movement or series of movements away from the Nile and the Western Desert by groups speaking proto-Amazigh. The large range of this language group, however, might imply that the process was fairly rapid, taking place between the end of the second millennium BC and the first two centuries of the first. Again, a movement into the Tell along the coast from the Delta along the Syrtis seems far less likely than a more southerly dispersal of large numbers of small, mobile bands, moving into Fazzan and the northern Sahara, and thence northwards towards the Mediterranean, in parallel to the Punic settlement of the coast. On this model, we might expect the settled occupation of the oases on the Saharan fringe, and the Saharan Atlas, to be earlier than those of the coastal mountains. The combination of nucleated settlement with a high degree of mobility that we can see in Fazzan will have been equally effective in the Tell: it is just this mobility that convinced the early geographical writers to characterise the Libyans as purely nomadic. The distribution of the languages that descend from proto-Amazigh shows their wide range, throughout the desert and right up to the coast in the Kabylie, and, presumably, northern Tunisia. In the Tunisian Tell, nucleated settlements were also developing at this time, as the recent excavations at Althiburos have shown.55 Ninth-century BC rectangular domestic structures and occupation layers there show that what became later on an important Numidian settlement may have had had its origins at that time, although there may have been a hiatus between the first settlement and later occupation.56 There seems little possibility that such a settlement emerged from the Neolithic of Capsian tradition, given the extraordinary conservatism seen at Haua Fteah cave in the same period and the real absence of evidence for a teleological development towards complexity in the Tell in the second millennium BC. However, the necropolis of dolmens associated with this site,57 together with the cultivation of legumes, olives and grapes (though the latter may be wild) as early as the ninth century (compared to the fourth century BC for olives and the first century AD for legumes in Fazzan)58 might suggest that its inhabitants had strong Mediterranean contacts, related to the spread of dolmens 55 56

57

Kallala and Sanmartí 2011. The excavators see a period of abandonment between the beginning of the seventh century and the middle of the sixth, Kallala and Sanmartí 2011, 159. Jornet et al. 2011, Sanmartí et al., Chapter 9, this volume. 58 Pelling 2005, 400.

https://doi.org/10.1017/9781108634311.016 Published online by Cambridge University Press

16 The Archaeological and Genetic Correlates of Amazigh Linguistics

throughout north-eastern Tunisia via Sardinia, southern Italy and Malta, which would have begun in the second millennium BC. This, for Cyprian Broodbank, might be the first moment when we can see clear connections across the narrow waters that separate Sicily from Tunisia.59 Certainly the absence of horse engravings and tumuli, and the total absence of ochre or ostrich-shell beads in the burials, seem to distance Althiburos from what was going on in the desert, although only the pottery in the earliest layers, ornamented with horizontal ridges and friezes of triangles, seems in any case different from the general run of Numidian hand-made pottery.60 What language was spoken there thus remains entirely obscure.

Genetic Considerations We have, then, a very complex situation in North Africa in the first centuries of the first millennium, which would be reflected in the genetics of the Maghrib.61 Over a dispersed, but perhaps rather thin genetic substrate relating to the Mechtoid populations would be found, in north-west Morocco, the descendants of those who crossed the straits of Gibraltar in the Cardial period. In north-eastern Tunisia there was perhaps some admixture from Sardinia, Malta and Sicily, people who might have been influential in the creation of sites like Althiburos, although they would naturally have interacted with the local population, and, potentially, new arrivals from the south-east, whose primary role cannot be excluded. Sporadic western Phoenician settlement along the coast as far as Lixus was also part of this mix.62 In the pre-desert, and increasingly in the Tell, arrived the new wave of proto-Amazigh-speaking immigrants, coming in from the Nile through the Sahara. It is just this heterogeneity that is shown by every genetic study of modern populations, however different their conclusions as to North African ‘origins’.63 None of these proposed ‘origins’ is really convincing, in that real questions remain about their samples, but their agreement on the heterogeneous nature of the population is total. However, recent work on the Tuareg has shown that those of Fazzan have 59

60 62 63

Broodbank 2013, 568. Camps 1961, 150–54, writing over 50 years ago, concluded that dolmens were introduced, probably in the late Bronze Age, but before Punic colonisation, from Sardinia. However, it is unclear whether their introduction entailed the arrival of substantial numbers of immigrants. If the earliest evidence from Althiburos contains the largest quantity of possibly exogenous pottery types, subsequent contexts show pottery similar in all ways to Numidian pottery elsewhere. Kallala and Sanmartí 2011, 155. 61 Frigi et al. 2010. Although, see Papi, Chapter 9, this volume. Arredi et al. 2004; Henn et al. 2012; Coudray et al. 2009, Fadhlaoui-Zid et al. 2013.

https://doi.org/10.1017/9781108634311.016 Published online by Cambridge University Press

507

508

Elizabeth Fentress

a remarkably homogeneous mitochondrial DNA, with strong East African affiliations, as well as some trace of the European component in Berber populations in north-west Africa.64 On the other hand, Y-chromosome data from the Tuareg elsewhere shows a high incidence of demic diffusion from the east.65 We may thus combine the evidence from genetics with the wide spread of proto-Amazigh, incorporating large-scale but dispersive movement of populations whose mobility was based now on both horses and donkeys, both of which originated to the east of Cyrenaica. However, this is the first moment that we may be certain that a Berber language, the ancestor of modern Kabyle, was being spoken in the Tell.

Proto-East-West Amazigh: c.500 BC (Fig. 16.4) The next ‘pulse’ in this process again seems largely to relate to desert populations, one of whose languages came to dominate the others. Here, the Kabylie region is left out of the process, as its isolation from more recent languages shows. Indeed, it seems plausible that the language spoken at this time in the whole of Numidia was related to Kabyle, which would represent a remnant of an originally much more widely dispersed language spoken throughout the central Maghrib by the Numidian and Mauretanian populations, and increasingly used epigraphically. Starting around 500 BC, East-West Amazigh might relate to the consolidation of the Garamantian state at the end of the millennium. This is the time of the foundation of the proto-urban phase at Jarma, whose influence spread south-westwards, possibly controlling sites as far as the Tadrart Akakus. During the Classic Garamantian period, it acted as a state in the Wadi alAjal, perhaps controlling more peripheral locations to the south-west like the settlement at Aghram Nadharif and the slightly earlier Fewet.66 We can see the contacts of these communities stretching far to the west: volcanic lamps were acquired from the south Tassili, between 250 and 450 km 64 65

66

Ottoni et al. 2009, 445. Ottoni et al. 2011, 121 suggest that ‘it is likely that most of the Tuareg E1b1b1b Y chromosomes are related to an expansion event that took place about 2.6 kya in an ancestral population inhabiting a region between Tunisia and the Central Sahara. This event may have coincided with an expansion that led to the formation of derived Tunisian and Central Saharan populations.’ See also Pereira et al. 2010, who imagines a demic diffusion from the Middle East, based on his view of the origin of Afro-Asiatic languages. Liverani 2006; Mori 2013.

https://doi.org/10.1017/9781108634311.016 Published online by Cambridge University Press

16 The Archaeological and Genetic Correlates of Amazigh Linguistics

Figure 16.4. The diffusion of Proto-East-West Amazigh, c.500 BC, into the territories said to be occupied by the Gaetuli.

away.67 Mediterranean imports were numerous.68 The society continued to have both a sedentary and a pastoral component, the latter ranging widely and presumably involved in regional exchange. The sheer weight of this state in the context of the Sahara may have made its version of the Berber language the dominant language throughout the whole of the region. The new grouping of languages spanned an arc almost as wide as the last, and includes the oasis language of Siwa westward all the way to the Atlantic with the Atlas and the Souss, where Tashelhit is spoken.69 These languages cover a wide area, all of it on the margins of the lands later dominated by Rome, or, indeed, by the earlier Numidian and Mauretanian kingdoms, where languages related to Kabyle probably continued to be spoken. The area involved coincides well with that associated with the Gaetuli, groups with the same mixture of sedentary sites and pastoral mobility that we have seen in the case of the Garamantes. The Gaetuli were in permanent opposition to the kingdoms in the north during the last centuries BC,70 although by the second century AD, Apuleius could jokingly refer to himself

67 68 69

Gatto et al., Chapter 3, this volume. See Trans-Saharan Archaeology volume, Mattingly et al. 2017, 323–33. Ehret, Chapter 15, this volume. 70 Fentress 1982.

https://doi.org/10.1017/9781108634311.016 Published online by Cambridge University Press

509

510

Elizabeth Fentress

as ‘half-Numidian and half-Gaetulian’.71 The position of the town of his birth, Madauros, gives us an idea of the northern extent of their territory. What this grouping of languages seems to imply is that the influence of the Gaetuli extended far to the south, comprising the languages of the oases, as well as to the Moroccan coast, where their presence is signalled by Pliny, citing the periplus of Polybius.72 How different they were, culturally, from the Garamantes is difficult to establish, and there are no specific archaeological ‘markers’ for the Gaetuli. What we do see is a general cultural unity, combining intensive oasis agriculture in places such as Biskra/Vescera and Capsa with nomadic pastoralism, signalled by numerous engravings showing horsemen, chariots and bovines all the way to the Atlantic coast.73 The horse was particularly important in the culture of the Gaetuli, reflected both in their engravings and in Strabo’s assertion that 100,000 foals were born to them every year.74 There was in all probability a dense population in the Saharan Atlas by this time. This is suggested not only by Sterry’s work75 but also by the subsequent extension of the Severan limes along the range towards Laghouat. The oases, too, were probably densely inhabited. Malika Hachid has called the Sahara the ‘autoroute culturelle de l’antiquité’, insisting on the mobility of pastoral peoples there well before the introduction of the camel.76 Their cultural unity is both a cause and an effect of this mobility. The clarity of the definition of the Gaetuli by the Romans might thus derive from their unified language and origins as well as from their loyalties and lifestyle.77 The Gaetuli could be distinguished from the Numidians primarily because they spoke a different language. The situation will hardly have remained static during the Roman period. The languages of the northern Tell will have been going through the normal processes of division, with separate communities speaking what were 71

72

73 74

75 76

77

Apuleius, Apologia 24. Note that at the time of the Jugurthine wars the Gaetuli, according to Sallust, Bellum Jugurthinum 81,1, were ignorant of the name of Rome. Pliny, Natural History, 5.9; 5.10. A group of the Gaetuli, the Darae, were apparently based in the valley of the Draa. On the tribes of the Gaetuli and their distribution: Desanges 1980, 342–6. Camps and Hachid 1982, Bokbot et al. 2011. Strabo, Geography, 17.3.17. The bibliography for ‘Libyco-Berber’ or ‘caballline’ engravings is as vast as their distribution: see generally Muzzolini 1995, Salih and Heckendorf 2002. Interestingly, by the end of the first millennium AD, they are found as far south as Burkina Faso: Barbaza 2012. Sterry and Mattingly Forthcoming. Hachid 2011, 200. Note Ehret’s discussion, Chapter 15, this volume, of the introduction of the word for camel at the end of the millennium. One thinks of the nationem Gaetulicarum sex quae sunt in Numidia (Inscriptiones Latinae Selectae I8966).

https://doi.org/10.1017/9781108634311.016 Published online by Cambridge University Press

16 The Archaeological and Genetic Correlates of Amazigh Linguistics

probably becoming mutually intelligible dialects. The new epigraphic habit allows us to see divisions between what was roughly the Numidian Kingdom and the Mauretanias, with possible other variants.78 These probably reflect divisions in dialect, although there is no way of knowing more about what these dialects were. A related script, ‘ancient Tifinagh’, is found all over the Sahara: its letters appear on blocks reused for the construction of the tomb of Tin Hinan in the fourth or fifth century AD, as well as on the stele of Azibs n’Ikkis in the High Atlas.79 Recent studies have shown distinctive concentrations, of which the most notable is found in the Irlarlaren pass in the Akakus Mountains, where a palimpsest of multiple Libyan inscriptions includes some that are certainly written in ancient Tifinagh.80 A similar concentration is found in the Saharan Atlas.81 In the Tell, the 1,125 inscriptions recorded by Chabot contain little information besides onomastics, and tell us very little about the language in which they were written, although it was certainly a Berber language. The important point is that, at least in the Tell, none of these is later than the third century AD. The primary linguistic force at work was by now the generalised diffusion of the Latin language, which seems to have created, in most areas, the disappearance of the various Berber dialects. Indeed, in the countryside around Hippo the earlier Punic dialects still survived, as we know from Saint Augustine and others,82 and neo-Punic inscriptions are found as far south as Constantine and in Tripolitania.83 The Berber languages of the Tell, with the single exception of the area of the Kabylie, seem to have given way to the economically and politically more important language of government.84 They probably survived in some places, particularly Mauretania Caesariensis, but even in the far west Latin seems to have been the language of ordinary people, as the thousands of tombstones and stele to Saturn show. There is no reason to think that Latin spread south of the limes, however, and we might assume that Volubilis marked the southern limit of the Latin language in Morocco. In the Atlas and the Souss, the use of Tashelhit will have continued. 78 79

80

81 82

83

Chabot 1940; Galand 1989. For a convincing redating of the latter inscription to no earlier than the second century BC. El Khayari 2009, with previous bibliography. He argues that the engraving of the horseman is cracked by fissures which must post-date it, while the inscription carefully works around the fissures, and is therefore later. di Lernia and Zampetti 2008, 300; on the age of the inscriptions, and their clear relationship to a Berber language (contra Blench, Chapter 14, this volume), personal communication Ali Ait Kaci. Amara 2003. For example, Augustine, Epistulae, 209.3, CSEL 57, p. 57: for discussion and previous bibliography Lepelley 2005. Jongeling and Kerr 2005. 84 Camps 1992; Lepelley 2005.

https://doi.org/10.1017/9781108634311.016 Published online by Cambridge University Press

511

512

Elizabeth Fentress

Elsewhere in the territory occupied by Rome, there may have been pockets of people speaking Berber languages, but they will have been isolated from one another and perhaps mutually incomprehensible. The language today spoken at the southern Tunisian oasis of Matmata, which may have split off from the Saharan languages around AD 100, may reflect a new settlement from the south.85 The Kabylie probably remained the one area where a Berber language was continuously spoken, in a form which, as we have seen, was already archaic in respect to that spoken in the northern desert.

Proto-Central Amazigh: c.AD 400 (Fig. 16.5) It is into this situation that we find the final ‘refoundation’ of the Berber language, distinguished by Ehret as proto-Central Amazigh, although also known as Zenatic.86 This parent language engenders all of the subsequent Western Berber languages except Kabyle, and Tashelhit in the Atlas and the Souss. It seems to be dated to the period around AD 400. Given the wide extension of these languages today, from the Rif to the south-western oases, the spread is remarkable, and must be linked to a major series of events, powerful enough to create a unified language over a wide area. Andrew Wilson and I have in a recent paper linked this refoundation to an invasion of the Tell from the south, just before the Vandal interregnum.87 I will summarise here the evidence fairly briefly, but it is worth revisiting because this is one of the clearest instances of the processes involved in this sort of refoundation, or levelling, to use Blench’s term.88 The major archaeological correlate for a movement from the desert is the appearance of Saharan tomb types north of the Aurès Mountains. It was first pointed out by Pierre Morizot that there are numerous tumuli built from, and often on top of, Roman ruins.89 Often these have a drum, and sometimes other features. A good example comes from the fort at 85 86

87 88

89

Ehret’s ‘proto-Wide-Central Amazigh’, Chapter 15, this volume. Destaing 1920. See the caveats on the use of the name in Ehret, Chapter 15, this volume. However, although there are clearly some exceptions, most of the speakers of these languages coincide with those referred to by Ibn Khaldun as ‘Zenatic’ tribes. As in the case of the Gaetuli, the distinction he perceived may have been as much linguistic as anything else. Fentress and Wilson 2016. Blench, Chapter 14, this volume. He curiously asserts that the ‘levelling’ took place between AD 100–200, without justifying his dating other than to assert that it would fit with the creation of the Roman limes. This is a purely circular argument (the levelling occurred, the limes occurred, therefore they are related and happened at the same time). Why the limes should create a unified Berber language remains entirely obscure. Morizot 2003.

https://doi.org/10.1017/9781108634311.016 Published online by Cambridge University Press

16 The Archaeological and Genetic Correlates of Amazigh Linguistics

513

Figure 16.5. The diffusion of Proto-Central Amazigh, c.AD 400, from the Saharan oases into the Aurès Mountains and beyond.

Ausum, at the foot of the Saharan Atlas, where eight huge tumuli were built inside the fort.90 The drums can have a small niche on their east face, like one from El Esnam, near Msila, or extended antennae, as at the Roman town of Sila. On a villa site in the territory of Zana occupied in the fifth century, I found a small circular tomb, built out of the orthostats of the farm it covers, with a clear offering niche, as well as a U-shaped structure which recalls desert structures associated with tombs.91 Far more elaborate structures are found in the region of Négrine, south-east of the Aurès, where two large drum tumuli have carefully built trefoil chapels on their eastern sides. Very similar tombs come from the Tafilalet in Morocco, while simpler tumuli found in the Wadi Tifariti, western Sahara south of Morocco, yield radiocarbon dates of calAD 430–769 and calAD 432–809.92 The tomb of El Gour, near Meknès, is the best known of this series, dating as late as the eighth century AD (calAD 565–945).93 Finally, the Jedars, the great monumental tombs of Tiaret, are clear statements of power, and have justly been identified as princely tombs of a dynasty that ruled Mauretania, using local builders to create monuments in a very different cultural 90 92 93

Fentress and Wilson 2016, Fig. 3.3. 91 Fentress 2013. Brooks et al. 2009, 928; Clarke and Brooks, Chapter 11, this volume. Camps 1974, though see Papi, Chapter 9, in this volume for the view that the date comes from an Islamic intrusion.

https://doi.org/10.1017/9781108634311.016 Published online by Cambridge University Press

514

Elizabeth Fentress

tradition.94 The point is an obvious one. The origin of these tombs is clearly Saharan, and their spread through the Tell as far as Sila, near Constantine, echoes that of the new language grouping. It leaves out the Kabylie, and the whole of Libya and the eastern oases, but spreads right into northern Morocco, covering the area of Roman occupation there. Interestingly, it leaves out the Wadis Souss, Draa and Noun, where such tombs are found rather earlier,95 concentrating in the north. Like the language, the tombs seem to suggest an influx of new people, whose origins could be found deep in the Sahara. Saharan cemeteries roughly contemporary with these sites, like that at In-Aghelachem,96 are their closest parallels, though the antenna tombs of an earlier date, so visible in the Sahara, may also have been deliberately recalled.97 The historical context of this influx is pretty clear. Our argument is based on a number of indications, both textual and archaeological, that the end of the first quarter of the fifth century saw a massive invasion of ‘Moors’ in the region of the Hodna. In a well-known letter of St Augustine to Boniface, comes of Africa, he recalls a visit to Boniface at Tobna in the Hodna, where the latter was a tribune, presumably praepositus of the limes Tubunensis, where he was protecting the province from the incursions of barbarians, relying on auxiliary troops.98 But now, in AD 427, Augustine feels that Boniface’s inaction, in spite of his sizable army, makes him responsible for: the devastation of Africa at this hour by hordes of African barbarians, to whom no resistance is offered, while you are engrossed with such embarrassments in your own circumstances, and are taking no measures for averting this calamity. Who would ever have believed, who would have feared . . . [that Boniface] should now have suffered the barbarians to be so bold, to encroach so far, to destroy and plunder so much, and to turn into deserts such vast regions once densely peopled? 94

95

96

97

Camps 1965; Kadra 1983; Laporte 2005. A radiocarbon date from the wooden coffin in Jedar B (Kadra 1983, 332) gives a date of 1440 ± 50 BP or calAD 410–615 (no lab information given). Bokbot et al. 2011, 317–20, publish two cylindrical tumuli with niches and reduced antennae that date respectively to calAD 120–330 and calAD 140–390 (above note 35). These early dates for western tumuli of this type, which strongly resemble the post-Roman tumulus near Zama, ‘might’ support the idea that the influx initiated in the west, although the distance between the dialect of the Draa, Tashilhet and the Proto-Central Amazigh languages would argue against this. There are two relevant dates for the In-Aghelachem site, given by di Lernia et al. 2002, 115 that can be taken to imply a third- or fourth-century date. These are: 1700 ± 40 BP (GX-27385-AMS) for a date stone from the main burial, calAD 246–416; and 1850 ± 50 BP (GX-27383) for charcoal from SF 25 Group 1, an altar or fireplace in front of the tomb, 18 calBC – calAD 380. Heddouche 2011; Gauthier and Gauthier 2007. 98 Augustine, Epistulae, 220.3.

https://doi.org/10.1017/9781108634311.016 Published online by Cambridge University Press

16 The Archaeological and Genetic Correlates of Amazigh Linguistics

Even allowing for rhetorical exaggeration, it is clear that the limes had given way. Other, widespread traces of destruction are detailed elsewhere, but the point is that the epicentre of this attack is just that of the new wave of Saharan language, stretching from the Aurès, where Shawiya is still spoken, as far north as Constantine, into the region of Tiaret and Altava and eventually through to the Rif and northern Morocco. That this barbarian invasion is not as well known as, say, that of the Vandals is evidently because the Vandals, who arrived only a couple of years later, eclipsed it from the point of view of the Mediterranean. By the time Procopius and Corippus report on Africa in the sixth century, the ‘new’ Moors were simply ‘the Moors’, with the camel-herding Louata as the latest in a series of waves. The concentration of the latter in the central Tunisian plain and Tripolitania might, in fact, show them as having very separate origins from the Moors of the Aurès, but the heavy Arabisation of this area would have eliminated any trace of the language they spoke. What sent these various waves north might, as Wilson and I suggest, be based on the sort of systems collapse that we see in Fazzan at that time, when the Garamantian state appears much weaker, perhaps because of the failure of its irrigation. But we know so little about the Western Sahara that proposing a systems collapse there is purely hypothetical. Perhaps they just started north because they could. The use of the dromedary may have helped. Although we know of examples of the dromedary in the Tell from the time of the battle of Thapsus onwards, the first proof of their presence in the Central Sahara is, as Hachid points out, the engraving from Abalessa.99 The use of the camel by the Garamantes is probably even earlier; as the thirdcentury ostraka of Gholaia/Bu Nijim show, the army used camelarii who carried wheat.100 The use and breeding of the camel will have given new impetus to an already mobile society, making movements of large amounts of people far easier. It is perhaps around this time that regular caravans began to be organised, like that shown on an engraving from the Monts des Ksour in the Saharan Atlas.101 In fact, the camel seems to have appeared in number around the time that oasis agriculture, based on the now-failing foggara irrigation systems, was no longer adequate to support a large population, including the slaves that maintained them.102 Its use will have added an important resource to desert communities, providing a new source of protein in its milk, and adding mobility for the whole group, including women, children and the 99 101

Hachid 2006; 2011, 204. Hachid 2011, Fig. 1–3.

100 102

Rebuffat 1992, 1631; Marichal 1992. Fentress and Wilson 2016.

https://doi.org/10.1017/9781108634311.016 Published online by Cambridge University Press

515

516

Elizabeth Fentress

sick. It also creates striking power over great distances, allowing attacks by large groups.103 It was an essential component of the early modern (and, presumably, medieval) hierarchical structure of the desert, allowing trading and raiding by the élite, with tributary relationships to the oasis communities, where Bedouin tribes owned lands worked by haratin. A shift to camel pastoralism will have enabled the Garamantes, and other oasis communities, to increase their mobility while maintaining their complex social structures. The move northward, which may have united a number of chiefdoms into confederated armies, may have been occasioned by a period of greater aridity, causing increasing stress to these desert communities, which then flowed into areas from which Roman defence was already disappearing. The leaders of this army, or their descendants, can be seen in the builders of the Jedars. All of this rather unfashionable emphasis on what can be seen as a straightforward invasion, whether in one coordinated mass or on several occasions, is intended to underline what I am proposing is a constant Saharan dynamic. The desert, particularly the Saharan Atlas and the southern mountain ranges like the Ahaggar, could have supported far more people than we have imagined. These depended for their mobility, by turn, on the donkey, the horse and the camel, and used a combination of pastoral resources, and intensive oasis agriculture, whose produce was supplemented by a network of regional and sub-regional trade.104 Their mobility reinforced their interaction, with the result they were able to coordinate their movements on the sort of scale that Corippus and Procopius’s descriptions of the massed hoards of the Louata suggests.105 Indeed, it can be argued that nomadism itself creates war zones, or areas of privileged mobility and the violent exercise of power.106 This goes hand in hand with predation and the creation of dependent, settled communities whose language would, in time, shift towards that of their masters.

The Oasis Languages The subsequent splitting of this parent language into a number of distinct ‘daughters’ is more straightforward, but, in a couple of cases, requires some comment. Shawiya, as we have seen, remains firmly rooted in the Aurès and the plain to the north of it as far as Constantine, while the languages of 103 105

Sweet 1965; Hachid 2011. 104 For a study of these in the modern period: Scheele 2012. For these texts: Modéran 2003, 153–207. 106 Retaillé 1998, 77.

https://doi.org/10.1017/9781108634311.016 Published online by Cambridge University Press

16 The Archaeological and Genetic Correlates of Amazigh Linguistics

the Rif and western Algeria – Rif in the north, Tamazight in central Morocco, with Sanhaja sandwiched between the two and Beni Snous around Tlemcen to the west – are closely related to one another, and to the oasis language of Figuig, but less so to Shawiya. The separation of the languages of Wargla and the Mzab is, of course, based on their isolation, but we should remember that both oases received a substantial influx from the Kharajite town of Tihert in the eleventh century: it would be interesting to see if, at a micro-level, they are closer to those languages than to Shawiya. The separation of the language of the Tuareg in the Ahaggar from that of Ghat is even more recent, and may speak to a weakening of the east-west connection through the Irlarlaren pass, and to some progressive isolation of the Ahaggar.

Conclusions This brings me to a final point. From the beginning of this chapter I have stressed the east-west connections through the Sahara over those running north-south. Liverani’s fundamental 2000 paper showed how the Garamantes could connect to the Niger. But the real point is that they could connect to the Saharan Atlas and Morocco by travelling south of the Tunisian chotts, and along the spring line south of the Aurès. Alternatively, though with more difficulty, they could move from Ghat through the Irlarlaren pass, with its extraordinary palimpsest of traveller’s graffiti, into the Ahaggar. It would be difficult even for the most mobile pastoralists to cross the great Erg, but it can be manoeuvered around, in an arc that would then lead down the Saharan Atlas from the Hodna to Figuig, Bechar and eventually the Anti-Atlas, the Souss and the Draa. This roughly east-west arc may also have been travelled far more than the narrative of north-south trade would suggest. If there is an area of North Africa that is crying for closer investigation, it is the Saharan Atlas: Fazzan remains the area of the Sahara that is lit by the proverbial street lamp, and thus the only place we have looked, and it would be good to light up the rest. As will have been clear throughout this paper, and particularly in the comparison between it and those of Blench and Ehret, I am proposing, with great uncertainty, a model for the diffusion of Berber languages which sees new arrivals and incursions as a series of pulses originating, initially, far to the east and then moving to the oases of the Central Sahara and the Saharan Atlas. This seems to coincide with what we know of the archaeological data,

https://doi.org/10.1017/9781108634311.016 Published online by Cambridge University Press

517

518

Elizabeth Fentress

although it is clear that we do not know nearly enough, and that what we do know is generally not properly dated. The reconstruction of this history proposes migrationary models that are virtually demanded by the diffusion of prehistoric language stocks.107 Other reconstructions could fit these data. In the end, they will have to be judged by their fit, both to the linguistic and the archaeological data. Ehret’s work, however, provides us with an important new set of evidence, and should be taken thoroughly into account in subsequent reconstructions.

References Amara, I. 2003. Nouvelle approche de l’art rupestre de l’Atlas saharien: les figurations de la période tardive. L’Anthropologie 107: 533–57. Anthony, D., Bogucki, P., Comșa, E., Gimbutas, M., Tovanović, B., Mallory, J. P. and Milisaukas, S. 1986. The Kurgan culture, Indo-European origins and the domestication of the horse: a reconsideration (and comments and replies). Current Anthropology 27: 291–313. Arredi, B. Poloni, E. S., Paracchini, S., Zerjal, T., Dahmani, M., Fathallah, M., Makrelouf, M., Pascali, V. L., Novelletto, A. and Tyler-Smith, C. 2004. A predominantly Neolithic origin for Y-chromosomal DNA variation in North Africa. American Journal of Human Genetics 75.2: 338–45. Ballouche, A. and Marinval, P. 2003. Données palynologiques et carpologiques sur la domestication des plantes et l’agriculture dans le néolithique ancien du Maroc septentrional (site de Kaf Taht el-Ghar). Revue d’archéométrie 27: 49–54. Barbaza, M. 2012. Libyco-Berber rock engravings: from one shore of the Sahara to the other, palaeoethnology of Africa. http://www.palethnologie.org/wp-content /uploads/2012/en-GB/Palethnology-2012-GB-07-Barbaza.pdf. Barich, B. 2010. Antica Africa. Alle origini delle società. Antica Africa 1. Rome: L’Erma di Bretschneider. Barich, B. 2016. The introduction of Neolithic resources to North Africa: A discussion in light of the Holocene research between Egypt and Libya, Quaternary International 410: 198–216. Barker, G. 2013. The Neolithisation of Northeastern Africa: reflections on knowns, unknowns, and unknown unknowns. In N. Shirai (ed.), Neolithisation of Northeastern Africa, Studies in Early Near Eastern Production, Subsistence, and Environment 16, Berlin: ex Oriente, 249–56. Blažek, V. 2010. On classification of Berber. Unpublished paper presented at the 40th Colloquium of African Languages and Linguistics. Leiden.

107

Anthony et al. 1986, 300.

https://doi.org/10.1017/9781108634311.016 Published online by Cambridge University Press

16 The Archaeological and Genetic Correlates of Amazigh Linguistics

Blench, R. 2013. Language, linguistics and archaeology; their integration in the study of African prehistory. In P. Mitchell and P. Lane (eds), Oxford Handbook of African Archaeology, Oxford: Oxford University Press, 49–64. Bobrowski, P., Czekaj-Zastawny, A. and Schild, R. 2014. Gebel el-Muqaddas (Site E-06–4). The early Neolithic tumuli from Nabta Playa (Western Desert, Egypt). In J. R. Anderson and D. A. Welsby (eds), The Fourth Cataract and Beyond. Proceedings of the 12th International Conference for Nubian Studies, British Museum Publications on Egypt and Sudan 1, Leuven: Peeters, 293–301. Bokbot, Y., Onrubia-Pintado, J. and Salih, A. 2011. Néolithique et protohistoire dans le bassin de l’Oued Noun (Maroc Présaharien). Quelques données préliminaires. In Actes du colloque international préhistoire maghrébine, Travaux du centre national de recherches préhistoriques, anthropologiques et historiques ns 11.2, 306–21. Broodbank, C. 2013. The Making of the Middle Sea. A History of the Mediterranean from the Beginning to the Emergence of the Classical World. Oxford: Oxford University Press. Brooks, N., Clarke, J., Garfi, S. and Pirie, A. 2009. The archaeology of Western Sahara: results of environmental and archaeological reconnaissance. Antiquity 83: 918–34. Camps, G. 1961. Monuments et rites funéraires protohistoriques, Paris: A.M.G. Camps, G. 1974. Le Gour, mausolée berbère du VIIe siècle. Antiquités africaines 8: 191–208. Camps, G. and Hachid, M. 1982. Un quadrige peint dans la région de Djelfa (Algérie). In G. Camps and M. Gast (eds), Les chars préhistoriques du Sahara. Archéologie et techniques d’attelage. Actes du colloque de Sénanques 21–22 mars 1981, Aix-en-Provence: LAPMO, 153–60. Camps, G. 1992. Punica lingua et épigraphie libyque dans la Numidie d’Hippone. Bulletin archéologique du comité des travaux scientifiques et historiques ns 23B [ 1990–1992]: 33–49. Camps, G. 1995. Djedar. Encyclopédie berbère 16 : 2049–422. Chabot, J. B. 1940. Recueil des inscriptions libyques. Paris: Imprimerie Nationale. Coudray, C., Olivieri, A., Achilli, A., Pala, M., Melhaoui, M., Cherkaoui, M., ElChennawi, F., Kossmann, M., Torroni, A. and Dugoujon, J. M. 2009. The complex and diversified mitochondrial gene pool of Berber populations. Annals of Human Genetics 73.2: 196–214. Desanges, J. 1980. Pline l’ancien, Histoire Naturelle V, 1–46. Paris: Les Belles Lettres. Destaing, E. 1920. Note sur la conjugaison des verbes de forms C1 et C2. Mémoires de la société de linguistique de Paris 22: 139–48. Destaing, E., 2002 (1894). Essai de classification des dialectes berbères de Maroc. Etudes et Documents Berbères 19–20 : 85–101. di Lernia, S., Bertolani, G., Castelli, R., Merighi, F. and Palombini, A. 2002. From regions to sites: the excavations. In S. di Lernia and G. Manzi (eds), Sand, Stones

https://doi.org/10.1017/9781108634311.016 Published online by Cambridge University Press

519

520

Elizabeth Fentress

and Bones. The Archaeology of Death in the Wadi Tanezzuft Valley (5000–2000 BP), AZA 3, Florence: All’Insegna del Giglio, 69–156. di Lernia, S. 2006. Building monuments, creating identity: cattle cult as a social response to rapid environmental changes in the Holocene Sahara. Quaternary International 151: 50–62. di Lernia, S. and Zampetti, D. 2008. Le pitture rupestri dell’Acacus tra passato e futuro. Florence: All’Insegna del Giglio. Douka, K., Grün, R., Jacobs, Z., Lane, C., Farr, L., Hunt, C., Inglis, R. H., Reynolds, T., Albert, P., Aubert, M., Cullen, V., Hill, E., Kinsley, L., Roberts, R. G., Tomlinson, E. L., Wulf, S. and Barker, G. 2013. The chronostratigraphy of the Haua Fteah cave (Cyrenaica, northeast Libya). Journal of Human Evolution 10: http://dx.doi.org/10.1016/j.jhevol.2013.10.001. Dunne, J., Evershed, R., Cramp, L. Bruni, S., Biagetti, S. and di Lernia, S. 2013. The beginnings of dairying as practiced by pastoralists in ‘green’ Saharan Africa in the 5th millennium BC. Documenta Praehistorica 40: 119–30. Ehret, C. 2010. Language and the Testimony of History. Berkeley: University of California Press. El Khayari, A. 2009. Nouvelles remarques épigraphiques et chronologiques sur l’inscription des Azibs n’Ikkis (Haut Atlas, Maroc). Bulletin d’archéologie marocaine 21: 128–36. Fadhlaoui-Zid, K., Haber, M., Martínez-Cruz, B., Zalloua, P., Benammar Elgaaied, A. and Comas, D. 2013. Genome-wide and paternal diversity reveal a recent origin of human populations in North Africa. PLoS ONE 8(11): e80293. doi:10.1371/journal.pone.0080293. Fentress, E. 1982. Tribe and faction: the case of the Gaetuli. Mélanges d’archéologie et d’histoire de l’école française de Rome 94: 325–34. Fentress, E. 2013. Diana Veteranorum and the dynamics of an inland economy. In L. Lavan and W. Bowden (eds), Local Economies? Production and Exchange of Inland Regions in Late Antiquity, Late Antique Archaeology 6, Leiden: Brill, 315–42. Fentress, E. and Wilson, A. 2016. The Saharan Berber diaspora and the southern frontiers of Byzantine North Africa. In S. T. Stevens and J. P. Conant (eds), North Africa under Byzantium and Early Islam, Dumbarton Oaks Byzantine Symposia and Colloquia 7, Washington, DC: Dumbarton Oaks Research Library and Collection, 41–63. Frigi, S., Cherni, L. and Fadhlaoui, K. 2010. Ancient local evolution of sub-Saharan mtDNA haplogroups in Tunisian Berber populations. Human Biology 82.4: 367–84. Galand, L. 1989. Les alphabets libyques. Antiquités africaines 25: 69–82. Gatto, M. C. 2006. The Nubian A-Group: a reassessment. Archéo-Nil 16: 61–76. Gauthier, Y. and Gauthier, C. 2007. Monuments funéraires sahariens et aires culturelles. Cahiers de l’association des amis de l’art rupestre saharienne 11: 65–78.

https://doi.org/10.1017/9781108634311.016 Published online by Cambridge University Press

16 The Archaeological and Genetic Correlates of Amazigh Linguistics

Grébénart, D. 1975. Matériaux pour l’étude de l’Épipaléolithique et du Néolithique du littoral du Maroc saharien. In L’Épipaléolithique méditerranéen (Actes du Colloque d’Aix-en-Provence, 1972), Paris: CNRS, 151–188. Gsell, S. 1901. Les monuments antiques de l’Algérie. 2 vols. Paris: Fontemoing et Cie. Gsell, S. 1902/1911. L’Atlas archéologique de l’Algerie. Algiers, Paris: A. Jourdan, Fontemoing et Cie. Guilaine, J. 2011. Asie, Europe, Afrique au néolithique: la Méditerranée, lien ou frontière culturelle? In Actes du colloque international préhistoire maghrébine, Travaux du centre national de recherches préhistoriques, anthropologiques et historiques ns 11.2, 77–88. Hachid, M., 1992. El-Hadjra el-Maktouba. Les pierres écrites de l’Atlas saharien. Algiers: ENAG. Hachid, M., 2006. Du nouveau sur le monument d’Abalessa (Ahaggar, Algérie). De la date de l’introduction du dromadaire au Sahara central, du personnage d’Abalessa et des inscriptions rupestres dites ‘libyco-berbères’. Sahara 17: 95–120. Hachid, M., 2011. Strabon; Idrissi, La Guerba et un Libyque plus tardif que les Ve/ VIe siècles? In Actes du colloque international préhistoire maghrébine, Travaux du centre national de recherches préhistoriques, anthropologiques et historiques ns 11.2: 191–226. Heddouche, A. 2011. Sur l’apport des monuments funéraires à la connaissance du peuplement et de l’environnement holocène de l’Ahaggar. In Actes du colloque international préhistoire maghrébine, Travaux du centre national de recherches préhistoriques, anthropologiques et historiques ns 11.1, 259–74. Henn, B. M., Botigué, L. R., Gravel, S, Wang W, Brisbin, A., Byrnes, J. K., Fadhlaoui-Zid, K., Zalloua, P. A., Moreno-Estrada, A., Bertranpetit, J., Bustamante, C. D. and Comas, D. 2012. Genomic ancestry of North Africans supports back-to-Africa migrations. PLoS Genet 8.1: e1002397. doi:10.1371/ journal.pgen.1002397. Hutterer, R., Linstädter, J., Eiwantger, A. and Mikdad, A. 2014. Human manipulation of terrestrial gastropods in Neolithic culture groups of NE Morocco. Quaternary International 320: 83–91. Jongeling, K. and Kerr, R. M. 2005. Late Punic Epigraphy: An Introduction to the Study of Neo-Punic and Latino-Punic Inscriptions. Tübingen: Mohr Siebeck Jornet, R., Fadrique, T, and Sanmartí, J. 2011. La fouille dans la zone 3. In N. Kallala and J. Sanmartí 2011, 393–99. Kadra, F. 1983. Les Djedars, monuments funéraires berbères de la région de Frenda. Algiers: University of Algiers. Kallala, N. and Sanmartí, J. 2011. Synthèse des résultats: deux mille ans d’histoire d’Althiburos. In N. Kallala and J. Sanmartí 2011, 31–38.

https://doi.org/10.1017/9781108634311.016 Published online by Cambridge University Press

521

522

Elizabeth Fentress

Kallala, N. and Sanmartí, J. (eds). 2011. Althiburos I. La fouille dans l’aire du capitole et dans la nécropole méridionale, Documenta 18. Tarragona: Institut Català d’Arqueologia Clàssica. Kossmann, M. Forthcoming. Berber subclassification. In: R. Vossen (ed.), The Oxford Handbook of African Languages, Oxford: Oxford University Press. Laporte, J. P. 2005. Les djedars, monuments funéraires berbères de la région de Frenda et de Tiaret. In C. Briand-Ponsart (ed.), Identités et cultures dans l’Algérie antique, Rouen: Publications des Universités de Rouen et du Havre, 321–406. Leahy, M. A. 1985. The Libyan period in Egypt: an essay in interpretation. Libyan Studies 16: 51–65. Lepelley, C. 2005. Témoignages de saint Augustin sur l’ampleur et les limites de l’usage de la langue punique dans l’Afrique de son temps. In C. Briand-Poinsart (ed.), Identités et cultures dans l ‘Algérie antique, Rouen, Mont-Saint-Aignan: Publications des Universités de Rouen et du Havre, 137–53. Linstädter, J. 2003. Le site néolithique de l’abri Hassi Ouenzga. Beiträge zur allgemeinen und vergleichenden Archäologie 23: 85–138. Liverani, M. 2000. The Garamantes: a fresh approach. Libyan Studies 31: 17–28. Liverani, M. 2004. Rediscovering the Garamantes. Libyan Studies 35: 191–200. Liverani, M. (ed.). 2006. Aghram Nadharif: The Barkat Oasis (Sha Abiya of Ghat, Libyan Sahara) in Garamantian Times, AZA 5. Florence: All’Insegna del Giglio. Luccarini, G., Radini, A., Barton, H. and Barker, G. 2015. The exploitation of wild plants in Neolithic North Africa. Use-wear and residue analysis on non-knapped stone tools from the Haua Fteah cave, Cyrenaica, Libya. Quaternary International 410: 77–92. Marichal, R. 1992. Les ostraca de Bu Njem. Tripoli: Libya Antiqua, supplément VII. Mattingly, D. J. (ed.). 2010. The Archaeology of Fazzan. Volume 3, Excavations carried out by C. M. Daniels. London: Society for Libyan Studies, Department of Antiquities. Mattingly, D. J., Leitch, V., Duckworth, C. N., Cuenod, A., Sterry, M. and Cole, F. (eds). 2017. Trade in the Ancient Sahara and Beyond. Trans–Saharan Archaeology Volume I. Series editor D. J. Mattingly. Cambridge: Cambridge University Press, Society for Libyan Studies. Mitchell, P. 2017. The Donkey in Human History. An Archaeological Perspective. Oxford: Oxford University Press. Modéran, Y. 2003. Les Maures et l’Afrique romaine (IVe-VIIe siècle). BEFAR 314. Rome: École Française de Rome. Moreno Garcia, J. C. 2014. Inveders or just herders? Libyans in Egypt in the 3rd and 2nd millennia BCE. World Archaeology 46.4: 610-23. Mori, L. 2013. Life and Death of a Rural Village in Garamantian Times, Archaeological investigations in the oasis of Fewet (Libyan Sahara). AZA 6. Florence: All’Insegna del Giglio. Morizot, P. 1997. Archéologie aérienne de l’Aurès. Paris: CTHS.

https://doi.org/10.1017/9781108634311.016 Published online by Cambridge University Press

16 The Archaeological and Genetic Correlates of Amazigh Linguistics

Morizot, P. 2003. Les recherches en matière de protohistoire: l’état de la question pour l’Aurès. Indices cartographiques pour le Djebel Amour. In M. Khanoussi (ed.), Actes du VIIIe colloque international sur l’histoire et l’archéologie de l’Afrique du Nord (1er colloque international sur l’histoire et l’archéologie du Maghreb), Tunis: Institut National du Patrimoine, 65–98. Morkot, R. G. 2016. Before Greeks and Romans. Eastern Libya and the oases, a brief review of interconnections in the Eastern Sahara. In N. Mugnai, J. Nikolaus and N. Ray (eds), De Africa Romaque. Merging Cultures across North Africa, London: Society for Libyan Studies Conference volume I, 27–38. Muzzolini, A. 1995. Les images rupestres du Sahara. Toulouse: Pré histoire du Sahara 1. Ottoni, C., Martínez-Labarga, C., Loogväli, E. L., Pennarun, E., Achilli, A., De Angelis, F., Trucchi, E., Contini, I., Biondi, G. and Rickards, O. 2009. First genetic insight into Libyan Tuareg: a maternal perspective. Annals of Human Genetics 73: 438–48. Ottoni, C., Larmuseau, M. H. D., Vanderheyden, N., Martínez-Labarga, C., Primativo, G., Biondi, G., Decorte, R., and Rickards, O. 2011. Deep into the roots of the Libyan Tuareg: a genetic survey of their paternal heritage. American Journal of Physical Anthropology 145: 118–24. Paris, F. and Saliège, J. F. 2011. Chronologie des monuments funéraires sahariens. Problèmes, methodes et résultats. In Actes du colloque international préhistoire maghrébine, Travaux du centre national de recherches préhistoriques, anthropologiques et historiquesns 11, 275–79. Pereira, L., Cerny, V., Cerezo, M., Silva, N. M., Hajek, M., Vasikova, A., Kujanova, M., Brdicka, R. and Salas, A. 2010. Linking the sub-Saharan and West Eurasian gene pools: maternal and paternal heritage of the Tuareg nomads from the African Sahel. European Journal of Human Genetics 18: 915–23. Pelling, R. 2005. Garamantian agriculture and its significance in a wider North African context: the evidence of the plant remains from the Fazzan Project. Journal of North African Studies 10: 397–411. Quinn, J. 2011. The Syrtes between East and West. In A. Dowler and E. R. Galvin (eds), Money, Trade and Trade Routes in Pre-Islamic North Africa, London: British Museum Research Publication, 11–20. Rebuffat, R. 1992. Bu Njem, Enciclopédie berbère XI: 1626–42. Renfrew, C. 2002. Genetics and language in contemporary archaeology. In B. Cunliffe, W. Davies and C. Renfrew (eds), Archaeology. The Widening Debate, Oxford: Oxford University Press, 42–90. Retaillé, D. 1998. L’espace nomade. Revue de géographie de Lyon 73.1: 71–81. Riemer, H. 2004. News about the Clayton rings: long distance desert travellers during Egypt’s Predynastic. In S. Hendrickx, R. F. Friedman, R. F. Cialowicz and M. Chlodnicki (eds), Egypt at Its Origins. Studies in Memory of Barbara Adams, Proceedings of the International Conference ‘Origin of the State. Predynastic

https://doi.org/10.1017/9781108634311.016 Published online by Cambridge University Press

523

524

Elizabeth Fentress

and Early Dynastic Egypt’, Krakow 28th August–1st September 2002, Leiden: Orientalia Lovaniensia Analecta, 972–89. Ritner, R. K. 2009. The Libyan Anarchy: Inscriptions from Egypt’s Third Intermediate Period. Leiden: Brill. Rossel, S., Marshall, F., Peters, J., Pilgram T., Adams, M. D. and O’Connor, D. 2008. Domestication of the donkey: new data on timing, processes and indicators. Proceedings of the National Academy of Sciences 105: 3715–20. Roubet, C. 2011. Le Mahrouguétien: activités forestières et agricoles probables durant le néolithique et la protohistoire, dans les Némencha (Algérie orientale). In Actes du colloque international préhistoire maghrébine, Travaux du centre national de recherches préhistoriques, anthropologiques et historiques ns 11.2, 13–26. Salih, A. and Hackendorf, R. 2002. L’art rupestre libyco-berbère au Maroc. Etat des connaissances. Beitrăge zur Algemeinen und Vergleichenden Archăologie 22: 65–94. Scheele, J. 2012. Smugglers and Saints of the Sahara. Regional Connectivity in the Twentieth Century, African Studies 120. Cambridge: Cambridge University Press. Sterry, M. and Mattingly, D. J. (eds). Forthcoming. State Formation and Urbanisation in the Ancient Sahara and Beyond. Trans-Saharan Archaeology Volume IV. Cambridge: Cambridge University Press, Society for Libyan Studies. Sweet, L. 1965. Camel raiding of North Arabian Bedouin: a mechanism of ecological adaptation. American Anthropologist 67: 1132–50.

https://doi.org/10.1017/9781108634311.016 Published online by Cambridge University Press

17

Concluding Discussion martin sterry, david j. mattingly, maria carmela gatto and nick ray

Saharan Connections, Identities and Funerary Practices At the start of this volume, we outlined three main themes: burial archaeology, migration and identity. In this concluding chapter, we revisit these themes in light of the varied case studies and examples explored in this book. We believe this volume illustrates that there is irrefutable evidence of links and connections that serve to define a Trans-Saharan zone at an early date. Migration and mobility created networks of connectivity and elements of a shared culture (koine). Berber languages are still spoken across large swathes of the Sahara today and Ehret and Blench in their chapters both argue that this was even more pronounced in the past, albeit they propose quite different reconstructions as to when and how Berber migration occurred.1 Ehret’s analysis of the family of Berber languages and his ordering of these in relative chronology, in terms of degrees of similarity and difference, is an important contribution. But the absolute dating in his chapter is somewhat uncertain, as too the relationship between linguistic changes and other forms of evidence, especially the archaeological data. Language in itself is just one aspect of culture and identity, but we have also noted the broadly similar spatial distributions of ‘horse period’ rock-art and Libyan (Berber) inscriptions in the Sahara.2 That would seem to support the sort of model suggested by Fentress that whereas there were already Proto-Berber elements widely disseminated within the Sahara by the second millennium BC, it was the diffusion of Proto-Amazigh that created a Berber-dominated Sahara. She correlates this with the spread of oasis settlements in the first millennium BC. 1

2

Blench, Chapter 14, this volume, in particular argues for a far more heterogeneous linguistic prehistory. Gauthier and Gauthier 2011.

https://doi.org/10.1017/9781108634311.017 Published online by Cambridge University Press

525

526

Martin Sterry et al.

It remains uncertain to what extent this represented a large-scale migration of ‘Mediterranean’ Africans taking over spaces formerly occupied by the Neolithic herders or on the other hand an amalgamation of surviving elements of the Neolithic pastoralists with smaller Berber elements.3 The rock art evidence suggests that the pastoralists associated with the early phases of Neolithic rock art were primarily black, but in the Late Pastoral period there are also images of light-skinned people of different physiognomy.4 Some of the evidence from the Early Garamantian phase supports the amalgamation theory, rather than the notion of direct replacement. In particular, the Early Garamantian funerary structures and settlements like Zinkekra featured material elements that were very similar to the Late Pastoral ones, alongside things that were obvious innovations brought from outside the Sahara. Whilst some, such as the developed agricultural package,5 attest to connection with the Nilo-Mediterranean world, others, such as the ceramic technology with roulette decoration and grog temper, attest to connection with West and Sub-Saharan Africa.6 The latter reminds us that the amalgamation process relied on multidirectional connectedness. How might we relate this evidence to other key themes of the Trans-SAHARA Project? Less geographically widespread, but still cross-cutting the Sahara, is the evidence of trade in items such as carnelian, glass, ceramics, metals and metal objects, textiles and slaves.7 A further category is the adoption of technologies, notably those associated with metalworking, agriculture and animal husbandry.8 However, these varied aspects can all be argued to be ephemeral or potentially due to short-term connections and they do not in themselves demonstrate the existence of a common Saharan culture (koine). Funerary archaeology therefore is an obvious form of evidence from which to look at societal practices and biological links across the Sahara, as it is both durable and easily recognisable with monuments recorded in most regions for the period 3000 BC–AD 1000. Urbanisation and state formation are also concepts that relate to major societal changes and in a Saharan context these seem to have appeared first within oasis communities.9 3 4 5 6 8 9

Mattingly 2003, 342–43. See for example, Lutz and Lutz 1995, 140–44; van Albada and van Albada 1993, 547–54. For extended discussion, see Mattingly 2003, 332–62; 2010, 78–84, 523–30; 2013, 525–44. Gatto 2006; 2010; 2013; Forthcoming; Castelli et al. 2005. 7 See Mattingly et al. 2017a. See Cuénod et al. Forthcoming. Mattingly and Sterry 2013; Sterry and Mattingly Forthcoming.

https://doi.org/10.1017/9781108634311.017 Published online by Cambridge University Press

17 Concluding Discussion

In terms of funerary practices and structures, among interesting aspects to emerge from the volume are the difficulty in comparing data, especially old data, and the inconsistency in typologies of funerary structures. Particularly difficult are typologies that employ terms like ‘bazina’ or ‘chouchet’ (where the parameters of each category are vague and often differently interpreted by researchers, as is clear from the chapters in this book where these terms are used). A strong recommendation arising from this volume is that all future work needs to develop clear descriptive typological categories, based on practical rules for discriminating one type from another (and that are still to a degree useable even with badly damaged monuments). An additional problem is that most typologies derived from survey work emphasise superstructure and external appearance, whereas excavation tends to reveal additional complexities of substructure and subsurface features. Typologies need to be sensitive to regional specificities – it is not straightforward to simply apply a typology developed in one zone to a new area of study. Bioturbations and looting activities also create added difficulties for consistently identifying the type of tumuli. The importance of radiocarbon (and other forms of absolute) dating will be clear from the preceding chapters. The conspicuous consumption patterns of the Garamantes aside, Saharan burials rarely contained material culture that is readily diagnostic. Assumptions about the approximate age of burials have on several occasions been shown to be wildly inaccurate. Scientific dating programmes are vital in studying Saharan funerary archaeology. New technologies clearly have huge potential to transform our knowledge and to allow us to record burials in new ways. Satellite images and drone photography are essential tools in mapping, but ground-based 3-D recording techniques, like structure from motion (SfM) are equally useful for understanding the morphology of individual monuments.10 In terms of burial rite, inhumation was the most commonly encountered practice across the Trans-Saharan zone, with cremation very rare and mainly limited to the coastal areas of Libyphoenician and Roman settlement. Sanmartí et al. open an important discussion regarding a possible Numidian practice of excarnation prior to selected remains being inhumed. Given the disturbed state of many monuments, the relatively small sample of well-excavated burials and the collective element of many monuments in that area (with clear indications of successive burials in the same structure), we are less convinced that the evidence points to

10

Mattingly et al. 2017b for example of SfM used on Moroccan desert tombs.

https://doi.org/10.1017/9781108634311.017 Published online by Cambridge University Press

527

528

Martin Sterry et al.

secondary inhumation after excarnation elsewhere. It seems equally plausible that the observed processes of disarticulation and selection of body parts could relate to disturbance of earlier interments by secondary burials being added to chambers. The curation and careful treatment/selection of the disturbed body parts is nonetheless a very interesting phenomenon and one that may well provide further insights into veneration of ancestors. The chapters in this volume have identified a number of traits that we offer as underlying Saharan cultural traditions of funerary practice between c.1000 BC and AD 1000: • Normally individual burials beneath circular stone-built monuments; • A wide distribution of drum tombs as well as cairns; • Some addition of annexes to monuments to contain subsidiary burials with presumed kinship links; • Tightly flexed corpses lain on the right or left side; • Use of leather or textile shrouds; • Application of ochre to body, tomb or stelae; • Important associations with the east side of monument both in terms of the alignment of the body and items in the tomb and of the placement of offering spaces and funerary furniture against that side of the tomb (although sometimes we can note association with the west side also); • Some more evolved structures on the east side of tombs probably served as funerary chapels used for the practice of communing with dead ancestors; • Some gender differentiation (including sometimes differences in orientation of male and female corpses in the grave); • Material culture as burial goods was generally limited to items of personal adornment, with pottery rare inside the grave; • Relatively egalitarian appearance of most burials, but with signs of increased hierarchy in some areas over time. It is clear from dated monuments that many of the key structural characteristics were already present at an earlier date among Late Pastoral communities between 3000 and 1000 BC.11 That may suggest some continuities with past populations and there is indication of a general westward trend from the eastern Sahara (where the earliest forms of funerary stone structures are dated to the fifth millennium BC) to the western Sahara. Nonetheless, there were also changes after 1000 BC, for instance, with more individual than collective monuments, at least within the Sahara itself. Some intriguing cultural cross-overs are detected, like the mudbrick pyramids of Fazzan, which may reflect complex patterns of movement of funerary ideas. Others, 11

Sivilli 2002.

https://doi.org/10.1017/9781108634311.017 Published online by Cambridge University Press

17 Concluding Discussion

instead, like the Egyptian funerary practice of mummification, did not travel far beyond the oases closest to Egypt. Within this framework, we would suggest that there are several aspects where variation and experimentation was common, most notably in tomb architecture.12 Although circular monuments of 5 m (or less) diameter predominate in the majority of Saharan regions, most areas also exhibit a range of more elaborate funerary monuments. These include larger tombs, or the addition of arms and antennae to tombs, encircling rings of stone and the use of different construction materials, such as mudbrick. In the well-researched case of the Garamantes (especially within the Wadi al-Ajal), we can outline a more coherent, regionally distinct set of traits for the period AD 1–600 that sat alongside the Saharan traits described above: • Introduction of rectilinear burial structures (including stepped monuments, mausolea and pyramids) and rectangular offering tables; • Some new types of circular burials – dense nucleated shaft burial cemeteries; • Use of mudbrick for some oasis-centre tombs; • Addition of imported artefacts and materials to grave goods included in the tomb and a dramatic increase in the range and number of these; • Emphasis on conspicuous consumption (feasting and drinking) indicated by choice of imported vessels to include in tombs; • Strong sense of hierarchy reflected in funerary architecture and burial rituals; • Many of the Saharan traits (shrouds, ochre, eastern orientation, etc.) continued alongside the innovations, but with more internal variation. Interestingly, the closest parallels for many of these Garamantian funerary traits and experimentations with form and monumentality come from the areas of the nascent Numidian and Mauretanian kingdoms of the first millennium BC.13 The work at Althiburos is particularly important, with dolmenic mounuments there taking both circular and quadrangular forms.14 In the Tanzzuft/Tadrart Akakus region, we see echoes of these traits (including occasional imported artefacts), but also a distinct regional identity portrayed through such things as vesicular basalt bowls.15 Elsewhere in 12

13 14

15

Camps 1961 remains indispensable still in this regard; see also di Lernia and Manzi 2002; di Lernia and Tafuri 2013 for the Neolithic precursors. See Sanmartí et al., Chapter 8, and Papi, Chapter 9, this volume. As illustrated in Kallala et al. 2018, a superb synthesis and the most important book on the Protohistoric burials of the Maghrib since Camps 1961. Gatto et al., Chapter 3, this volume.

https://doi.org/10.1017/9781108634311.017 Published online by Cambridge University Press

529

530

Martin Sterry et al.

southern Morocco and Western Sahara, we see further regionalism and variation on the Saharan traits.16 While Clarke and Brooks argue for significant Atlantic connections and influences alongside Saharan ones in the Western Sahara, we think that the overlap with broader Saharan traditions is the more predominant aspect. However, much of what makes the Garamantian burials distinctive relates to their increased complexity and the display of wealth and inequality that to some extent were dependent on (and reflective of) Saharan trade and migration. Notwithstanding some structural differences, parallels can also be drawn with the burials of Tin Hinan and In-Aghelachem and we might think of these as nodes in the new sphere of interaction that emerged in the late first millennium BC. The new styles of burials, especially the introduction of rectilinear elements and of monumental (elite) types, and the association with enhanced assemblages of grave goods may thus be indicators of the spread of trading and oasis farming communities within the wider Saharan world. In the preceding chapters we have encountered evidence of recognisable Saharan burial traditions, but also of distinctive regional practices. While the environment had its part to play in facilitating contact or separating different areas, it was fundamentally people who created Trans-Saharan connectivity – whether mobile pastoralists, specialist traders or oasis cultivators. As addressed by Edwards, there were commonalities in the funerary evidence from the Middle Nile region and the Sahara, mainly based on past shared experiences and themes. The east-west trajectory of interaction was less pronounced than the Nilotic links, because Egyptian influence on Nubian funerary architecture and rituality was stronger. However, this does not mean a complete absence of connectivity, rather a difficulty in discerning it archaeologically. In the Maghrib, to the north of the Sahara, Sanmartí and Papi, both building on the foundational work of Camps, emphasise the increasing hierarchy and complexity found in burial practices of the first millennium BC.17 In Northern Morocco, we find the same Saharan traits described above from the fourth/third centuries BC alongside distinctive regional practices (for instance, rectangular cists and the wide-array of locally sourced grave goods). Whilst much of the innovation and complexity is similar to what we have described for the Garamantes, some of the Moroccan evidence dates several centuries earlier to the seventh-third 16 17

Bokbot, Chapter 10, this volume; Clarke and Brooks, Chapter 11, this volume. Camps 1961; Sanmartí et al., Chapter 8, this volume; Papi, Chapter 9, this volume.

https://doi.org/10.1017/9781108634311.017 Published online by Cambridge University Press

17 Concluding Discussion

centuries BC. Expanding on the work of Bokbot, our current work in the Draa Valley south of the Atlas Mountains has revealed a different pattern there, with classic Saharan drum tombs and mound and corbel cairns dated predominantly to the first millennium AD.18 In the eastern Maghrib, we find a dizzying diversification of tomb types and funerary practices, again in the mid-first millennium BC, but the central features of Saharan burials continued to persist. Several case studies relate to societies that seem to have been undergoing state formation processes (Garamantes, Numidia, Mauretania). The Maghribian societies also demonstrate a strong interaction with the material culture of Phoenician and Roman civilisations, mainly promulgated via coastal colonies and emporia. Such goods only reached the Garamantes in any quantity after a considerable time lag and the relative rarity of Roman goods in other oases suggests that the access of the Garamantes to Mediterranean material culture was fairly exceptional. In the Tripolitanian pre-desert to the north of Fazzan, the impact of Libyphoenician and Roman culture on funerary traditions was also pronounced, but not before the late first century AD.19 Given this broad social koine, it is significant that the practices south of the Sahara are so abruptly different. To the south of Lake Chad there are a few similarities, such as some instances of flexed burials, use of textile shrouds, wrapping of bodies in matting and occasional uses of Saharan material (for instance, carnelian beads) in burials, but there are no really close parallels. The prevalence in burials of a western orientation is interesting given some of the examples of this from Fewet, but such practices as jar burials and burial within settlements, along with the lack of surface funerary monuments, are not paralleled in the Sahara. In Burkina Faso, the cemetery of Kissi, although a distinct burial area, has no surface monuments and there are few comparisons with the north, apart from the burial of materials traded in along Saharan routes (textiles, carnelian and copperalloy objects) within the graves. Apart from this strong break in funerary practices, the other notable aspect of these funerary groups is how in both cases there is a diversification of forms and some sense of stratification that partly parallels increasing contacts (albeit still poorly understood) with the Sahara. On the other hand, the finds from Kissi emphasise its connectedness with areas that were trading in copper alloys, glass and carnelian beads and textiles. More excavations of the quality of the Kissi work are needed to

18

Mattingly et al. 2017b.

19

Barker et al. 1996, 142–49; Mattingly 1995, 162–67.

https://doi.org/10.1017/9781108634311.017 Published online by Cambridge University Press

531

532

Martin Sterry et al.

fully evaluate the impact of such long-range contacts on local customs and practices. Although there are signs of increasing complexity and hierarchy over time across the Trans-Saharan region, we should be clear that we do not suggest that this related to an importation of complexity only from the north (Mediterranean) to south (for example, West Africa).20 Rather, we would highlight that this may reflect the increasing connectivity and movement of people taking place in the Trans-Saharan sphere. The movement of people, goods and technologies was multi-directional and while it is true that a number of technologies appear to have entered the Sahara from the Nile/Western Egyptian Desert (the agricultural package, foggara irrigation technology, horses, donkeys and camels, chariots), this was not invariably the case. Some other datasets, like handmade pottery traditions and the spread of cultivation of Sub-Saharan crops (sorghum, pearl millet and cotton), suggest also a south to north direction for connectivity.21 These multilateral connectivities may have fostered experimentation and innovation, but may also have led to changing power dynamics and inequalities. Both processes may be reflected in the funerary record.

Burial Landscapes Excavation of Saharan tombs is both costly and time-consuming, with more complex and better preserved tombs taking longer to fully examine and record. The Desert Migrations Project excavated a little over 142 tombs over five seasons of fieldwork, while the Italian Mission in Fewet excavated just 24 in three seasons.22 It is therefore challenging within the constraints of a single project to sample fully the wide array of funerary types (in the Wadi al-Ajal alone there were 10 main tomb types and more than 20 subtypes identified). It is also difficult to cover the full range of periods (especially those with no diagnostic artefacts to assist dating) or to recover a sample that is large enough to be considered representative. This has further effects on how to 20 21

22

For instance, S. McIntosh 1999; R. McIntosh 2005. On the handmade pottery tradition, see Gatto 2010, 35–36; Forthcoming; on sub-Saharan crops in Fazzan, see Mattingly 2013, 482, 511. Mori 2013. The discrepancy largely relates to the predominance of the simpler shaft burials in the Wadi al-Ajal, which can be more rapidly examined than a cairn structure, and the better state of preservation in Fewet.

https://doi.org/10.1017/9781108634311.017 Published online by Cambridge University Press

17 Concluding Discussion

contextualise data analysis, such as that from isotopes. It is on this basis that David Edwards draws attention to the value of analysing burial landscapes, particularly through the use of satellite remote sensing and geographical information systems (GIS). While we would agree with the general argument, particularly given the current inaccessibility of much of the Sahara, we would argue that a multi-scalar approach is most effective, as for many Saharan regions we are almost bereft of reliable contextual data or dated tombs. This approach combines excavation, scientific dating programmes, field survey and remote-sensing to generate a broad range of data types, although we should still be mindful of the scale of work this implies. In addition to the excavations mentioned above, the Italian Mission surveyed 1,324 tombs around Fewet and the British Mission more than 10,000 in a relatively small area of the Wadi al-Ajal, providing a measure of statistical back-up for conclusions drawn from a relatively small number of excavated tombs. This type of survey is particularly effective in the Sahara where a small team can record several hundred tombs in a day, collecting evidence of architecture, finds, dating and recovering some osteological data. Satellite remote sensing correspondingly can easily generate datasets of hundreds of thousands of tombs although the potential is yet to be fully realised in the Sahara.23 The greatest strengths of this research will be to understand better the variability of different funerary monuments, their numbers and arrangement across the Sahara. It is hard to overstate the importance of obtaining radiocarbon dates for burials, as so many Saharan burials contain little diagnostic material culture and only with greater chronological precision can we link funerary behaviours with other societal changes. In Fazzan, the dense spread of cairns and shafts are strongly associated with the spread of oasis agriculture and corresponding population growth in the late first millennium BC. A similar pattern may be observed in the Wadi Draa of Morocco in which thousands of cairns cluster around groups of hilltop settlements, some of which are now dated to the pre-Islamic era.24 Yet, we cannot create a dichotomy between dense cemeteries of agriculturalists and dispersed cairn fields of pastoralists. Some agricultural communities may have utilised less formal models of cemeteries. Merlo 23

24

Some recent examples from other parts of the world include: Bradbury and Philip 2011; Bradbury et al. 2016; Löwenborg 2010. See now, Bewley et al. 2016; Rayne et al. 2017 for the Endangered Archaeology of the Middle East and North Africa (EAMENA) Project. See also, Bokbot, Chapter 10, this volume; Mattingly et al. 2017b, for a first set of AMS dates.

https://doi.org/10.1017/9781108634311.017 Published online by Cambridge University Press

533

534

Martin Sterry et al.

et al. have noted in the Wadi ash-Shati how cairns there were not arranged into discrete cemeteries, but rather extended in linear lines along wadi banks.25 Equally, there appear to be some examples of dense funerary landscapes and high-status burials that are better interpreted in relation to pastoral or mobile groups, as at In-Aghelachem on a desert route between the Wadi Tanzzuft and Jarma.26 One of the densest cairn cemeteries in the Wadi ash-Shati lies north-west of the main agricultural zone at a key jumping-off point for a route across the Ubari sand sea towards Ghadamis. Another very large cluster of tombs is at the most south-easterly point in Fazzan – Tajirhi – from where caravans face a difficult journey to the next major oasis at Seguedine. In the Wadi Draa, Morocco, dense concentrations of cairns occur at Foum Larjam, at the point where the resources for oasis agriculture stop and from where a route runs south-east towards the Tuwat oases. Hence, some of these larger nucleated areas of funerary activity seem directly connected to forms of movement, or else they may represent places of refuge from which mobile groups operated, or places of strong interaction between sedentary and mobile groups. Some Saharan oases are currently bereft of visible signs of preIslamic funerary activity, such as the Tuwat and Ghardaia groups of central Algeria. This could relate, of course, to a lack of pre-Islamic activity, but could also be due to local variance in funerary rite that renders tombs and cemeteries less prominent in the landscape. Investigating and explaining such absences should be priorities of future research.

Migration In the introduction to this volume, we adopted a loose definition of migration within which we could identify a number of different forms, encompassing a range of mechanisms – pastoral mobility, large-scale population movements (including refugees, conquest), endogamy, trade, slaving, financial migration and so on (see Chapter 1, Table 1.1). Different examples of past ‘migrations’ need to be carefully considered against this sort of interpretational matrix. In this section, we summarise by periods our view on the potential role of migration within different parts of the Trans-Saharan zone from late prehistory to the early modern period. As Magnavita’s chapter reminds us, in the Sub-Saharan area population 25

Merlo et al. 2013.

26

Gatto et al., Chapter 3, this volume.

https://doi.org/10.1017/9781108634311.017 Published online by Cambridge University Press

17 Concluding Discussion

migration is a recurrent theme in interpreting disjunctures in past societies, though difficult to trace in detail archaeologically.

The Neolithic (Pastoral) Phase Pastoralism probably provided the major form of mobility in this phase, evidence for which stretches back in the Sahara to the Mid-Holocene, although during this time it had gradually shifted from being cattle-based to caprine-based.27 Alongside intra-regional transhumant mobility, however, there is evidence for superregional mobility across the Sahara, with directionality mainly driven by ecological factors. Long-distance mobility of women is suggested by the isotope analysis of funerary remains from the Wadi Tanzzuft and the Takarkori Shelter in the southern Akakus Massif of Fazzan.28 While the formation of a cultural koine, which encompassed the Central and Western Sahara, is attested by the spatial distribution of the peculiar ‘key-hole’ and ‘antenna/crescent’ tombs, broadly dated to fourth/ third millennium BC.29

The First Oasis Settlements of the First Millennium BC In the first millennium BC, there is a great deal of evidence for shortdistance forms of mobility and migration, but far less for longer distance migrations. Pastoral populations no doubt had a continuing role in connecting up the vast spaces of the Sahara. Berber language, rouletted ceramics, rock art and funerary practices all suggest that there are shared traditions across the Sahara at this time. Ehret and Fentress have relied on this as evidence for the large-scale movement of people, but this is far from conclusive. Either way, it seems unlikely that long-distance forms of migration were common during this period. Forms of forced movement and slavery were certainly possible or indeed probable, but the low population density in the Sahara may well have favoured short distance communication and raids between groups rather than the long-distance slaving that developed later. Towards the end of the first millennium BC and start of the first millennium AD, there is evidence for a much broader range of forms of mobility and migration. Pastoralism likely continued, but the introduction of the camel around this time would have encouraged new forms of 27 29

di Lernia and Merighi 2006. 28 Tafuri et al. 2006a; 2006b; di Lernia and Tafuri 2013. Berkani et al. 2015; Gauthier 2015

https://doi.org/10.1017/9781108634311.017 Published online by Cambridge University Press

535

536

Martin Sterry et al.

husbandry, probably over longer ranges.30 Blench suggests this was the point for the expansion of Berber language, but in any case we have strong evidence for an explosion in agricultural populations in the Sahara, both in established oases and in newly colonised oases.31 In Fazzan, the earliest settlements of the Garamantes reveal combined evidence of the adoption of a Mediterranean/Nilotic package of crops and an underlying assemblage of lithics and handmade pottery that represents a continuation of the preexisting Late Pastoral groups.32 That suggests some sort of merger between established pastoral groups and migrants from outside the area, with access to and expertise in new agricultural crops and technologies. A prime model for the spread of oases is thus that migrant groups and mobile technologies played a role in sedentarising elements of the established pastoral communities across the Sahara. While the presentation of the new evidence relating to the Garamantes has placed emphasis on the significance and size of the oasis agricultural element within their society, this is not to deny the existence of significant more mobile pastoral groups within their wider community. A range of activities will have depended on the navigational and survival skills of these pastoral elements – notably of course, desert travel and trade, slaving and warfare, the breeding and raising of pack animals. Pastoral groups were surely a key element of Garamantian society, potentially crucial in terms of military muscle and both guides and guards of caravans. They were also no doubt the main providers of pack animals, meat and secondary animal products for the oases communities. The relations between oasis cultivators and pastoralists were, of course, symbiotic ones and the composition of the communities in and around individual oases was always liable to be quite mixed.

The Roman Period The heyday of the Western Roman Empire, between the first and early fifth centuries AD, was a period of profound impact on the Sahara by its northern neighbour. Roman interest in the desert clearly began at a comparatively early date, with the campaigns of Cornelius Balbus c.20 BC coinciding with similar long-range desert expeditions into upper Egypt and Arabia in the same period.33 30 31

32 33

Bernus 2006. This is a key theme of Sterry and Mattingly Forthcoming, but we may note how the chronology proposed by Fentress, Chapter 16, this volume, fits around this. Mattingly 2003, 341–46; 2010, 78–84. Mattingly 2003, 76–90, for discussion of the historical sources relating to the Garamantes.

https://doi.org/10.1017/9781108634311.017 Published online by Cambridge University Press

17 Concluding Discussion

On the northern fringes of the Sahara, the Roman conquests of North Africa led to the implantation of military communities at many key oases, for example Bu Njim, Ghadamis, Gemellae.34 It is not possible in most cases to date the earliest activity at these oasis sites, but there are clear indications in some cases at least that the oasis pre-dated the military garrison. Military garrisons at oases in the northern Sahara thus created another route for migration of people and material culture into the Sahara. The Roman sources indicate that there was regular contact across the desert frontiers with Saharan peoples, such as the Pharusiens who with the aid of water-skins traversed the area of the Chotts on horseback to reach as far north as Cirta.35 At Jarma, in the central Sahara, trade with the Mediterranean intensified during the early first millennium AD and the frequency of contacts implies the establishment of some kind of trading community specialising in longdistance mobility. The provinces of Roman North Africa provide evidence of facilitating forms of economic or wage-labour migration, such as during harvest periods.36 The oasis agriculture in the Sahara is also likely to have functioned in a similar way at certain times. Finally, the liminal nature of the Sahara for the Mediterranean, and possibly also Nilotic and Sahelian communities, led some people to use it as a place of refuge or exile. This is most apparent in the oases of Egypt’s Western Desert, but the Garamantes may well have been an attractive prospect for those fleeing persecution or who were religious or political dissenters in neighbouring territories.37

The Late Antique and Early Medieval Periods One of the outcomes of this volume is a realisation that migration was a continuing dynamic in successive phases of Saharan history. Indeed, the late antique and early medieval period appears to have been a time of dramatically increased mobility and population movements across the region. One needs only to cite the Vandal, Byzantine and Islamic invasions of North Africa as examples of major incidents that sparked political 34 36

37

Mattingly et al. 2013a. 35 Strabo, Geography, 17.3.7. The account of the raids of the Austuriani people into Tripolitania in the mid-fourth century AD indicate that one of the causes was the execution of one of their chieftains who had been attempting to enter the province with other seasonal labourers, Amm Marcellinus 28.6.2–4; Mattingly 1995, 176–77. Boozer 2013; Jackson 2002 on the Western Desert. When the Nasamones people rebelled against Domitian, a Roman army was first sent against their oasis bases and afterwards, it seems, a column was also dispatched to Garamantian lands to prevent refugees finding safe haven there, Mattingly 1995, 33, 72–73; 2003, 83–85.

https://doi.org/10.1017/9781108634311.017 Published online by Cambridge University Press

537

538

Martin Sterry et al.

change with attendant consequences for populations in Maghrib and Sahara. But the indications of increased mobility go beyond these highlevel political events. A late antique Saharan people known as the Laguatan played a major part in destabilising the Roman and Byzantine frontier in Tripolitania, though exactly where they came from has been controversial. The clearest evidence links them with the oases south of the Syrtic gulf – broadly the old territory of the Nasamones, but as time went on they seem to have become a confederation of allied peoples spanning a large area of the Sahara and the pre-desert zone.38 In a recent paper, Fentress and Wilson have argued for just such a model of increased mobility of desert peoples in relation to the collapse of the Roman frontier further west.39 Such ideas, of course, have echoes of the old French colonial view of the menace of ‘nomades chameliers’,40 though of course we know now that the camel was present at far earlier dates. The obvious question is what caused such an upturn in mobility and confrontation with the Roman/ Byzantine empires? The evidence of Roman contacts with the Garamantes and, for instance, the oasis of Ghadamis suggests that the late antique unrest followed a prolonged period of more stable relations between Rome and the desert peoples of the northern Sahara. On the other hand, our fieldwork in southern Libya and more recently in southern Morocco has revealed that after about AD 300 there was a significant upsurge in both areas in the construction of fortified settlements.41 This correlates with a pronounced emphasis on fortified settlements within the Roman frontier zone.42 So what exactly had changed? One possible interpretation is to see the Sahara as a more connected place by the early first millennium AD in which the network of incipient oases and routeways provided foci for more mobile groups in the surrounding desert areas. For some time – at least in certain privileged areas – these alliances were channelled towards trade and treaty relations with the Roman Empire. Desert peoples were certainly allowed entry into the empire as seasonal labour and probably also as transhumant groups accessing recognised grazing lands. Trade seems to have diminished in the late Roman period and was perhaps further disrupted after the Vandal conquest. There must have been implications for the oasis communities and 38

39 41

See Mattingly 1983 for an early study that suggested a link with renewed mobility running all along the east to west oasis chain from the Nile to southern Tunisia. Modéran 2003 was critical of this and preferred a much more localised focus in Syrtica. But the geographical range of Laguatan/Lawata does seem to require consideration as part of a broader pattern of increased mobility and migration. Fentress and Wilson 2016. 40 Notably, Rachet 1970. Mattingly 2017b; Sterry and Mattingly 2013; Sterry et al. 2012. 42 Mattingly et al. 2013b.

https://doi.org/10.1017/9781108634311.017 Published online by Cambridge University Press

17 Concluding Discussion

their associated pastoral groups (who had presumably provided much of the personnel and animals of the caravans). It is conceivable that some of the latter turned from traders to raiders and, though the camel had been a part of Saharan caravans for some time, there may have been an increasing emphasis on camels over other pack animals in this period – a circumstance that may also have enhanced the political importance of the pastoral groups who raised them over their sedentary oasis neighbours. The Islamic conquest and the power struggles of the early caliphate and its Islamic successor states in North Africa further complicated the picture. In particular, religious nonconformists played a significant role in many revolts against the Islamic powers in control of Ifriqiya. Refugees from these conflicts are attested at many Saharan oases, such as the Ibadites at Zuwila or the Kharijites/Rustamids at Wargla.43 The role of such people in the development of Islamic-era trade across the Sahara was highly significant.44 Much more work is needed on the Saharan cultural comparisons in this period. The later history of the Sahara contains many attested long-range movements of people. Among others, the forced extraction of Sub-Saharan people via the Trans-Saharan slaving routes45 and the east to west migration of the Banu Hillal and Banu Sulayman.46

The Early Modern and Contemporary Eras Among more recent examples of migration in the Sahara, we can list the continued slave trade until the late nineteenth century, the west to east movement of Moroccan murabitin into Fazzan,47 the presence of Turkish officials and janissaries in the desert,48 and routine patterns of smuggling and trading across ill-defined borderlands,49 and most recently the economic migrants risking the hazardous Saharan crossing for economic opportunities in Europe. Migration remains a key leitmotif of the Trans-Saharan zone.

Identities Identity is the theme that is perhaps least well developed in the case studies in the volume. This may in part be due to a decreasing confidence among 43 44 46 49

See Haour 2007; 2013; Prevost 2010; Savage 1997. Austen 2010 ; De Moraes Farias 2003; Lydon 2009. Thiry 1995. 47 El-Hesnawi 1990. 48 Lyon 1821. McDougall and Scheele 2012; Scheele 2012.

45

Savage 1997; Wright 1998.

https://doi.org/10.1017/9781108634311.017 Published online by Cambridge University Press

539

540

Martin Sterry et al.

archaeologists in simplistic identity affiliations. The greater recognition of the complexity (indeed plurality) of identities has taken us a long way from the cultural-historical approach to ethnicity and community. Increasingly sophisticated methods of scientific analysis can highlight the complications and inconsistencies of ascribed identity based on material culture and behaviours. Burials and funerary practices offer an important window in many societies on such constructed ideas about identity and how this relates also to biological identity (age, gender, genetics, etc.). Isotope analyses provide a powerful tool with which we can study biological identity and the movement of individuals through their lifetime.50 However, the method is not without limitations at present, in part because of the relative small numbers of data points across the regions. The preservation conditions of the Sahara pose some challenges, notably in relation to the extraction of samples suitable for analysis. Bone collagen degenerates badly in hyper-arid burial conditions and this impacts negatively on aDNA and isotope work. Only two out of 30 collagen samples from Fewet were considered to be reliable and none could be obtained from samples in the Wadi al-Ajal.51 This effectively rules out the use of nitrogen isotopic analyses, but carbon, oxygen and strontium signatures can still be obtained. The studies presented in this volume set an important baseline for research, but to take these analyses further it is clear that a far greater number of samples with a wider distribution is required. The dataset of 308 δ18O values from the Nile Valley collected by Buzon et al. far exceeds the total from the entirety of the rest of North Africa and the Sahara.52 Furthermore, each of the major studies reported in this volume has used a different set of isotopic analyses making direct comparisons impossible at this time.53 A key conclusion of each of the chapters that focus on isotopes is that isotopic analysis alone is inadequate to provide clear interpretation of the funerary data and that an integrated bioarchaeological approach is needed. A further important conclusion of this volume is to advocate that future work should always seek to integrate the isotopic, osteological and funerary data and to build links between the different excavated assemblages. Again, 50

51 52 53

The rapid development of this field and impressive results that have been achieved in Africa and elsewhere are well demonstrated in Power et al., Chapter 4; Ricci et al., Chapter 5; and Buzon et al., Chapter 7, this volume. See Ricci et al., Chapter 5, and Power et al., Chapter 4, this volume. Buzon et al., Chapter 7, this volume. Compare Power et al., Chapter 4 (δ13C, δ18O); Ricci et al., Chapter 5, this volume (87/86Sr); see also Keenleyside et al. 2009 (δ13C, δ15N).

https://doi.org/10.1017/9781108634311.017 Published online by Cambridge University Press

17 Concluding Discussion

the studies in this volume demonstrate the potential, but also highlight the constraints imposed by the still small datasets available. Given the current lack of access to the majority of the Sahara and the difficulty of obtaining new funerary material, this is a suitable juncture at which to consider the relationship between these analyses and the interpretative framework in which they are embedded. It seems clear from a range of factors that there were people and groups in the Sahara who were mobile at varying spatial and temporal scales. Edwards in his paper discusses the complicated nature of ethnicity in the Upper Nile. We suggest several directions that can be pursued in future studies of biological identity from the historic and protohistoric periods. Slavery, captives and other forms of forced movement represent a problematic realm of study in archaeology, but one that has remained important in the wider literature as well as in the Sahara.54 The problems are readily encountered in thinking about how to identify a slave, as bar an inscription of some kind it is almost impossible to be certain of the free/ enslaved status of a buried individual. However, most ancient societies have been shown to make use of captives in some form, and in the historic period the Sahara has been considered to be one of the major avenues of slave trade.55 It should be possible to use a combination of biological indicators to identify individuals who had moved long distances and who lived the remainder of their lives under greater levels of stress than the rest of the community where they died. If we were able to correlate certain types of funerary contexts with people displaying such low status immigrant markers, we might be able to postulate that these people were captives or slaves. Given that our confidence is likely to be greater in historic periods, it may help to start with examination of medieval and early modern cemeteries and use these as comparanda, against which to consider the evidence from earlier periods. However, we should also recognise that we may need to focus our efforts in those locations where slaves were trafficked to, rather than the places that they were traded through. This is not to downplay the human cost of trafficking, but slaves are more likely to be identifiable in the funerary record if they came to form a distinctive minority within the local community. An interpretive challenge of the Garamantes case study is that although we suspect the presence of significant numbers of slaves in their society, they do not immediately stand out in the funerary record. That may suggest a relatively rapid assimilation 54 55

Savage 1992. On captives: Cameron 2016; on the historic Saharan slave trade: Wright 1998.

https://doi.org/10.1017/9781108634311.017 Published online by Cambridge University Press

541

542

Martin Sterry et al.

of such people into Garamantian society – something that will have been much easier if the Garamantes were from the start a diverse population. One possibility for identifying slave (or low-status) burials is the category of shaft burials (Type 2, discussed in Chapter 2 above). These were a distinctive development of the late Proto-Urban and Early Classic Garamantian period, but had a long life thereafter. The richness of the Garamantian heartlands in general meant that these burials contain comparatively rich grave goods when compared to the Saharan norm, but they are among the least well furnished of Classic Garamantian date and the form is far from the ostentation of the more monumental elite burials. The inclusion of some Roman imports in these burials would suggest in any case that these evidently low-status members of Garamantian society were nonetheless well integrated in a society that placed value on Mediterranean goods in funerary rituals. Trading communities represent perhaps the flip side of slavery. In the historic period, there were a number of famous trading communities across the Sahara such as the Ghadaminsi (from the oasis of Ghadamis), Saharan Jews and the Kharajite networks of the early medieval period.56 All of these were known for the construction of enclaves of traders and agents in cities to the north and south of the Sahara. They have also been implicated in a number of dual centres, such as at Gao, where the urban community appears divided in two. These seem ideal locations at which we may look to distinguish between a local community who were mobile only within their region and a trading community who were characterised by Trans-Saharan mobility.

Beyond Burials, Migration and Identity This volume has demonstrated the utility of examining issues like identity and migration through the funerary record of societies across the Trans-Saharan zone. However, many questions remain unanswerable on the basis of presently available evidence for past societies. These are highlighted as priorities for future research. The extremely poor preservation of aDNA in Saharan skeletal material is a major obstacle to gaining a full understanding of the ethnic/genetic complexities of Saharan societies. Isotopes and physiological studies can partly fill this gap, but the available data remain small. 56

Haarman 1998; Boum 2011.

https://doi.org/10.1017/9781108634311.017 Published online by Cambridge University Press

17 Concluding Discussion

While we have evidence that attests to migration and different sorts of mobility and to changes in the presentation of personal or group identities, the linkage between these is harder to demonstrate without further work. How did migration change Trans-Saharan communities? Ideally, we need to evaluate the relative contributions of migrants of different sorts and of how the communities whose lifestyle was by definition more mobile contributed to the fashioning and refashioning of Saharan identities as a result of repeated contacts and assimilations. What was life like for people who had migrated, especially for second- and third-generation migrants? How quickly did such individuals become absorbed in the wider community they found themselves part of? These are questions that are difficult to answer categorically for the remote past, but we can see the evidence of these processes in more contemporary accounts of Saharan communities.57 These may provide further clues as to how we should approach the archaeological evidence. If the norm in Saharan societies has been for progressive mixing, through incorporation of new elements, we also need to take account of the exceptional groups like the Tubu/Teda who proved more resilient to such changes and whose language and social customs remained very distinctive down to modern times. Given that they have been significant actors in Trans-Saharan trade, their social cohesion and resilience is intriguing. In the end, there is no neat and predictable correlation between the three processes we have been exploring in this volume, but there are plenty of hints that these were often interconnected. While more detailed evidence at some future point may bring patterns into clearer focus, there do seem to have been important material consequences of the mobility/migration of the historic-era Saharan populations and these are clearly marked in funerary rites and traditions.

References Austen, R. A. 2010. Trans-Saharan Africa in World History. Oxford: Oxford University Press. Barker, G. W., Gilbertson, D., Jones, B. and Mattingly, D. J. (eds). 1996. Farming the Desert. The UNESCO Libyan Valleys Archaeological Survey, vol. 1 Synthesis (Senior Editor G. Barker), Paris: UNESCO. 57

As discussed in nineteenth-century accounts of European travellers, like Barth 1857; Nachtigal 1974; Rohlfs 2001; and so on, or more recent ethnographers, such as Nicolaisen and Nicolaisen 1997.

https://doi.org/10.1017/9781108634311.017 Published online by Cambridge University Press

543

544

Martin Sterry et al.

Barth, H. 1857. Travels and Discoveries in North and Central Africa. (Reprint 1965), London. Bernus, E. 2006. Le dromadaire, le cheval et l’ane. Sages et symbols chez ls Tuareg. In M. H. Fantar (ed.), Le Sahara et l’homme. Un savoir pour un savoir-faire. Tunis: Univiversitè de Tunis, 45–54. Bewley, R., Wilson, A. I., Kennedy, D., Mattingly, D. J., Banks, R., Bishop, M., Bradbury, J., Cunliffe, E., Fradley, M., Jennings, R., Mason, R., Rayne, L., Sterry, M., Sheldrick, N., and Zerbini, A. 2016. Endangered Archaeology in the Middle East and North Africa: Introducing the EAMENA Project. In S. Campana, R. Scopigno, G. Carpentiero, and M. Cirillo (eds), CAA2015. Keep the Revolution Going: Proceedings of the 43rd Annual Conference on Computer Applications and Quantitative Methods in Archaeology, Oxford: Archaeopress, 919–32. Berkani, H., Zazzo, A. and Paris, F. 2015. Les tumulus à couloir et enclos de la Tassili du Fadnoun, Tassili Azger (Algérie) : Premières datations par la méthode du radiocarbone. Journal of African Archaeology 13.1: 59–70. Boozer, A. 2013. Frontiers and borderlands in imperial perspectives: exploring Rome’s Egyptian frontier. American Journal of Archaeology 117.2: 275–92. Boum, A. 2011. Saharan Jewry: history, memory and imagined community. Journal of North African Studies 16.3: 325–41. Bradbury, J. and Philip, G. 2011. The world beyond Tells: pre-classical activity in the basalt landscape of the Homs region, Syria. In Pierres levées, stèles anthropomorpheset dolmens/Standing stones, anthropomorphic stelae and dolmens. Oxford: Archaeopress, 169–80. Bradbury, J., Davies, D., Jay, M., Philip, G., Roberts, C. and Scarre, C. 2016. Making the dead visible: problems and solutions for ‘big’ picture approaches to the past, and dealing with large ‘mortuary’ datasets. Journal of Archaeological Method and Theory 23.2: 561–91. Cameron, C. M. 2016. Captives: How Stolen People Changed the World. Lincoln: University of Nebraska Press. Camps, G. 1961. Aux origines de la Berbérie. Monuments et rites funéraires protohistoriques. Paris: AMG. Castelli, R., Cremaschi, M., Gatto, M., Liverani, M. and Mori, L. 2005. A preliminary report of excavations in Fewet (Libyan Sahara). Journal of African Archaeology 3: 69–102, Cuénod, A., Duckworth, C. and Mattingly, D. J. Forthcoming. Mobile Technologies in the Ancient Sahara and Beyond. Trans-Saharan Archaeology Volume IV, Series editor D. J. Mattingly. Cambridge: Cambridge University Press and The Society for Libyan Studies. de Moraes Farias, P. 2003. Arabic Medieval Inscriptions from the Republic of Mali: Epigraphy, Chronicles and Songhay-Tuareg History. Fontes Historiae Africanae, new series. Oxford: Oxford University Press.

https://doi.org/10.1017/9781108634311.017 Published online by Cambridge University Press

17 Concluding Discussion

di Lernia, S. and Manzi, G. (eds). 2002. Sands, Stones and Bones. The Archaeology of Death in the Wadi Tanezzuft Valley (5000–2000 BP). AZA 3. Firenze: All’Insegna del Giglio. di Lernia, S. and Merighi, F. 2006. Transitions in the later prehistory of the Libyan Sahara as seen from the Acacus Mountains. In D. J. Mattingly, S. McLaren, E. Savage, Y. al-Fasatwi and K. Gadgood (eds), The Libyan Desert: Natural Resources and Cultural Heritage, London: Society for Libyan Studies, 111–21. di Lernia, S., and Tafuri, M. A. 2013. Persistent deathplaces and mobile landmarks. The Holocene mortuary and isotopic record from Wadi Takarkori (SW Libya). Journal of Anthropological Archaeology 32: 1–15. el-Hesnawi, H. W. 1990. Fazzan under the Rule of the Awlad Muhammad. A Study in Political, Economic, Social and Intellectual History. Sabhā. Fentress, E. and Wilson, A. 2016. The Saharan Berber diaspora and the southern frontiers of Byzantine North Africa. In S. Stevens and J. P. Conant (eds), North Africa under Byzantium and Early Islam, Washington, D.C.: Dumbarton Oaks Byzantine Symposia and Colloquia 7, 41–63. Gatto, M. C. 2006. The local pottery. In Liverani (ed.) 2006, 201–40. Gatto, M. C. 2010. The Garamantes of the Fazzan: imported pottery and local production. Bollettino di Archeologia on line I, Volume speciale A/A10/3: 30–38. www.archeologia.beniculturali.it. Gatto, M. C. 2013. Ceramics from Fewet. In Mori (ed.) 2013, 79–92. Gatto, M. C. Forthcoming. Garamantian ceramic technology: Change and exchange in a Trans-Saharan perspective. In Cuénod et al. (eds) Forthcoming. Gauthier, Y. 2015. Pre-Islamic dry-stone monuments of the central and western Sahara. In C. L. N. Ruggles (ed.), Handbook of Archaeoastronomy and Ethnoastronomy. Heidelberg: Springer-Verlag GmbH, 1059–77. Gauthier, Y. and Gauthier, C. 2011. Des chars et des Tifinagh: étude aréale et corrélations. Les Cahiers de l’AARS 15: 91–118. Haarmann, U. 1998. The dead ostrich. Life and trade in Ghadames (Libya) in the nineteenth century. Die Welt des Islams ns 38.1: 9–94. Haour, A. 2007. Rulers, Warriors, Traders, Clerics. The Central Sahel and the North Sea, 800–1500. Oxford: Oxford University Press. Haour, A. 2013. Outsiders and Strangers: an Archaeology of Liminality in West Africa. Oxford: Oxford University Press. Jackson, R. B. 2002. At Empire’s Edge. Exploring Rome’s Egyptian Frontier. New Haven: Yale University Press. Kallala, J., Sanmartí, J. (directors) and Belarte, M. C. (ed.). 2018. Althiburos III. La nécropole protohistorique de Althiburos (Mmassif du Ksour). Taragona: Institut Català d’Arqueologia Clàssica. Keenleyside, A., Schwarcz, H., Stirling, L. and Ben Lazreg, N. 2009. Stable isotopic evidence for diet in a Roman and late Roman population from Leptiminus, Tunisia. Journal of Archaeological Science 36: 51–63.

https://doi.org/10.1017/9781108634311.017 Published online by Cambridge University Press

545

546

Martin Sterry et al.

Liverani, M. (ed.). 2006. Aghram Nadharif. The Barkat oasis. (Sha’abiya of Ghat, Libyan Sahara) in Garamantian times. AZA 5. Firenze: All’Insegna del Giglio. Lutz, R. and Lutz, G. 1995. The Secret of the Desert: The Rock Art of Messak Sattafet and Messak Mellet, Libya. Innsbruck: Umversitatbuchhandlung, Golf Verlag. Lydon, G. 2009. On Trans-Saharan Trails: Islamic Law, Trade Networks and Crosscultural Exchange in Nineteenth-century Western Africa. Cambridge: Cambridge University Press. Lyon, G. F. 1821. A Narrative of Travels in Northern Africa in the Years 1818–1819 and 1820. London. Löwenborg, D. 2010. Excavating the Digital Landscape. GIS Analyses of Social Relations in Central Sweden in the 1st Millennium AD. Uppsala: Arkeologiska Institutionen. Mattingly, D. J. 1983. The Laguatan: a Libyan tribal confederation in the late Roman Empire. Libyan Studies 14: 96–108. Mattingly, D. J. 1995. Tripolitania. London: Batsford. Mattingly, D. J. (ed.). 2003. The Archaeology of Fazzan. Volume 1, Synthesis, London: Society for Libyan Studies, Department of Antiquities. Mattingly, D. J. (ed.). 2010. The Archaeology of Fazzan. Volume 3, Excavations Carried Out by C. M. Daniels, London: Society for Libyan Studies, Department of Antiquities. Mattingly, D. J. (ed.). 2013. The Archaeology of Fazzan. Volume 4, Survey and Excavations at Old Jarma (Ancient Garama) Carried Out by C. M. Daniels (1962–69) and the Fazzan Project (1997–2001). London: Society for Libyan Studies, Department of Antiquities. Mattingly, D. J. and Sterry, M. 2013. The first towns in the Central Sahara. Antiquity 85 (336): 503–518. Mattingly, D. J., Rushworth, A., Sterry, M. and Leitch, V. 2013a. The African Frontiers. Frontiers of the Roman Empire series. Edinburgh: Hussar Books. Mattingly, D. J., Sterry, M. and Leitch, V. 2013b. Fortified farms and defended villages of Late Roman and Late Antique Africa. Antiquité Tardive 21: 167–88. Mattingly, D. J., Leitch, V., Duckworth, C. N., Cuénod, A., Sterry, M. and Cole, F. (eds). 2017a. Trade in the Ancient Sahara and Beyond. Trans-Saharan Archaeology Volume I. Series Editor D. J. Mattingly. Cambridge: Cambridge University Press, Society for Libyan Studies. Mattingly, D. J., Bokbot, Y., Sterry, M., Cuénod, A., Fenwick, C., Gatto, M., Ray, N., Rayne, L., Janin, K., Lamb, A., Mugnai, N. and J. Nikolaus, 2017b. Long-term history in a Moroccan oasis zone: the Middle Draa Project 2015. Journal of African Archaeology 15: 141–72. McDougall, J. and Scheele, J. (eds). 2012. Saharan Frontiers: Space and Mobility in Northwest Africa. Bloomington: Indiana University Press. McIntosh, R. 2005. Ancient Middle Niger: Urbanism and the Self-Organizing Landscape. Cambridge: Cambridge University Press.

https://doi.org/10.1017/9781108634311.017 Published online by Cambridge University Press

17 Concluding Discussion

McIntosh, S. 1999. Beyond Chiefdoms: Pathways to Complexity in Africa. Cambridge: Cambridge University Press. Merlo, S., Hackenbeck, S. and Balbo, A. L. 2013. Desert Migrations Project XVIII: The archaeology of the northern Fazzan: a preliminary report. Libyan Studies 44: 141–61. Modéran, Y. 2003. Les Maures et l’Afrique romaine IVe-VIIe siècle. Bibliothèque des Ecoles françaises d’Athènes et de Rome 314. Rome: Ecole française de Rome. Mori, L. (ed.). 2013. Life and Death of a Rural Village in Garamantian Times. Archaeological Investigations in the Oasis of Fewet (Libyan Sahara). AZA 6. Firenze: All’Insegna del Giglio. Nachtigal, G. 1974. Sahara and Sudan, vol I, Tripoli and Fezzan, Tibesti or Tu. Translated from the German by A. G. B. and H. J. Fisher, London. Nicolaisen, J. and Nicolaisen, I. 1997. The Pastoral Touareg, Ecology, Culture and Society. 2 vols. London: Thames and Hudson. Prevost, V. 2010. Les Ibadites. De Djerba à Oman, la troisième voie de l’Islam. Turnhout: Brepols. Rachet, M. 1970. Rome et les Berbères. Un problème militaire d’Auguste à Diocletian. Brussels. Rayne, L., Sheldrick, N. and Nikolaus, J. 2017. Endangered Archaeology in Libya: tracking damage and destruction. Libyan Studies 48: 23–49. Rohlfs, G. 2001. Voyages et explorations au Sahara. Tome II, Tripoli – Rhadamès – Fezzan – Kaouar – Bornou 1865–1867, (translated by J. Debetz). Paris: Karthala. Savage, E. 1997. A Gateway to Hell, a Gateway to Paradise: The North African Response to the Arab Conquest. Princeton: Princeton University Press. Savage, E. (ed.). 1992. The Human Commodity: Perspectives on the Trans-Saharan Slave Trade. London: Frank Cass and Co. Scheele, J. 2012. Smugglers and Saints of the Sahara: regional connectivity in the twentieth century. Cambridge: University Press. Sivilli, S. 2002. A historical background: mortuary archaeology in the Sahara between colonialism and modern research. In S. di Lernia and G. Manzi (eds), Sand, Stones and Bones: The Archaeology of Death in the Wadi Tannezzuft Valley (5000–2000 BP), Firenze: All’Insegna del Giglio, 17–24. Sterry, M. and Mattingly, D. J. 2013. Desert Migrations Project XVII: further AMS dates for historic settlements from Fazzan, South-West Libya. Libyan Studies 44: 127–40. Sterry, M. and Mattingly, D. J. Forthcoming. Urbanization and State Formation in the Ancient Sahara and Beyond. Cambridge: Cambridge University Press, Society for Libyan Studies. Sterry, M., Mattingly, D. J. and Higham, T. 2012. Desert Migrations Project XVI: Radiocarbon dates from the Murzuq region, Southern Libya. Libyan Studies 43: 137–48. Tafuri, M. A., Bentley, A. R., Manzi, G. and di Lernia, S. 2006a. Mobility and kinship in the prehistoric Sahara: Strontium isotope analysis of Holocene

https://doi.org/10.1017/9781108634311.017 Published online by Cambridge University Press

547

548

Martin Sterry et al.

human skeletons from the Acacus Mountains (south western Libya). Journal of Anthropological Archaeology 25: 390–402. Tafuri, M. A., Bentley, A. and di Lernia, S. 2006b. Consuming food, embodying places. Dietary and social resources in the Holocene Acacus Mountains (Libyan Sahara). Proceedings of the18th Biennial Meeting of the Society of Africanist Archaeologists, Calgary. Thiry, J. 1995. Le Sahara Libyen dans l’Afrique du Nord Medievale. Leuven: Orientalia Lovaniensia Analecta 72. Van Albada, A. and van Albada, A. M. 1993. L’art rupestre du Messak Libyen, La montagne des hommes-chiens, Paris: Seuil. Wright, J. 1998. Murzuk and the Saharan slave trade in the 19th century. Libyan Studies 29: 89–96.

https://doi.org/10.1017/9781108634311.017 Published online by Cambridge University Press

Index

Page numbers in italics are figures; with ‘t’ tables; with ‘n’ notes. 3D geometric morphometrics 140–41, 143–45, 144t Abalessa (Algeria) 17, 102–3 Adrar Zerzem (Morocco) 325, 326, 327, 328 Afroasiatic languages 434, 436, 438, 454, 464–65, 498 Aghram Nadharif (Libya) 53, 101, 117, 120, 121 loom weights 124 straw matting 117 agriculture 135 Kissi (Burkino Faso) 375–78 Lake Chad Basin 399, 490, 492t and language 485–87, 486t Middle Nile 197–98 Numidia 506–7 Wadi Tanzzuft 125, 125, 129, 135 Ahaggar Tuareg 497 Aïn Dahlia Kébira (Morocco) 283, 287, 288, 295, 297 Aïn el Bey (Algeria) 250, 266 Aissa Dugjé (Cameroon) 406–7, 410–11, 412 Algeria Abalessa 17, 102–3 Aïn el Bey 250, 266 Beni Messous 250, 261, 266, 267 Bou Chène 250, 261, 266 Djebel Mazela 250, 252n9, 259, 260, 263–64, 266–67, 268, 270 animal bone 270 pottery 273 El Khroub 250, 256, 262 Fedj el Koucha 250, 266 Gastel 250, 253, 266, 271, 272 Jorf Torba 15–16 Roknia 250, 261, 266, 300 Sétif region 250, 253 Sigus 250, 266 Sila 250, 264, 265, 267, 268, 270, 274, 275–76 Tebessa 250, 253

Tiddis 253, 260, 264, 267, 268 children’s burials 265 pottery 270, 271 Tiffech 250, 266 Tin Hanan 17–19, 103, 530 Wadi Gir 15 Allain, Charles 321–24 Althiburos (Tunisia) 14, 249, 250, 276, 506–7, 529 bone in domestic areas 268–69 pottery 271, 272, 273 Amazighs see Berbers amazonite 95–96, 117, 124, 406, 408, 417, 420 ancestor worship 5, 118, 128–29, 330–31, 528 animals in burials 269–70 burials of 7, 364, 406–7, 411, 412, 418 names for 452–53, 482–87, 484–85t, 489–90 see also livestock; pastoralism antennae tombs 61–62, 363, 514 Western Sahara 16, 355–56 Apuleius 509–10 Arabic 200, 432, 433t, 434, 495–96, 502 Hassānāya 431, 432, 433t, 434, 450–51, 456 Shuwa 433–34 archaeology, funerary 3, 4–19, 526, 527 architecture, Wadi Tanzzuft 120–22, 122 Arnaud, R. 450 artefacts 27 Wadi Tanzzuft 123–25, 124 see also grave goods Asbytes 431 Atarantes 108, 125–29, 432, 504–5 Atlantic Europe 362, 363, 366–68 Augustine, St 514 Awjila (Libya) 9, 478, 481, 497 Ayoub, M. S. 12, 54 Azayr/Azer 444, 450–51 Bach, J. S. 438 bags, leather 86–87 Baistrochi, M. 359

https://doi.org/10.1017/9781108634311.018 Published online by Cambridge University Press

549

550

Index

Al-Bakri 447 Banasa (Morocco) 285–86, 303 Bangsgaard, P. 208 banquets, funerary 273, 330, 332 Barkat (Libya) 116, 152–53, 154, 163 bazinas 252n9, 257, 258, 260, 261, 267, 277, 345–49, 346–48t, 350, 351 artefacts 270, 272 children’s burials 265 dating 264, 355 Djebel Mazela Bazina XXII 263–64, 266–67 Gour Mausoleum 276, 290, 291–92t, 293, 294, 333, 513 treatment of the body 268 see also chouchet beads Burkino Faso 386 Fewet (Wadi Tanzzuft) 117, 124 Kissi (Burkino Faso) 385–87, 388, 392, 395 Morocco 297, 300, 327, 328, 333 Nigeria 409 southern Lake Chad basin 406, 409, 411, 414, 415, 417–18, 419, 420 Tin Hanan burial 17 Wadi al-Ajal 69–71, 77, 78, 88–89t, 91, 94, 96, 97 Bechhaus-Gerst, M. 434–35 Beja 213, 215, 432 Bell Beaker burials 366 bells, bronze/copper 318 Beni Messous (Algeria) 250, 261, 266, 267 Beni-Snous, language 466, 467, 470, 471, 473, 476, 478, 481, 497, 517 Berbers burial practices 332 and Christianity 331 identity 25–26 languages 23, 25–26, 27, 431, 434–38, 435, 452–54, 453t, 456, 464–66, 466, 467, 470, 471, 473, 476, 478, 481, 490–93, 497, 525 and economies 482–87 Oasis languages 516–17 proto-Amazigh 480, 483, 484–85t, 486t, 503–8 Proto-Berber 436, 447, 449t, 465, 482–87, 486t, 498–502 proto-Central Amazigh 512–16 proto-East-West Amazigh 508–12 origins 138 Beria (=Zaghawa) 196, 431, 434, 440, 451 Berti 196, 204 Biagetti, S. 446

https://doi.org/10.1017/9781108634311.018 Published online by Cambridge University Press

biological identity, Garamantes 134–56, 137, 144t, 146, 147, 150, 151, 152, 154 blacksmiths 202, 452 Blažek, V. 448, 449t, 496 Blé (Cameroon) 406 ‘Blemmye’ burials 213 bone worked 301 see also skeletal analysis Boniface 514 Boozer, A. 21 Borno (Nigeria) 408–9 Bou Chène (Algeria) 250, 261, 266 Bou Mimoun (Morocco) 290, 291–92t Bouïa (Morocco) 319, 320, 321, 330, 335, 336 ‘bound’ bodies 204, 411 Braber (Morocco) 323–24, 324, 330 Bradley, R. 62 Breunig, P. 401–3 Britain 21 Brocca, Paul 281 Broodbank, C. 506–7 Brooks, N. 449 Brosset, D. 450 Brubaker, Rogers 25 burial landscapes 532–34 burial rite see ritual, funerary Burkino Faso 386, 395 Kissi 375–78, 376, 377 burial practices 379, 380, 381, 382, 383, 384, 385, 387, 388, 389, 390, 395, 531–32 migration 395 modern parallels 390–95, 392, 393, 394 cairns 10–12, 60, 65, 80t, 101 Fazzan 57–60, 64, 65, 69, 70t, 100–1 beads 94 pottery 92 Fewet 100–1 Iwelen 18 types corbelled 57–60, 64, 65, 65, 72t, 78, 80, 98, 100–1 crater 57 crevice 57, 72t kerbed 57, 65, 72t mound Central Sahara, Wadi al-Ajal 64, 65, 71–77, 72t, 80t Numidia 257–60, 258, 259 Western Sahara 16, 345, 349, 355 quadrangular 60, 64–65, 65, 72t

Index

with vertical outer walls 60, 72t see also monuments camels 3, 197–98, 434, 437, 515–16 terms for 483, 484–85t Cameroon Aissa Dugjé 406–7, 410–11, 412 Blé 406 Doulo Igzawa 406 Doulo Kwovré 410, 410 Houlouf 406 Mandara (Cameroon) 410, 412, 412, 413, 416 Camps, Gabriel 12–13, 249–50, 251–52, 257–60, 262–63, 267, 269, 272, 281, 331, 505 Aux origines de la Berbérie. Monuments et rites funéraires protohistoriques 251 Camps-Fabrer, H. 272 Camus, A. 319 Canary Islands 435, 438–39, 455 Caputo, G. 54 caravan routes 112 Carcopino, Jérôme, Le Maroc antique 281 carnelian beads Durbi Takusheyi (Nigeria) 415 Morocco 333 Nigeria 409 southern Lake Chad Basin 406, 408, 409, 411, 414, 415, 417–18, 420 Carthage 250, 250, 282–83 bells 318 burials 253, 260 red ochre 89 cowrie shells 300 glass paste beads 297 Punic 446, 448 see also Djebba; El Hkayma castes 202, 452 cattle 2, 7, 436, 449 Lake Chad Basin 405, 489 caves, burial 253, 268 Çelebi, Evliya 439 cemetery morphology 63–64 Central Amazigh languages 466, 467, 470, 471, 473–76 Central Sahara 10–12 Chabot, J.-B. 511 Chad 19 Mdaga 413–14 Mege 406 see also Lake Chad Basin Chadic languages 465, 487–93, 491t, 492t, 498

chalcedony 297, 298, 300, 386 chapels, funerary 9 Western Sahara/Morocco 16, 316, 320, 323, 325n27, 328–31, 330 chariots 18, 28, 326, 440, 505–6, 510 chefferies 284, 285, 293–94 children’s burials Kissi (Burkino Faso) 383, 384 Numidia 264–65, 276 southern Lake Chad Basin 411 chouchet 254, 255, 268 see also bazinas Christianity 331, 419 ‘Christian’s Tomb (Kbour Roumia) 250, 257, 293, 294 ‘Círculo del Estrecho’ 283 Cirta 250, 256, 537 citadels, fortified, Wadi Tanzzuft 120, 121 Civrac, M. A. 452 Clapperton, Hugh 451 Cleaveland, T. 444 climate 2–3, 23, 530 and linguistics 431 Middle Nile 199, 201, 206 Southern Lake Chad Basin 400, 405–6, 421–22, 490 Western Sahara 343–44, 367–68 see also ‘green Sahara’ concept Coarelli, F. 257 collective burials see multiple burials colonisation 20t, 22, 23 in Morocco 281, 285 Nubia by Egypt 9, 209–10, 212 see also Roman period ‘complex megalithic monuments’ 254–56, 255 complex monuments 346–48t, 354, 355, 356–57 conical tumuli Fewet 112–14, 113 Morocco 319, 325 Connah, Graham 403, 408–9 connectivity xxiv, 2, 21–22, 525, 530–31, 532 and the Western Sahara 358 see also trade conquest 20t, 22, 23, 378, 495–96, 537–38 Islamic 539 and language 470 Roman 437, 537 Songhay of Timbuktu 442 Vandal 538 copper/copper alloy 201, 305, 449t Kissi 31, 386–87, 403, 531–32

https://doi.org/10.1017/9781108634311.018 Published online by Cambridge University Press

551

552

Index

copper/copper alloy (cont.) Lake Chad Basin 409, 414, 415, 418 Morocco 297, 298, 299, 318, 327–28, 332, 333 corbeille monuments 346–48t, 349–52 Corippus 515, 516 covered galleries of Kabylia 256 cowrie shells Kissi (Burkino Faso) 386–87, 388 Morocco 299–300, 305 southern Lake Chad Basin 406, 409, 415 craftworkers, status of 202 craniometry Egyptian and Nubian 240 Fewet 168–72, 169, 170, 171 Wadi al-Ajal 140–41, 144t, 145, 146–48, 147, 149–55, 151, 152, 154 cremation 262–69, 272, 305, 527 crescent-shaped monuments 10, 15–16, 346–48t, 352–53, 359, 360–61t, 362–63 Crews, D. 231 cults ancestor 118, 128–29, 273–74 Astarte 285, 298 Daima sites (Nigeria) 408–9 Daniels, Charles 12, 54, 60, 69, 77, 85, 86, 180, 504 Darfur (Sudan) 196–97, 198 date palm cultivation 198n8, 212 Dawada 451, 455–56 defended sites 505 Lake Chad Basin 403, 404, 420–21 Lower Nubia 225 Morocco 336–37 Wadi Tanzzuft 111n17, 120 demographic profile, Wadi al-Ajar 79–81, 80t Desert Migrations Project (DMP) 12, 54, 55, 62, 532 di Lernia, S. 359, 436, 449 diaspora 31 diet Garamantian 134, 141–42, 148–49, 153–55, 180, 183 Morocco 301, 302 Proto-Berber 485–87, 486t Tombos 239 see also agriculture; livestock diffusion 20–21, 289, 495 discrepant identity 26–27 Djebba (Tunisia) 250, 252, 253, 260, 261, 262, 264, 266, 267 children’s burials 265

https://doi.org/10.1017/9781108634311.018 Published online by Cambridge University Press

grave goods 269, 270, 272 red ochre 272 secondary burials 263, 264, 267–68 Djebel Mazela (Algeria) 250, 252n9, 259, 260, 263–64, 266–67, 268, 270 animal bone 270 pottery 273 Djebila (Morocco) 283, 286–87, 289, 290, 295–97, 296, 299–300, 301–2 djedars/jedars 293, 294, 513–14, 516 dogs 411, 499, 501 dolmens 12–13 Northern Sahara/Maghrib 12–13, 254–56, 255, 268, 281 zenithal access 259, 260, 264, 277 domestic evidence, Wadi Tanzzuft 119–25, 121, 122, 123, 124, 125 domesticated animals see livestock Doulo Igzawa (Cameroon) 406 Doulo Kwovré (Cameroon) 410, 410 Drake, N. A. 453 drum tombs 60, 65, 71–77, 72t, 80t, 92, 101 Abalessa 102–3 beads 94 Egypt/Nile Valley/Eastern Sahara 9 Fewet 101, 113, 115, 119, 127 funerary furniture 98 Iwelen (Niger) 18 Western Sahara 16 drum-shaped monuments 60–61, 62, 64, 65, 72t square-rectangular 60–61, 62 see also bazinas Durbi Takusheyi (Nigeria) 414–15 East-West Amazigh 476–82, 478 Eastern Sahara 7, 8, 10 Egypt 8, 10 isotopic analysis of interaction in Nile Valley 223–41, 224, 229, 234t, 235t, 236 and Libyans 503 El Hkayma (Tunisia) 250, 252, 253, 260, 261, 262, 264–66, 267, 269 grave goods 270, 272 El Khroub (Algeria) 250, 256, 262 El Ksour (Tunisia) 252n9, 259, 260–61, 264, 266, 267, 269, 275 and animal bones 270 funerary banquets 273 funerary cults 274 pottery 272 elephants 302–3 see also ivory

Index

Ellès (Tunisia) 250, 252, 254, 255, 261 ‘emergent-hierarchy’ 207–8 enamel hypoplasia 179–80 enclosure type monuments, Numidia 252–53 Enfida (Tunisia) 250, 259, 266 entanglement 110–11 environment see climate Erfoud (Morocco) 316, 317–18, 318 excarnation 262–63, 269, 293, 527–28 exogamy 20t, 23 Fazzan 10–12, 11, 503–04, 504–6, 533 see also Garamantes; Wadi al-Ajal; Wadi Tanzzuft feasts, funerary see banquets, funerary Fedj el Koucha (Algeria) 250, 266 Fewet (Wadi Tanzzuft, Fazzan) 53, 101–3, 163 burials 60, 80–81, 100–1, 111–19, 113, 136 orientation 101–2 shrouds 101, 116 domestic architecture 120, 122, 123 osteology 162–64, 237 fish/fishing, Morocco 301 foggara irrigation systems 125, 125, 129, 135 forced movement 535, 541–42 see also slavery Fortier, C. 451 fortifications see defended sites Foum Larjam (Morocco) 16, 324, 325, 331–32, 534 Francis, E. D. 126 Free Zone see Western Sahara frontiers 110–11 Fuchs, P. 451 Fulani 378, 391 funerary archaeology 3, 4–19, 526, 527 funerary banquets 273, 330 funerary cults 273–74 furniture funerary 5, 63 Fewet (Wadi Tanzzuft) 101, 102, 117–18, 127 Wadi al-Ajal 56, 97–99, 98 Gabès (Tunisia) 250, 266 Gaetuli 509–10 Gajiganna culture 399, 401–3, 402, 404–5, 407–8 Garamantes xxiii–xxiv, 10–12, 108–11, 125–29, 515–16, 517, 526, 536 burials 9–10, 53–56, 54t, 55, 100–3, 529–30 excavated burials 68–71, 70t

AMS dates/chronological trends in burial practice 71–79, 73–76t demographic profile 79–81, 80t grave furnishings 88–89t, 93, 95, 96, 97, 98, 99–100 orientation and treatment of the body 81–89, 82, 83, 84t tomb survey 64–68, 65, 67, 68 tomb typologies 57–63, 58, 59 identity 28 biological 134–56, 137, 144t, 146, 147, 150, 151, 152, 154 language 446, 447, 508–10 and the southern Lake Chad Basin 418–22 Gastel (Algeria) 250, 253, 266, 271, 272 Gattefossé, J. 336 Gauthier, Y. 362–63 gender and burials 94, 98–99, 115 Middle Nile 204, 207 task-specialisation 136 genetics 22, 497, 498–500, 507–8 geology, Nile Valley 228–29 Gerteiny, A. G. 450, 451 Ghadamis (Libya) 497, 538, 542 language 473, 474–76, 476, 477, 478, 480, 481 Ghat (Libya) 125, 163, 497 language 473, 474, 476, 478, 481, 517 textiles 116 Ghwa Kiva (Nigeria) 406 Giddens, A. 26 globalisation 21–22 theory 24 Gobin, C. 353 gold 17, 71, 88–89t, 91–92, 297–99, 298, 305, 415, 421 goulets 346–48t, 353, 354, 356–57, 363 Gour Mausoleum (Morocco) 276, 290, 291–92t, 293, 294, 333, 513 grave goods Fazzan 12 Fewet (Wadi Tanzzuft) 102–3, 116, 117, 127 Iwelen (Niger) 18 Kissi (Burkino Faso) 385–90, 387, 388, 389, 390, 392, 393, 394–95 Morocco 291–92t, 295, 296, 298, 304, 305, 327, 328, 333, 337 bells 318 Numidia 270–72 southern Lake Chad Basin 409, 414, 417–18 Southern Sahara, tomb of Tin Hinan 17–18

https://doi.org/10.1017/9781108634311.018 Published online by Cambridge University Press

553

554

Index

grave goods (cont.) Wadi al-Ajal 69–71, 77–78, 88–89t, 92, 93, 95, 97, 98, 99–100 Winchester (Britain) 21 ‘green Sahara’ concept 2–3, 400, 405–6, 431, 432, 448–50, 450, 454, 501 grinding equipment 123–24 Gronenborn, Detlef 409 Gsell, Stéphane 318, 337 Guanche 435, 438–39, 455 Haaland, G. 198 Hachid, Malika 510 Hackendorf, R. 505 Haddad 452 haouanet 14, 253, 254, 260 Haour, A. 419, 421 Harvati, K. 140 Hassānāya Arabic 431, 432, 433t, 434, 450–51, 456 headrests 86, 117 Henchir Mided (Tunisia) 250, 252, 261, 264, 267, 270, 272 Hennessy, R. J. 141 Herodotus, Histories 108, 126, 330–31, 431, 432, 440, 504–5 Holl, Augustin 413, 414 Holy, D. 204 horses 483, 484–85t, 503, 504–5, 508, 510 southern Lake Chad Basin 406–7, 411, 412, 418 Houlouf (Cameroon) 406 Human Mobility and Identity workgroup 137–38 hunter-gatherers 2, 470 hunting 448–49, 450–51, 455, 491 hypogea 9, 253, 260, 265, 267, 275–76 Iberian Peninsula 281, 283, 303, 305–6 Ibn Khaldun 446 Icard, F. 267 Ichoukkane 505 identity 4, 24–28, 539–42 Imraguen 451 In-Aghelachem 111–14, 112, 530 Royal Tumulus 111, 112, 118, 119 infant burials see children’s burials inhumation see treatment of the body in burial inscriptions 444–46, 445, 447, 455, 456 iron tools 124 working 201–2 Burkino Faso 375

https://doi.org/10.1017/9781108634311.018 Published online by Cambridge University Press

Morocco 297–99, 298, 305 southern Lake Chad Basin 403 see also grave goods; jewellery irrigation systems Middle Nile 198n8 Wadi Tanzzuft 125, 125, 129 Islam 333, 395, 419, 456, 496 Islamic remains 61, 72t, 336 isotopic analysis 143, 540–41 Egypt and Nubia 223–41, 224, 229, 234t, 235t, 236 Fewet 180–83, 182 and southern Lake Chad Basin 418 Wadi al-Ajal 141–42, 144t, 145–48, 146, 147, 154–55 ivory 302–3, 366 Iwelen (Niger) 18–19 Jabal Bouïa (Morocco) see Bouïa (Morocco) Jacques-Meunié, D. 316, 321–24 jar burials 411, 413, 414, 416, 419 Jarma see Wadi al-Ajal jedars see djedars/jedars jewellery Kissi (Burkino Faso) 385–87, 387, 388, 392–93 Morocco 299, 300, 305, 332 southern Lake Chad Basin 409 see also beads Jewish communities 331–32 Jodin, A. 366 Jorf Torba (Algeria) 15–16 Kabyle 476, 478–79, 480, 481, 497, 508 Kabylie (Algeria) 256, 512 Kalaat es-Snam (Tunisia) 250, 260 Kallala, N. 252 Kanuri-Kanembu 433, 439 Kapteijns, L. 212 Kbour Roumia (the ‘Christian’s Tomb) (Morocco) 250, 257, 293, 294 Kerma (Middle Nile) 198–99, 205–6, 208–9, 225 isotopic analysis 233, 234t, 235t language 200 keyhole tombs 10, 356, 363 Kienlin, T. L. 199–200 kingdoms 200, 505 Middle Nile 200, 207, 208 Kissi (Burkino Faso) 375–78, 376, 377 burial practices 379, 380, 381, 382, 383, 384, 385, 387, 388, 389, 390, 395, 531–32 migration 395 modern parallels 390–95, 392, 393, 394

Index

Kitchen, Andrew 465 Knudson, K. 237 Kopytoff, I. 214 Kordofan (Sudan) 196–97 Kossmann, M. 437, 438, 448, 472n8 Laguatan 538 Lake Chad Basin (southern) 399–400, 402, 531–32 archaeological/historical background 400–4 connections/separations 404–7, 418–22 see also Chadic languages Lalla Ghano (Morocco) 290, 294, 297 Lalla Mimouna (Morocco) 290, 291–92t lamps, vesicular basalt 113, 118 landscapes funerary 5–6, 532–34 and siting of monuments 363, 364, 365 Atlantic Europe 366–67 languages and linguistics 431–32, 454–56 former ethnolinguistic diversity 448–52, 450 inscriptional languages 444–46, 445, 447, 455, 456 language change 495–96, 497–98 Middle Nile 200–1 and migration 22, 23 Nilo-Saharan (Eastern Sudanic) 200 and the present-day 432–44, 433t, 434, 435, 440, 441, 443, 443t substrate words in Sahara languages 452, 453t, 454t see also Berbers, languages Latin 436, 437–38, 445, 445–46, 487, 495, 511–12 Le Quellec, J. L. 452 leather bags 86–87 Kissi (Burkino Faso) 389 pillows, Fewet (Wadi Tanzzuft) 117 shrouds Fewet (Wadi Tanzzuft) 101, 116, 117 Iwelen (Niger) 18 tomb of Tin Hinan 17 Wadi al-Ajal 85–87 Lebeuf, Jean-Paul 412–13 legumes, Morocco 302 ‘Leiterband Complex’ 205 Leptiminus 14 Leschi, L. 152 Lhote, H. 18 Libya Aghram Nadharif (Barkat) 53, 101, 117, 120, 121, 163

loom weights 124 straw matting 117 Awjila 9, 478, 481, 497 Barkat 116, 152–53, 154, 163 Ghadamis 497, 538, 542 language 473, 474–76, 476, 477, 478, 480, 481 Murzuq 152–53, 154 Murzuq region (Libya) 101n67, 125, 138 Saniat bin Huwaydi 54, 61, 69, 70t Watwat 54, 65, 77–78, 87–89, 149–55 osteology analysis 149, 150, 153 Zinkekra 54, 60, 69, 70t, 77, 504 osteology analysis 149, 150, 153, 154 see also Fazzan; Garamantes Libyan Valleys Survey 6n24, 14–15 Libyans, proto-Amazigh 503 Libyco-Berber language 316, 431, 444–46, 445, 447, 455 lip plugs 69, 91, 155, 419 Liverani, M. 108, 517 livestock 436, 449, 482, 484–85t, 498, 504 Lake Chad Basin 403, 405, 489–90 Middle Nile 199 in burials 208 Lixus (Morocco) 281, 285–86, 302, 303–4, 507 loom weights, Aghram Nadharif 124 Louata 515, 516 Maban 454t, 488 MacMichael, H. A. 440 Maghreb 12–15, 507–8 see also Algeria; Mauretania; Morocco; Numidia; Western Sahara Magnavita, Carlos 408 Makthar 250, 252, 256, 266, 274, 275 malaria 232 Malášková, Z. 448, 449t Mandara (Cameroon) 410, 412, 412, 413, 416 Manzi, G. 359 Margat, J. 319 Marion, J. 336 Masalit 197 Matmata 469, 470–72, 471, 473, 476, 477, 478, 481, 516–17 matting in burials Fewet (Wadi Tanzzuft) 117 Kissi (Burkino Faso) 389–90 Wadi al-Ajal burials 85 Mauretania 6, 12–13, 15, 26, 299, 433t, 449, 511, 529

https://doi.org/10.1017/9781108634311.018 Published online by Cambridge University Press

555

556

Index

Mauretania (cont.) and jedars 513–14 Siga 250, 256, 286 Zanega 435, 478–80, 502 mausolea 61, 67 Mdaga (Chad) 413–14 Mdidi (Tunisia) 14 Medracen (Morocco) 250, 257, 294 megalithic structures see dolmens Mege (Chad) 406 memorialisation 5–6, 30–31 see also furniture; monuments; ritual Merlo, S. 533–34 Meroë/Meroitic kingdom 9, 209–12, 226 language 200, 201 metals/metalworking Lake Chad Basin 403 Middle Nile 201 Morocco 281–82, 297–99, 298, 333 see also gold; silver microliths 449–50 Middle Nile 195–97, 196, 215–16, 530 agriculture 197–98 burials 9–10, 203–15, 226–27 craft production 201–2 ethnicity 198 language 200–1 pastoralism 197–98 social status 202–3, 206–10 trade 199–200 migration 3–4, 19–24, 20t, 134–35, 395, 534–35, 543 Neolithic (Pastoral) phase 535 first oasis settlements of the first millennium BC 535–36 Roman period 536–37 late antique and early medieval periods 537–39 early modern and contemporary eras 539 Milburn, M. 362 ‘milking stools’, Fewet (Wadi Tanzzuft) 117–18, 127 millet farming Kissi (Burkino Faso) 375–78, 391, 395 Lake Chad Basin 401 mobility 20t, 22–24 Mogador (Morocco) 285, 298–99, 303, 318 Monod, T. 353 Monteil, C. 444 monuments

https://doi.org/10.1017/9781108634311.018 Published online by Cambridge University Press

funerary 7, 512–14, 527, 528–29 Central Sahara, Fazzan 57–63, 58, 59 Numidia 251–60, 253, 254, 255, 256, 257, 258, 259 Western Sahara 15–16 see also antennae tombs; bazinas; cairns; conical tumuli; corbeille; crescentshaped; dolmens; drum tombs; enclosure type; keyhole tombs; platform tumuli; pyramid tombs; pyramids; ridge; shaft burials; siloshaped tombs; stepped tombs ‘Moors’ 514–15 Moreno García, J. C. 431–32 Morocco north-west 281–85, 282, 284, 286–87, 305–6 cemeteries burial practices 287, 288, 289, 530–31 chronology 294 grave goods 295, 296, 298, 304, 305 tomb typologies 290, 293, 294 tumuli 289–93, 290, 291–92t settlements 285–86 pre-Saharan 315, 317, 332–38 funerary zones/monuments 317–28, 324, 325, 333 Adrar Zerzem 325, 326, 327, 328 Bouïa 319, 320, 321, 330, 335 Braber 323–24, 324, 330 Erfoud 316, 317–18, 318 Foum Larjam 16, 324, 325, 331–32 Taouz 321–23, 322, 323, 330 settlement 335, 335–36 tumuli 315–16 with cruciform chapels 328–31, 330 with windows 331–32 Tamazight 466, 468, 469, 470, 472, 478, 479, 483, 517 morphometric analyses, Garamantes 143–44, 144t mound type monuments see cairns, mound mud-brick architecture 120–23, 126, 127 multilevel funerary towers 256 multiple burials Fewet (Wadi Tanzzuft) 115 Kissi (Burkino Faso) 384 Middle Nile 210, 226–27 Morocco 334–35 Numidia 267–68, 277 Wadi al-Ajal 79–81 mummification/mummies 9, 232, 438, 529 natural 77, 87 Murzuq (Libya) 152–53, 154

Index

Murzuq region (Libya) 101n67, 125, 138 Mzab 467, 467, 470, 471, 474, 476, 478, 481, 497, 517 Mzora (Morocco) 290, 291–92t, 293, 294 Nafusa 472n8, 473, 476, 476, 477, 478, 481, 497 Napatan dynasty 223, 226–27, 230–31, 233, 235–37, 240–41 Nasamones 330–31, 537n37 Nemadi 450, 451, 455–56 Neumann, K. 401–3 Nicolaisen, I. 452 Niger, Iwelen 18–19 Niger-Congo languages 444 Nigeria Borno 408–9 Daima sites 408–9 Durbi Takusheyi 414–15 Ghwa Kiva 406 Thuliva Kwacha 411–12, 413 Zilum 402, 406, 408, 415, 420–21 Nile Valley 6–7, 8, 10 drum tombs 9 isotopic analysis 223–41, 224, 229, 234t, 235t, 236 strontium isotope analysis 228–31, 229 Nilo-Saharan (Eastern Sudanic), languages 200, 440–44, 453, 454t nomadism 23, 24, 337, 516 Non-Tuareg Berber languages 433 Nubia see Middle Nile; Tombos Numidia 249–50, 250, 275–77, 506–7 funerary ritual 261–74, 529 language 455 research history 251–52 structure of cemeteries 260–61 types of funerary monuments 252–60, 253, 254, 255, 256, 257, 258, 259 oasis languages 472n8, 475, 516–17 Beni-Snous 466, 467, 470, 471, 473, 476, 478, 481, 497, 517 Ghat 473, 474, 476, 478, 481, 517 Moroccan Tamazight 466, 468, 469, 470, 472, 478, 479, 483, 517 Mzab 467, 467, 470, 471, 474, 476, 478, 481, 497, 517 Rif 467, 470, 471, 473, 476, 478, 479, 481, 517 Sanhaja 466, 467, 470, 471, 472–73, 473, 476, 478, 479–80, 481, 497, 517 Shawiya 467, 467, 470, 471, 473, 476, 478, 481, 497, 515, 516–17

Wargla 467, 467, 470, 471, 471, 473, 474, 476, 478, 481, 497, 517 ochre 13, 18, 19, 528, 529 Fewet (Wadi Tanzzuft) 115–16 Morocco 287 Numidia 272, 276 tomb of Tin Hinan 17 Wadi al-Ajal 69, 84t, 87–89 offering tables 63, 320 ‘Old North African’ (ONA) 455–56 olive cultivation 302, 449t, 506 orientation burial Fewet (Wadi Tanzzuft) 101–2, 115 Iwelen (Niger) 18 Kissi (Burkino Faso) 379, 382, 383, 384, 390 Middle Nile 203–4 Morocco 287, 288 southern Lake Chad Basin 409 Wadi al-Ajal 81, 82, 83 see also treatment of the body of monuments 362–64, 363, 365 Osborne, R. 27 osteology see skeletal analysis ostrich eggs beads made from southern Lake Chad Basin 409 Wadi al-Ajal 77, 78 Morocco 303, 304, 327–28 Oujda (Morocco) 284, 286, 293, 303, 306, 336 Ounia 440, 451 oxygen isotope analysis 231–37, 234t, 235t, 236 painted decoration in tombs 16 ‘Pan-Grave’ burials 213 Parenti, R. 152 Paris, F. 17, 356, 362 Park, D.P. 444 pastoralism 20t, 23, 24, 526, 535–36 and the Berber language 436–37, 456, 500–2 Lake Chad Basin 399, 405 Middle Nile 197–98, 204–5, 211–13 and Proto-Berber 499–502 see also livestock pathologies, dental, Fewet 179–80 Pauphilet, P. 274 Phoenicians 14, 276, 281–83, 302, 446, 507, 531 cremation 262 and Morocco 315 Punic 446–48, 449t pillows, leather 117

https://doi.org/10.1017/9781108634311.018 Published online by Cambridge University Press

557

558

Index

pits, burial 253 placement/position platform tumuli 345, 346–48t, 349, 355 Plautus, Poenlus 446–47 Pliny 446, 510 POLISARIO 341–42 porotic hyperostosis 177–79 pottery 126, 127, 526 Middle Nile 213 Morocco 296–97 Numidia 270–72 southern Lake Chad Basin 407, 417 Wadi Al-Ajal 65–66, 72, 88–89t, 90–91, 92, 93, 99, 100 Wadi Tanzzuft 102–3, 117, 118, 127 preservation, bone 143–44 primary deposition 262, 263–65, 268, 334 Fewet 115 Procopius 515, 516 Proto-Amazigh 480, 483, 484–85t, 486t, 503–7 genetics 507–8 Proto-Berber 436, 447, 449t, 465, 482–87, 486t, 498–502 proto-Central Amazigh 512–16 proto-Chadic language/society 465, 487–93, 491t, 492t proto-Chado-Berber 465, 492–93 proto-East-West Amazigh 508–12 Punic 446–48, 449t, 495, 511 putrefaction of bodies 86 pyramid tombs 61, 67, 69, 70t, 72t, 80t, 92 pyramids, Egyptian 8–10

Renfrew, Colin 495 reproductive behaviour, Fazzan 136 Reygasse, M. 17, 18, 102–3 ridge monuments 355, 356, 357 Rif 304, 497, 512, 515 language 467, 470, 471, 473, 476, 478, 479, 481, 517 ritual 5 funerary 527–28 Fewet (Wadi Tanzzuft) 117–18, 127 Numidia 261–74 ‘Royal Tumulus’ (In-Aghelachem) 118 sleeping with the dead 329–31 Wadi al-Ajal 127 rock art 127, 456, 525–26 animals in 453 Iwelen (Niger) 18 Morocco 16, 282, 316 Roknia (Algeria) 250, 261, 266, 300 Roman period 26–27, 456, 509–12, 536–37, 538–39 burial practices 6n26 imports 114–15, 128–29, 531 and Latin 436, 437–38, 445, 445–46, 487, 495 Leptiminus 14 Roset, J.-P. 400 Royal Cemeteries 54, 61, 102 beads 95 osteology analysis 149, 150 ‘Royal Tumulus’ (In-Aghelachem) 111, 112, 118, 119 Ruhlmann, Armand 316, 317–18

quartz 357, 359–62, 360–61t, 364, 366–67 querns 18, 123–24, 128

sacrifices 119, 274, 320, 328, 330, 332, 334 human 119, 206, 290–93 Saguia al-Hamra 342, 344, 367 Saharan languages 431, 439, 440 Sala (Morocco) 285, 297 Salih, A. 505 Sanhaja 466, 467, 470, 471, 472–73, 473, 476, 478, 479–80, 481, 497, 517 Saniat bin Huwaydi (Libya) 54, 61, 69, 70t ‘Sao’ sites 412–14, 416 Saxe, A. 6 Schuhmacher, T. X. 303, 366 secondary deposition 528 Morocco 290–93, 334 Numidia 262–64, 266–67, 268 southern Lake Chad Basin 408 sedentarisation 198, 401–3, 536 see also agriculture; pastoralism Sen, A. 25 Sétif region (Algeria) 250, 253

radiocarbon/AMS dating 62, 527 Central Sahara 501n32 Chad 413–14 Kissi 380, 381 Libya 436–37 Garamantian burials 69, 71–79, 73–76t Morocco 16, 325, 328, 333, 501–2, 513–14 Tunisia 273 Western Sahara 341, 344, 355–56 Rakob, F. 257 red ochre see ochre refugees 20t, 22, 367, 421, 537n37 Reitsema, L. 231 religion 419–20 Christianity 331, 419 Islam 333, 395, 419, 456, 496 Jewish communities 331–32

https://doi.org/10.1017/9781108634311.018 Published online by Cambridge University Press

Index

shaft burials 60, 66, 67, 68, 69, 72t, 77–78, 80t, 84t, 84–85, 501 grave goods 92, 94 Shawiya 467, 467, 470, 471, 473, 476, 478, 481, 497, 515, 516–17 shells 299–300 Shenwa 467, 467–68, 471, 478, 481, 497 Shilh 469, 470, 473, 476, 478, 481, 497 shrouds 85 Fewet (Wadi Tanzzuft) 101 Kissi (Burkino Faso) 389–90 leather 17 Fewet (Wadi Tanzzuft) 101 Iwelen (Niger) 18 Wadi al-Ajal 85–87 Wadi al-Ajal burials 83–85, 84t Shuwa Arabic 433–34 sickle-cell disease 231–32 Sidi Allal el Bahraoui (Morocco) 284, 290, 291–92t, 293, 294 Sidi Bchir (Morocco) 290 Sidi Jari (Morocco) 290, 291–92t Sidi Khelili (Morocco) 290, 291–92t, 293, 294 Sidi Latrech (Tunisia) 250, 254 Sidi Slimane (Morocco) 290, 291–92t, 293, 294, 302 Siga (Mauretania) 250, 256, 286 Sigus (Algeria) 250, 266 Sila (Algeria) 250, 264, 265, 267, 268, 270, 274, 275–76 silo-shaped tombs 253 silver 17, 71, 88–89t, 91–92, 297, 298, 299, 305 Simitthus (Tunisia) 250, 258, 261 Simoneau, André 332, 336 Sivilli, S. 355–56, 362 Siwa 497, 504 languages 476, 478, 481, 509 mummification 9 skeletal analysis Egyptian/Nubian 223–41, 224, 229, 234t, 235t, 236 Garamantes Fewet 162–64, 163, 183 craniometry 168–72, 169, 170, 171 dental dimensions 172–74, 173 mortality profiles 164–68 postcranial skeleton 174–77, 176, 178 sample 164, 165–66t stress markers/dental pathologies 177–80 Wadi al-Ajal 135–37, 138–40, 144t, 145 slavery 20t, 22, 23–24, 539 Lake Chad Basin 404, 407, 420–22 and language 439

Middle Nile 202–3 Smirat (Tunisia) 250, 272 Smith, H. F. 140, 141 social memory 5–6 social mobility 21–22 social status 530 Durbi Takusheyi (Nigeria) 414–15 and isotopic analysis 141 Middle Nile 202–3, 206–10 craftworkers 202 Morocco 333–34 southern Lake Chad Basin 414, 417–18 Wadi al-Ajal 152, 154–55 Wadi Tanzzuft 119, 129 solar alignment 62–63 Songhay languages 433, 434, 440–44, 441, 443, 443t, 488 Souk el Gour see Gour mausoleum Southern Lake Chad Basin, settlement 399, 401–4, 402, 420–21 Southern Sahara 17–19 Souville, G. 283, 289–90, 293 Spain see Iberian Peninsula Spaulding, J. 212 standing stones 346–48t, 352, 357, 358, 365 stelae 63, 97–99, 98 funerary 102 Burkino Faso 379–81, 381 Fewet (Wadi Tanzzuft) 117 Moroccan 316, 330 of Pasenhor 503 stepped tombs 9–10, 61, 66–68, 67, 69, 70t, 72t Fewet 114 pottery 92 Stockhammer, P. W. 110 Stojanowski, C. 237 Strabo 510 Stringer, C. B. 141 strontium isotope analysis, Nile Valley 228–31, 229 structuration theory 26 Sudan 198 Darfur 196–97, 198 Kordofan 196–97 Tombos 223–27, 224, 239–41 burials 240–41 porotic hyperostosis 178 strontium isotope analysis 229–31, 233, 235t tables, offering 63, 320 Tadrart Akakus see Wadi Tanzzuft/Tadrart Akakus

https://doi.org/10.1017/9781108634311.018 Published online by Cambridge University Press

559

560

Index

Tafilalat area (Morocco) 15–16 chapel monuments 276 see also Bouïa; Braber; Erfoud; Taouz Tamazight, Moroccan 466, 468, 469, 470, 472, 478, 479, 483, 497, 517 Taouz (Morocco) 321–23, 322, 323, 330 Taqallit (Fazzan) 67 burials 81–82, 82, 98–100 osteology analysis 149, 150, 153, 154 Tashelhit 509, 511 task-specialisation 136 Tayadirt (Morocco) 284, 286, 290–93, 300, 303 Tebessa (Algeria) 250, 253 Teboursouk (Tunisia) 250, 266 Teda-Daza 432, 434, 440 teeth, see under skeletal analysis textiles Fewet (Wadi Tanzzuft) 116–17 Kissi (Burkino Faso) 387, 390 see also shrouds Thamusida (Morocco) 285–86, 297, 300, 302, 303–4 Thébert, Y. 257 Thugga 250, 256, 256, 261, 286 Thuliva Kwacha (Nigeria) 411–12, 413 Tiddis (Algeria) 253, 260, 264, 267, 268 children’s burials 265 pottery 270, 271 Tiffech (Algeria) 250, 266 Tifinagh inscriptions 128, 444–45, 511 Tin Hanan (Algeria) 17–19, 103, 530 Tissot, Charles 281 Tombeau de la Chrétienne 250, 257, 293, 294 Tombos (Sudan) 223–27, 224, 239–41 burials 240–41 porotic hyperostosis 178 strontium isotope analysis 229–31, 233, 235t tools 299 iron 124, 403 trade 20t, 56, 126–27, 136, 155, 526, 531, 532, 542 caravan routes 112 Durbi Takusheyi (Nigeria) 415 Egypt/Nubia 225 (Kissi) Burkino Faso 378 Middle Nile 199–200 Morocco 333, 338 and Iberian Peninsula 297, 301, 303, 305–6 southern Lake Chad Basin 406–7, 417–18, 420 Western Sahara 343–44, 366 see also slavery

https://doi.org/10.1017/9781108634311.018 Published online by Cambridge University Press

trading chiefs, Middle Nile 207 trading communities 20t, 22 Trans-Sahara (defined) xxiv, 8 Trans-SAHARA Project xxiv–xxvi, 12, 134–35, 137 transhumance 20t, 20, 23, 205, 211, 213 transportation, of bodies to burial 86, 204 treatment of the body in burials Iwelen (Niger) 18 Morocco 287–89, 288 Numidia 262–69, 276–77 southern Lake Chad Basin 415–16, 531 Daima/Borno 408–9 Gajiganna Culture 407–8 Mandara 410, 412, 413, 416 ‘Sao’ sites 414 Wadi al-Ajal 83–89, 84t trephinations 136–37 Tripolitania 61, 446, 511, 515, 531, 537n36, 538 Tuareg 26, 378, 379, 435, 452–53, 453t, 455–56, 478, 481 and genetics 507–8 language 473, 474–78, 476, 480, 484–85t, 517 and Tin Hinan 17 Tuareg Berber, language 432, 434 Tubiana, Joseph and Marie-Jose 440, 451 Tubu see Teda-Daza tumuli see monuments, funerary Tunisia Althiburos 14, 249, 250, 276, 506–7, 529 bone in domestic areas 268–69 pottery 271, 272, 273 Ellès 250, 252, 254, 255, 261 Enfida 250, 259, 266 Gabès 250, 266 Henchir Mided 250, 252, 261, 264, 267, 270, 272 Kalaat es-Snam 250, 260 Mdidi (Tunisia) 14 Sidi Latrech 250, 254 Simitthus 250, 258, 261 Smirat 250, 272 Teboursouk 250, 266 see also Djebba; El Ksour; Numidia urbanisation 443t, 444, 505, 526 V-shaped monuments 346–48t, 352, 352–53, 355–56, 363 Valencia de la Concepción (Spain) 366 Vandals 515 Vernet, R. 359

Index

vesicular basalt bowls, Wadi Tanzzuft 124, 125 lamps, Wadi Tanzzuft 113, 118 Volubilis (Morocco) 283, 284, 285, 286, 290, 299n48, 303, 511 von Cramon-Taubadel, N. 141 Wadi al-Ajal 53, 111, 532 biological identity 134–56, 137, 144t, 146, 147, 150, 151, 152, 154 burials 60, 64–68, 65, 67, 68, 102 and the Garamantes 108, 125–29 Wadi Draa (Morocco) 15–16, 337–38, 531, 533, 534 horse/chariot engravings 505 see also Adrar Zerzem; Foum Larjam Wadi Gir (Algeria) 15 Wadi Tanzzuft/Tadrart Akakus (Fazzan) 11–12, 108–11, 109, 529 burials 111–19 cairns 57–60 drum-shaped monuments 60 domestic evidence 119–25, 121, 122, 123, 124, 125 Wadi Tifariti (Western Sahara) 364, 367, 513 wage labour 20t, 23, 537 Wargla 467, 467, 470, 471, 471, 473, 474, 476, 478, 481, 497, 517 water, availability of 3 water courses 448–49 Western Sahara 344, 358, 367–68 Watwat (Libya) 54, 65, 77–78, 87–89, 149–55 osteology analysis 149, 150, 153

weapons 455 as grave goods 337 Kissi (Burkino Faso) 387, 389, 393 microliths 449–50 Weaver, T. D. 140 West Africa and Chad 19 Western Sahara 15–17, 341–44, 342, 343, 530 monuments 344–58, 349, 350, 351, 353, 354, 355, 356, 357, 358, 364–66, 513 locations/alignments 362–64, 363, 365 use of colours/materials 362 use of quartz 357, 359–62, 360–61t, 364 Western Sahara Project 344 Western Sahelian languages 488 Wilson, Andrew 56, 496, 512 Wölfel, D. J. 438–39 word borrowing 469–70, 479 and Berber 442, 469–70 and Latin 436, 437, 487 and Punic 447–48, 449t and Songhay 444 Wynne-Jones, S. 419 Al-Ya’qubi 440 Zaghawa see Beria Zenaga 435, 448, 452–53, 476, 478–80, 481, 483, 484–85t, 486–87, 497, 499, 501–2 Zenati language 472–73 zenithal access dolmens 259, 260, 264, 277 Zilum (Nigeria) 402, 406, 408, 415, 420–21 Zinkekra (Libya) 54, 60, 69, 70t, 77, 504 osteology analysis 149, 150, 153, 154

https://doi.org/10.1017/9781108634311.018 Published online by Cambridge University Press

561

https://doi.org/10.1017/9781108634311.018 Published online by Cambridge University Press