Trade and Travel in the Red Sea Region: Proceedings of Red Sea Project I Held in the British Museum October 2002 9781841716220, 9781407326894

Table of contents :
Front Cover
Title Page
Copyright
Table of Contents
List of Maps and Figures
List of Tables
Acknowledgements
Contributors
Introduction
The Red Sea: the wind regime and location of ports
Arabian trade with Ethiopia and the Horn of Africa: from ancient times to the sixteenth century
The elusive land of Punt revisited
Pharaonic Egypt and the Red Sea arms trade
Possible connections in antiquity between the Red Sea coast of Yemen and the Horn of Africa
Ancient interaction across the southern Red Sea: new suggestions for investigating cultural exchange and complex societies during the first millennium BC
The ‘pre-Aksumite’ state in northern Ethiopia and Eritrea reconsidered
Pre-Aksumite Aksum and its neighbours
Adulis to Aksum: charting the course of antiquity’s most important trade route in East Africa
The Egypto-Graeco-Romans and Panchaea/Azania: sailing in the Erythraean Sea
Reflections of ethnicity in the Red Sea commerce in antiquity: evidence of trade goods, languages and religions from the excavations at Berenike
Gold dinars and silver dirhams in the Red Sea trade: the evidence of the Quseir documents
The merchants’ diet: food remains from Roman and medieval Quseir al-Qadim
‘What the devil are you doing here?’ Arabic sources for the arrival of the Portuguese in the Red Sea and Indian Ocean
Mamluk and Ottoman activity in Yemen in the sixteenth century: coastal security and commercial significance
Quseir Fort and the archaeology of the Hajj
Les échanges commerciaux entre les rives africaine et arabe de l’espace mer Rouge golfe d’Aden aux seizième et dix-septième siècles
Luxury wares in the Red Sea: The Sadana Island shipwreck
Index

Citation preview

Society for Arabian Studies Monographs No. 2 Series editors D. Kennet & St J. Simpson

Trade and Travel in the Red Sea Region Proceedings of Red Sea Project I Held in the British Museum October 2002

Edited by

Paul Lunde Alexandra Porter

BAR International Series 1269 2004

ISBN 9781841716220 paperback ISBN 9781407326894 e-format DOI https://doi.org/10.30861/9781841716220 A catalogue record for this book is available from the British Library

BAR

PUBLISHING

Table of Contents LIST OF MAPS AND FIGURES .......................................................................................................................................... ii LIST OF TABLES ............................................................................................................................................................ iii ACKNOWLEDGEMENTS ................................................................................................................................................. iv CONTRUBUTORS ............................................................................................................................................................ v INTRODUCTION .............................................................................................................................................................. 1 MAP OF THE RED SEA ................................................................................................................................................... 6 THE RED SEA: THE WIND REGIME AND LOCATION OF PORTS William Facey.................................................................. 7 ARABIAN TRADE WITH ETHIOPIA AND THE HORN OF AFRICA: FROM ANCIENT TIMES TO THE SIXTEENTH CENTURY Richard Pankhurst ....................................................................................................................................................... 19 THE ELUSIVE LAND OF PUNT REVISITED K.A. Kitchen................................................................................................. 25 PHARAONIC EGYPT AND THE RED SEA ARMS TRADE D. M. Dixon .............................................................................. 33 POSSIBLE CONNECTIONS IN ANTIQUITY BETWEEN THE RED SEA COAST OF YEMEN AND THE HORN OF AFRICA Edward J. Keall ........................................................................................................................................................... 43 ANCIENT INTERACTION ACROSS THE SOUTHERN RED SEA: NEW SUGGESTIONS FOR INVESTIGATING CULTURAL EXCHANGE AND COMPLEX SOCIETIES DURING THE FIRST MILLENNIUM BC

Matthew C. Curtis ................................... 57

THE ‘PRE-AKSUMITE’ STATE IN NORTHERN ETHIOPIA AND ERITREA RECONSIDERED Rodolfo Fattovich .................... 71 PRE-AKSUMITE AKSUM AND ITS NEIGHBOURS Jacke Phillips...................................................................................... 79 ADULIS TO AKSUM: CHARTING THE COURSE OF ANTIQUITY’S MOST IMPORTANT TRADE ROUTE IN EAST AFRICA Walter Raunig ............................................................................................................................................................. 87 THE EGYPTO-GRAECO-ROMANS AND PANCHAEA/AZANIA: SAILING IN THE ERYTHRAEAN SEA Felix Chami ............. 93 REFLECTIONS OF ETHNICITY IN THE RED SEA COMMERCE IN ANTIQUITY: EVIDENCE OF TRADE GOODS, LANGUAGES AND RELIGIONS FROM THE EXCAVATIONS AT BERENIKE STEVEN E.

Sidebotham ...................................................... 105

GOLD DINARS AND SILVER DIRHAMS IN THE RED SEA TRADE: THE EVIDENCE OF THE QUSEIR DOCUMENTS Li Guo.. 117 THE MERCHANTS’ DIET: FOOD REMAINS FROM ROMAN AND MEDIEVAL QUSEIR AL-QADIM Marijke Van der Veen............................................................................................................................................... 123 ‘WHAT THE DEVIL ARE YOU DOING HERE?’ ARABIC SOURCES FOR THE ARRIVAL OF THE PORTUGUESE IN THE RED SEA AND INDIAN OCEAN Paul

Lunde................................................................................................................................ 131

MAMLUK AND OTTOMAN ACTIVITY IN YEMEN IN THE SIXTEENTH CENTURY: COASTAL SECURITY AND COMMERCIAL SIGNIFICANCE Clive

Smith......................................................................................................................................... 137

QUSEIR FORT AND THE ARCHAEOLOGY OF THE ДAJJ Charles LeQuesne................................................................... 145 LES ÉCHANGES COMMERCIAUX ENTRE LES RIVES AFRICAINE ET ARABE DE L’ESPACE MER ROUGE GOLFE D’ADEN AUX SEIZIÈME ET DIX-SEPTIÈME SIÈCLES Michel

Tuchscherer................................................................................... 157

LUXURY WARES IN THE RED SEA: THE SADANA ISLAND SHIPWRECK Cheryl Ward .................................................. 165 INDEX ........................................................................................................................................................................ 172

i

List of Maps and Figures Figure 1: Red Sea winds in April. Mean barometric pressure (mb) and dominant winds (mean force). ........................... 8 Figure 2: Red Sea winds in July. Mean barometric pressure (mb) and dominant winds (mean force). ........................... 10 Figure 3: Red Sea winds in October. Mean barometric pressure (mb) and dominant winds (mean force)...................... 10 Figure 4: Red Sea winds in January. Mean barometric pressure (mb) and dominant winds (mean force) ...................... 11 Figure 5: The Red Sea and Gulf of Aden according to Ibn Mājid, after Tibbetts 1971 ................................................... 14 Figure 6: Painted relief from the temple of Deir el-Bahri (Naville 1898: pl. 72)............................................................. 36 Figure 7: Painted relief from the temple of Deir el-Bahri; the Pharaonic envoy meets the Chief of Punt and his wife (Naville 1898: pl. 69).............................................................................................................................................. 36 Figure 8: Egyptian axes were used for ‘cutting hbny-wood’ (Naville 1898: pl. 7) .......................................................... 39 Figure 9: Location of archaeological sites mentioned in text........................................................................................... 44 Figure 10: The sites and landscape setting of al-Midamman ........................................................................................... 44 Figure 11: Modelled bull’s head from pottery vessel and monkey figurine..................................................................... 45 Figure 12: Pottery figurine of a human ............................................................................................................................ 45 Figure 13: Neck of decorated pottery jar.......................................................................................................................... 45 Figure 14: Composite reconstruction drawing of typical ‘libation pail’ .......................................................................... 46 Figure 15: Incised stone panel of chevron-bordered snake bodies................................................................................... 47 Figure 16: Snake heads .................................................................................................................................................... 48 Figure 17: Oryx heads and water jars............................................................................................................................... 48 Figure 18: Cross hatched design; ibex horns.................................................................................................................... 49 Figure 19: Gold beads ...................................................................................................................................................... 49 Figure 20: Painted ostrich egg fragments......................................................................................................................... 49 Figure 21: Miniature horse head carved in bone .............................................................................................................. 50 Figure 22: Symbols scratched in clay (before firing) of incense burner tops................................................................... 51 Figure 23: Map of Southern Red Sea region showing sites mentioned in the text........................................................... 58 Figure 24: Ona flaked stone lithics .................................................................................................................................. 61 Figure 25: Map of northern Ethiopia and Eritrea showing the location of Aksum (‘D site’)........................................... 79 Figure 26: Sherd from the ‘D site’ at Aksum with a single ‘alpha’ sign ......................................................................... 81 Figure 27: Sherds with ‘thumbnail’ rocker decoration from the ‘D site’ at Aksum......................................................... 82 Figure 28: Pottery with orange fabric from the ‘D site’ at Aksum................................................................................... 83 Figure 29: Lower body and base of a splay-footed vessel ............................................................................................... 83 Figure 30: Ruins of ancient Adulis; Gulf of Zula in Eritrea............................................................................................. 88 Figure 31: Temple-ruin on the Qohaito plateau. .............................................................................................................. 88 Figure 32: Ancient water reservoir on the Qohaito plateau.............................................................................................. 89 Figure 33: Map of routes to Qohaito plateau. .................................................................................................................. 90 Figure 34: Ascending the plateau of Qohaito................................................................................................................... 90 Figure 35: Pre-Roman expeditions to Africa south of the Sahara. ................................................................................. 94 Figure 36: Roman expeditions and knowledge of eastern Africa. ................................................................................... 96 Figure 37: Harappa and South Indian Megalithic 10th -4th century BC found in Mafia cave excavations..................... 98 Figure 38: Marl clay pottery from the Nile Valley excavated from Mafia cave excavations .......................................... 98 Figure 39: Stamped ceramic found in association with early centuries AD ceramics from Mafia Island, Tanzania. ...... 99 Figure 40: Map showing the location of Berenike ......................................................................................................... 106 Figure 41: Map showing routes and sites between the Nile and the Red Sea ................................................................ 106 Figure 42: Berenike and surroundings. .......................................................................................................................... 107 Figure 43: Site plan, Berenike........................................................................................................................................ 109 Figure 44: Site view with places of interest numbered 1-15. ......................................................................................... 109 Figure 45: RN 970a, recto.............................................................................................................................................. 109 Figure 46: RN 970b, verso. ............................................................................................................................................ 109 Figure 47: Quseir and the Indian Ocean......................................................................................................................... 124 Figure 48: The location of Quseir al-Qadim and other sites mentioned in the text........................................................ 124 Figure 49: Artichoke bracts (mid second century AD). ................................................................................................. 125 Figure 50: Hazelnut shells (Mamluk period). ................................................................................................................ 126 Figure 51: Pistachio shells (Mamluk period). ................................................................................................................ 126 Figure 52: Parrot fish (Mamluk period; photo Sheila Hamilton-Dyer and Wilfried Van Rengen). ............................... 126 Figure 53: Chicken foot and feathers (Roman period; photo Sheila Hamilton-Dyer).................................................... 126 Figure 54: Peppercorns imported from India (early first century AD)........................................................................... 127 Figure 55: Rice kernels imported from India (early first century AD)........................................................................... 127 ii

Figure 56: Epicarp and fibrous husk of coconut imported from India (mid second century AD).................................. 127 Figure 57: Shell of coconut imported from India (mid second century AD). ................................................................ 127 Figure 58: Aubergine calyces (Mamluk period). ........................................................................................................... 128 Figure 59: Map of the northern Red Sea showing the location of Quseir ...................................................................... 146 Figure 60: Map of modern Quseir showing buildings in existence in 1931 and approximate boundaries of the town in 1799. ........................................................................................................................................................ 147 Figure 61: Early nineteenth century view of Quseir from the south painted by Robert Moresby.................................. 147 Figure 62: Part of Napoleonic bastion at the south corner of the fort incorporating blocks taken from a monumental Graeco-Roman building. ...................................................................................................................................... 148 Figure 63: Plan of the fortress showing location of archaeological trenches. ................................................................ 149 Figure 64: North tower retaining a portion of the original sixteenth century masonry beside the curtain wall – note also the break in the Napoleonic ashlar refacing of the curtain wall short of the tower to accommodate a planned bastion. .................................................................................................................................................... 150 Figure 65.1-4: Tobacco pipes......................................................................................................................................... 151 Figure 66.1-7: Examples of sherds from grey-brown water-carriers (qulal).................................................................. 151 Figure 67: Late eighteenth century structures in Trench B overlying sixteenth century stone cross-wall. .................... 152 Figure 68: View of seaward defences taken in 1960 with partially collapsed Napoleonic bastion around original Ottoman tower visible in the foreground and Napoleonic entrance redan in front of the south-eastern curtain wall in the background (courtesy of Gisela Kircher)............................................................................................ 153 Figure 69: The Napoleonic entrance redan with reconstructed arch from the reign of MuΉammad ΚAli...................... 154 Figure 68: Excavation of the 50-meter-long Sadana Island ship began in 1995. Two archaeologists work in the area of the ship’s main deck. (Photograph A. Flanigan) .............................................................................................. 166 Figure 69: Chinese export porcelain on the ship was typical for the mid-18th century preferences of people living in the Islamic Middle East. (Photograph N. Piercy) ................................................................................................. 167 Figure 70: Although the cobalt blue underglaze is easy to identify on many of the porcelain pieces, recovering the original enameled design is much more difficult. Pencil lines on this porcelain plate demonstrate part of the recovery process. (Photograph N. Piercy) ............................................................................................................ 167 Figure 71: Only this cup and a fragment of one other similar cup depict living animals; there are no examples of human portraiture in the Sadana Island shipwreck assemblage. (Drawing N. Piercy) ......................................... 168 Figure 72: Arabic inscriptions from copper serving and cooking wares in the stern provide the date of 1765 CE or later for the ship’s sinking. a) Damaged monogram including the numerals 1178, presumably a Hijri date equivalent to 1764 CE b) Inscription including the Hijri date 1169, equivalent to 1754/5 CE c) Inscription reading ‘Sahib el ra’is Moussa Mahmoud’ followed by what seems to be an illegible date. (Photographs H. Wellman; tracing author)...................................................................................................................................... 169 Figure 73: Qulal, spouted jugs, clay goblets or incense burners, and pipe bowls from the Sadana Island shipwreck. (Photograph M. Kato)........................................................................................................................................... 169

List of Tables Table 1: Recent 14C dates from Neolithic Zanzibar and Mafia caves, off the coast of Tanzania.................................... 97 Table 2: Imported ceramics from India and Egypt excavated from Mafia cave, off the coast of Tanzania..................... 98

iii

Acknowledgements

The Society for Arabian Studies most warmly thanks the organiser of the Red Sea Project 1, Dr Francine Stone, and the editors of the Proceedings, Paul Lunde and Alexandra Porter, for contributing their time and expertise to this enterprise. The Society for Arabian Studies is most grateful to the following sponsors of this first volume of Proceedings of Red Sea Project 1: Seven Pillars of Wisdom Trust Leigh Douglas Foundation British Yemeni Foundation Without all this help this valuable volume would not have been possible.

iv

Contributors publishing, research and museum consultancy company.

Felix Chami is an archaeologist with a first degree in Sociology from the University of Dar-es-Salaam, an MA in Anthropology from Brown University, USA, and a Ph.D. in Archaeology from Uppsala, Sweden. He has been employed by the University of Dar-es-Salaam since 1986 and is currently an Associate Professor. He is also the co-ordinator of the African Archaeology Program involving thirteen countries of southern, eastern and western Africa. Most of his work has focused on the coast and islands of East Africa, and he has recently started a project in the Great Lakes Region. He is married with three sons.

Rodolfo Fattovich is professor of Ethiopian Archaeology at the University of Naples 'L’Orientale' (formerly Istituto Universitario Orientale). He has conducted research in Egypt, Sudan, Ethiopia and Eritrea. He has directed or co-directed archaeological projects at Naqada (19771984) and Tell el-Farkha (1987-1990) in Egypt and Kassala (1980-1995) in the Sudan. Since 1993 he has directed, with K. A. Bard of Boston University, the Joint Archaeological Expedition of the Oriental Institute, Naples and Boston University at Bieta Giyorgis (Aksum, Ethiopia). Since 2001 he has directed the Archaeological Expedition of the Istituto Italiano per L’Africa e l’Oriente (Is.I.A.O.), Rome, at Wādī Gawasis on the Red Sea coast of Egypt, also in collaboration with Boston University. His research interests include the process of state formation in northeastern Africa and the development of the Red Sea trade in late prehistoric and ancient times.

Matthew C. Curtis is a Ph.D. candidate and instructor in Anthropology at the University of Florida with a specialisation in Holocene African archaeology. His research interests include the archaeology of ancient complex societies and urbanism in East Africa and the Horn of Africa, regional settlement analysis in archaeology, culture contact and interaction and cultural resource management and education in Eritrea. He has conducted archaeological research in the highlands of Eritrea since 1997. He was a Fulbright fellow to Eritrea 1999-2000 and directed the Greater Asmara Regional Archaeological Survey Project (GARASP). He returned to Eritrea in 2001 to co-teach the University of Asmara’s archaeological field school and continue regional archaeological research with the University of Asmara and National Museum of Eritrea.

Li Guo was born in China and studied Arabic in Yemen and Egypt. After receiving a Ph.D. in Arabic and Islamic Studies from Yale University in 1994, he taught at the University of Chicago (1995-1999). He is currently an assistant professor of Arabic at the University of Notre Dame. The author of Early Syrian Mamluk Historiography: al-Yunini’s Dhayl Mir’at al-zaman (Leiden: E.J. Brill), published in 1998, he has written extensively on medieval Arabic historiography and literature. His most recent book (2004) is Commerce, Culture and Community in a Red Sea Port in the Thirteenth Century: The Arabic Documents from Quseir (Leiden: E.J. Brill).

David M. Dixon is an independent researcher. He studied at London University. He has travelled in Egypt, Cyrenaica and Sudan and worked on archaeological excavations in these countries. His main fields of interest are Nubia, the Sudan and the Red Sea region, and the history of Egyptology with particular reference to the role of the French and English military in the archaeology of Egypt and the Sudan, 1798-1900. This latter topic is the subject of a forthcoming book. Among his recent publications are papers on the Gordon Relief Expedition and the part played by river gunboats during the First (1884-85) and Second Sudan wars (1896-99).

Ed Keall was never formally trained as an archaeologist, but his involvement on many different sites in the Middle East for forty years has given him broad experience. Born in the UK and trained in Greek and Latin literature, he first visited the Middle East on an adventure in 1962. In 1970, after learning the trade through practical experience in countries like Iran, Iraq and Turkey, he acquired a Ph.D. in Islamic art at the University of Michigan. His first Canadian appointment was as an Assistant Curator in the West Asian section of the Royal Ontario Museum in 1971. He was cross-appointed to the University of Toronto at the same time. He is currently Senior Curator as well as Head of the ROM’s Near Eastern and Asian Civilizations Department, territory which incorporates everything in the art and archaeology of the Old World from North Africa to Japan (Casablanca to Kyoto). He has directed the Canadian Archaeological Mission of the ROM (CAMROM) in Yemen since 1982 (for more information see the ROM website: www.rom.on.ca/neac).

William Facey was born in Zambia in 1948 and brought up in England. He read classics, philosophy and art history at Oxford before becoming involved in the Arabian Peninsula in 1974. Since then he has worked as a planning and research consultant on numerous projects to set up museums of the archaeology, history and natural history of the Arabian states. He has also carried out exhibition and museum consultancy work in England, Europe, Central America and the Far East. His books cover the history, architecture and early photography of the countries of the Arabian peninsula. Since 1994 he has been director of the London Centre of Arab Studies, a v

Egypt with its neighbours both in the Aegean and the Horn of Africa regions, as well as interaction between the Red Sea and its hinterland. She developed and directs the Shire District Archaeological Survey in north-western Ethiopia, investigating the east-west land corridor of the Red Sea and the Nile Valley; its pilot season was in 2001. She has been a Research Fellow of the McDonald Archaeological Research Institute and Honorary Fellow of the British Institute in Eastern Africa since 1998.

Kenneth Kitchen is emeritus Professor of Egyptology at the University of Liverpool, but also has long-term expertise in ancient Near Eastern disciplines. He has published, lectured and travelled widely. Charles Le Quesne graduated with an MA from Edinburgh University in 1988. From 1989-1992 he worked for the Ben Ezra Synagogue Restoration Project in Old Cairo (Fustat). He has continued to work there and is presently part of an American Research Centre in Egypt (ARCE), an archaeological team monitoring the insertion of a major waste-water system in the area. From 1993-1995 he was employed by Gifford and Partners in Chester to work on a series of development-related archaeological projects. Since 1995 he has worked independently in the UK and abroad, combining small developer-funded projects with larger research projects, including the writing up of twenty years of fieldwork on the Roman defenses of Chester for Gifford and Partners. In 1996 he carried out an archaeological study, as part of a larger Environmental Impact Assessment, in Aden harbour. In 1997-1998 he directed archaeological investigations at Quseir Ottoman Fort on the Red Sea coast as part of an ARCE project to turn the fort into a Red Sea Visitors Centre. Since 1999 he has directed excavation and post-excavation at The Grove, a large multi-period settlement site near Watford, Hertfordshire. He is presently preparing the final publications of both Quseir Fort and The Grove.

Walter Raunig was born in Innsbruck in 1936. He studied agriculture in Vienna from 1956-1959 (Hochschule für Bodenkultur), followed by cultural and physical anthropology at the University of Vienna from 1959-1964, with a D.Phil. Thesis on 'Die kulturellen Verhältnisse Nordost-und Ost-Afrikas im ersten nachchristlichen Jahrhundert-entworfen an Hand des Periplus des erythräischen Meeres' (Vienna 1964). He has lectured on African history, anthropology and art history at the University of Bayreuth and Augsburg (1986-1989). He was appointed a research assistant at the Museum für Völkerkunde Basel (1964-1968), followed by research assistant and vice-director at the Völkerkundemuseum der Universität Zürich (1968-1977) and finally director of the Staatlisches Museum für Völkerkunde in Munich (1978-2001). Since 1958 he has travelled extensively in Africa, the Near and Middle East and in eastern Asia; he is a member of various scientific societies and co-edits the journals Mare Erythraeum (Munich) and Nubica et Aethiopica (Warsaw).

Paul Lunde was raised in Saudi Arabia and is a graduate of the University of California (Berkeley) and S.O.A.S. He has lived and studied in the Middle East and for many years conducted research in the Vatican Libary and Propaganda Fide in Rome and the Archivo de Indias in Seville. Specialising in Arabic geographical literature, he has written extensively on related topics. His most recent book is Islam: Culture, Faith and History (2002) (London: Dorling Kindersley).

Steven Sidebotham, a professor in the History Department at the University of Delaware, has conducted archaeological projects since 1972, both on land and underwater, in Italy, Greece, Tunisia, Libya, Israel, North Yemen, India and Egypt. He is currently co-director of the Berenike Project, director of the Sikait Project and director of the Eastern Desert Survey Project. His main interests are Ptolemaic-Roman economic activity in the Red Sea-Indian Ocean and Eastern Desert of Egypt. He has conducted fieldwork in Egypt since 1980.

Professor Richard Pankhurst was the founder and first director of the Institute of Ethiopian Studies of Haile Sellassie I (later Addis Ababa) University. He has published widely on Ethiopian history and culture. Currently attached to the Institute of Ethiopian Studies, his books include Introduction to the Economic History of Ethiopia (1961) and Economic History (1968).

Clive Smith spent five years as a District Officer in Kenya, after reading for the Classical Tripos at Cambridge University; this was followed by a period in Mauritius as Private Secretary and A.D.C. to the Governor. Subsequently, he spent twenty-eight years with the British Council, twenty-one of which were directly concerned with the Arab world and nine of which were in the Arabian peninsula, as Director in Yemen (1973-1978) and Saudi Arabia (1988-1992). He has visited all but one of the Arabic speaking countries. In 1972-1973 he studied Arabic at the Foreign and Commonwealth Office Middle East Centre for Arabic Studies in Shemlan, Lebanon. He was awarded the O.B.E. in 1979 and the Order of the Two Niles in 1984.

Jacke Phillips studied at the University of Toronto, receiving her Ph.D. in 1991. In 1993 she moved to the UK as the full-time Researcher for the Aksum Archaeological Research Project (Ethiopia, 1993-1997) and contributed heavily to its two-volume final report, chiefly for the ceramics. She also has worked on both excavations and survey in the Sudan, at Hambukol (19851992), Old Dongola (1993), Suakin (2002) and the South Dongola Survey (1998-). An Egyptologist by training, she has never worked in Egypt but rather specialises in 'Egypt outside Egypt', the interaction of (mainly) ancient

Michel Tuchscherer is a professor of modern and contemporary history of the Middle East at the Université vi

de Provence (Aix-en-Provence). He is also associated researcher at the Institut d’etudes et de recherches sur le Monde arabe et musulman (IREMAM). He spent fifteen years in Egypt, Saudi Arabia, Yemen and Turkey, first as a teacher, then as a researcher in Cairo at the Institut français d’archéologie orientale (IFAO) and in Istanbul at the Institut français d’etudes anatoliennes (IFEA). He works on the early modern social and economic history of Egypt and the Red Sea area. Cheryl Ward teaches nautical archaeology, the history and construction of wooden ships, archaeobotany and archaeological conservation in Florida State University’s Program in Underwater Archaeology. Dr Ward’s most recent projects include: serving as nautical archaeologist on the Institute for Exploration expeditions in the Black Sea, acting as Chief Scientist for the investigation of a fifth millennium BC wooden object in an Oman lagoon, funded by the National Geographic Research Committee, and working as the ship reconstructor in the excavation of the world’s oldest planked boats at Abydos, Egypt. Ongoing research also includes publication preparation of the eighteenth century Sadana Island Shipwreck excavation report. Archaeobotanical studies of shipwrecks off the coast of Florida and Turkey are also underway.

vii

viii

Introduction The genesis of the project On February 15th, 2001, a nascent steering committee of the Society for Arabian Studies met over coffee in Bloomsbury to consider an event that had begun as just a notion among executive committee members in late 2000. This was a brain-storming session. Imagine. We had a blank canvas on which we could paint whatever we wanted – an epic scene or a closely observed still life. We had a good basic idea – let’s convene a study meeting on the Red Sea – and we had some fine minds amongst us to add bold strokes of energy and ambition. What we lacked was a focus. That could be fixed. However we also lacked money, and without someone in the Society ready to fund raise specifically for the venture, this was more difficult to remedy. So it was an immense boost to hear that day that the British Museum might host the event for the Society since it combined wiell with the exhibition on the ‘Queen of Sheba’ planned for the summer and autumn of 2002. This opportunity immediately took the onus off the Society to find substantial sums, and it gave the project an undisputed prestige venue in London. There was no looking back. The notion that we could collaborate with the British Museum instantly transformed a workshop or seminar concept into a British Museum Study Day. These Saturday lecture events are organised by the British Museum’s Department of Education for the Friends of the British Museum and a sophisticated general public with enquiring minds but average attention spans. If we were to gear the Workshop – now a Study Day – to such an audience we would need an accessible theme, and that of Trade was broad, understandable, and it would incorporate the alluring clichés about the Red Sea across the span of time – expeditions to the fabled land of Punt, coffee merchants, grain and sheep-laded dhows, pirates and Portuguese among them. Undoubtedly the idea of a captive audience determined the main theme of Trade at the first Red Sea study event. But there were strong feelings amongst us that the opportunity should not be missed to tackle little studied issues of Red Sea trade and to exchange multi-disciplinary views back and forth as well as up and down the Red Sea with as wide a gathering of first class scholars as possible. Certain of us wanted to limit the time frame to Antiquity in order to assemble a choice selection of pre-Islamic scholars to share recent work. This triggered a sense of caution amongst us that showcasing archaeological campaigns, no matter what period, should be kept to a minimum. We saw no reason to compete with other seminars where excavation reports are meat and potatoes to the gathering. A consensus emerged that we should broaden the topic and the time span to discuss Trade down throughout the ages in all its facets. It was hoped that human commodities such as slaves would figure in this brief. And pilgrims? one asked. How can we include the fascinating question of the pilgrimage as an engine of Red Sea trade and culture? Before this could be answered, someone warned that the notion of a Red Sea ‘culture’ should be discouraged, since current precepts of social anthropology deemed the idea of culture itself to be a fallacy. Again it looked as if we would get bogged down, as committees do. So we decided to adopt the joint theme of ‘Trade and Travel’ to satisfy those with an interest in movements of humans not just goods. We had to refrain from trying to shape the programme aggressively until we saw what papers might come forward. But we determined that the event should represent all the major eras from the New Kingdom, pre-Aksum, Aksum/Sabaean, Graeco-Roman, Medieval and pre-Modern. From Punt to Coffee was the cry. And we concluded that those with a strictly pre-Islamic interest could meet together outside the main programme, if more recent historical periods did not appeal. Talk turned to the future scope of the project. Our natural inclination was to foster much more wide-ranging discussions than ones geared for the general public. As we sat around our now empty coffee-cups in that canteen, we let our imagination roam across the entire spectrum of human endeavour and the natural world of the Red Sea basin as well as its hinterland and its outreach, and we came to the conclusion that if handled well, the project we were creating could occupy at least five years worth of study events from ‘Navigation’ to ‘Peoples to Natural Resources’ to ‘Material Culture’ and finally to ‘The Wider World of the Red Sea’ sphere. It became obvious that we had let the genii out of the bottle and that the Society for Arabian Studies, if it had the courage and the stamina, could become a matrix for unprecedented and much needed study in this key 1

realm that encompasses Arabia and East Africa, the Mediterranean and the Indian Ocean. The prospect of inaugurating such an ambitious long-term project was more a quaff of champagne than caffeine for many of us that day. But hard work had to begin. As the convenor, I wanted to start by creating a network of interested and interesting scholars to ask their suggestions for crafting the project. By the end of the summer of 2001, we had nearly a hundred potential participants and advisers. This response was stupendous, and it proved to me that the single Study Day event (where we were restricted by the Department of Education to eight speakers) would preclude the participation of many of those who had come forward initially with exciting topics and in several instances with whole teams of researchers ready and eager to contribute. The Society agreed to expand the event into a full weekend affair – a Study Day and a Workshop Day. But there was intense concern about allotting meagre Society resources for this larger undertaking – and it did not help that an eventual appeal to the British Academy to sponsor the Society proved unsuccessful. It was apparent that I would have to get on with my expanded plans on a shoe-string. I am afraid I never got over my frustration about this, but never mind. The project would live on energy alone. By September 2001 we were ready to send out the formal call for papers for an October 2002 weekend at the British Museum. This gave us over a year to prepare, and just enough time at that. The steering committee was enlarged to become a vetting committee and we enlisted some of the best researchers on African and Arabian affairs that Britain had to offer. I name them here to recognise each and everyone for their outstanding work in vetting the abstracts and selecting the two day programme: Dionisius Agius, Lucy Blue, Greville Freeeman-Grenville (a towering mentor to the project), Paul Lunde, Carl Phillips, Jacke Phillips, Sarah Searight, StJohn Simpson, Caroline Stone, John Sutton, Shelagh Weir, and in reserve David MacDowall and Trevor Marchand. While I am paying tribute, let me also acknowledge the incomparable Irena Fairless who volunteered her secretarial services one day a week for the next year, and the officers of the Society who pitched in to bring the event to fruition on that proverbial shoestring. In the final two months’ lead-up to the event, staff members of the British Museum’s Department of West Asian Antiquities took over much of the production work – website, abstracts, maps, technical coordination, etc. – that were beyond me and Irena Fairless to provide on a volunteer basis. Here we have StJohn Simpson and Alexandra Porter primarily to thank. By February 2002 some thirty-five abstracts were whittled down to eighteen papers. Three further submissions were accepted as Brief Papers to be distributed but not read at the Workshop for lack of ‘floor’ time, or due to scheduling conflicts with the speakers. The chosen papers ran the gamut of topics from commodities, ports and ships, to trade relations and regional economic strategies. Inevitably there were some disappointments. Three speakers chose to withdraw their papers. We never did receive submissions dedicated to pilgrims or to slavery. So much for the Travel part of our chosen theme. No Arab scholars came forward and only one African. Several excellent abstracts on trade beyond the Red Sea had to be respectfully declined in order to keep the focus on the Red Sea itself. We bent the rule to include a controversial paper based on recent archaeological finds on the Tanzanian coast and islands by Felix Chami, a scholar from Dar Es Salaam. The paper had thought-provoking evidence of Roman trading presence through the Red Sea to the shores of East Africa where it had not heretofore been thought to operate. Prof Chami’s participation was sponsored by the Society and the generosity of the British Institute in East Africa, Bristol University and the University of Ulster thanks to the fund-raising efforts of Paul Lane, John Sutton, Mark Horton, Colin Breen and Greville Freeman-Grenville at my behest. Perhaps in future the Society will ear-mark one of its own annual travel grants for a foreign scholar who otherwise would not have the opportunity to attend forthcoming Red Sea study days and workshops. Once the steering committee had selected the speakers, we were obliged to coordinate with the British Museum’s Department of Education to assure that the number, the length and the content of the Study Day papers would not exceed established Study Day norms. In effect the Study Day programme went through a second vetting process. While this might have been awkward, it was accomplished with grace and intelligence by the newly appointed director of the Department’s Arab World division, Nicholas Badcott.

2

Nick was a delight to work with at every stage thereafter and without him the Study Day would not have been such a success. And a success it was. On Saturday October 5th, 2002 in the BP Theatre of the Clore Education Centre the Study Day was attended by some 180 people; a record for the Arab World division of the Department of Education. We had worked hard to introduce the best speakers, the best illustrations and that ineluctable element of fun for our audience. Judging by the energy in the hall that day, we achieved our goals. The papers were presented chronologically in three main periods: Antiquity, Medieval and Pre-Modern. The programme (reproduced below) had only one change due to a serious accident that found Marijke van der Veen in hospital rather than at the British Museum that morning. Her new work on the diet of the Kārimī merchants of Quseir al-Qadim has had to wait until this publication to be ‘heard’ at last. Daniel M. Varisco’s paper on ‘The Sailing Seasons in the Red Sea and Indian Ocean: The View from Rasulid Aden’ does not appear in this publication because its author judged it too whimsical for a scholarly volume. Intended as a personal tribute to Prof Rex Smith’s work on the Red Sea traveller Ibn al-Mujāwir, Varisco’s talk took the form of a fictional account of a seafaring journey to the port of Aden by a thirteenth century merchant teasingly named Ibn Mujabbir. It was replete with anecdotes and tirades and sea lore that imaginatively replicated the medieval maritime world which is the subject of Dr Varisco’s own scholarship and that of his colleague, Prof Smith. It was a tour de force and it helped make the Study Day highly enjoyable, ephemeral though it was. The Workshop day on Sunday October 6th, 2002 was held in the more intimate setting of the Stevenson Theatre, where 94 attended. Here a moderator for each session was able to comment on the speakers’ short summaries and to take questions from the audience. Without wishing to single out contributions from a uniformly superb field of contributors, one of the high points of the Workshop was the cluster of papers on pre-Aksumite and Aksumite studies. These posed challenges to long-held concepts of Sabaean cultural determinants at play on the western side of the Red Sea and highlighted the question of the need to assign new terms to the indigenous culture. Equally rich was the opportunity to compare the port site of Berenike (presented by Steven Sideboham at the Study Day) with that of Quseir al-Qadim/Myos Hormos, studied in depth by the University of Southampton archaeology team under the leadership of David Peacock. A final measure of the success of the event could be described as the tapestry of knowledge woven between the disciplines and the historical periods. It was so intricately threaded that it held one’s attention tightly on both days. What Michel Tuchscherer had to say about the coffee trade of the sixteenth and seventeenth centuries, for example, was relevant toWalter Raunig’s coastal routes on the African side in the first century. Ed Keall’s early Bronze Age monoliths of al-Midamman on the Yemeni coast managed to invite comparisons with the feat of engineering in the giant amphora causeway of Roman Quseir al-Qadim/Myos Hormos illustrated by the Southampton team. The fables about the land of Punt studied by Kenneth Kitchen and David Dixon echoed fantastical Arab explanations about the arrival of the Portuguese in the Red Sea researched by Paul Lunde. And these are only some of the strands of the event’s intellectual raiment. Few if any of the specialists ‘ducked out’ of sessions that did not apply to their areas of expertise. Everyone had something to learn from the whole. That was extremely satisfying. The main workshop programme managed to finish on time despite its density and thus there was time for a Plenary Session in which all participants, scholars and attendees alike, could make suggestions for the next phase of the Red Sea project. Clearly the topic of Trade and Travel had barely scratched the surface of possible themes for future events. There were calls for attention to be paid to coinage, to textiles, to navigation and cartography, to the natural history of the region, to the languages and customs of its people, and to the outreach of the Red Sea sphere beyond its own shores. There was interest in facilitating small study groups to meet and share information informally at the fringes of the future events. There was a plea for more time to be allotted to speakers and for later starting times for non-London dwellers. While some of these thought might need to be modulated, the overriding sense was that the project itself merited a vibrant future. It is my hope that this future will be carried forward with intellectual rigour. Subsequent phases should receive adequate financial support to assure a high standard of preparation, and every courtesy to those involved, especially the speakers and authors. I know that the Society are proud of the Red Sea project and I want to thank them for the privilege of having been involved in this, its seminal phase. 3

The publication of the Proceedings The selection and preparation of the proceedings in this volume have been undertaken by editors Paul Lunde and Alexandra Porter, and the supervision of Sarah Searight and StJohn Simpson. The contents differ somewhat from the ‘Trade and Travel’ programme of October 5th and 6th, 2002 at the British Museum. I regret not having been able to take an active role in the effort. Painstaking work, worthy of high praise, has ensued since the ‘Trade and Travel’ weekend was convened. It is a credit to all the authors, the editors and the Society for Arabian Studies that such a volume now exists. The original ambitions for the Red Sea project that stirred that day amid the coffee cups are on the way to being realised, and it is reasonable to believe that a series of volumes shall one day contain the fruits of the entire project in its several phases. If this happens, we shall benefit by a collection of studies that ought to stand as a monograph on the Red Sea. It will be an enticing and a catalytic achievement, fostering nothing less than a significant contribution to our body of knowledge about this unique region.

Francine Stone

4

5

6

The Red Sea: the wind regime and location of ports William Facey As a sweeping historical generalisation, one might say that the Red Sea is an extreme example of a sea on the way to somewhere else. It is a sea that bisects the globe’s most extensive arid zone, the Saharo-Arabian. No large river systems flow into it, giving direct access to riparian civilisations. Apart from Ethiopia/Eritrea and Yemen at its southern end, its shores were not lined with civilisations presenting desirable commercial destinations in their own right. Only with Islam and the Muslim pilgrimage did that change at all. In this it contrasts greatly with, say, the Gulf, the western Indian Ocean and the Mediterranean: the coasts of these seas were dotted with maritime, mercantile civilisations in direct contact with the sea.

Winds and the location of ports in the northern Red Sea From the map alone one would have thought that Suez, or somewhere very close by, would have been the major port in the northern Red Sea throughout the centuries. But it is not so: the place now known as Suez and in antiquity as Arsinoë and Clysma/Qulzum has played a relatively minor role in Red Sea trade. It is true that Suez was used as the shipyard for Ottoman fleets in the Red Sea in the sixteenth century and continued in use as a sailing port by the Egyptians until MuΉammad ΚAli’s time in the early nineteenth century,1 but it was to become really dominant only with steam and the Suez Canal. In reality ports through the ages serving the Nile Valley exhibit a notable southward creep, thereby sparing ships the inconvenience of sailing up the Gulf of Suez. An early navigator to explicitly record this tendency was that hardy seadog and marine surveyor, Lt. J. R. Wellsted of the Indian Navy. While coasting under sail down the northern Hijaz in March 1831 he was struck by the advantages of ΚAynūnah as an anchorage and watering stop. In suggesting ΚAynūnah might be the real location of the first-century BC port of Leuke Kome, he makes the following comments on the Gulfs of ΚAqabah and Suez:

Leaving aside Yemen, Tihāmah and Eritrea, the civilisations in closest contact with the Red Sea were separated from it by mountain and desert: the Nile Valley; Edom/Palestine (the Nabataeans); even, to stretch a point, the non-commercial destinations of Mecca and Medina, inland from Jiddah and YanbuΚ respectively. This generalisation applies less to the southern part of the Red Sea – the Yemen and Eritrea. Even here, however, one might argue that it was the highlands, inland and away from the coast, that were the real commercial attraction. But these coasts did have certain attractions of their own in terms of population and trade. It was the climate that made the difference to the southern part of the Red Sea: unlike the northern half, it benefits from the monsoon winds and rains of the western Indian Ocean. It is that same climatic regime which also made the difference for seafarers, as it accounts for the wind regime of the Red Sea, with which this paper is concerned.

…as the navigation of this arm of the sea [sc. the Gulf of ΚAqabah], which, even at the present day, is considered perilous, must have presented insurmountable difficulties to them [the ancients], it is known that a port was fixed upon near the entrance, but outside the gulf, where the vessels coming from the south discharged their cargoes, and from which depôt, the merchandize was transported by land to Elath and Eziongeber. Thus the tedious passage up the gulf was avoided. It may be observed that the same motive for shortening a dangerous and tedious passage, has at different periods operated in causing the transfer of the trade from the port of Arsinoë, near the modern Suez, successively, to Myos-Hormos, Berenicé, Adulis, and, lastly, to Aden, without the Straits of Báb-el-mandeb (Wellsted 1836: 54-55).

This paper will demonstrate three points of general relevance to all periods of Red Sea history. First, it will be suggested that the separation between the coasts and the centres of trade and civilisation has given ports in the northern part of the Red Sea a certain mobility in their location along the coasts. A southward ‘creep’ is evident, for which the wind regime provides an explanation. Next, an examination of Jiddah and ΚAydhāb will show that, for sailors, the difference of seasonal winds in the northern and southern halves effectively divided the Red Sea into two seas. Finally, the author will illustrate the nautical problems of the Red Sea by reference to the experiences of some Arab and Portuguese mariners from the twelfth to the sixteenth centuries AD.

He could have left Adulis and Aden off his list, because to them other considerations apply, but his point is certainly valid as far as Myos Hormos/Quseir al-Qadim2 1

Quseir and Suez seem to have been used about equally in the early nineteenth century. In 1811 MuΉammad ΚAli’s son КūΒūn embarked his infantry for Yanbu and the Hijaz campaign from Suez, while in 1816 Ibrahīm Pasha used Quseir, not Suez, to embark his cavalry for YanbuΚ (Sabini 1981: 89, 164). 2 Whitcomb & Johnson (1979: 1–10) provide a useful summary of the history of Quseir al-Qadim, with the caveat that they identify the classical site with Leukos Limen/ Albus Portus (‘White Harbour’),

7

TRADE AND TRAVEL IN THE RED SEA REGION

Figure 1: Red Sea winds in April. Mean barometric pressure (mb) and dominant winds (mean force). Normal position of Intertropical Convergence Zone (ITCZ) and Berenike are concerned, and he might have added the Safagah area, as well as ΚAydhāb far to the south. Thus, from north to south, Mersa Gawasis, Quseir al-Qadim, Quseir itself, Berenike and ΚAydhāb have all at one time or another enjoyed prominence as ports, despite the somewhat alarming rigours of the desert crossing to the Nile.3 This southwards creep finds at least part of its explanation by what we know of the actual seasonal wind patterns, which divide the Red Sea into two main zones, divided roughly by a line between Jiddah and ΚAydhāb .

Jiddah and ΚAydhāb Jiddah Let us take a look first at Jiddah. For hundreds of years it has been the chief port on the Red Sea. This is of course partly because it is the port of Mecca and the pilgrimage. But there is another reason too for its rise to prominence, which tells us a great deal about trade and navigation in the Red Sea as a whole. It is often assumed that the pilgrimage created an enormous amount of commerce, in the sense that the pilgrims themselves absorbed and transacted a large volume of trade goods. The argument goes that it was this trade, a by-product of the Ήajj, that made Jiddah an important entrepôt.4 This needs to be qualified in two ways. First, pilgrims themselves certainly did go in for a good deal of petty buying and exchange along the routes and at Mecca itself, but it is most unlikely that this alone

whereas we are now certain from the investigations of David Peacock that Quseir al-Qadim is to be identified with Myos Hormos. 3 For the rigours of the overland crossing between the Nile and the Red Sea, see e.g. James Silk Buckingham who, going from Quseir to Qeneh in 1816, ‘was stripped naked, among the mountains, plundered of money, paper, arms and instruments and abandoned to my fate. I had to trace the rocky path naked and barefoot, scorched by day and frozen by night … I continued for two days without food or water, and the first nourishment of which I partook was some raw wheat from a sack, which, swelling in the stomach, nearly proved fatal to me.’ (Searight 1979: 152). Moses Maimonides’ brother David had a lucky escape in c. 1170 because he had failed to join up with a caravan from Qus to ‘Aydhāb which was pillaged en route (Goitein 1973: 209–11). In June 1183, Ibn Jubayr found the route between Qus and ‘Aydhāb relatively safe, with goods unguarded en route, but woe betide any pilgrim who strayed into the hands of local tribesmen (Broadhurst 1952: 61, 64).

4

Pilgrimages do bring economic benefits to pilgrim centres, but do not necessarily make them internationally significant commercial hubs. The idea that Mecca, for example, was an important trading city in preIslamic and early Islamic times has been effectively debunked by Patricia Crone (1987).

8

WILLIAM FACEY: THE RED SEA: THE WIND REGIME AND LOCATION OF PORTS could account for a large volume of trade.5 Secondly, ship captains from India certainly did combine carrying pilgrims to Jiddah with other cargoes, as Wellsted (1836), for example, describes, but that too, while undoubtedly contributing to the economic rise of Jiddah would not in itself be enough to sustain Jiddah’s status as a hub of international commerce.6

The ships from Suez seldom proceed further than this port; and those from India are not suffered to advance to Suez. The master of a vessel from Surat, being driven one year too far north to enter the harbour of Jidda, proceeded to Suez, and there discharged his cargo. But he was put into prison, next year, at Jidda, and obliged to pay the full dues that would have been charged at Jidda, upon the goods which he had disposed of at Suez (Niebuhr 1792, i: 235).

So it is clear that the pilgrimage did serve to some extent to make Jiddah a focus of trade. However, the pilgrimage takes place during only part of one month of the year. Furthermore, that year is a lunar one, so the timing of the pilgrimage precesses through the seasons by ten days or so each solar year, with the result that at least half of the pilgrimages in a 36-year cycle will fall outside the October–March sailing season, in the latter part of which ocean-going vessels find it possible to sail north into the Red Sea. No doubt the pilgrimage enriched the Jiddah merchants to the extent that for the remainder of the year they had some resources available for engaging in trade. That too would have contributed to the virtuous cycle by which Jiddah became a prosperous port. Nor should we ignore the economic advantage the Hijaz cities derived from the pious patronage of the rulers of Egypt and Muslim India, and the regular food subsidies that were sent there through Jiddah. Nonetheless, it is the author’s belief that it is wrong to think of Islam and the pilgrimage as being the only factors in Jiddah’s prosperity.

Burckhardt is equally emphatic. Having described how the coffee ships arrive at Jiddah from the Yemen all the year round and the fluctuations in the coffee trade, he goes on to describe Jiddah’s India trade in 1814: The fleets, principally from Calcutta, Surat and Bombay, reach Djidda in the beginning of May, when they find the merchants already prepared for them … Large sums are also sent hither by the Cairo merchants to purchase goods on their account; but the cargoes for the greater part are bought up by the merchants of Djidda, who afterwards send them to Cairo to be sold for their own advantage. The India fleets return in June or July … It is the nature of this commerce that renders Djidda so crowded during the stay of the fleet. People repair thither from every port on the Red Sea (Burckhardt 1829, i: 32-34).7

In other words, by the mid-eighteenth century (and perhaps earlier) Jiddah’s position in the India trade had been institutionalised, to its advantage, by an enforceable tax regime stipulating that customs dues on the India– Egypt trade were to be paid at Jiddah. That amounted to artificially imposed discrimination in favour of the Jiddah merchants, one might say, and a strong incentive for Jiddah to maintain good relations with Egypt. Indeed, Niebuhr’s description could be taken to imply that Jiddah was treated as if it were actually an Egyptian port, given that Egypt seems to have gone along with the enforcement of customs dues on the India trade there rather than at Suez. However, there have to be reasons why a regulation such as this can grow up and find acceptance in the first place. The author’s belief is that it reflected a very important natural factor in Jiddah’s growth as an entrepôt: its latitude in the wind system of the Red Sea.

Another way of looking at it is proposed. The pilgrimage explains the existence of a port at Jiddah or somewhere in the vicinity – after all, with Mecca being where it is, there had to be a port for pilgrims somewhere within reach. But the pilgrimage cannot wholly explain Jiddah’s prominence as a port. Why, then, should Jiddah have been more than simply a local port for the pilgrimage coming to life for a month a year, rather like the purely pilgrim-season settlement of Minā between Mecca and Jabal ΚArafāt? Both Niebuhr (1762) and Burckhardt (1814) give quite detailed accounts of the trade at Jiddah and make much of a feature that Wellsted does not mention. They make it very clear that Jiddah’s relative prosperity was based on its role as the Red Sea entrepôt between India and Egypt – a role that for example Aden and ΚAydhāb had played in previous times. Niebuhr writes:

Jiddah was as far north in the Red Sea as large oceangoing sailing ships could comfortably reach at the optimum sailing time – the end of the north-east monsoon season around April. The thing to know about the Red Sea is that it is very easy to sail out of it southwards for most of the year; and correspondingly difficult to sail northwards up it (see Figs 1–4: Red Sea winds during the four quarters of the year). That is because, in the northern half of the Red Sea above Jiddah, the prevailing wind blows from the north the

Although the trade of Jidda is so considerable, yet this city is no more than a mart between Egypt and India. 5

‘The commodities which are now brought to Jiddah from India are either disposed of during the hajj to pilgrims, who again distribute them through Turkey, Syria &c., or they are such as are required at Mecca, Jiddah, and other cities in Hejáz.’ (Wellsted 1836: 90). 6 ‘Ships seldom leave Bombay direct for the Red Sea, unless they are small, and intended for the coasting trade. If they obtain a sufficient number of pilgrims to defray the greater part of the freight, they ballast with sugar; but the usual practice is to proceed to the Malabar coast, where they take in cargoes of the same articles as the Bengal ships …’ (Wellsted 1836: 90).

7

Burckhardt‘s observation of the merchantmen from India reaching Jiddah around the beginning of May, at the end of the NE monsoon sailing season, finds corroboration in Ahmad ibn Mājid’s Kitāb alfawāΜid, where it is stated that March–April is the best time for sailing from the Indian coast to Arabia (Tibbetts 1971: 225, 230–1).

9

TRADE AND TRAVEL IN THE RED SEA REGION

Figure 2: Red Sea winds in July. Mean barometric pressure (mb) and dominant winds (mean force). Normal position of Intertropical Convergence Zone (ITCZ)

Figure 3: Red Sea winds in October. Mean barometric pressure (mb) and dominant winds (mean force). Normal position of Intertropical Convergence Zone (ITCZ) 10

WILLIAM FACEY: THE RED SEA: THE WIND REGIME AND LOCATION OF PORTS

Figure 4: Red Sea winds in January. Mean barometric pressure (mb) and dominant winds (mean force) whole year round. In the southern half of the Red Sea the wind blows from the north for most of the year. It is only during a relatively short time during the period from October to March/April – the season of the north-east monsoon winds in the Arabian Sea – that a southerly wind blows in the southern half of the Red Sea. It blows reliably only as far north as the latitude of Jiddah and of ΚAydhāb on the African side (Royal Navy 1980: 19-24).

yard round from one side of the mast to the other while you wear round before the wind.9 Hence it is not as if the northern half of the Red Sea was completely unnavigable for those trying to voyage northwards. However, it would have required muscular seamanship and special local navigational knowledge of weather, winds and coastal hazards as one zigzagged laboriously northwards. Thus, for captains of ocean-going trading dhows, it made sense not to attempt to go north of Jiddah if there were a port that could relieve you of the inconvenience of having to do so.

Smaller coasting vessels could take advantage of off- and on-shore breezes to go further up the Red Sea at most times of the year, so long as they were not in too much of a hurry. Most such voyages were coastal, and made only by day, with stops in creeks and on beaches overnight. This was usually done even when sailing south, as is demonstrated by Wellsted’s 1831 voyage. Fleets of Ottoman galleys made their way from Suez all the way down the Red Sea and back again, but the significant thing about a galley is that it can be rowed, and one suspects that getting back to Suez often required a very great deal of elbow grease.

It is this fact, that it is easy to sail south out of the Red Sea but hard to sail north, that provides some explanation why, in antiquity and Islamic times, ports on the Egyptian side show a tendency to be some way down the coast. Clysma/Suez at the far northern end might have been in an obvious position geographically, but in navigational terms it was ill suited because of the difficulties of sailing into the wind. The author believes that this is why we find, under the Ptolemies and Romans, Myos Hormos and Berenike (Ras Banas), both of them quite a way down the coast, developed as ports and served by well-maintained routes from the Nile Valley.10

Larger ships that were prepared to brave the prevailing north wind in the open sea could resort to takkīyyah, as Ibn Mājid describes for the southern half of the Red Sea, which probably means making way against the wind in a general sense.8 In a dhow, sailing into the wind necessitates the very arduous process of working the sail

9

For the principles and techniques of wearing round stern-to-wind in a dhow, and the inadvisability of trying to tack into the wind, see Facey (1979:115–16). 10 See e.g. Ball (1942: 83–84) for the location of well-maintained staging-posts or hydreumata for the twelve day journey between

8

Tibbetts (1971:310–12) deals at length with takkīyyah in the Red Sea; see also John Carswell’s note in Hourani (1995:149).

11

TRADE AND TRAVEL IN THE RED SEA REGION ΚAydhāb was about 120 miles farther south down the coast from Berenike, and the letters show that the organisation of trade to ΚAydhāb must have been similar to that in classical times between Coptos and Berenike. Merchants went up the Nile to Qus, a few miles south of Qift/Coptos, and then undertook the seventeen day journey overland to ΚAydhāb via a more or less insecure route provided with wells.14

ΚAydhāb Similarly, the main port of medieval trade on the African side of the Red Sea was ΚAydhāb, even further south than Berenike, on a latitude just north of Jiddah. ΚAydhāb seems to have all but vanished from the map. Its site is north of Marsa Halaib and just south of Ras Abu Fatma at 22˚ 20’ N, 36˚ 32’ E, inside Sudan.11 Although ruins are mentioned in the Admiralty Handbook Western Arabia and the Red Sea and plotted on a map, as far as the author has been able to ascertain there is very little to be seen in this area today to suggest that this was, from the eleventh to the fourteenth century,12 the most important Red Sea entrepôt in the trade between India and Egypt.

Ibn Jubayr, making the pilgrimage in 1183, tells us of an alternative route which went from Qus further up the Nile to Aswan, before taking to the land and joining the Qus– ΚAydhāb route at the well of Dinqash (Broadhurst 1952: 59-60). He also provides much detail on travelling the overland Qus–ΚAydhāb route and on ΚAydhāb itself, which he reached towards the end of June 1183. In the following extract he shows ΚAydhāb to have been largely built of Κuššah-type dwellings, perhaps not dissimilar to those found on the Tihāmah coast of Saudi Arabia and Yemen, which might explain why surface settlement remains have not been noted, and also explicitly states that it was a port of the India trade as well as for pilgrims crossing to Jiddah:

ΚAydhāb’s position obviously had something to do with the location of Jiddah, for as much as anything else it served as a port for pilgrims from Africa. However, in the one set of sources we have in which ΚAydhāb is frequently mentioned, it is clear that its chief role was in the trade between India and Egypt. These sources are the correspondence of Jewish merchants found in the geniza, or storeroom, of a synagogue in Fustat, old Cairo, dating from the eleventh to the early thirteenth centuries. An intriguing example of medieval multiculturalism, these letters were written in the Arabic language using Hebrew script.13

…we entered ΚAydhāb, a city on the shores of the Jiddah sea. It has no walls and most of its houses are booths of reeds. It has, however, some houses, newlybuilt, of plaster. It is one of the most frequented ports of the world, because of the ships of India and the Yemen that sail to and from it, as well as the pilgrim ships that come and go. It is in the desert, with no vegetation and nothing to eat save what is brought to it. Yet its people, by reason of the pilgrims, enjoy many benefits, especially at the time of their passing through, since for each load of victuals that the pilgrims bring, they receive a fixed food tax, light in comparison with the former customs duties which, as we have said, have been raised by Saladin. A further advantage they gain from the pilgrims is in the hiring of their jilab: ships which bring them much profit in conveying the pilgrims to Jiddah and returning them when dispersing after the discharge of their pious duty. There are no people of easy circumstances in ΚAydhāb but have a jilabah or two which bring them an ample livelihood (Broadhurst 1952:62-63).

Coptos/Qift and Berenike according to Pliny the Elder in the first century AD. 11 On the location of ‘Aydhāb see Naval Intelligence Division 1946:108 and map facing page 107; although elsewhere in the same publication (p. 245, n. 1), it is stated that ‘authorities do not agree about its position’. Broadhurst (1952: 371, n. 35) places it 12 miles north of Halaib at 22˚ 20’ N, 36˚ 32’. Kammerer (1936: 119–20) places ‘Aydhāb at Marsa Halaib itself, at 22º 15’ N. The Red Sea and Gulf of Aden Pilot (Royal Navy 1980), which usually mentions coastal ruins if they are to be seen, e.g. Berenike (Royal Navy 1980: 104), makes no mention of ruins either at this spot (p. 107), or at Marsa Halaib (pp. 107–8). Goitein’s statement (1973: 183, n.8) describing ‘Aydhāb as ‘corresponding approximately to Port Sudan of today’ is misleading in implying that ‘Aydhāb was located at or near Port Sudan. 12 According to Broadhurst (1952: 371, n. 35), when ‘Aydhāb was destroyed in AD 1422 by the ruler of Egypt, Suakin took over its role in the Red Sea trade. According to Naval Intelligence Division (1946:108, 245), Suakin [Sawākin] flourished between 1058 and 1368, and sank into oblivion ‘after Indian trade reverted to the north of the Red Sea at the end of the fourteenth century’. For a brief history of Suakin, see Greenlaw 1995:13–16. On the rise of Suakin, Greenlaw writes: ‘Between the years 1048 and 1300 Suakin had a powerful rival in the Port of Aidhab, a few hundred miles [actually c. 230 miles] north and nearer to Jedda. This rivalry came to an end with the fall of Aidhab, along with Jedda, as a result of misappropriation by their inhabitants of merchandise from Egypt destined for Mecca. From then on, Suakin became the principal port for Egypt on the African coast of the Red Sea until the building of Port Sudan at the beginning of the present [i.e. 20th] century.’ 13 The author has referred only to Goitein (1973) for the purposes of this article. No doubt much more on ‘Aydhāb can be gleaned from his fivevolume A Mediterranean Society: the Jewish Community of the Arab World as Portrayed in the Documents of the Cairo Geniza (1967-1983). It is most unfortunate that Goitein died before completing the ‘India Book’ to which he frequently refers.

Ibn Jubayr then goes on to describe the local pearl fishery, the dangers of missing ΚAydhāb when returning by sea from Jiddah, and the method of boat construction used: as one would expect at this period, the jilāb were of stitched construction. Timber was brought from India and the Yemen, and the cordage for stitching was made from coconut fibre. Sails were woven from the leaves of the muql tree, according to Broadhurst ‘the Theban palm or bdellium, a kind of gum-tree’. The local boat-owners had a nasty habit of packing in as many pilgrims as they possibly could for the voyage to Jiddah, ‘like chickens crammed in a coop’ as Ibn Jubayr says, excusing 14 See Goitein (1973: 65–70, 181–85, 197–99, 209–10), for a description of the route, dangers from brigands, and the merchandise carried in the twelfth and early thirteenth centuries.

12

WILLIAM FACEY: THE RED SEA: THE WIND REGIME AND LOCATION OF PORTS antics of the Crusader lord of Karak and Shawbak, Reynauld de Châtillon, who in 1181 had contrived to build and launch a small fleet of five galleys and various smaller craft at ΚAqabah. Reynauld’s plan was said to have been to raid the Holy City of Medina and remove the body of the Prophet Muhammad from his tomb. In this he failed, but in 1182 he did manage to terrorise the coasts of the Hijaz, and even at one point ravaged ΚAydhāb itself, slaughtering a merchant caravan and taking ships. This arresting sideshow of the Crusades came to an end only in 1183. Reynauld‘s mastery of southern Jordan had rendered the land route from Egypt to the Hijaz temporarily impassable for Muslims, thereby making ΚAydhāb for the time being the preferred route.16

themselves by saying: ‘Ours to produce the ships: the pilgrims’ to protect their lives.’ (Broadhurst 1952: 65). As with Jiddah, the location of such an important port at ΚAydhāb can only partly be explained by its relation to the pilgrimage. Its position in the wind regime gave it an advantage in the India trade, by enabling traders to bypass the troublesome northern half of the Red Sea. In the end it seems that it could not compete with both Jiddah and Suakin, despite Suakin’s lack of ΚAydhāb’s advantage in being linked by an overland route to the centres of Egyptian population on the Nile. By 1500 and perhaps before, ‘Aydhāb had been forgotten: it is not even mentioned by Ibn Mājid in c. 1500, nor by João de Castro in 1541.15

In the summer of 1183 therefore there were extra reasons for Ibn Jubayr to be apprehensive as he embarked. Winds were light at first but on the evening of the second day, when they were almost within sight of the Hijaz, a terrible storm got up and blew them two days off course along the coast, probably to the north of Jiddah. They took refuge on an island, where they stayed for a day, as the wind was contrary. On the evening of the sixth day they anchored in Ubhur creek, about nine miles north of Jiddah, and today the favourite seaside resort of Jiddah’s well-to-do. On the seventh day they sailed close to Jiddah but were unable to put in on the following morning because the wind got up again. Eventually at midday on the eighth day, 26 July, they managed to put in. At this point Ibn Jubayr is full of praise for the skill of the sailors:

Given the importance of the Jiddah–ΚAydhāb latitude to the long-distance India trade, we may draw the conclusion that the pilgrimage only explains why the major commercial port of the Red Sea during Islam was on the Arabian and not on the African side. Islam changed the pattern of the centuries-old India trade by making the Hijaz coast of the Red Sea a destination for the first time in its own right; in doing so, it raised Jiddah to a position where it could act as a magnet for the India trade too. The experiences of mariners, twelfth to sixteenth centuries AD. We have had a glimpse of how an early nineteenth century British sailor experienced the northern half of the Red Sea above Jiddah. Now let us look at a selection of medieval voyages, to see what light they shed on the experience of sailing in the Red Sea, the skills required to do so, and the idea that the Red Sea was actually two seas as far as sailors were concerned.

The entry into it [Jiddah] is difficult to achieve because of the many reefs and the windings. We observed the art of these captains and the mariners in the handling of their ships through the reefs. It was truly marvellous. They would enter the narrow channels and manage their way through them as a cavalier manages a horse that is light on the bridle and tractable (Broadhurst 1952: 69).

Ibn Jubayr’s voyage from ΚAydhāb to Jiddah, July 1183

What he describes here is precisely the kind of very detailed knowledge of reefs and other coastal features that ship-masters and pilots would need to sail a dhow anywhere along the Red Sea coast of Arabia and Africa, particularly in the northern half where the prevailing northerlies in the open sea made northward voyages so difficult.

Ibn Jubayr finally embarked for Jiddah in mid-July on one of the frail, overcrowded jilāb he describes. Although his voyage was a short one across the Red Sea and relates to the pilgrimage rather than the India trade, it is worth describing because it provides a vivid insight into the perils of Red Sea sailing and the skills needed to negotiate its reef-ridden coasts (Broadhurst 1952: 66-70).

AΉmad bin Mājid’s instructions for mariners, c. 1500

Ibn Jubayr himself much regretted that he had to make a sea voyage at all. He warns pilgrims off going via ΚAydhāb (Broadhurst 1952: 55-56), and it seems that he himself had been compelled to go that way only by the

AΉmad ibn Mājid al-Najdī, the greatest of all Arab navigators, compiled his massive Kitāb al-fawāΜid (‘Book of Profitable Things’) around the end of the fifteenth century. An early example of what we would call a Pilot and the Portuguese a Roteiro, its aim was to

15 Tibbetts (1971: 413) puts the northernmost point on the Sudanese coast mentioned by Ibn Mājid at RaΜs DawāΜir which he identifies as Ras Abu Shagarah, 21º 04’ N, 37º 19’ E, i.e. well south of ‘Aydhāb‘s location. See also Tibbetts’ folding map in pocket at the back of the book entitled ‘The Red Sea and Gulf of Aden‘, and Red Sea and Gulf of Aden Pilot (Royal Navy 1980: 109). De Castro makes no mention of ‘Aydhāb at this part of his voyage (Kammerer 1936:109).

16 On Reynauld‘s adventures in AD 1181–3 see Broadhurst 1952: 51–3, 371 n. 31; Naval Intelligence Division1946: 245; Wright 1923: 294–95, 472, n. 190. Moritz (1923: 119–20) gives the sources for Reynauld‘s Red Sea campaign, which seem to be entirely Arabic ones.

13

TRADE AND TRAVEL IN THE RED SEA REGION

Figure 5: The Red Sea and Gulf of Aden according to Ibn Mājid, after Tibbetts 1971 give comprehensive sailing instructions for the coasts of the western Indian Ocean, the Gulf and the Red Sea. But for the purposes of the argument presented here, its most significant aspect is what it leaves out.

ruin for Venice, to realise the damage the Portuguese would inflict. In one of his last poems – he died in about AD 1505 – he laments their arrival as a catastrophe (Shumovsky 1957: facsimile, ff. 92v-93v.).

As a consummate long-distance navigator, Ibn Mājid was at pains to record everything he knew for his fellow practitioners in the India trade, and the southern half of the Red Sea is covered in detail. But the very farthest north that he describes is Jiddah (Tibbetts 1971: 35, 24368, 398-421). It seems that for him the northern half of the Red Sea simply does not exist beyond Jiddah. We can conclude that for Ibn Mājid and other ocean-going mariners the northern half of the Red Sea was another clime altogether; they never went there and as far as they were aware, alien conditions prevailed and different rules applied (see Fig. 5).

Thus, though he seems never to have sailed north of Jiddah, he undoubtedly knew that it was a vital link in the international India trade with Egypt and the Mediterranean, not just a destination in its own right. But it was not seen as the role of Indian Ocean captains to carry their cargoes beyond it. Ibn Mājid and his like were experts in the monsoon wind regime of the Indian Ocean, and the northern Red Sea fell outside that sphere. Afonso de Albuquerque’s voyage into the Red Sea in 1513 Fortunately for the local traders, the Portuguese succeeded only partially in their designs. Their great empire-builder Afonso de Albuquerque arrived soon after Vasco da Gama, and judged at once that there were three key points in the eastern trade: Malacca, Hormuz and Aden. The Portuguese were to succeed at Malacca and Hormuz, but Aden eluded their grasp. Paradoxically, as Albuquerque himself notes, the Portuguese disruption of

AΉmad ibn Mājid lived just long enough to witness the effects of the Portuguese irruption into the arena of Muslim sea-borne trade. One of his poems describes the astonishment the ifrānj aroused among the peoples of the Indian Ocean coasts. Most significantly, he was just as quick as the Venetian banker in Lisbon who in the early years of the sixteenth century predicted sudden financial 14

WILLIAM FACEY: THE RED SEA: THE WIND REGIME AND LOCATION OF PORTS the Egyptian side used for the India trade, and India business was transacted there by traders such as Jews who could not use Jiddah. Suakin too was important. On the basis of his new intelligence he decided that the Portuguese ought to return to capture Aden and establish their main Red Sea base at Massawa (Earle & Villiers 1990: 239-241, 247, 259-261).

the India trade had given Aden a considerable boost: cargo ships were no longer able to reach the Red Sea at the right season, and so could not reach Jiddah. Jiddah suffered, while the ships took to off-loading at Aden and the further shipment of their goods was assigned to the Aden merchants and agents (Earle & Villiers 1990: 257). Albuquerque made his attempt on Aden in March 1513.17 Having failed there he decided to sail into the Red Sea to investigate it for potential bases – this despite the fact that he knew, presumably from local pilots whom he always took pains to consult, that ‘the easterly monsoon was nearly over’ and that therefore time for sailing into the Red Sea was getting short. Professional pilots (rubbān), whom he specifically says charged for their services as far as Jiddah, were to be had at the mouth of the Red Sea, and were duly taken on board (Earle & Villliers 1990: 221-223, 243).

Albuquerque had learned the hard way that it is no good trying to sail north up the Red Sea, even up its southern half, outside the north-east monsoon season of the western Indian Ocean. He had also learned how very delicate the seasonal pattern of the India trade was, and hence how easily disrupted: if the ships missed the right sailing season, economic disaster followed for the chief ports of the trade. All sorts of news reached his ears about how Jiddah and Mecca were suffering, how there were disturbances at Cairo, and how Aden was flourishing as a result, chiefly it seemed at the expense of Jiddah. Albuquerque knew that Jiddah was prosperous because of Mecca and the pilgrimage. But one can also deduce from his account that it played the pivotal role in the India trade to and from Egypt. Now that ΚAydhāb was gone, it alone provided the gate through which goods passed between the upper and the lower Red Sea.

They made first for Kamarān Island, which Albuquerque had hopes of as a base because it had fresh water. There among others they found two large merchantmen ‘belonging to the Sultan of Cairo’, but it becomes clear that they were based at Jiddah and not in Egypt, so supporting the argument put forward in this paper: ‘Their factor, who is in Jiddah, uses them on the Sultan’s business’ (Earle & Villiers 1990: 227, 237). At this point he says: ‘Though the easterly monsoon was coming to an end, my Muslim pilots and the rubbans, who are local men, led me to expect that we would have easterlies to carry us to Jiddah, Suez and al-Tur, and urged me to take on water as quickly as possible.’ He soon found that he had been misinformed: ‘Days went by and I saw, greatly to my annoyance, that there was no east wind.’ Instead the wind was veering to the west and north, and Albuquerque felt that his pilots had deceived him. They set sail and fretted off some islands near Kamarān for more than three weeks until their water ran out, waiting for easterlies which never came. There was talk of going to Massawa and Dahlak instead, reinforced by a convenient vision of a cross in the sky over the land of Prester John, but nothing came of these plans. At the end of May they returned to Kamarān, having given up hope of further progress. The fleet then spent the summer there preparing for their return voyage. They left for Aden on 15 July, aiming to be able to depart from there for India ‘at the right time … which is four days after the August full moon’ (Earle & Villiers 1990: 249, 253, 261).

He collected some most interesting information on navigational lore in the Red Sea. Muslim seamen divided it north-to-south into three zones of equal width, one each along the coasts, the third up the centre. Each zone was considered to be a day’s sail wide. Both coastal zones were full of reefs, islands and shallows, making for difficult sailing and demanding the hire of local pilots. The central zone, the open sea, was not the sphere of local pilots, unless contrary winds made tacking from coast to coast necessary. ‘In the middle of the Red Sea the pilots who have made the voyage from India are in sole charge of the ships and the course … The ships going to Jiddah sail through the open sea, past a group of islands in the middle called Jabal Zukar and another nearer to Jiddah called Ceibão.’ Jabal Zukar is Ibn Mājid‘s Zuqar (today still Jabal Zuqar), and Ceibão his Saybān. Saybān is today known as Jazīrat al-TāΜir, and was perhaps the most important landmark in the whole of the Red Sea as far as Ibn Mājid was concerned (Tibbetts 1971: 419-420, Map). Thus it is noteworthy how closely Albuquerque‘s information tallies with Ibn Mājid’s. But the really significant point is that the long-distance navigators making the voyage up the middle of the Red Sea are nowhere said to be going beyond Jiddah – further corroboration that the northern Red Sea was perceived as a zone where other sailing conditions prevailed (Earle & Villiers 1990: 243-245).

On Kamarān he made a point of discovering as much as he could about the navigation of the Red Sea and the lands on either side of it, especially about Jiddah, Massawa and Dahlak. Jiddah was clearly the chief entrepôt. As for Suez, it was ‘once a great city; now it is deserted, but its large buildings, all in ruins, suggest that it once was an important place’. Quseir was the port on

The Red Sea voyage of João de Castro to Suez and back in 1541

17

In 1517, soon after Albuquerque’s attempt on Aden and the Red Sea, the Ottoman Turks took power in Egypt and

See Naval Intelligence Division (1946: 259) for a summary of Albuquerque’s doings in 1513. Otherwise this account of his Red Sea voyage is taken from Earle & Villiers (1990: 221–61).

15

TRADE AND TRAVEL IN THE RED SEA REGION The fleet set out from Goa with variable breezes on 31 December 1540. From Socotra they sailed past Aden on 27 January 1541 with an easterly breeze, and reached the Bab al-Mandab two days later, anchoring on the African side. As they set out from there, the easterlies veered, true to form, round to the south-east, providing a following breeze. In early February, with variable easterly and south-easterly breezes, they sailed on, passing the Dahlak Islands and reaching Massawa on the 9th, staying there till the 18th. At this point de Castro pauses to record much of interest about the Negus (‘Prester John’) and the manners and customs of the Abyssinians. The fleet’s big ships were left at Massawa to await the return of the smaller vessels from Suez under da Gama and de Castro.

the Hijaz, taking control of Yemen and Aden too. Although the Ottoman presence in the Red Sea was never more than patchy, their arrival did seem to have a deterrent effect on any sustained Portuguese influence thereafter. In 1538 João de Castro, future Viceroy of India (1546– 47), cousin of the then Viceroy Garcia de Noronha and all-round Portuguese renaissance man, arrived in India just in time to help mop up after the unsuccessful Ottoman naval assault on the Portuguese at Diu in Gujerat. The Ottoman fleet, under the command of the eunuch Suleyman Pasha had sailed back to Suez having failed to dislodge the Portuguese from any of their existing strongholds. But they had at least managed to take Aden once again and, as they thought, secure the Red Sea against the dreaded Christians.18

As they left Massawa on February 20th, there came the first hint of difficulties ahead: a light breeze from the north-east forced them to take to their oars for the first time. They were able to set sail once it veered round to the east, but then they were becalmed, having to row once more. A mixture of rowing and sailing brought them off Suakin on March 1st, where they spent eight days before struggling on past Port Sudan (‘Dradate’), Quseir (which they ravaged), and al-Кūr in Sinai.

The Portuguese response was to prepare a large fleet for the Red Sea under Estevão da Gama, Vasco’s son, with several aims in view: •

To punish the Ottomans by taking their Red Sea bases (in this they were to fail);

•

To attack Suez and bombard the Ottoman fleet and arsenal (they reached Suez but retired in the face of superior odds);

•

To return Ethiopian ambassadors to their homeland (this they accomplished).

•

To help the Negus against his enemies (which they also succeeded in doing).

To cut a long story short, they did not reach Suez until the 27th of April 1541. They had got there mostly by rowing into more or less contrary breezes. Only twice did they get a day of sailing, around the latitude of ΚAydhāb on the 8th of April and off Ras Banas (ancient Berenike) on the 10th. On seeing the size of the Ottoman fleet at Suez (‘41 ‘royal ships’ and five large vessels’) and the defences, they decided that discretion was the better part of valour and prepared to retire (Kammerer 1936: 153154). On leaving Suez on the 28th of April de Castro sighs: ‘ … we were at death’s door from exhaustion.’ (Kammerer 1936: 160). They had fallen foul of the unpredictable on-shore and off-shore breezes of the coasts and had had reefs and sand-banks to contend with as well.

João de Castro was captain of one of the large ships in the fleet. For our purposes his record of the expedition is of the greatest significance. He did not write a detailed account of events, which we have from other sources, but a day-by-day Roteiro of such thoroughness that it must count as the first scientific European exploration of the Red Sea. As a mark of his wide erudition he refers frequently to the writings of Pliny the Elder and Ptolemy. The Portuguese fleet sailed out and back along the entire African coast of the Red Sea, calling at all the ports, making this is a very valuable record indeed.

It had taken them 66 days to get from Massawa to Suez. It took them a mere 25 to make the return voyage, almost entirely under sail, to Massawa where, on the 22nd of May, they rejoined the fleet and remained until the 9th of July for a well-earned rest. By the 9th of August they were back in Goa.

De Castro recorded the wind conditions every day, systematically noting when they had to row and when they could sail. The fleet did not venture into the open sea, the preserve of the ocean-going Arab navigators, so we learn nothing of conditions there. Nonetheless, we are able to draw conclusions about sailing conditions in the Red Sea.

The Portuguese had gained a very great deal of detailed knowledge about the Red Sea’s African coast, but it seems they were never tempted to pay a return visit to this difficult waterway. They had entered it at a good time (January) for sailing north, but did not have the navigational know-how to take advantage of the prevailing southerly wind in the open sea, sticking instead to the coastline and its erratic breezes. Even if they had braved the open ocean, they would have found that the southerlies would not have carried them much north of Jiddah. So in the northern half of the Red Sea

18

See Naval Intelligence Division (1946:260–61) for a summary of the eunuch Suleyman Pasha‘s proceedings in 1538 and the 1541 Portuguese expedition to Suez. Otherwise the detail of this account is drawn from Kammerer (1936).

16

WILLIAM FACEY: THE RED SEA: THE WIND REGIME AND LOCATION OF PORTS Greenlaw, J.-P. 1995. The Coral Buildings of Suakin. London: Kegan Paul International. Hourani, G.F. 1995. Arab Seafaring in the Indian Ocean in Ancient and Early Medieval Times. (Second edition) (ed. John Carswell). Princeton/Chichester: Princeton University Press. Kammerer, A. (ed. and transl.) 1936. Le Routier de Dom Joam de Castro. L’Exploration de la Mer Rouge par les Portugais en 1541. Librairie Orientaliste. Paris: Paul Geuthner. Moritz, B. 1972. Arabien. Studien zur physikalischen und historischen Geographie des Landes. (Reprint of 1923 edition). Osnabrück: Biblio Verlag. Naval Intelligence Division. 1946. Western Arabia and the Red Sea. Geographical handbook series, B.R. 527. London. Niebuhr, C. 1792. Travels through Arabia and other Countries in the East. (2 volumes). (transl. R. Heron). Edinburgh: R. Morison and Sons. Pesce, A. 1977. Jiddah. Portrait of an Arabian City. Falcon Press. Royal Navy, UK. 1980. Red Sea and Gulf of Aden Pilot (12th edition). Taunton: Hydrographic Department. Searight, S. 1979. The British in the Middle East. London: East-West Publications. Shumovsky, T.A. 1957. Tri neizvestnye lotsii Akhmada ibn Madzhida, Arabskogo lotsmana Vasko da-Gamy, v unikal’noi rukopisi Instituta vostokovedeniia AN SSSR. Moscow: Izd-vo Akademii nauk SSSR. Tibbetts, G.R. (ed. and transl.) 1971. Arab Navigation in the Indian Ocean before the Coming of the Portuguese, being a Translation of Kitāb al-fawāΜid fī usūl al-bahr wa Μl-qawāΜid of AΉmad b. Mājid alNajdī. London: Royal Asiatic Society. Treat, I. 1930. Pearls, Arms and Hashish. Pages from the Life of a Red Sea Navigator, Henri de Monfried. London: Gollancz. Wellsted, J.R. 1836. Observations on the Coast of Arabia between Ras Mohammed and Jiddah. London Geographical Journal (= Journal of the Royal Geographical Society) 6: 51–96. Whitcomb, D.S. & Johnson J. H. 1979. Quseir al-Qadim 1978. Preliminary Report. (American Research Center in Egypt Reports 1). Cairo/ Princeton: American Research Center in Egypt. Wright, J.K. 1923. The Geographical Lore of the Time of the Crusades. New York: American Geographical Society.

they would have had to resort in any case to the laborious coastal method which had given them so much trouble: rowing by day and finding an anchorage by night. Conclusion As a final illustration of the difficulties of heading north along the northern stretch of this coast, let us leave the last word with the French dhow master, smuggler and allround seadog Henri de Monfried. Making his way north from Djibouti and Massawa in July 1915 – admittedly not a sensible choice of season for the voyage – with a cargo of hashish destined to be landed near Suez, he writes: From Koseir on we had the wind against us, the northwest wind which at all seasons sweeps the upper half of the Red Sea. It blew with tempest strength, belabouring us with short, vertical seas. One would have said as much water passed over the deck as under the keel. We beat our way painfully day and night, in interminable zigzags; soaked in our wet cottons, and suffering from cold as a few days earlier we had suffered from the sun. Each tack gained barely one-fourth of its length in the wind. In twenty-four hours we made only nine miles (Treat 1930:184).

Little wonder, then, that there has been a tendency throughout history for Suez to be replaced as a commercial port by locations farther south on the African side of the Red Sea. Quseir was ideally placed, being almost as close to the Nile Valley as Suez. But for ships sailing north, even Quseir could be improved upon, and the farther south the port for Egypt was, the better. This was undoubtedly an important factor in the emergence of ports in otherwise seemingly unpromising locations, as far south as ΚAydhāb. Only with the coming of steam in the 19th century did the wind regime of the Red Sea become irrelevant to the location of ports, and it is only from then that Suez rose to pre-eminence. References Ball, J. 1942. Egypt in the Classical Geographers. Cairo: Government Printing Press, Bulaq. Broadhurst, R.J.C. 1952. The Travels of Ibn Jubayr. London: Jonathan Cape. Burckhardt, J.L. 1829. Travels in Arabia. (2 volumes). (ed. W. Ouseley). London: H. Colburn. Crone, P. 1987. Meccan Trade and the Rise of Islam. Oxford: Blackwell. Earle, T.F. & Villiers J. (ed. and transl.) 1990. Albuquerque, Caesar of the East. Selected Texts by Afonso de Albuquerque and His Son. Warminster: Aris and Phillips. Facey, W. 1979. Oman – A Seafaring Nation. Oman. Muscat: Ministry of Information. Goitein, S.D. (ed. and transl.) 1973. Letters of Medieval Jewish Traders. Princeton: Princeton University Press. 17

18

Arabian trade with Ethiopia and the Horn of Africa: from ancient times to the sixteenth century Richard Pankhurst still further east, had a ‘harbourless shore’ (Periplus 1980:24).

Ancient trade and ports Trade across the Red Sea and Gulf of Aden probably dates back to the dawn of history and, being governed by the largely unchanging facts of geography, maintained a remarkable degree of continuity over the centuries. Evidence of the antiquity, and extent, of this commerce is provided by the Periplus of the Erythraean Sea, a Graeco-Egyptian trade manual dating from the first century AD. This work indicates that ports on both sides of the Red Sea and Gulf of Aden were in close commercial contact across the sea. The principal ports on the Arabian side, as mentioned in the Periplus (and mapped by the subsequent ancient cartographer Ptolemy) were Muza and Okalis. Muza is identified with modern MawzaΜ, twenty-five miles north of Mocha. Muza is described in the Periplus as ‘an established mart beside the sea ... full of Arabs, shipmasters and sailors’. The port, we are told, hummed with business, for the traders used their own ships ‘for commerce with the opposite coast’, i.e. the Horn of Africa as well as with Barugaza (Broach), in north-west India (Periplus 1980: 32, 102). Okalis lay further south-east near the southern tip of Arabia, opposite, and about four miles away from Perim Island (Periplus 1980: 23, 102).

Ancient exports from Arabia Arabia supplied the above ports on the African coast with a variety of manufactured articles. Some were made in Arabia itself, while others were shipped there from India and other lands further afield. Goods from Arabia, many of them specially made for export, were manufactured at Muza, and included spears, small swords, axes, awls and several kinds of glassware. Articles from India transshipped from Arabia comprised mainly cotton, silk and other textiles (Periplus 1980: 30). Ancient imports from Arabia In exchange for such manufactured goods the ports on the African side of the Red Sea and Gulf of Aden, and their extensive African hinterland, supplied Arabia with a variety of raw materials, and slaves. Adulis thus exported ivory, rhinoceros horns and tortoise shell, in exchange for imported cloth, spears, axes and swords (Periplus 1980: 21-22, 30). Avalites exported spices, some ivory and a little myrrh, but better, the Periplus claims, than that obtained elsewhere. Some of these exports were transported on rafts navigated by the local people of the area and were shipped to both Muza and Okalis (Periplus 1980: 23, 102). Exports from Malao likewise included myrrh, as well as incense, some slaves, and makeir, an Indian medical root transshipped to the port (Periplus 1980: 23-24, 134).

The ports on the African side referred to in the Periplus were Adulis, or Adule, the principal port of the Aksumite empire and south-east of modern Massawa; and along the more arid northern Horn of Africa coast, from west to east the two no less important ports of Avalites, or modern ZaylaΚ, and Malao, or modern Berbera. Further east again, towards the mouth of the Gulf, and just beyond it, were four lesser ports: Mondou, modern Hays or Bandar Harshau; Mosullon, possibly modern Bandar Kassim; the so-called Market of Spices, which cannot be identified; and Opone, or modern Ras Hafun.

Goods exported from ports to the east were not dissimilar from those to Avalites and Malao. Mondou is said to have thus exported the same kind of goods as those at the above-mentioned ports, as well as fragrant gum. Mosullon exported similar goods, in addition to a large amount of cassia, or cinnamon, apparently transshipped to the port from eastern Asia (Periplus 1980: 24).

Adulis lay by an inland stretch of territory inhabited by a ‘quantity of elephant and rhinoceros’, which were also ‘occasionally seen’ by the nearby sea (Periplus 1980: 20, 87-89). Avalites, further east, was situated near the shortest sea crossing from the Arabian coast and could therefore be easily reached from across the sea in rafts and small boats (Periplus 1980: 23, 90). Malao had a harbour ‘exposed to the sea’ but ‘sheltered by a projection running out from the east’ (Periplus 1980: 23, 90-94). Mondou, further east again, had a good anchorage where ships could ‘anchor more safely by an island lying very close to the land’, whereas Mosullon,

The Christian Topography Trade between Arabia and the opposite coast of Africa continued on similar lines for the next five hundred years. This is evinced, albeit briefly, in another GraecoEgyptian work, the Christian Topography of Kosmas Indicopleustes which dates back to the early sixth century AD. This text recalls that Ethiopia abounded in elephants with large tusks which the inhabitants exported 19

TRADE AND TRAVEL IN THE RED SEA REGION ZaylaΚ – the Avalites of the Periplus. Medieval documentation on the trade of this port dates back to the tenth century, when the Arab geographer Ibn Дawqal stated that ZaylaΚ served as a place of embarkation for Christian Abyssinian slaves who were taken across the Red Sea to the great Yemeni slave depot at Zabīd. The latter area, he states, also imported panther skins and oxhides from the Abyssinian side of the sea (Ibn Дawqal 1964, i: 54). Two centuries later the twelfth century Arab geographer al-Idrīsī noted that ZaylaΚ was visited by ‘many foreigners’ and exported slaves (probably largely to Arabia) and, on occasion, gold (al-Idrīsī 1866: 29-30).

by sea to various countries, including the Homerite [Himyarite] country, i.e. southern Arabia. The Red Sea that separated the Asian and African continents was at that point, according to the Topography, so narrow that it could be ‘crossed in a couple of days’. The Horn of Africa, which Kosmas refers to as Barbaria, continued to be in close commercial contact across the sea with Arabia. Merchants of Barbaria, according to Kosmas, thus travelled far into the Ethiopian interior and returned with ‘many kinds of spices’, including frankincense, and ‘other articles of merchandise’ some of which, it is expressly noted, were shipped by sea to the Homerite country, presumably as had been the case since the time of the Periplus five hundred years earlier (Kosmas 1929: 51-52, 372).

Berbera, the Malao of ancient times, was likewise much involved in commerce across the Gulf of Aden and was described by the thirteenth century Arab writer Ibn SaΚīd as a major source of slaves shipped to Arabia (Devic 1893: 52-53).

Medieval commerce and the slave trade Early Arab sources suggest that the ensuing centuries witnessed significant commerce across the Red Sea and Gulf of Aden, and that many slaves were at this time exported from various parts of Abyssinia to Arabia. The nineteenth century British historian William Muir reminds us in his Life of Mahomet that caravans from Mecca, in the seventh century (as perhaps earlier), left for Abyssinia every year (Muir 1878: 212). One of those thus trading across the Red Sea was none other than the Prophet MuΉammad’s grandfather ΚAbd al-MuΓΓalib (Watt 1953: 30). Such trade, which included the shipment of slaves from the African side of the sea, led to the emergence at Mecca, and other parts of Arabia, of a sizable Ethiopian community. The best known of its members at this time was Bilāl, who will forever be remembered as MuΉammad’s muezzin who called the early Muslims to prayer, and who was referred to by the Prophet as ‘the first fruit of Abyssinia’.

The fifteenth and sixteenth centuries and the development of Aden Trade across the Red Sea and Gulf of Arabia continued to be important throughout the fifteenth and sixteenth centuries. This commerce was largely based on at least eight ports: half on the Arabian side and half on the African. Those on the Arabian coast, from south-west to north-east, were Aden, Mocha, Zabīd and Jiddah; those on the African, from north-west to south-east, were Massawa, Hergigo (also known as Arkiko), ZaylaΚ and Berbera (the two latter being the former ports of Avalites and Malao). Much of this inter-continental trade centred on the port of Aden, which enjoyed an important position in the trade of Arabia, India, and Africa. Visited by both Egyptian and Indian merchants, as the Moroccan traveller Ibn BaΓΓuΓa noted in the early fourteenth century, the port‘s merchants possessed ‘enormous wealth’ (Ibn BaΓΓuΓa 1962, ii: 372). This was subsequently confirmed by the mid-fifteenth century Venetian traveller Niccolò de’ Conti who called Aden an ‘opulent’ town (Conti 1857: 21).

The export of slaves from the African coast was subsequently reported, in the tenth century, by the Arab author Ibn Дawqal. He stated in 976-77 AD that the then ruler of Yemen received slaves, as well as amber and leopard skins from the chief of the Dahlak islands (off the coast from Massawa). These slaves were reported to number 1,000, half of them Abyssinian and Nubian women (al-Дakamī 1892: 8). Subsequently, in 985, alMuqaddasī listed Abyssinian slaves as among the principal imports at the great Arab emporium of Aden (al-Muqaddasī 1906: 97). Later again, in 1021, an Abyssinian slave called NajāΉ who had been purchased on the other side of the sea by another ruler of Yemen seized power to the north, at Zabīd, where there were reportedly 5,000 Abyssinian spearmen. A subsequent Yemeni ruler is said to have sent messengers across the Red Sea for the purchase of a further 20,000 (al-Дakamī 1892: 83, 87, 92).

The commercial importance of Aden was likewise noted by several early sixteenth century observers. An Ethiopian monk, Brother Thomas, was quoted by the Venetian author Alessandro Zorzi as stating that Aden was ‘the gateway for all the spice and cloth and other things that come by land to Barara’, then the Ethiopian capital (Crawford 1958:184-85). Another Venetian, Andrea Corsali, subsequently described Aden in the early sixteenth century as ‘the principal port of Arabia and Ethiopia‘ and ‘the noblest and richest city of the Orient’ comparable, he claimed, to Venice and Cairo (Corsali, in Alvares 1558: 20-21). The port’s trade was described by several early sixteenth century European visitors, among them the Bolognese, Ludovico di Varthema. He referred to Aden as the ‘rendezvous’ for ships from India ‘Major and Minor’, Ethiopia and Persia, as well as for vessels

The Port of ZaylaΚ Much of the commerce between Arabia and the Gulf of Aden coast was at this time conducted through the port of 20

RICHARD PANKHURST: ARABIAN TRADE WITH ETHIOPIA AND THE HORN OF AFRICA subsequently underlined this point when, writing to the Portuguese governor of the Indies, Diego Lopes de Sequeira, he declared ZaylaΚ ‘a port of much food for Aden, and all parts of Arabia, and many countries and kingdoms; and those kingdoms and lands have no other supplies except what come to them from Zeila’ (Alvares 1961, ii: 479).

bound, in the Red Sea, for Mecca and Jiddah (Varthema 1863: 31, 37-38, 60). The involvement of Ethiopia, the Gulf of Aden, and the African interior, in this Aden-based commerce was shortly afterwards noted by the Portuguese apothecary Tomé Pires. He states that cloth and many other things transshipped from Aden found a good market at the Gulf of Aden ports of ZaylaΚ and Berbera, which he describes as together constituting an ‘outlet for the whole of Abyssinia.’ The СanΚāΜ area, he says, produced roses and rosewater which were ‘much prized’ in Abyssinia. The Abyssinians were visited, he adds, by merchants from Aden, and Shihr in Hadramawt, while Adeni traders sent coarse cloth and ‘various trifling things’ to Berbera and ZaylaΚ in exchange for gold, ivory, slaves and horses, while these ports supplied the Arabian port of Jiddah with much meat, fish, wheat, barley and millet (Pires 1944: 7-8, 11, 17-18, 43).

This picture was confirmed at around the same time by the great Portuguese mariner Afonso de Albuquerque. He saw, he reports, many ships from ZaylaΚ and Berbera, trading on the Arabian coast. They were bringing provisions and slaves, he says, to Mecca and Jiddah. The latter port was obliged, he says, to import all its supplies by sea, mainly from ZaylaΚ, Berbera and Massawa (Albuquerque 1875-84, i: 58; iv: 27-28, 35, 58). ZaylaΚ and Berbera exports The export trade of ZaylaΚ was later referred to by several early sixteenth century travellers. Barbosa describes the port as a ‘great store of millet, barley and fruits of divers sorts’, all of which, he says, were carried to Aden (Barbosa 1918-21: 35-36). Corsali states that ZaylaΚ provided meat and other supplies to Aden and Mecca (Corsali, in Alvares 1558: 32-33), while Ludovico di Varthema tells of a ‘very great number of slaves’ being exported through ZaylaΚ to Arabia Felix, Mecca and elsewhere (Varthema 1863:86-88). Evidence as to the extent of the foreign trade of ZaylaΚ is likewise provided by another Portuguese observer, Bernado Pereira, who claims that a thousand camels and other beasts of burden arrived from the interior in 1624, bearing grain, ivory and slaves. He also tells of sizable slave exports, no less than six vessels laden with slaves that year leaving for Mecca (Beccari 1903-17, xii: 67).

A not dissimilar picture was presented a little later by the Portuguese official Duarte Barbosa. He observes that ‘all the goods for Prester John’s country’ were taken to Aden and that this port carried out a considerable trade with ZaylaΚ and Berbera, as well as with Massawa (which had replaced Adulis as the principal port of northern Ethiopia) and the ‘Country of Prester John’, i.e. Ethiopia. These ports, he explains, were all in contact with the interior, whence they received ‘a great store’ of gold, ivory, honey, wax and slaves. From the above and other ports on the African coast, he adds, ‘many ships’ sailed to Aden ‘with foodstuffs in abundance’, which were sold in exchange for ‘a great store of spices and drugs, cotton cloth and other goods of the great Kingdom of Cambaya’, i.e. Cambay in north-west India (Barbosa 1918-21, i: 5557).

Trade between Berbera and Arabia was also considerable. The port, according to Zorzi‘s Ethiopian informants, had a ‘good harbour’ through which Ethiopia was supplied with ‘various merchandise’ from Aden, as well as from Persia and India (Crawford 1958:185, 189, 193). Corsali likewise stated that Berbera was visited by numerous ships from Aden and the Indies, laden with ‘much merchandise’, principally incense from Dhofar, pepper and cloth, which were sent into Ethiopia by camel caravans. The port, he adds, also provided Aden and Jiddah with a ‘large quantity’ of provisions (Corsali, in Alvares 1558: 32). Berbera, Barbosa agrees, was a ‘place of great traffic’, frequented by ‘many ships’ which brought ‘much merchandise’ from Aden and India, and took away gold, ivory and ‘divers other things’. The inhabitants of Berbera, he adds, had many horses and cattle, and accordingly ‘butter in plenty, milk and flesh,’ as well as a ‘great store of millet, barley and fruits of all sorts’, which they took to Aden. Merchants from the latter port in particular took on board extensive provisions, which included meat, honey and wax, for which the port was reportedly ‘exceedingly fruitful’ (Barbosa 1918-21, i: 24- 25). Similar trade with Aden

Arabian dependence on the African ports Arabia, as evident from such early sixteenth century evidence, thus imported considerable quantities of provisions from the opposite side of the Red Sea and Gulf of Aden. The extent of this commerce was already noted by the sixteenth century Portuguese diplomat Francisco Alvares. He reported, in the 1520s, that the Debarwa area, on the northern rim of the Ethiopian plateau, produced ‘numerous grain crops of every kind and nature’ which were carried down to the coast at Massawa, and then shipped across the sea to Arabia for sale at Mecca, Zabīd and Jiddah (Alvares 1961, i: 117). The situation to the east, towards the Gulf of Aden coast, was, as in ancient times, not so dissimilar. Exports from the Horn of Africa to Arabia, as already suggested, passed largely through ZaylaΚ and Berbera. The importance of ZaylaΚ was well known to the Ethiopian Emperor Lebna Dengel (1508-1540). Alvares quotes him as stating that the port was well supplied with provisions, which it exported to Aden, Jeddah, Mecca and ‘all Arabia’ (Alvares 1961, ii: 305). The emperor 21

TRADE AND TRAVEL IN THE RED SEA REGION ZaylaΚ and Berbera imports

also took place, as in ancient times, still further to the east at Cape Guardafui, according to Barbosa. (Barbosa 1918-21, i: 33).

Arab commerce with northern Ethiopia based on the port of Massawa had its parallels further to the south-east, along the Horn of Africa. Imports from Arabia, or transshipped there, were thus taken to ZaylaΚ and Berbera, both of which handled much of the trade of central and southern Ethiopia, as well as that of the Horn of Africa as a whole.

The slave trade The Red Sea and Gulf of Aden ports, as in the past, probably all handled slave exports. Alvares, in the early sixteenth century, asserted that Arabia and other regions, including Persia and India, were ‘full of slaves’ from Ethiopia and that they ‘made very good Moors and great warriors’ (Alvares 1961, ii: 455 ). Many such slaves were acquired by merchants in the course of trade, but others, as Pires notes, were seized by Arab slave-raiders who carried out constant slave raids into Abyssinia with the result, he claims, that ‘large numbers’ of Abyssinians were taken for sale in Arabia (Pires 1944: 14).

ZaylaΚ received a wide range of imports from Arabia. Corsali states that the port was visited by many ships from Aden and India, laden with much merchandise, principally cloth, pepper, and incense from Dhofar in Arabia. These articles, he explains, were then transported inland by camel caravans, as far as Ethiopia, the ‘land of Christian churches’ (Corsali, in Alvares 1558: 32-33). This picture is confirmed by Brother Thomas, who spoke of ZaylaΚ as ‘the chief port’ of the area, and explains that this was because it lay ‘opposite Aden and Arabia Felix and India’, while Barbosa states that ‘many ships’ visited the port ‘to dispose of their wares’ (Crawford 1958: 18485; Barbosa 1918-21, i: 35).

Exports from Arabia to Ethiopia and Horn of Africa Arabia, as in former times, was a major source of Ethiopian and Gulf of Aden imports. Massawa Imports

Berbera, as in the past, was also involved in the ongoing commerce between Arabia and the Horn of Africa. Barbosa states that the port was frequented by ‘many ships’, from Aden and Cambay in India which brought in ‘much merchandise’, including incense, pepper and cloth (Barbosa 1918-21: 24-25).

Articles imported from, or by way of, Arabia into northern, and to some extent central, Ethiopia were shipped mainly to the Red Sea port of Massawa. Such imports, according to early seventeenth century Portuguese Jesuit Manoel de Almeida, included ‘Mecca brocades’, and ‘a thousand other things’ from Mecca and Jiddah (Beckingham & Huntingford 1954: 43).

Arab-Ethiopian trade and the spread of Islam Commerce across the Red Sea and Gulf of Aden was intimately connected with the spread of Islam into Ethiopia, which had formerly basically been a Christian state. Muslim expansion resulted in a characteristically Ethiopian religious division of labour in which the dominant Christian population tended to shun trade which was largely monopolised by Muslims (and to some extent Armenian Christians). This situation facilitated commercial activity across the Red Sea and Gulf of Aden, in that trade was largely carried out by merchants, on both the Arabian and African sides, belonged to a single faith: Islam.

Similar imports from Arabia were later reported, in the eighteenth century, by the Scottish ‘explorer’ James Bruce. He recalls that ‘the goods imported from the Arabian side’ of the Red Sea consisted of: ‘blue cotton, Surat cloth, and cochineal ditto, called Kermis, fine cloth from different markets in India; coarse cotton cloth from Yemen; cotton unspun from ditto in bales; Venetian beads, crystal, drinking, and looking-glasses; and cohol, or crude antimony. These three last articles come in great quantities. From Cairo, first in the coffee ships to Jiddah [i.e. ships transporting coffee from that port], and then in small barks over to this port [i.e. Massawa]. Old copper too is an article on which much is gained, and great quantity is imported’ (Bruce 1790, iii: 54).

The Jesuit Almeida, looking at this religious division of labour through Christian eyes, recalled that ‘Moors’, i.e. Muslims, were ‘mixed up’ with Christians throughout the entire Ethiopian empire. Christians, he claims, were not, however, allowed to go to the Arabian sea ports, such as Mocha, Jiddah and al-Hudaydah, and that even at Massawa, Moors were ‘better received and more welcome’, with the result that they were ‘left in all the important trade of Ethiopia. The great and the rich men of this empire’, he adds, ‘all have many of these Moors as their agents, and carry gold to the sea for them and bring them silks and clothing’ (Beckingham & Huntingford 1954: 55). This statement was supported by Job Ludolph, the illustrious seventeenth century German

The import into Ethiopia of glass bottles transshipped in Arabia from Venice, is also mentioned by Bruce who observes that ‘thousands of packets’ of such bottles were taken from Arabia to the then Ethiopian capital, Gondar (Bruce 1790, ii: 678). One other Ethiopian import reported by Bruce came from Arabian coastal waters. This was a univalve shell found near Qunfudhah and LuΉayyah. Shipped across the Red Sea, it served not only for decorative purposes but also as ‘primitive money’ among the Jawi and other Oromo tribesmen of the Ethiopian interior (Bruce 1790, iii: 54-55).

22

RICHARD PANKHURST: ARABIAN TRADE WITH ETHIOPIA AND THE HORN OF AFRICA Society, second series, volumes 44, 49). London: The Hakluyt Society. Beccari, C. 1903-17. Rerum Aethiopicarum Scriptores Occidentales inediti a saeculo XVI ad XIX. (14 volumes). Rome: Casa Editrice Italiana. Bruce, J. 1790. Travels to Discover the Source of the Nile. (5 Volumes). Edinburgh: J. Ruthven. Crawford, O.G.S. 1958. Ethiopian Itineraries, circa 1400- 1524. (The Hakluyt Society, second series, volume 109). Cambridge: Cambridge University Press. Devic, L.M. 1883. Les pays des Zendjs ou, la côte orientale d’Afrique au moyen-age. Géographie, moeurs, productions, animaux légendaires d’après les écrivains arabes. Paris: Librairie Hachette. Ibn BaΓΓūΓa/Gibb, H.A.R. (ed. & transl.). 1958-2000. The Travels of Ibn BaΓΓūΓa, A.D. 1325-1354. (4 volumes). (The Hakluyt Society, second series, volumes 110,117,141,178), Cambridge: Cambridge University Press. al-Idrīsī, Abū ΚAbd Allāh MuΉammad/Dozy, R. P. A. & de Goeje, M. J (eds. & transls.). 1866. Description de l’Afrique et de l’Espagne. Leiden: E.J. Brill al-Дakamī, Najm al-Dīn ΚUmārah/Kay, H. C. (ed. and transl.) 1892. Yaman, its Early Medieval History by Najm al-Din ΚOmarah al-Hakami. Also the abridged history of its dynasties by Ibn Khaldun, and an account of the Karmathians of Yaman by Abu ‘Abd Allah Baha ad-Din al-Janadi: the original texts, with translation and notes. London: E. Arnold. Ibn Дawqal/Kramers, J.H. & Wiet, G. (eds. & transls.) 1964. Configuration de la Terre. (2 Volumes). (Collection UNESCO d’oeuvres représentives. Série arabe). Beirut/Paris: Aubenas Kosmas/McCrindle, J. W. M. (ed. and transl.) 1897. The Christian Topography of Cosmas, an Egyptian Monk. (The Hakluyt Society, first series, volume 98). London: The Hakluyt Society. Ludolph, Job [Ludolphus, Hiob]. 1682. A New History of Ethiopia.(Second Edition). London: Samuel Smith Major, R. H. (ed.) 1857. India in the Fifteenth Century. (The Hakluyt Society, first series, volume 22). London: The Hakluyt Society Muir, W. 1878. The Life of Mahomet: from original sources. (2 volumes). London: Elder Smith al-Muqaddasī/Goeje, M.J. de (ed.).1906. Descriptio imperi Moslemici. Kitāb AΉsān al-Taqāsīm fī MaΚrifah al-Aqālīm. (2nd edition). (Bibliotheca geographorum arabicorum 3). Leiden: E.J.Brill. Periplus/Huntingford, G. W. B. (ed. and transl.) 1980. The Periplus of the Erythraean Sea. (The Hakluyt Society, second series, volume 151). London: The Hakluyt Society. Pires, T./Cortesão, A. (ed. & transl.). 1944. The Suma Oriental of Tomé Pires. (2 volumes). (The Hakluyt Society, second series, volumes 89, 90). London: The Hakluyt Society. Varthema, L. di/Badger, G. P. (ed.) 1863. The Travels of Ludovico di Varthema in Egypt, Syria, Arabia Deserta and Arabia Felix, in Persia, India, and Ethiopia, A.D.

scholar of Ethiopian affairs, who quotes his Ethiopian informant, Abba Gregory, as declaring that ‘the Muslims live in our country in small numbers ... they are our servants for trade’. Ludolph’s own conclusion was that: The Habessines are in no way addicted or expert in the Art and Intreagues of Merchandising, for they that will not travel into Foreign Parts must yield their gains to others... Therefore, the Arabians who inhabit the Ports of the Red Sea, especially the Mahometans scattered over the Kingdom, are the Chief Merchants, for being of the same Religion they have the free liberty of all the Ports of the Red Sea (Ludolph 1684: 390-91).

Conclusion Trade across the Red Sea and Gulf of Aden, which dates back to ancient times and changed only slowly in the period under review, was long of major importance to both Western Arabia and the Ethiopian-Horn of Africa region. Ethiopia, over the centuries, was a major source of raw materials, particularly ivory and gold, as well as slaves. The more arid lands of the Horn on the other hand provided Arabia with frankincense and other kinds of incense. Ethiopia and particularly the Horn also supplied the Arabian ports with sizable quantities of cereals, livestock and foodstuffs of all kinds. Arabia for its part provided the opposite side of the Red Sea and Gulf of Aden with vitally important manufactured and luxury goods, including textiles, weapons, tools and glassware. Some of the textiles were spun and woven locally in Arabia, while others were imported from India and the East. Such articles thus formed part of a major intercontinental trade network which linked Ethiopia and the Horn of Africa since ancient times with the Indian subcontinent. References Albuquerque, Afonso de/Birch, W.de G. (ed. and transl.) 1875-84. The Commentaries of the Great Afonso Dalboquerque, second Viceroy of India. (4 volumes). (The Hakluyt Society, first series, volumes 53, 55, 62, 69). London: The Hakluyt Society. Almeida, Manoel de/Beckingham, C. F. & Huntingford, G. W. B. (ed. and transl.) 1954. Some Records of Ethiopia 1593-1646. (The Hakluyt Society, second series, volume 107). London: The Hakluyt Society. Alvares, Francisco/Bellere, J. (ed.) 1558. Historial description de’Ethiopie. [Preceded by] ‘Lettre de la royne Helène à Emanuel roy de Portugal’ and ‘Lettre d’André Corsale touchant ses voyages’. Antwerp: C. Plantin. Alvares, Francisco/Beckingham, C.F. & Huntingford, G.W.B. (eds.) 1961. The Prester John of the Indies. (2 volumes). (The Hakluyt Society, second series, volumes 114, 115). Cambridge: Cambridge University Press. Barbosa, Duarte/Dames, M. L. (ed.) 1918-21. The Book of Duarte Barbosa. (2 volumes). (The Hakluyt

23

TRADE AND TRAVEL IN THE RED SEA REGION 1503 to 1508. (The Hakluyt Society, first series, volume 32). London: The Hakluyt Society. Watt, W. M.1953. Muhammad at Mecca. London: Clarendon Press.

24

The elusive land of Punt revisited K.A. Kitchen The land of ‘Punt’ (probably more correctly Pwanet) was first found in Egyptian texts in the nineteenth century; its association with incense and myrrh led at first to identification with Arabia. Additional references indicated East Africa but suggested locations which varied widely from Sudan to Somalia

assembly and use on the Red Sea. This was probably the intention here. Second phase: millennium BC

Kingdom,

early

second

During the Eleventh and Twelfth Dynasties, a reunified Egypt again opened relations with Punt. Under Mentuhotep III Sankhkare (Year 8, c. 1985 BC), the high official Henenu left a stela in the middle of Wādī Hammamat, telling how he took a 3,000-man expedition from Coptos through the wadi to the sea, where he assembled a ship and sent it off to Punt and back (translation and further references, Kitchen 1993a: 589 and n. 6). Later under Sesostris I, seemingly in his twenty-fourth year (c. 1930 BC), we have confirmation of this procedure. At Mersa Gawasis on Egypt’s Red Sea coast (about 22 km south of Safagah) was found a Twelfth Dynasty port, with inscriptions of two officials mentioning the making of ships at Coptos, their (re)assembly at Mersa Gawasis itself (under its ancient name of Sawaw), and their despatch to ‘the miningregion of Punt’, involving over 3,700 expeditionaries, including five hundred sailors. The slightly later inscription of Khenty-khety-wer dates to Amenemhat II (Year 28, c. 1884 BC), set up ‘when he had returned safely from Punt, his expedition with him, safe and sound, and his ships resting at Sawaw.’ (cf. Sayed 1977, 1978; Kitchen 1993a: 590-591.)

Egyptian data referring to Punt range from c. 2500 to 600 BC, giving Punt a long history, an exotic range of products, and a political system of chiefdoms. Here, we review the basic geographic indicators to establish a northern limit and main area for Punt, then explore its possible southern extent, and deal with some radically different recent suggestions as to its location. First phase: Egypt’s Old Kingdom, third millennium BC The earliest indisputable mention of Egypt’s relations with Punt refers to the time of King Sahure (c. 2500 BC), and is recorded on the Palermo Stone (latest publication, Wilkinson 2000). Here, products brought from Punt comprised 80,000 measures of Κntyw (myrrh), 6,000 measures of electrum (a naturally occurring gold/silver alloy), 23,030 staves, plus 2,900 of an untranslatable item. The occurrence of aromatics and a form of gold is consistent with later sources for Punt, but no indication is given of the route by which these goods travelled from Punt to Egypt.

The picture that emerges is consistent and uniform. In order to conduct direct trade with Punt, the Egyptian authorities constructed suitable ships at the Nile docks at Coptos, at the closest point to the Red Sea. They then sent fully equipped and manned expeditions by the shortest route (i.e. Wādī Hammamat and its branches) to a designated coastal inlet and port of their making on that seacoast. There, the prefabricated ships were reassembled, launched, and sent off via the Red Sea to Punt and back. On return, the expeditions and the goods obtained went back overland to Coptos (then on to Thebes, Memphis or both). The case of Pepynakht in the Old Kingdom shows that this procedure had good precedent. That the ‘mining-region of Punt’ produced gold or electrum (besides aromatics) may be inferred from Sahure. What neither the Old nor the Middle Kingdom references tell us is how far down the Red Sea these argosies went, and where they landed: in East Africa or South Arabia? However the Old Kingdom data about pygmies or dwarfs sent down the Nile via Punt suggests an African rather than an Arabian location for Punt.

Some sixty years later, a pygmy or dwarf was ‘brought from Punt in the time of (king) Isesi’, an exploit reported and repeated by the later explorer Harkhuf under Pepy II, c. 2270 BC.1 Harkhuf brought his man down the Nile from Iam, probably the later land of Irem, locatable in Upper Nubia, most likely south-west of the Bayuda desert, along the Shendi stretch of the Nile. So, Punt would be (East) African and further south still. Goods and people might descend the Nile from Punt, but this was not the normal way for Egyptians to go to Punt themselves. Also under Pepy II, Pepynakht had to retrieve the body of another expedition leader murdered by local desert tribesfolk ‘in the land of the Asiatics‘ while building a ship there to visit Punt (text, Sethe 1932: 134; transl., Breasted 1906, i: 163). The only ‘land of the Asiatics’ so far south was the Eastern Desert, between the Nile and the Red Sea; this term excludes the Nile Valley itself completely, for it was populated solely by Egyptians and Nubians. But in later times, boats might be constructed at Coptos to be taken through the desert via Wādī Hammamat (this being the shortest route) for 1

Middle

Text, Sethe 1932: 128-131; translation, e.g. Breasted 1906, i: 160-161.

25

TRADE AND TRAVEL IN THE RED SEA REGION Commiphora grow (or grew) extensively in East Africa. So, in the eastern Sudan, in Kassala province (adjoining Eritrea and Ethiopia) and east and south of Roseires (south Funj district), a 1961 survey showed the presence of well over a million trees of Boswellia. Commiphora (myrrh) is found along the Red Sea coast at about 22o north latitude, and up to 250 km inland on the present Egyptian-Sudanese border (Herzog 1968: 64). Further Boswellia are reported in the Nuba mountains west of the White Nile, and beyond in Darfur (Kitchen 1971: 186). Northern Somalia has long been famed for its aromatics, as is Dhofar in southern Arabia. Thus, incense and myrrh are too widely distributed to decide the question of the location of Punt; but ‘Egyptian ebony’ and doum palms with hamadryas baboons favour the north-eastern Sudan and its immediate neighbours.

Third phase: New Kingdom, later second millennium BC Under the Empire, the most informative new source is the set of relief scenes in the great memorial temple of Queen Hatshepsut at Deir el-Bahri in western Thebes. The main scene is a huge rectangular tableau occupying the south end wall, depicting the land of Punt (the fullest representation is in Smith 1962). Directly next to it, at the south end of the rear wall, another tableau in three registers (bottom to top) shows the arrival of a flotilla of five Egyptian ships at Punt, then their loading with goods and departure homewards, and finally a presentation in Egypt of the goods obtained before a symbol of the queen (Kitchen 1993a: 593, fig. 35.2). We shall consider briefly these splendid scenes. If we look at the second tableau, the ships float upon water that is peopled with several varieties of fish; this phenomenon recurs in the two strips of water that appear in the first tableau. These depictions of fish have been analysed, and almost exclusively represent species found in the Red Sea and Indian Ocean (cf. Danelius & Steinitz 1967). Of these, five are too vague to identify, but of the remaining forty, all but three are Red Sea and Indian Ocean species (and clearly so for the most part). The other three might be freshwater species (tilapia? catfish?), but the tilapia identification is not certain, and such catfish can penetrate salt water. So, the depictions of fish at Deir el-Bahri overwhelmingly favour Punt having a Red Sea coastline, and confirm the very clear Middle Kingdom data for Egyptian ships sailing to Punt from Red Sea ports such as Sawaw.

Generally, gold (and probably its silver alloy, electrum) can be found so widely in East Africa and western Arabia that it is not a precise geographical indicator. However, the ancients sometimes specified named territories as the origin of their gold. From Punt, the Egyptians sometimes named ‘gold of Amau’. The desert land of Amau evidently lay between Punt and Egypt’s own Nubian imperial domain. It did not belong to Egypt; otherwise she would not have traded it in from both Amau and Punt. Amau may feasibly be located in the gold-bearing region west of the Red Sea mountains that run along behind Ras Shagarah – Port Sudan – Suakin (cf. Kitchen 1999: 174-176, and fig. 1). There have been doubts as to whether the two dng men brought back to Egypt in the Old Kingdom were truly pygmies or simply dwarfs. Natural dwarfs were not unknown in ancient Egypt, being given special roles in the Old Kingdom (cf. Seyfried 1986), but these were not confused with the pygmies brought in as extreme rarities from far-south African regions, and dedicated to ‘the dances of the god’; young Pepy II wanted to see Harkhuf‘s pygmy (or dwarf) ‘more than the tribute of the mining-region of Punt!’ and commanded that extraordinary precautions be taken to ensure his safe arrival in Memphis. This suggests that this was no ordinary dwarf. Hence these dng men brought from Punt (or, likely, from beyond it) are indeed to be considered to have been true pygmies, as would be required by fuller evidence used by ethnographers familiar with real pygmies (Herzog 1968: 56-58). Thus this may favour an African Punt, not an Arabian one.

As for land-based fauna, the fragments of the grand tableau of Punt include a giraffe and a rhinoceros. These are native to Africa in particular; and are certainly foreign to Arabia. Baboons and doum palms are to be found in the eastern Horn of Africa (eastern Sudan/Eritrea/north-eastern Ethiopia/Somalia), but the overlap or symbiosis of these two fits best with Eritrea and eastern Sudan (Kitchen 1971: 186f. and nn. 11,15). On this basis, Punt would be somewhere in these latter two lands (if not further south), accessible from their coastlines on the Red Sea. In Hatshepsut‘s land of Punt, the Egyptians obtained what they called hbnyw, ‘ebony’. As is attested from actual surviving examples (Lucas 1962: 434-436), what the Egyptians called and used as ebony was and is a dark, handsome wood known botanically as Dalbergia melanoxylon. This tree grows naturally in north-west Ethiopia on lowlands that run down onto the border of north-east Sudan, along with doum palms and acacias, as well as west of the White Nile between 12o and 14o north latitude (Kitchen 1971: 187, n. 13). Frankincense and myrrh belong to the same botanical family, Burseraceae, frankincense to the genus Boswellia, myrrh to the genus Commiphora (formerly Balsamodendron) (see Hepper 1969). A number of species of Boswellia and

As for Parehu, the only named chief of Punt, the consonant p in his name and that of Punt itself also firmly excludes Arabia. The Old South Arabian languages (like classical and later Arabic) possess an f, but not a p. So, Arabia would have had a *Farehu, chief of *Funt! Egyptian has both p and f, so its transcript is reliable in this distinction. It is interesting to notice that, in the long lists of African names left us by Tuthmosis III (Sethe 1930: 797-806 passim), some fifteen names include a p (nn. 33, 40, 46, 47, 99, 130, 150, 152, 192, 203-F(?), 20326

K.A. KITCHEN: THE ELUSIVE LAND OF PUNT REVISITED route followed, while the Old Kingdom and New Kingdom data clearly favour an East African location, in part at least in the north-east quadrant of the Sudan, extending into Eritrea and Ethiopia. Along its northernmost front, Punt lay next to Amau and probably Irem, and thus had no border contact of any kind with Egypt. No pharaonic army ever marched to invade it, or the wild, desertic terrain of Amau; Egypt battled with and often subdued Irem, but did not occupy it or go beyond it. The southern extent of Punt is another question entirely, discussed below.

G, 206, 218, 221, 253) besides Punt itself, but only one has an f (n. 74). Parehu wore a close-fitting set of metal rings up one leg; this usage is specifically African, well attested among Sudanese tribes like the Bongo, Dinka, Djur, and Niam-niam (cf. Kitchen 1971: 188 and n. 20). Parehu had a remarkably plump wife, and a daughter who showed already the same tendency; some regard this as a medical condition, but this is not necessary. Such a mode was a mark of beauty in East Africa and the Sudan (Herzog 1968: 59-61). The great tableau at Deir el-Bahri shows two kinds of Puntites – medium brown ones, almost Egyptian in appearance, and very dark-skinned folk with negroid build and features; buildings on piles or stilts are prominent. The population is clearly to be understood as African; likewise the use of stilted buildings is East African (such usage in East Asia is too far away to be relevant).

Fourth phase: the late period, early first millennium BC. After this long history, we have no further record of any more official Egyptian expeditions to Punt, or of Puntites visiting Egypt. The whole phenomenon simply disappears from view. The reasons for this are not clear, and could be deceptive.

The Hatshepsut trade contact was so appreciated by the Puntites that they began sending their own sea-borne missions to Egypt in the succeeding reigns of Tuthmosis III to Amenophis III (Kitchen 1993a: figs 35.5-8). Theban tomb-chapel number 143 shows the Puntite boats, while its text mentions ‘travelling to [Thebes], going by road’, evidently from the boats (landing at Sawaw or Quseir?) via Wādī Hammamat to reach Thebes. It should be emphasised that the royal officials receiving the goods on the pharaoh’s behalf were all Theban officials - they were the responsible authority for whatever was transhipped to the Nile via Hammamat and nearby Coptos. This cuts out the theory (Bradbury 1996) that the Puntites had sailed down the Nile; if they had, they and their trade-goods would have come under the authority of the viceroy of Nubia, not his Theban colleagues. The products are, as usual, aromatics (myrrh), gold and ebony; likewise in entries in the Annals of Tuthmosis III for Years 33 and 38 (c. 1447 and 1442 BC). Intermittent contacts are known under Akhenaten, Haremhab, Sethos I and Ramesses II, but these are rather fleeting allusions (cf. Kitchen 1993a: 600-601 and references); Punt was no longer headline news, more a regular trading partner.

We have just one document that mentions Punt in a narrative context. It is attributed to the Twenty-Sixth Dynasty, reign unknown, which means within the years 664-525 BC, say about 600 as a mean rough date. Lines 11-15 run as follows: … the army-force(s) thanked His Majesty (12)…, [saying]: ‘(How) great is Your Majesty, O Victorious King! A great marvel has happened in Your Majesty’s time; (13) this has not been seen or heard (before) rainfall upon the mountain of Punt - (for) rain was a scarcity in the Southern provinces (14) … this month in which its rain was, (when) it was not the season (for it), even in the Delta towns. (15) Your mother, the goddess Neith, has brought it for you - a Nile-flood (Hapi) to sustain your forces! (Kitchen 1993a: 602).

The context associates ‘rainfall upon the mountain of Punt’ with (as its consequence) Hapi, normally a Nileflood to benefit Egypt but here it sustains the army. In other words, the rain that fell on Punt’s mountains drained into the Nile, causing it to carry the resultant flood-water downstream through Egypt, and in this case, out of the normal season. This is certainly possible, if the Ethiopian mountain massif is in what was Punt, and the rain drained west and north, into the Atbara and even the Nile proper, north of their confluence. But neither Somalia nor any other location away from the Sudan/Ethiopia/Eritrea continuum would fit this evidence.

The last significant New Kingdom Punt event took place under Ramesses III (c. 1184-1153 BC). In the Great Harris Papyrus, he speaks of sending transport-boats with goods ‘on the great sea of the inverted water’. These returned with products of Punt and ‘reached the desertland of Coptos in safety. They moored safely, carrying the goods brought; these were (re)loaded overland, on donkeys and men, then loaded into ships on the River (=Nile) at the quay of Coptos, being sent on downstream’ i.e. to Memphis, the capital. As clearly as one might wish, we see here a flotilla sent out down the Red Sea, whose currents flowed south (‘inverted’ in direction, compared with the northward flow of the Nile). They evidently landed back at Sawaw or nearby; their goods came overland via Hammamat to Coptos, and were finally sent downstream to the king. All the data, down to Ramesses III, give a clear and consistent picture of the

Punt: the geographical nucleus The net result of our investigations both in previous studies and in the present overall summary is clear and consistent. Punt was reached from Egypt via the Red Sea, going south. Expeditions thither would have landed along the Sudanese and/or Eritrean coast along a general stretch from Port Sudan south to Aqiq, within a range of roughly 200 km. They would have gone inland (presumably with 27

TRADE AND TRAVEL IN THE RED SEA REGION changes). This would accord with Fattovich’s suggestion. Or else these Genebtyu were in fact from Somalia, and had simply come north along the Red Sea’s western coastline to share in the trade with Egypt, which is also a possibility (Kitchen 1999b: 105). So, at least the door lay open for vistas beyond ‘minimal Punt’.

Puntite guides), through the Red Sea hills by one of several feasible routes (Kitchen 1971: 203). That would enable them to tap the natural resources in terms of aromatics, timber, and other products, in the areas all the way to Kassala if need be. All this (like the general location for Punt established so far) fits remarkably well with the geo-archaeological map of the region presented by Fattovich (Fattovich 1991: 264, abb. 1, cf. 260, 262). Consideration of the sequence and distribution of the archaeologically attested cultures in this large region can help to establish the mode of life during the centuries through which Punt is known historically, and also to trace possible circuits of early trade (Fattovich 1991, 1993).

At this conference three factors in particular led me to think over more carefully the southern possibilities. One was the suggestion made in David Dixon’s paper that if the Egyptians had over-exploited the north-east Sudanese aromatic growth zones, they might well have sought pastures new, to find fresh supplies; the logical move would simply be to push farther south along the coasts of Eritrea to Djibouti and Somalia where such supplies would indeed be found. This possibility began to look more realistic, in the course of profitably wider-ranging discussions on Somalia with Hamish Wilson in the light of his extensive first-hand knowledge of the country. Then, in turn, Professor Felix Chami from Tanzania was able to report on the finding of genuinely Egyptian-Nile Valley ceramics in the course of his work on sites there, dating to at least the 6th century BC if not before (as verified by Hans Nordstrom) (Chami, infra). The archaeology of Somalia (thinking here mainly of the northern region) is not yet fully developed, so it is premature to attempt to use it at present.3 But there is obviously scope for a fuller picture in the future.

So far so good – but how much further? The above presentation encapsulates all the principal evidence, and a minimum view of the Punt that can be envisaged when collating and using it. But the minimum is just that – it may not be the whole story. Already, for several decades before Herzog’s epochal monograph of 1968, it had been quite commonplace to locate Punt in northern Somalia. This suited the obviously African locale demanded by the data in the Hatshepsut reliefs, as well as the Red Sea route, and the classical and modern reputation of Somalia and environs as producers of goodquality incense and myrrh, as is clear, for example, from the Periplus Maris Erythraei in the first century AD.2 But further studies from 1968 onwards have, as seen above, uncovered data on Punt that do not suit an exclusively Somalian location. The result has been an ‘either-or’ situation.

In such a context, one may, with due reserve, begin to suggest the model of an ‘expanding concept’ of Punt, the use of the term covering progressively larger areas as the Egyptians reached out ever farther for a Puntite range of products. Thus, Punt would cover first the zones along the Sudanese and Eritrean coast and inland/westward from that coast as in the ‘minimal’ model; then to it would be added the coastal strip along which the quest was pursued for fresh aromatic sources, and then northern Somalia in particular with myrrh and incense reserves. Ras Hafun was probably the limit here. Southward along the sandy shores of southern Somalia and on to Tanzania, we reach the Azania of the classical sources (cf. Periplus/Casson ed. 1989: 136, §15:5.18), and we are no longer dealing with Punt. But the attestation of Egyptian/Nile pottery in a Tanzanian coastal context (in Azania, therefore) brackets northern Somalia into an Egyptian trading horizon from the sixth century BC and possibly earlier, in archaeological terms.

But if Punt by any standards has northern and northwest/western frontiers set by Amau and Irem respectively (cf. Kitchen 1999: 175, fig. 1), and is bounded by the Red Sea along its east flank, what is to be said of its possible southern extent? The present author has never previously discussed this, simply because there seemed to be no answer. However, the impact of this conference, combined with other hints, has enabled the author to review the possibilities afresh. Long ago, Fattovich had suggested that, in the New Kingdom, South Arabian aromatics producers and traders might have crossed the Red Sea, presumably at the Bab al-Mandab, and gone north along its western coast to sell to the Egyptian expeditionaries (Fattovich 1991: 267d). Much earlier, Professor A.A. Saleh (1972a, 1972b) drew attention to a curious entry in the Annals of Tuthmosis III (Year 32; c. 1448 BC) reporting that the Genebtyu had brought myrrh and other products (now lost in lacuna); the name recurs in a list of foreign lands under Ramesses II, which led me to the review the matter (Kitchen 1999b: 104-105). It is possible that these Genebtyu were early precursors of the Old South Arabian Gebbanitae of Pliny, assuming a simple metathesis of n and b in his case (but not of Qatabān/Qitbān, which would involve more drastic 2

What about practicalities? As pointed out long since, it would be practical to allow ships of Hatshepsut’s type to travel by day at about three knots (or three and a half to four miles an hour) for eight or nine hours a day, travelling at least thirty miles a day (Kitchen 1971: 196). The 600 miles or so from Sawaw or Quseir to Suakin would thus have implied a twenty to twenty-five day 3

Overview, cf. Fattovich 1995; finds from Ras Hafun, first century BC to the fifth century AD, mentioned in passing by Sinclair et al. 1993: 419.

See Periplus/Casson ed. 1989 for text and invaluable commentary.

28

K.A. KITCHEN: THE ELUSIVE LAND OF PUNT REVISITED solution. He too pulls in the irrelevant subject of copper supplies; major modern world suppliers of copper and tin are, in most cases, different from the ancient sources.5 He makes major errors, such as sending the luckless Harkhuf to Punt and by sea. But Harkhuf never went to Punt, but Iam, and via the Western Desert oases, as his inscriptions make clear! He also confuses three separate issues: Punt, Ophir and the origin of the Phoenicians. Use of routes over the vast Ethiopian massif is simply an unnecessary extravagance, not justified by the ancient evidence. On Punt, he seems unaware of the fundamental study by Herzog (1968), my later papers, the discoveries by Sayed (1977), or any other modern treatment.

journey by sea. We may now extend this concept farther. From Suakin south to Massawa or possibly Zula (near Adulis) is a distance of about 300/350 miles, or ten to fifteen days sailing. Then, on to Obock or even round west into the Djibouti (Tadjura) inlet to Ghubbat-Kharab we have a further 400/450 miles, or another thirteen to fifteen days of seafaring. Then, travelling on eastwards to Berbera (classical Malao) in western Somalia would cover another 200 miles or so, say seven more days. Beyond that point, up to some 360 miles of Somalian north coast as far as Cape Guardafui would provide landing-places to go and trade for the local aromatics. Thus, from Suakin to Berbera at between 900 and 1000 miles of quite leisurely sea travel (walking speed!) would have cost at most thirty to forty days, little more than a month or so, while Somalia’s entire north coast could be reconnoitred in twelve days to its eastern limit.

Very different to these is the attempt to move Punt back to Arabia by D. Meeks (2002, and 2003 in English). It is a brilliant, most impressive tour de force, but risks forcing the evidence improbably. It is impossible to do full justice to it here, but some concise, cautionary notes may illustrate the point. Some evidence is inconclusive. The distribution of the relevant aromatic flora in both Arabia and East Africa is too wide to serve as an indicator of the location of Punt, and the same is true for gold. Deities attributed to Punt by expeditionaries in Sinai (Meeks 2002: 291ff, 320) merely illustrate the wide travels of those men; and south-east Arabia’s aromaticproducing region is no nearer Sinai than East Africa; it is in fact further away. The brown-red skin of some Puntites, who are portrayed as much like the Egyptians, no more proves that Puntites lived in Arabia than it proves that the Egyptians did. Not all East African peoples are, or were, negroid; the Somalis, for example, are not. Religious and other such texts of secondary relevance may link Punt with the ‘east/sunrise’; but south-west Arabia is no more ‘east’ of Egypt than is East Africa. It is the same phenomenon as Dilmun being said to be ‘east’ of Mesopotamia, when in fact it lies south and south-west (cf. among many others, Kitchen 1994: 140). Travelling from Suez to Punt is no different, regardless of whether Punt lay in south-west Arabia or in East Africa; the same applies to the return to Sawaw or Quseir (because of bad wind conditions up to Suez). That an officer might mention Byblos and Punt in one breath does not imply that Punt was in Asia (hence, Arabia), any more than this writer alluding to his own travels ranging from Aberdeen to Abu Dhabi would put the latter in Britain!

Thus it is possible that from the New Kingdom or at some later date (but before 500 BC), the Egyptians had learned of, and sought to visit directly, further suppliers of aromatics beyond their northern Punt. Halfway to what we now call Berbera, the landing-places of (modern) Massawa and Zula were within some 100 km of a goldbearing zone around Asmara (cf. Fattovich 1991: 264, map), a theoretical gold source from a ‘middle’ Punt? Then from Berbera eastwards (in Northern Somalia), a new, southern Punt would provide additional supplies of aromatics. What the attraction was still further south, in classical Azania, is not so clear; the Periplus lists as exports from Rhapta abundant ivory and various horn and shell products (Periplus/Casson ed. 1989: 61). It remains unclear whether the Genebtyu of Tuthmosis III and Ramesses II were ancestral to the Gebbanitae of southwest Arabia, a view noted above, or were in fact a people from this more ‘southern’ Punt. Punt: a note on diverging suggestions Here we may mention recent suggested locations for Punt, and why they seem not to fit the total data. First, F.D.P. Wicker (Wicker 1998), who proposes to locate Punt in Kenya, but does so on the flimsiest grounds. He denies that Egyptian ships went by sea, which is directly contradicted by the evidence of 98% Red Sea fish depicted in the water under Hatshepsut’s ships at Deir elBahri, and by the existence of the Middle Kingdom Red Sea port at Mersa Gawasis, whose remains and texts make it clear that Egyptian expeditions went that way. No attention is paid to the true location of Iam/Irem, or to any modern study of this name. His consideration of copper is an irrelevance; copper is never claimed to be a Puntite product; it is well known that electrum occurs naturally in Egypt.4 All Egyptian vessels in antiquity were square-rigged, and every bit as able to sail by sea to Punt as their Middle Kingdom counterparts indubitably did; he appears to know none of the literature on Punt of the last 35 years. J.E. Dayton (2002) proffers a similar 4

Against the Arabian hypothesis stand other points. Punt, Parehu and fifteen of the twenty-nine Puntite names (nn. 49-77) – half the total – contain the letter and phoneme p, unattested in the Semitic languages of the Arabian peninsula (ancient and modern) which have f instead (found only once in the Punt list). This would be inexplicable with an Arabian Punt, especially as ancient Egyptian has both p and f, and so can write them separately at need. As Meeks admits (2002: 276), giraffes are exclusively East African, not Arabian; the thesis that 5

See Lucas 1962, of which Wicker appears not to know.

29

See Lucas 1962 for Egypt and Moorey 1994 for Near East.

TRADE AND TRAVEL IN THE RED SEA REGION compact sequence of narrative that has rain falling on the Mountain of Punt, allegedly when it is not doing so elsewhere (cities of Egypt), and Neith bringing Hapi (normally the Nile flood) for the king’s forces. Even if one allowed that this is a heavenly Hapi rain (as Akhenaten once expressed it), on the king’s army, the whole matter is clearly happening in a Nile Valley context. Rain in a south-west Arabian Punt is so far distant as to have no relevance to towns in Egypt proper. In short, despite its brilliance (and it deserves more attention than can be paid it here), Meeks’s case is not convincing. I would in closing still, on balance, more easily locate Punt in eastern Africa, well south of Egypt, conveniently accessible from the Red Sea’s western shores.

some obscure chief deep inside East Africa sent such a creature all the way to a mere chief in the Dhofari uplands of south-west Arabia is, surely, somewhat bizarre. Punt was not a centralised kingdom but had a multiplicity of local chiefs, as the repeated Egyptian phrase ‘chiefs of Punt’ indicates. The same criticisms also largely apply to the rhinoceros; the presence of but one horn is of a piece with similar examples from Kerma and the stela Louvre C.14; all these examples are simply abbreviated, not of Indian origin (cf. Kitchen 1971: 187, and nn. 16-17). Against Arabia is the role of Amau, between Punt and Egypt’s southernmost Nubian domains, as clearly evident from the Luxor Minerals list of Ramesses II and the Sabu graffito of Userhat. If Punt were in Arabia, no solution is possible whereby gold from Amau could be retailed from a south-west Arabian Punt (or even one in central west Arabia at Mahad alDhahab!) to a man working at least 300 miles away not so far east of the third Nile Cataract. If Punt is in Arabia, why are its twenty-nine toponyms systematically included in the heart of the detailed African list of Tuthmosis III? This is very different from its occurrences in much more summary (and largely derivative!) lists under Sethos I and Ramesses II. The isolation of Punt in ‘Syrian’ lists at Aksha and Amarah West finds instructive comparison with the Karnak lists of Sethos I. Here, we find the ‘Tribesfolk of Nubia’ (Iuntyu-Ta-Sti) heading Asiatic lists (Kitchen 1993c: 25, n. 20; 27, n. 8; cf. 29, n. 9). This does not put Nubia in Arabia. Also in these lists, Punt actually heads the African toponyms, coming straight after Asia (Kitchen 1993c: 23, n. 40) with Shasu and Arzawa (interversion, Kitchen 1996: 38, Nos. 33-34 wrongly between Nos. 32 and 35ff.). South and north toponyms can even be alternated (Kitchen 1996: 52 end, Nos. 1-9)! And at Abydos, Ramesses II has ‘Puntites’ at the end of a genuine all-African list (Kitchen 1996: 57, §33(a)). Likewise, Libyan Meshwesh can occur with Asiatic names (Kitchen 1996: 58 end, 59 top). Finally, the indubitably Puntite place Wet(j)enet occurs solidly among African names not only with Tuthmosis III (n. 64) but also under Sethos I (Kitchen 1993c: 23 end, No. 56, and 26, No. 53). In short, the isolated occurrences of Punt at Aksha and Amarah West are simply abbreviations from a longer listing, and do not place Punt in ‘Asia’ only in Ta-netjer, ‘God’s Land’, which covers a vast swathe of terrain north-east, east and south-east from Egypt (Meeks 2002: 272). Its mention by Ptolemy X is merely a late abuse of the name Punt, of no first-hand value, using ‘Punt’ as simply a stylistic variant for Tanetjer, ‘God’s Land’. Under Ptolemy II it is not used, as real Punt had long since disappeared as a political entity.

Postscript, July 2003 Suddenly, a newly announced Egyptian mention of Punt cuts out all attempts to locate original Punt in either Arabia or anywhere beyond the basic zone from the Red Sea (Suakin/Port Sudan) through the area adjoining the Sudan-Ethiopia border west to the Nile above or below Khartoum. Dr Vivian Davies (Davies 2003: 18-19; and forthcoming) and team have discovered a text in the sixteenth-century tomb-chapel of Sobek-nakht at El-Kab that indicates an invasion northward into Upper Egypt led by the ruler of Kush [based in area from 2nd to 4th cataracts of the Nile], who recruited levies from the tribes of Wawat, the Khent-hen-nufer districts, the desert-folk of the Medjaw, and from ‘the land of Punt’. Levies brought down as allies from the areas eastward of the Nile above the 5th cataract [western Punt] into Kush makes excellent sense. That such levies came over 2,000 miles of difficult travel from Arabia, or Somalia for a purely local war between Kush and the kings of Thebes in the sixteenth century BC would, clearly, have been a totally bizarre, not to say nonsensical, concept. So, this brief text indicates that basic and original Punt stays where it has been successfully located on good evidence for the last 30 years, and fantasy solutions can be discarded. Its name may through time have later been extended to reach increasingly further south (eventually even Somalia), as very guardedly suggested above, but this was long after the 16th century BC. References Bradbury L. 1996. Kpn-boats, Punt Trade, and a Lost Emporium. Journal of the American Research Center in Egypt 33: 37-60. Breasted J.H. 1906. Ancient Records of Egypt, I-V. (5 volumes). Chicago: University of Chicago Press; London: Luzac. Danelius E. & Steinitz H. 1967. The Fishes and other Aquatic Animals on the Punt-Reliefs at Deir el-Bahri. Journal of Egyptian Archaeology 53: 15-24. Davies, W.V. 2003. Sobeknakht’s Hidden Treasure. British Museum Magazine 46 (Summer 2003): 18-19.

Other items may oppose, rather than support, an Arabian Punt. The term mu-qed, ‘inverted (flow of) water’ is clearly the Red Sea under Ramesses III, and the land equivalent in the papyrus of Ramesses IX (Meeks: 329) is equally clearly the adjoining Red Sea coast, as it deals with desert expeditions sent out from Thebes, not in the north. The other mu-qed is the Euphrates. As for the Tell Defenneh stela (Twenty-Sixth Dynasty), there is a clear, 30

K.A. KITCHEN: THE ELUSIVE LAND OF PUNT REVISITED Periplus/L. Casson (ed. and transl.) 1989. The Periplus Maris Erythraei. Text with Introduction, Translation, and Commentary. Princeton: Princeton University Press. Saleh, A.A. 1972a. The Gmbtyw of Thutmosis III’s annals and the south Arabian Geb(b)anitae of the classical writers. Bullletin de l’Institut français d’archéologie orientale au Caire 77: 245-262. ----------1972b. Some problems relating to the Pwenet reliefs at Deir el-Bahari. Journal of Egyptian Archaeology 58: 140-158. Sayed A.M.A.H. 1977. Discovery of the Site of the 12th Dynasty Port at Wadi Gawasis on the Red Sea Shore. Revue d’Egyptologie 29: 140-178, pls 8-16. ----------1978. The Recently Discovered Port on the Red Sea Shore. Journal of Egyptian Archaeology 64: 6975. Sethe K. 1930. Urkunden der 18. Dynastie. (Urkunden des Ägyptischen Altertums 4/3). Leipzig: Hinrichs. ----------1933. Urkunden des Alten Reiches. (Urkunden des Ägyptischen Altertums 1). Leipzig: Hinrichs. Seyfried K.J. 1986. Zwerg. Columns 1431-1435 in W. Helck & E. Otto (eds), Lexikon der Ägyptologie (6/9). Wiesbaden: Otto Harrassowitz. Sinclair P.J.J., Morais J.M.F., Adamowicz L. & Duarte R.T. 1993. A perspective on archaeological research in Mozambique. Pages 409-431 in T. Shaw, P. Sinclair, B. Andah & A. Opoko (eds), The Archaeology of Africa: food , metals and towns. (One world archaeology 20). London/New York: Routledge. Smith W.S. 1962. The Land of Punt. Journal of the American Research Center in Egypt 1: 59-60. Wicker F.D.P. 1998. The Road to Punt. The Geographical Journal 164: 155-167. Wilkinson T.A.H. 2000. Royal Annals of Ancient Egypt. The Palermo Stone and its associated Fragments. London: Kegan Paul International.

----------forthcoming. Proceedings of the Colloquium on the British Museum and Ancient Egypt, 15-16 July, 2003. Dayton J.E. 2002. [Poster-presentation on Punt, privately issued.]. Fattovich R. 1991. The Problem of Punt in the Light of Recent Fieldwork in the Eastern Sudan. Pages 257272 in S. Schoske (ed.), Akten des Vierten Internationalen Ägyptologen Kongresses München 198. (4 volumes)/vol. 4., Hamburg: Helmut Buske Verlag. Fattovich R. 1993. Punt: the archaeological perspective. Pages 399-405 in Atti di Sesto Congresso Internazionale di Egittologia. (2 volumes)/vol.1. Torino: International Association of Egyptologists. ----------1995. L’archeologia del Mar Rosso: problemi e prospettive. Note in margine alla recente pubblicazione di due siti costieri della Somalia settentrionale. Annali dell’Istituto Universitario Orientale di Napoli 55: 158-176. Hepper F.N. 1969. Arabian and African Frankincense Trees. Journal of Egyptian Archaeology 55: 66-72. Herzog R. 1968. Punt. Abhandlungen des Deutschen Archäologischen Instituts Kairo, Ägyptologische Reihe, 6. Glückstadt: Augustin. Kitchen K.A. 1971. Punt and How to Get There. Orientalia NS, 40: 184-207. ----------1993a. The land of Punt. Pages 587-608 in T. Shaw, P. Sinclair, B. Andah & A. Opoko (eds), The Archaeology of Africa: food, metals and towns. (One world archaeology 20). London/New York: Routledge. ----------1993b. Ramesside Inscriptions Translated and Annotated, Notes & Comments, II. Oxford: Blackwell. ----------1993c. Ramesside Inscriptions Translated and Annotated, Translations, I. Oxford: Blackwell. ----------1994. Documentation for Ancient Arabia, I. Liverpool: Liverpool University Press. ----------1996. Ramesside Inscriptions Translated and Annotated, Translations, II. Oxford: Blackwell. ----------1999a. Further Thoughts on Punt and its Neighbours. Pages 173-178 in A. Leahy & J. Tait (eds), Studies on Ancient Egypt in Honour of H.S. Smith. (Occasional Publications, 13). London: Egypt Exploration Society. ----------1999b. Ramesside Inscriptions Translated and Annotated, Notes & Comments, II. Oxford: Blackwell. Lucas A. (ed.) 1962. Ancient Egyptian Materials and Industries. (Fourth edition revised and enlarged by J.R. Harris). London: Edward Arnold. Meeks D. 2002. Coptos et le chemin de Pount. Topoi, Supplément 3: 267-335. ----------2003. Locating Punt. Pages 53-80 in D. O’Connor, S. Quirke (eds), Mysterious Lands, Encounters with Ancient Egypt. London: UCL Press. Moorey, P.R.S. 1994. Ancient Mesopotamian Materials and Industries: The Archaeological Evidence. Oxford: Clarendon Press. 31

32

Pharaonic Egypt and the Red Sea arms trade D. M. Dixon Throughout history a major factor in producing conflict between groups ranging from tribes to empires has been, and still is, competition for access to, and control of, certain raw materials deemed by the parties involved to be vital to their interests. Of course, the nature of these materials has varied at different periods. At the present time the vital commodity is oil. During the sixteenth and seventeenth centuries it was – unbelievably – pepper (Boxer 1969: 59-61).1 In Pharaonic Egypt it was certain aromatic substances that were considered essential. Such substances, all of plant origin, had been widely used in religion, medicine and magic, cosmetics and other areas of daily life from a very early date. For example, in graves, of both sexes, dating from the Predynastic period (i.e. prior to c .3000 BC)2 small lumps of a resinouslooking material were very often placed near the deceased’s head and sometimes in proximity to toilet objects such as palettes and eye-paint materials (galena or malachite).

which was doubtless its native name as it sounded to Egyptian ears. The great importance of antyu to the Egyptians lay in the fact that, apart from its use among a relatively privileged few as an ingredient in medicinal and cosmetic preparations (von Deines & Grapow 1959: 99-104), it was employed in the service of the gods on whose continued benevolence the power and prosperity of Egypt was thought to depend. Its acquisition was therefore not a matter of luxury but of major importance for the state. In Pharaonic times the only geographical source of antyu was, in general, a region or area designated in Egyptian texts from the Old Kingdom onwards as the land of Punt (or Pun or Pwâne),4 also known as ‘God’s-land’ (TaNetjer). In very general terms, Punt lay in the Red Sea region primarily on the African coast and adjoining hinterland (Kitchen, supra).5 As far as the evidence goes, Punt was

As far as aromatic substances in general were concerned, it was not necessary to travel far afield to obtain them. For example, at least as late as the Middle Kingdom (2055-1650 BC) species of Pistacia trees, which yielded the aromatic resin known in Egyptian as sonter (sntr), grew wild in the deserts east and west of the Nile Valley in the latitude of Egypt itself (Loret 1949).

As the Egyptians travelled further afield, they may well have encountered other trees producing gum-resin, or other varieties of the same trees. Expeditions may have been accompanied by one or more specialists, perhaps priests, who would identify and assess the quality of various aromatic products. But these men would not, of course, be scientific botanists in any modern sense, and provided a particular material looked and felt, and above all smelt, like antyu, then as far as they were concerned it was antyu. We should not expect anything like total consistency in the ancient naming of plants and plant products, especially over a period of several millennia. 4 A few references do occur to antyu from elsewhere than Punt (Retjenu, in western Asia, and Nubia), but almost certainly the material had originally come from Punt (Dixon 1969: 55 n.5). 5 The possibility cannot be excluded that the Egyptians may also have visited the Arabian side of the sea. Vessels setting out from a port on the west coast of the Sinai peninsula may have travelled down the coast of Arabia. A number of the items said to have been obtained in Punt occurred, and occur, in southern and western Arabia – fragrant gumresins, an ebony sp., leopards, ostriches (Thomas 1932: 44 n.2, 55, 94, 147), Hamadryas baboons (Kummer 1995: 17-18, 233-5, 251-2), etc. – and in antiquity their range may have extended further west and northwest. In the southern part of the Red Sea south of about latitude 18°N there is a scatter of small islands that might, at some period, have served as ‘stepping-stones’ between the two coasts.

In Pharaonic times the choicest and therefore most sought-after aromatic material was an imported fragrant gum-resin, reddish-brown in colour, produced by one or more species of trees belonging, in all probability, to the Family Burseraceae. It was known as antyu (Κntyw),3 1

To the Portuguese in particular, pepper was very important. Even after the arrival in the East during the seventeenth century of the Dutch and English, ‘as late as 1611 it was officially stated at Lisbon that pepper was still the basic commodity of the Portuguese India-trade, and the only one which yielded a satisfactory profit to the Crown.’ (Boxer 1969: 59-61). 2 The dates in this paper are taken from Shaw 2000: 479ff. 3 Despite a number of studies based on textual data, the botanical identity or identities of antyu has yet to be determined. Some light might be shed on this question by the examination of the contents of a sealed container or containers labelled ‘antyu’, of known date and provenance, which has not been reused and which still retains a quantity of its original material. At least one such vessel does exist: one of a set of eight small cylindrical alabaster oil or ointment jars recovered from the tomb at Dahshur of the Twelfth Dynasty (1985-1773 BC) princess Itwer is inscribed on the lid in blue paint ‘top-quality antyu’ (tpt nt Κntyw). It is now in the Cairo Museum (von Bissing 1904-07: 143, Nº.18664 and Tafel A). To the author’s knowledge, the surviving contents, if any, have not been analysed. However, they might at best only furnish an indication of the identity of antyu at the period in question, for antyu may not have designated precisely the same species, or even genus, throughout Egyptian history; indeed it is unlikely to have done so.

Once landed on the Arabian coast, expeditions would not have needed to undertake lengthy journeys inland, even if they had been permitted to do so; all they would have needed to do was to wait on or near the coast for the various products to be brought to them – which is, for the most part, what they doubtless did on the African side – for the approximate arrival dates of expeditions would from experience have been known in advance. It is also possible, though perhaps unlikely, that the occasional vessel or vessels sailing in the inner channel on the African side may incautiously have ventured through one of the openings in the barrier reef into the open sea and been blown off course.

33

TRADE AND TRAVEL IN THE RED SEA REGION it is clear that a very substantial quantity of antyu was involved, and this is an indication that Sahure’s expedition was almost certainly not the first Egyptian mission to Punt.

always reached by sea from a port or ports on the Egyptian coast, and perhaps occasionally from Sinai. The Egyptians of Pharaonic times were, broadly speaking, not a naturally curious people as far as geographical exploration was concerned. For ordinary Egyptians, i.e. something like ninety per cent or so of the population, their world consisted of their villages and immediate environs. What lay beyond, even within Egypt itself, did not normally concern them greatly, unless they were traders or administrators etc. Travel abroad, that is, outside the confines of the Delta and the Nile Valley, was viewed with distaste and fear, as it usually entailed periods of compulsory service in the army or expeditions to the mines and quarries of the Eastern and Western Deserts, Nubia, and Sinai. The deserts were lonely, fearsome places, the domain of the evil god Set and the abode of potentially hostile nomads, wild animals such as lions and leopards and of strange creatures, evil spirits and goblins (cf. Arabic Κifrīt).

The demand for antyu was quite insatiable. However, in common with most ancient peoples, the Egyptians had little or no appreciation of the need for conservation.7 They ruthlessly exploited natural resources of all kinds. In the case of antyu, they were probably able to pressure local producers lying nearer to Egypt (wherever they were) into over-exploiting their trees, perhaps against their better judgement, until they were exhausted. Egyptian expeditions were then compelled, as time went on, to travel ever further afield to procure supplies, and such overbearing behaviour would become more difficult the further they were from home. ‘Pun(t)’ therefore did not designate a fixed locality or area, with more or less clearly defined boundaries, throughout Egyptian history, but was a shifting term. Originally it was probably a common noun in one or more of the languages of the region, perhaps meaning something like ‘strand’, ‘seashore’, ‘trading-place’, etc.8 Later one or more of these ‘puns’ became the Pun, par excellence as it were, and hence a proper name.9

Nevertheless, expeditions would travel, however reluctantly, to virtually anywhere, despite natural and man-made obstacles, if it was known or suspected that valuable items or raw materials were to be found there. There are, in fact, very few if any deposits of raw materials – stone, metals – in the areas mentioned which were not exploited by the Egyptians at some period or other. However, they would not go further than they needed or stay longer than was necessary6 and Punt was doubtless no exception. It was not that the Egyptians were poor sea-going sailors; on the contrary, the personnel of the Egyptian vessels that plied the waters of both the Mediterranean and the Red Sea in the New Kingdom and earlier were, as far as is known, mostly Egyptians. But it is broadly true that unlike the ancient Greeks, the Egyptians were not motivated to voyage over distant seas merely from curiosity.

As the Egyptians voyaged further afield, a major and inevitable consequence was that their lines of communication became increasingly stretched, and their 7

However, the care bestowed on incense-trees by the modern inhabitants of Dhofar which is doubtless the product of centuries of experience, indicates that in antiquity the forebears of these people did appreciate the need to conserve their most valuable commercial commodity. In the case of the ‘Shazari tree of mughur’, a variety of Dhofar frankincense which begins to bear in its third or fourth year, the gumresin is collected ten days after slight incisions have been made in its low, stout branches. The tree then continues to yield from these incisions, deepened as necessary at intervals of ten days, for a further period of five months. After that, however, the tree dries up, and must be left to recover, the period varying, according to its condition, from six months to two years (Thomas 1932: 122).

In their quest for antyu, therefore, they naturally exploited first the sources of supply nearest to Egypt – wherever they were. On the fragment of Old Kingdom royal annals, now known as the Palermo Stone, it is recorded that Sahure, the second king of the Fifth Dynasty (c. 2482 BC), obtained 80,000 units of antyu, as well as gold in unspecified quantity, from Punt (Sethe 1933: 39). This is the first mention of Punt in Egyptian texts. Whatever the size or capacity of an individual unit,

In Somaliland the tapping of frankincense trees (Boswellia spp.) is suspended every fifth or sixth year, ‘for it is realized a resting period is necessary for the trees if they are not to be exhausted.’ (Howes 1949: 151). 8 There is a Swahili word pwani, which in the ordinary dialect of Mombasa means ‘coast’, ‘seashore’ (Stigand 1913: 332; Meinhof 1942: 300-302). The author is not suggesting the presence in the Red Sea region in the New Kingdom of Swahili or ‘proto-Swahili’ speakers, though in view of our ignorance of the linguistic situation in the area, the possibility that one or more cognate languages or dialects may have been spoken in parts of the region cannot be entirely excluded. The resemblance of pwani to ‘Punt‘ may of course be entirely fortuitous, and one must in any case be wary of making comparisons on the basis of what may be superficial similarities presented by transliterated and conventionally vocalised forms. That said, however, the consonantal sequence p-n is interesting; the final -t in Egyptian Pwnt is merely the feminine ending. According to Meinhof (loc.cit.), pwani is pronounced with an aspirated p: phwani. 9 The name itself may have survived in the Oponê of the geographer Ptolemy, cf Arabic (Ras H)afun, a promontory in the former Italian Somaliland, facing the Indian Ocean (Alliot 1951).

The fact that some of the animals depicted in the Deir el-Bahri reliefs are African spp. is not necessarily significant, for not every expedition, even those which landed in Africa, would have encountered giraffes and rhinoceroses! That said, however, it is probable that the bulk of Egyptian activity in the Red Sea did take place on the African side. Apart from any other consideration, that is where Egypt itself lies, and it would have been so much easier to follow the same coastline. 6 Of course, allowance must always be made for exceptional individuals, such as some of the Sixth Dynasty barons of Elephantine, who may have been motivated to undertake their journeys deep into the Sudan as much by curiosity as by the command of Pharaoh.

34

D.M. DIXON: PHARAONIC EGYPT AND THE RED SEA ARMS TRADE Be that as it may, the policy adopted by the Egyptians to secure their aims was that subsequently pursued over the millennia by other powerful states geographically distant from sources of raw materials essential to their national interests (to this point we shall return): in the areas concerned they ensured the survival in power of a ruling group or groups, perhaps relatively small, who were welldisposed towards them. They did so by supplying them with modern military hardware – in the case of Punt, various kinds of metal, probably bronze, weapons – and they may well have had a small, more or less permanent military mission in Punt (probably relieved periodically) to train the forces of the local paramount chief.

capacity, if need ever arose, to exercise military force in or around ‘the land of Punt’ to protect their interests was progressively reduced. In fact, as far as is known, no part of Punt ever was conquered at any period. The problem, then, facing the Pharaonic government was how to ensure the continuance of antyu supplies, in other words, how to gain and retain the goodwill and cooperation of the ruler or rulers of the increasingly distant places they now had to visit, and how to ensure peace and stability in those places and safeguard vital supply routes. The area where the trees that yielded the best quality antyu grew is termed in New Kingdom texts ‘the antyuterraces’ of Punt and clearly designated elevated ground rising inland away from the coastal plain in a series of ‘steps’ or stages, either natural or to some extent artificially constructed. It is represented by the hieroglyphic sign depicting a stairway, which is used as the determinative of the word khetyu (htyw) ‘terraces’. The ‘terraces’ are nowhere depicted in surviving reliefs or paintings and their distance from the coast is not known, but it may not have been all that great.

The supply of military material, however, formed only part, albeit the most important part, of a wider Egyptian policy which in modern jargon could be described as a concerted ‘hearts and minds’ programme designed to secure long-term Puntite friendship and co-operation by attracting the population, or at least that part of it which counted, to the delights and benefits of Egyptian material culture. More on this later. The acquisition of antyu, Egyptian travel in the Red Sea area, and the export to Punt of military material were thus inextricably linked.

One obvious way of making certain that antyu would be readily available would have been to successfully acclimatise the tree(s) that produced it in Egypt, thus obviating the need to travel to Punt at all, or at least to go there so frequently. During the New Kingdom, from at least as early as the reign of Queen Hatshepsut (14731458 BC), determined and persistent efforts were in fact made to do this (Dixon 1969). Trees were imported, complete with roots and packed in large baskets, and transplanted in Thebes.

At what point the Egyptians adopted this multi-faceted policy is uncertain, but the decision to do so would obviously have been taken once they realised that they were geographically and therefore militarily overstretched. This may have occurred prior to the New Kingdom,11 but the earliest specific evidence dates from towards the middle of the Eighteenth Dynasty and relates to the expedition despatched to Punt in year nine of Queen Hatshepsut (c. 1465 BC). The expedition is recorded in the splendid painted reliefs and texts in her limestone terraced temple at Deir el-Bahri, Thebes (Naville 1898: pls 69-79).

However, despite some apparent initial, temporary success, in the long term the venture proved unsuccessful; indeed the continuance of the Punt voyages and the very persistence of the attempts at transplantation, which lasted until the reign of Ramesses ΙΙΙ (1184-1153 BC) in the Twentieth Dynasty, only serve to underline the Egyptians‘ failure. The reason(s) for this failure are probably complex. Climatic and environmental factors were doubtless important, but it is also possible, even likely, that the Puntites, fearful of losing a lucrative trade if the expeditions were discontinued, quietly sabotaged the Egyptians‘ efforts, perhaps by tampering with the roots of the trees and/or giving incorrect or misleading information regarding their care and cultivation.10

Among other events the arrival is shown of the Egyptian ships, five in number, at the water’s edge. From one of the vessels a smaller longboat is being loaded by the Egyptian crew with various items: large two-handled beer and wine-jars, each capped with a conical mud jarsealing, and sacks of various commodities such as dried dates and figs and salted meat, all mentioned by name in

11 This depends on how far south earlier expeditions may have penetrated. This question cannot be examined in detail here, but reference should be made to an interesting paper by G.A. Wainwright (Wainwright 1946: 31-38) in which he suggested that an island, clearly envisaged as being in the Red Sea, on which the hero (the ‘Shipwrecked Sailor‘) of a Middle Kingdom literary tale was cast ashore may be identified as St. John’s Island, or Zeberged/Zabarjad, a volcanic island lying 32 miles south-east of the promontory of Ras Banas in latitude 23º 26´N.

10 Some support for the ‘sabotage theory’ may be afforded by the fact that it is, apparently, possible, no doubt with great care, successfully to transplant some incense-trees to very different places. In the fourth century BC Theophrastus (Enquiry into Plants IX, 4, 7-9; Hort 1916: 240-1) describes a frankincense tree that was growing in a sacred precinct in Sardes in Asia Minor, that yielded frankincense. Not far from Salalah, in southern Arabia, in the 1950s, were ‘four hundred flourishing frankincense trees’ which had been transplanted from the Dhofar steppe. There are also descriptions of Boswellia plants from Somaliland growing in botanical gardens in Aden and Bombay (Groom 1981: 113-114).

If Wainwright’s well-argued identification be correct, and Egyptian expeditions of the Middle Kingdom did indeed reach latitude 23º and beyond, problems of over-stretched lines of communication may already have begun to surface at that time.

35

TRADE AND TRAVEL IN THE RED SEA REGION

Figure 6: Painted relief from the temple of Deir el-Bahri (Naville 1898: pl. 72)

Figure 7: Painted relief from the temple of Deir el-Bahri; the Pharaonic envoy meets the Chief of Punt and his wife (Naville 1898: pl. 69). 36

D.M. DIXON: PHARAONIC EGYPT AND THE RED SEA ARMS TRADE Few methods of gaining gratitude and hopefully lasting goodwill are more effective than the timely provision of medical assistance, particularly in a crisis or emergency; and the reputation of Egyptian doctors and Egyptian medicine stood high in the ancient world. There are, it is true, no references to medical staff on any of the Punt expeditions, but then very few details are available about the composition of the expeditions’ personnel. Medical and paramedical personnel were, however, a regular feature of missions to the desert mines and quarries, and it would have been strange if no provision had been made for the health and welfare of expeditions to so hazardous a region as the Red Sea. Egyptian medical skill would certainly have been made available to at least the higher echelons of Puntite society.

the accompanying texts along with ‘every good thing of the land of Egypt’. (Fig. 6). As the Pharaonic envoy, escorted by a small but impressively armed detachment of troops, comes ashore, he is met by the Chief (or Wr ‘Great One’) of Punt, Parehu who advances bowing slightly with arms raised in deferential greeting (Fig. 7). On his face is an ingratiating smile, well captured by the Egyptian artist accompanying the expedition.12 Behind the Chief is his enormously fat wife13 and other members of his family, and retinue. Parehu is wearing an Egyptian-style kilt doubtless of linen, in the waistband of which is tucked an Egyptian dagger, both items evidently the products of an earlier Egyptian mission.14 Placed on the ground before the Egyptian envoy is a small low table on which are displayed samples of the more prominent and valuable objects brought by the expedition: necklaces, probably of blue faience beads (highly prized by African rulers even as late as the nineteenth century (cf. Speke 1863: passim), and personal ornaments of metal, etc. But two items in particular must have brought a sparkle to Parehu’s eyes, namely a fine axe, presumably of bronze, identical with those carried by the Egyptian troops, and lying alongside it a sword, or large dagger, in its sheath. Round the base of the hilt is a length of cord, probably for attaching the weapon to the wearer’s waist. These weapons were no doubt part of a substantial consignment, which would also have included bronze (or copper) spears, daggers, arrowheads, superior quality bows, and shields.

As far as is known, none of the items, arms included, brought to Punt by the Egyptians were obtainable at this period by the recipients from any other source than Egypt. As a visible and permanent reminder to the Puntite populace of Pharaoh’s power and of the wisdom of staying on friendly terms with Egypt, Hatshepsut‘s expedition erected a red granite statue, presumably made in Egypt and at least life-size, of the Queen and the god Amun. It was set up ‘in the face of’ (or ‘in front of’, m hnt) that is, facing, the antyu-terraces (Sethe 1906: 316ff.), though at what distance from them is not clear. Where the statue, or the fragments thereof, now lie must be a matter of speculation. Over the millennia stretches of the Red Sea have gradually narrowed with the build-up of coral formation. If the statue was set up on or near the coast, it may now be some distance inland.

Many of the Puntites are shown wearing Egyptian kilts and wigs, with headband and filet, over their own hair. This garb (comparable perhaps to baseball caps and designer jeans) evidently appealed to them.

What, then, were the consequences for all concerned of Egyptian policy in the Red Sea and in particular of the arrival in the region of Egyptian weaponry?

They also wear a long, straight up-turned beard, the socalled ‘divine beard’ worn in Egypt only by deities and Pharaoh, himself a god. Was this too a feature acquired from Egypt? Did Puntites learn about the beard from their Egyptian visitors or observe it on imported statuettes of deities (no examples of which have so far come to light from Punt)? Or had this beard long been a feature of indigenous Puntite adornment? If the latter, what are the implications for the origins of the Puntites and their possible relationship to the Egyptians? Both peoples are depicted as physically virtually identical. These questions, however, cannot be pursued here.

Before considering the position in antiquity, it may be instructive to look briefly at circumstances in East Africa during the early and later nineteenth century AD. At this time thousands of firearms were flowing unrestricted into the region, into the hands of tribal chieftains. There was a close link between this inflow of weapons and the development and use of new types of firearms in Europe (Beachey 1962: 452). The brief Austro-Prussian war of 1866 saw the first use of the breech-loading Snider rifle, and in Africa it was the Snider in the hands of Napier’s troops on the march to Magdala in 1868 that gave the Abyssinians such an unpleasant surprise at Arogi.15 By 1878 all the major European powers had adopted various types of breechloaders – the Chassepot, Snider, or Martini-Henry. All these used metallic cartridge cases and steel barrels in place of iron as in the old muskets (Beachey ibid.). In consequence of the change to breech-loaders, huge

12 Not very visible in the reduced illustration reproduced here but clearer in Naville’s original publication (Naville 1898: pl. 69). 13 Although this is not the place to discuss in detail the possible reason(s) for her dimensions, it may be noted that before any further medical ‘diagnoses’ are offered of her supposed pathological condition, further consideration should be given to a possible cultural explanation (cf., e.g., Scholz 1984: 529-556; Speke 1863: 172, 189-90). 14 Hatshepsut has thus given the lie to her own boastful claim that she was the first ruler to open trade with Punt!

15

37

Hozier 1869: 193; Myatt 1970: 63-64, 139-41

TRADE AND TRAVEL IN THE RED SEA REGION with a wide variety of other commodities.16 There may, of course, have been some variation in the quantities and regularity – for example, possibly during the Amarna period and during some very brief New Kingdom reigns – but it appears that throughout the New Kingdom until and including the reign of Ramesses ΙΙΙ (1184-1153 BC), a period of, say, three hundred years, Egyptian policy in the region achieved its aims.

numbers of now obsolete muzzle-loading firearms became available for export – to Africa where there was an insatiable demand, on the part of African rulers and Arab slavers alike, for firearms of any kind. They conferred on their possessors’ enormous prestige and power in warfare. To begin with, the obsolete muzzle-loaders exported to Africa consisted of flintlock or percussion-type muskets, ‘gas-pipes’, as they were known. These were liable to explode in the hands and were therefore often as dangerous to the user as to his intended victim. When discharged, however, they made a great noise, which was in itself, along with the smoke, something of a deterrent. Many natives liked to use them for firing salutes, which they did as frequently as their supplies of ammunition allowed. When it came to fighting, however, many men, even when they carried a firearm, still preferred to use their tried and trusted traditional bows and arrows and spears.

Equally satisfied would have been Parehu and his successors. As already noted, as far as is known, there were no other sources from whence they could, at this period, have obtained modern weapons and the same was true for the other items delivered by the Egyptians. It paid the Puntite chiefs to keep on good terms with the Egyptians – just as it paid them, as noted earlier, quietly to frustrate their efforts to grow antyu-trees in Egypt! As certain modern states have found to their cost, one of the risks of arming distant surrogates or satellites is, and was, that a sudden change of ruler(s) or an internal upheaval, which the presence of a small military mission might have been unable to prevent, might result in weapons being used against the very people who had supplied them. There is, however, no evidence that this occurred in Punt and little likelihood that it did, since whoever was in power there at any given time would know that his or their continued prosperity, and even position, depended on the maintenance of friendly relations with Egypt.

As long as the firearms available consisted solely of these potentially dangerous muzzle-loaders, there was, as it were, a fairly ‘level playing-field’ between tribes as far as weapons were concerned. However, hardly had the breech-loader been perfected in Europe when various types of magazine or repeating rifles were introduced and by 1888 all the major powers had re-equipped their armies with them, thus rendering obsolete, and available for export, large quantities of breech-loaders. To the alarm of prescient European observers, breech-loaders began to arrive in East Africa only a year after they had been superseded in Europe. Breech-loaders conferred considerable superiority of firepower. In due course, inevitably the magazine-rifle too arrived in Africa. But let us return to the Red Sea in antiquity.

For the inhabitants of territories adjacent to Punt, or even further afield, the arrival of modern Egyptian weapons in the hands of the Puntites would have been a disturbing development. Captives from the interior of Africa had been arriving in Punt long before the New Kingdom, though probably on a relatively modest scale. Among the more exotic of these was the occasional pygmy or deneg (dng) (cf. Amharic denk, meaning in that language ‘dwarf’).17 As early as the reign of DjedkarêΚ-Isesi of the Fifth Dynasty (2414-2375 BC) the head of an Egyptian mission, the Treasurer of the God, Bawerded, had obtained a pygmy in Punt (wherever it then lay).18 He had probably come originally from deep within the interior, possibly even from as far away as the swamps of the Bahr el-Ghazal region in the Sudan. The function of this

While the various inhabitants of the area were all using weapons of stone, wood, and reeds (for arrows), some sort of balance of power, however fragile, may have prevailed, but once metal weaponry from Egypt (roughly comparable with the nineteenth-century breech-loaders) came into the hands of one group – the Puntites – the situation changed. Examination of bodies of battlefield casualties from ancient Egypt has revealed just how lethal even arrows tipped with hardwood points could be (Winlock 1945: plates, passim). How much more so, then, bronze arrowheads fired from powerful new Egyptian bows!

16 Including other aromatic substances: ihmut, an unidentified (?) incense-material (determined by three round pellets) and the already mentioned sonter (sntr), i.e. Pistacia resin. The Red Sea sonter was accepted simply because, along with a long list of other miscellaneous items, it was on offer; but the Egyptians would not have voyaged all the way to Punt for sonter alone, even if, as seems probable, the sontertrees that once grew in the deserts of Egypt had by the New Kingdom been exhausted. Not only was sonter, though sometimes coupled in the texts with antyu, less highly regarded, but it was also available during the New Kingdom in large quantities from Pistacia spp. growing in western Asia (Loret 1949). 17 Egyptian distinguished between deneg ‘pygmy‘ – the name doubtless came to Egypt with the people it designated – and nemiu ‘dwarf‘. Achondroplasiac dwarves occurred among the population of Egypt itself. On pygmies in Egypt see Dasen 1993: 27ff. and passim. 18 Sethe 1933: 129; Vycichl 1957: 48-49.

For the Egyptians the export of arms to Punt produced eminently satisfactory results: the security of the antyuterraces and the routes linking them to the coastal region seem to have been maintained; steady, and presumably sufficient, supplies of antyu arrived in Egypt together

38

D.M. DIXON: PHARAONIC EGYPT AND THE RED SEA ARMS TRADE

Figure 8: Egyptian axes were used for ‘cutting hbny-wood’ (Naville 1898: pl. 7). (mr(y)t,20 with man- and woman-determinatives) together with their children’ (Sethe 1906 329). These are, of course, recorded from the viewpoint of Egyptian scribes who probably knew little, and cared less, about the precise ethnic and geographical origins of these unfortunates. In this laconic reference is encapsulated the misery wrought among surrounding peoples with the help of imported Egyptian weapons.21

poor creature was to perform before Pharaoh ‘the dances of the god’. In the Sixth Dynasty another pygmy was brought to Egypt in the reign of Pepy II (2278-2184 BC), probably from the same general area as the first, but this one was conveyed, amid tight security and stifling solicitude, down the Nile through the Sudan. He had been acquired by the Yamites, a tribe living in the region of the Fourth Cataract, who had in turn obtained him from even further afield, ‘the land of the Horizon-dwellers’. They handed him over to the head of an Egyptian trading mission, the Elephantine frontier baron and caravanleader Harkhuf (Sethe 1933: 128).

The flow of captives to Egypt from the Red Sea region was probably relatively less than that from the Sudan and Nubia by the more direct Nile route, for the primary objective of the Punt voyages was, as already indicated, antyu; all the other goods listed, slaves included, were ‘optional extras’ as far as Punt was concerned, being readily obtainable elsewhere in Pharaoh’s African empire, and they would be accepted as space on board ship allowed. That said, however, probably few captives were refused and they were doubtless crammed in like the proverbial sardines. Puntite slave-raiding can only have had the same deleterious effect on the political, economic, and social structure of neighbouring tribes as such razzias had elsewhere in Africa in later centuries.

The arrival in the Red Sea area in the New Kingdom of metal weapons must have greatly enhanced the Puntites‘ capacity for intimidating and raiding their neighbours near and far. It was not only for ‘cutting hbny-wood in very great quantities’ (Fig. 8) that Parehu used his Egyptian axes etc. Fragments of painted reliefs at Deir elBahri show tall, slim Nilotic-looking negroes and negroids bearing logs of ornamental wood, probably hbny, and leading hunting-dogs on leash (Naville 1898: pl. 71).19 They had probably travelled to the coast, possibly under compulsion, to deliver products of their homeland. Apart from their products, however, these people were themselves very desirable items of commerce; for what Puntite ruler equipped with overwhelmingly superior weaponry could have resisted the opportunity to acquire, with relatively little effort or resistance, profitable items of human cargo for the Egyptian market?

20 mr(y)t ‘servants’, ‘underlings’; but here certainly ‘slaves‘, prisoners destined for permanent service in captivity in Egypt. (Bakir 1952: 2225). 21 The existence of such activities by the ancient Egyptians and/or their surrogates, as well as by later rulers of Egypt and the Sudan, (Lewis 1990: espec. 50ff.; Hill 1959: 62ff., 101ff.) must at least call into question the validity of recent demands by some African countries for an apology and financial ‘reparations’ from certain European states for the slave trade. While European participation in this odious traffic cannot be denied, it is clear that slavery had been in full swing for millennia before the first Europeans set foot in Africa. Perhaps applications should be addressed, in the first instance, to those who now hold sway in the successor-capitals to ancient Memphis, Thebes, and Napata.

At the very end of the long list of ‘wonders’ brought to Egypt from Punt by Hatshepsut’s expedition are ‘slaves 19

For an attempt to place these fragments in their original positions, see Smith 1962.

39

TRADE AND TRAVEL IN THE RED SEA REGION At one time the United States was self-sufficient in oil, but with rapid industrialisation and the arrival of the automobile, combined with prodigal consumption, homeproduced supplies of oil became quite inadequate and the American government had therefore to find supplementary sources. In so doing it followed, albeit unconsciously of course, the policy pursued millennia earlier by the Pharaohs: initially it was the nearer sources of oil that were exploited, in this case in central and northern Latin America and the Caribbean. When these too proved insufficient to meet ever-increasing American needs, the government turned its attention increasingly to the seemingly limitless resources that lie beneath the arid sands of the Near East.

At first glance the depiction of just one axe and one sword among the items presented to the chief of Punt might appear rather flimsy evidence from which to infer the existence of Egyptian arms exports to the Red Sea region. Might not the two weapons shown have been simply ceremonial or diplomatic gifts to a friendly ruler for his own use? Gifts of weapons were indeed occasionally made by Pharaoh to foreign rulers. In the present context, however, such gifts would by themselves hardly have furthered, in practical terms, Egyptian objectives in the area. Clearly all the items on display, weapons included, are but samples of larger quantities. Given the small size of the table, only the more prominent weapons could be displayed; one would not therefore expect to see small arrowheads, while at the other end of the scale large weapons like bows and cases and spears could not readily have been accommodated on a small surface. Of the categories of objects shown – beads, bangles and weapons – the last-named constitute a good third, surely an indication of the importance attached to them. Finally, given the circumstances already discussed, how could the Egyptians have successfully pursued their objectives in the region over a period of nearly three centuries other than by the policy described?

This region, however, is not only a very long way from the United States but has also long been, in varying degree, politically volatile. Peace and stability in the area, and the presence of friendly régimes, were and are therefore essential if American oil supplies are to be safeguarded; and here again the Americans in large measure have emulated, mutatis mutandis, the Egyptians. They supplied friendly, albeit in some cases very unsavoury, régimes in the region with modern arms and equipment together with American military personnel to train local forces in their use and inspire confidence among regional rulers that in times of trouble the United States would, hopefully, assist them.

In conclusion, it will not be without interest or relevance briefly to compare with the policy and methods of Pharaonic Egypt those of another great imperial power, widely separated in time and space, in the same very general region.22

Of course, the relatively recent advent of huge long-range aircraft each capable of rapidly transporting large numbers of troops and heavy equipment to virtually every corner of the globe has enabled the United States to do what was quite beyond the capacity of Pharaonic Egypt, namely to take forceful action against distant lands or groups judged to pose a threat to its oil supplies and strategic interests, or states and powerful individuals formerly well-disposed who were no longer fulfilling, for whatever reason(s), the rôle expected of them; and to intervene in areas in the exploitation of whose oil resources the Americans hope eventually to participate.

What antyu was to ancient Egypt, oil is to the United States,23 the vital raw material on which its industrial and military might is very largely dependent and the acquisition of which is therefore a major driving force behind much of American foreign policy in various parts of the world. As Mr Lawrence Korb, Assistant Defense Secretary, famously remarked at the time of the Iraqi invasion of Kuwait in 1990, “If Kuwait grew carrots we wouldn’t give a damn” – a public expression of candour he may subsequently have regretted! (International Herald Tribune, 21 August 1990).24

In its determined pursuit of its interests in the Near East, the Red Sea region, and the Horn of Africa, the American Empire is no different from any other imperial power from Pharaonic Egypt onwards. The latter, however, did have one advantage denied to the United States: however dimly aware the ill-used peoples of Africa and the Red Sea region in antiquity may have been of the ultimate source of their woes, they were not in a position to strike back and certainly never able, at this period, to threaten the peace and security of the distant super-power on the lower Nile.

22 Naturally, as with most, if not all, historical comparisons, it is not totally exact, for history rarely, if ever, repeats itself in precisely the same way; but the comparison is close enough to warrant notice in the present context. 23 It does admittedly require some effort to equate, or compare, the importance of antyu to the ancient Egyptians with that of oil to the Americans (and other modern peoples), but in the author’s opinion the comparison is a valid one. 24 A number of other issues are conveniently available which can be pursued with varying intensity from time to time as circumstances require – for example, human rights violations, corruption, undemocratic government, internal oppression of women and ethnic minorities, alleged possession of nuclear, biological or chemical weapons, etc. To some extent, these issues may be compared to the material ‘optional extras’ available to the Punt expeditions: matters of some significance, certainly, but not in themselves necessarily vital to the United States itself and until recently, not necessarily directly affecting adversely its peace and security. The attacks of 11th September 2001, however, changed this situation overnight and

‘homeland security’ has now permanently and indissolubly joined the quest for oil as a top priority.

40

D.M. DIXON: PHARAONIC EGYPT AND THE RED SEA ARMS TRADE Shaw I. 2000. The Oxford History of Ancient Egypt. Oxford: Oxford University Press. Smith W.S. 1962. The Land of Punt. Journal of the American Research Center in Egypt 1: 59-60. Speke J.H. 1863. Journal of the Discovery of the Source of the Nile. Edinburgh; London: William Blackwood and Sons. Stigand C.H. 1913. The Land of Zinj: being an account of British East Africa, its ancient history and present inhabitants. London: Constable. Theophrastus/Hort A. (ed and transl.). 1916. Enquiry into Plants and minor works on odours and weather signs. (Loeb series). (2 volumes). Cambridge, Massachusetts: Harvard University Press; London: Heinemann. Thomas B. 1932. Arabia Felix: Across the Empty Quarter of Arabia. London: Jonathan Cape. Vycichl W. 1957. Amharique denk “nain”, égyptien d-ng. Annales d’Éthiopie 2: 248-249. ----------1967. Punt, Opone, Hafun-Das Weihrauchland der alten Ägypter? Mitteilungen des Instituts für Orientforschung 13: 45-46. Wainwright G.A. 1946. Zeberged: the Shipwrecked Sailor’s Island. Journal of Egyptian Archaeology 32: 31-38. Winlock H.E. 1945. The Slain Soldiers of Neb - hepet-Rê’ Mentu-hotpe. New York: Metropolitan Museum of Art, Egyptian Expedition. Zyhlarz E. 1942. Das Land. Pwn.t.Zeitschrift für Eingeborenen-Sprachen 32: 302-311.

Acknowledgements I am most grateful to Mrs Betty Barber who has patiently typed and repeatedly re-typed my text to produce this final version. Bibliography Alliot M. 1951. Pount-Pwâne, l’Oponé du Géographe Ptolémée. Revue d’Egyptologie 8: 2-7. Bakir A.M. 1952. Slavery in Pharaonic Egypt (Supplement aux Annales du Service des Antiquités de l’Égypte.Cahier 18). Cairo: Institut français d’archéologie orientale. Beachey R.W. 1962. The Arms Trade in East Africa in the late Nineteenth Century. Journal of African History 3: 451-467. Bissing F. von. 1904-07. Steingefässe (Catalogue général des antiquités égyptiennes du Musée du Caire). Vienne: A. Holzhausen. Boxer C.R. 1969. The Portuguese Seaborne Empire 1415-1825. London: Weidenfeld and Nicholson. Dasen V. 1993. Dwarfs in Ancient Egypt and Greece (Oxford Monographs on Classical Archaeology). Oxford: Clarendon Press. Deines H. von & H. Grapow. 1959. Wörterbuch der Ägyptischen Drogennamen (Grundriss der Medizin der alten Ägypter,VI). Berlin: Akademie Verlag. Dixon D.M. 1969. The Transplantation of Punt Incense Trees in Egypt. Journal of Egyptian Archaeology 55: 55-65. Groom N. 1981. Frankincense and Myrrh. A Study of the Arabian Incense Trade. London: Longman. Hill R. 1959. Egypt in the Sudan 1820-1881. London: Oxford University Press. Howes F.N. 1949. Vegetable Gums and Resins. Waltham, Massachusetts: Chronica Botanica Company. Hozier H.M. 1869. The British Expedition to Abyssinia. London: Macmillan and Co. Kummer H. 1995. In Quest of the Sacred Baboon, A Scientist’s Journey. Princeton: Princeton University Press. Lewis B. 1990. Race and Slavery in the Middle East. An Historical Enquiry. New York; Oxford: Oxford University Press. Loret V. 1949. La Résine de Térébinthe (Sonter) chez les anciens Égyptiens. Cairo: Institut française d’archéologie orientale. Meinhof C. 1942. Pwani. Zeitschrift für EingeborenenSprache 32: 300-302. Myatt F. 1970. The March to Magdala. The Abyssinian War of 1868. London: Leo Cooper. Naville E. 1898. The Temple of Deir el-Bahari. (pt. III). London: Egypt Exploration Fund. Scholz P. 1984. Fürstin Ιti - “Schönheit aus Punt”. Studien zur altägyptischen Kultur 11: 529-556. Sethe K. 1906. Urkunden der 18. Dynastie. (Urkunden des Ägyptischen Altertums 4). Leipzig: Hinrichs. ----------1933. Urkunden des Alten Reichs. (Urkunden des Ägyptischen Altertums 1). Leipzig:. Hinrichs. 41

42

Possible connections in antiquity between the Red Sea coast of Yemen and the Horn of Africa Edward J. Keall highlands.5 A probable cause was population pressures and a drying climate which denuded the high slopes of vegetation and initiated massive floods.6 At least five metres of sediments were built up and deposited far out to sea in a classic delta formation pattern (Fig. 10). Beginning more than two millennia ago, when the sediments were no longer accumulating, the sea began to erode the delta, exposing a sandy cliff that has been subject to tidal erosion at its base. A small mosque at the southern end of the delta (at the northern end of the alFazza bay) is currently on the verge of being undermined by the sea. Twenty years ago it was nearly 50 metres inland from the tidal beach.

Introduction The observations presented in this paper stem from the results of archaeological survey and excavation on the Tihāmah coastal plain of Yemen, sponsored by the Canadian Archaeological Mission of the Royal Ontario Museum (CAMROM) (see map, Fig. 9). The investigation was initiated in 1982, but prior to 1997 mainly Islamic period remains were encountered,1 except for the notable discovery of a large pre-Islamic site ten kilometres north of Zabīd.2 Following the chance encounter in 1997 of the megalithic site of al-Midamman, there was a major shift in the focus of the enquiry. The overall CAMROM project was now driven by the need to define the varying ways by which people dealt with the different conditions of landscape and climate that were prevalent during the last five thousand years.3 Following ephemeral activity in the area during the Neolithic, major human settlement occurred beginning towards the end of the third millennium BC.4 Occupation of the area reached its final and abrupt end somewhere in the early first millennium. There is no further archaeological trace of settlement for another 2500 years.

In the central portion of the sea-cliffs in the study area there is now a stratum of peat exposed at the bottom of five metres of sediment. Recovered from the former peat bog was a log of a mature tree that had died and collapsed into the peat. Radiocarbon analysis of the log provides a calibrated 14C age of 5070-4820 BC for the living tree. The terminus post quem date of c. 5000 BC represents the earliest possible time for the beginning of the sediment deposition. However, one must acknowledge that it may have taken some time before the dead tree collapsed into the peat. The c. 5000 BC radiocarbon date, therefore, is by no means incompatible with the idea that the onset of a drier climate and massive run-off flooding started after around 3000 BC (cf.fn. 3).

Ecological setting We may start from the basic premise that the landscape of this part of the Yemeni Red Sea coast has changed drastically since the time of the extensive second millennium settlement. It can now be confirmed that massive deposits of alluvium were transported across the coastal plain from the Wādī Zabīd drainage system, coinciding with an environmental disaster in the

The floods brought nutrient-rich sediments and water to the coastal edge of the plain. The environmental disaster in the highlands provided a temporary window of opportunity for settlement at the coast, beyond the subsistence level of a fishing community. Here there were now resources from the sea to be harvested, and the flood sediments to be exploited. Following the classic model of early irrigation farming, where water is utilised far downstream away from the potentially destructive force of the spate, only minimal diversions needed to be made from flood streams.7

1

Keall 1983; 1999a; 1999b; 2001. The site was originally reported in 1983 simply as ‘Gas Station’, because the site lay behind a then newly constructed petrol station (maΉaΓΓah) alongside the Zabīd highway, but did not otherwise have a local place name. At the time of its discovery, the surface pottery was tentatively identified as ‘Himyaritic’ (Keall, 1983: 385). In 1998, the author presented a more appropriate attribution of the pottery, i.e. the second-first millennia BC (Keall 1998: 143). The site was still referred to as ‘STN’ in 1996 (Ciuk & Keall 1996: 4). Confusion resulting from others using different terms for ‘Gas Station’, such as ‘Petrol Station’ and ‘Filling Station’, prompted the adoption of the name al-Kašawba. Muhammad al-Kašawba is the name of the filling station owner, and those wishing to visit the site should ask for MaΉaΓΓat al-Kašawba. 3 A three-year award by the Social Sciences and Humanities Research Council of Canada was made in 1999 to the main applicant Keall, along with the cited collaboration of Dr. Ingrid Hehmeyer, in support of ‘The changing ecology of Southern Arabia in the Holocene.’ See also Hehmeyer & Keall 1993, 1998; Hehmeyer 1995. 4 Keall 1997; 1998; 2000. 2

5

For the effect of the drying climate in the highlands, see Edens & Wilkinson 1998; Wilkinson 1999; Wilkinson & Edens 1999. Specific reference to the continued existence of a humic soil in the highlands (reflecting moist conditions) until around 3000 BC, and the onset of erosion in the subsequent drier period, is given in Wilkinson 1999: 190. 6 The sediments at the coast have been recorded by CAMROM, but the reports submitted for publication have yet to appear in print. The full topic was presented in a conference paper, as Hehmeyer & Keall 2000. 7 The people of al-Midamman do not appear to have attempted any major engineering, as would be the case for anyone coming from a Sabaean background. Such a connection would undoubtedly have left tangible traces on the landscape in the form of canals and sluice gates. For the practice of irrigation technology already as early as the second

43

TRADE AND TRAVEL IN THE RED SEA REGION

Figure 9: Location of archaeological sites mentioned in text

Figure 10: The sites and landscape setting of al-Midamman

44

EDWARD J. KEALL: POSSIBLE CONNECTIONS IN ANTIQUITY The most dramatic aspect of the site of al-Midamman is a group of standing stones dating to the end of the third beginning of the second millennium BC, based on a typological assessment of finds associated with the setting up the megaliths (Giumlia Mair et al. 1999; 2002). In the second millennium many of the former standing stones were pulled down to form the building blocks of monumental structures.8 The inhabitants of the site lived in ephemeral dwellings but produced good quality pottery and made tools of imported obsidian (Keall 2000: 724-725). Obsidian microliths have been documented right across the site (Rahimi 2001). A complete ‘tool kit’ of four different tool types was recovered in 2002 from immediately outside one of the monumental buildings. From the same context, gold beads (Fig. 19), painted ostrich egg shells (Fig. 20), and a miniature piece of bone in the form of a horse’s head (Fig. 21), add to our current picture of the richness of the material culture.

Activity I. Arabian Bifacial Tool Use An ephemeral presence defined by the surface recovery of stone projectile points and scrapers produced in the ‘Arabian Bifacial Tradition’. These would normally be defined as belonging to a Neolithic culture, dating to before 4000 BC, and represent hunting activities.9

Figure 12: Pottery figurine of a human

Figure 11: Modelled bull’s head from pottery vessel and monkey figurine Site occupation overview The following entries consist of a brief overview of the material record observed at al-Midamman. It is presented with a site chronology in mind, but subdivisions called ‘Facets’ listed under each ‘Activity’ have no chronological significance. They may or may not be contemporary with one another. Figure 13: Neck of decorated pottery jar 9

No definitive study of these finds has been presented in print, but the Project’s Lithic specialist has referred to this material in passing, as belonging to the ‘Arabian Bifacial Tradition,’ cf. Rahimi 2001 (symposium paper). Reference to a ‘Neolithic technology’ present in these stray surface finds that can be attributed to the ‘Arabian Bifacial Tradition’ is also made, courtesy of Rahimi, by Keall 1998: 720; 725.

millennium BC at Mārib, in the interior of Yemen, see Brunner 1998: 63; 1999: 48. 8 Keall 1998: 143, figs 5 & 8; Keall 2000: 723-724, fig.4.

45

TRADE AND TRAVEL IN THE RED SEA REGION

Figure 14: Composite reconstruction drawing of typical ‘libation pail’ Facet IIB. Megaliths with metal and obsidian, pottery and gold beads

Activity II. Era of the standing stones This is the first monumental phase of the site. Whether individually placed as menhirs, or set up in some kind of alignment, these stone markers may be attributed in general to a time when commemorative markers were an important cultural expression. A distinction can be made between slim, naturally formed benzoidal-shaped pillars of basalt, and larger giant blocks of granite that can be called megaliths. However, there is no firm evidence to show that the pillars were set up earlier than the megaliths. The use of rhyolite was hypothetically present at this time, too, but later re-use of it makes this assertion difficult to verify. There is a single example of a giant, water-worn stone that gives the viewer the vague sense of an animal head. It might have been used originally as an idol. The order of Facets IIA-C, as presented here, has no chronological significance.

The megaliths may have been intended to form some kind of alignment. The line of the megaliths as seen in Keall (1997: 13; 1998: fig. 2), corresponds with the sun’s setting in mid-winter, behind the southern tip of the highly visible Zuqar Island. A group of copper-alloy tools set around a quarried, raw block of obsidian was found as a cache buried beneath one of the pillars.10 Other items appear to have been deliberately interred in this setting include a variety of grinding stones, though they are too generic in form to have specific cultural association. A modelled ceramic bull’s head (Fig. 11), now detached from the original vessel, and a human bust (Fig. 12), are important cultural indicators. But they are not much use as chronological indicators, because the broken condition of the bull’s head (detached from a vessel) may mean that it is displaced from its original context. The human figure is important because it came from a stratified context. The recent (2002) unearthing of the neck of a tall jar (Fig. 13) immediately adjacent to one of the megaliths, is equally significant. It is decorated with a punctate design. The same kind of design is also represented in the vessels that may be called libation pails (Fig. 14), and which were found in a cache not far away from the megaliths. Both

Facet IIA. Pillars with infant burials The evidence is firm that certain pillars were once set up with infants buried beneath them, yet without burial goods (Keall 1998: 142). Sacrificial interment is not out of the question. The traces are mostly ephemeral, however, because of later use of many of the stone pillars for other purposes.

10

Keall 1997: 14-15, figs; 1998: 144, and fig.7; Giumlia-Mair et al., 2002: figs 2, 3.

46

EDWARD J. KEALL: POSSIBLE CONNECTIONS IN ANTIQUITY vessel types were made in the same potting tradition, proving that the standing stones were revered by people who used the same kind of pottery as that which is found in the domestic settlement described below. Two gold beads were also recovered, similar to those found in one of the buildings described below.

Activity III. Monumental buildings and markers using recycled stone The second monumental phase of the site is represented by the re-use of stone, to construct both buildings and graves. The use of rhyolite predominates in this activity. A different construction technique used for one of three buildings suggests two different sub-phases at a minimum. One of the buildings includes in its footings what is suggested above to have been once an idol.

Facet IIC. Cluster of standing stones Numerous stones of rhyolite or tuff were once set up in a cluster, though it is not clear whether this represented some kind of alignment. All of these standing stones have been removed at a later time, for re-use elsewhere in the area of the site. Their position is marked by shallow depressions in the ground, surrounded by spalled stone (Keall 1997: 16, fig. right). No burial activity has been recorded.

Facet IIIA. Stone-lined graves (site HWN) Stone-lined graves were built using (hypothetically) reused former standing stones. The ceramic grave goods consist of whole vessels, of a kind known from the domestic settlement, though perhaps specially made in poorer quality for funerary purposes.

Facet IID. Adult burial Facet IIIB. Partitioned buildings (site HWA, Building A, and site BNF)

The lower portion of an adult male in flexed position was recovered from a context involving a burial set beneath a commemorative stone of rhyolite (Keall 1997: 17; 1998: 142-143). No grave goods were recovered. Inclusion of this feature under Activity II is purely hypothetical.

Two rectilinear buildings were constructed with foundation walls employing either roughly broken granite or largely intact basaltic pillars. The stone used for

Figure 15: Incised stone panel of chevron-bordered snake bodies

47

TRADE AND TRAVEL IN THE RED SEA REGION

Figure 16: Snake heads

Figure 17: Oryx heads and water jars 48

EDWARD J. KEALL: POSSIBLE CONNECTIONS IN ANTIQUITY

shallowly carved facade decoration (Figs 15-18). Found near the hypothetical entranceway were some gold beads (Fig. 19), gold necklace spacers, painted ostrich egg fragments (Fig. 20), and a tiny horse head pendant of bone (Fig. 21).

Figure 19: Gold beads

Figure 18: Cross hatched design; ibex horns building the walls above ground is of roughly dressed blocks of rhyolite. Partitions of mudbrick divide the interior space into narrow aisles and hypothetically were foundation benches for wooden post roof-supports (Keall. 2000: 723-724, and fig. 4). Facet IIIC. Decorated facing stone (site HWA, Building B, and Figs 15-18) A monumental structure consisting now of two long intersecting walls was built of roughly dressed rhyolite blocks. A trace of mudbrick suggests a partition, as with the other two buildings. But the heavy concentration of stone fallen in one area and not another is difficult to explain. Conceivably, the structure was an open compound with an elaborate entrance. Highly significant are some finely dressed facing stones, and pieces of

Figure 20: Painted ostrich egg fragments 49

TRADE AND TRAVEL IN THE RED SEA REGION as the megaliths are concerned, one must stress the fact that there is no rock of this kind in the immediate vicinity. The possibility that the Red Sea islands on the Yemeni side — which are volcanic — were the source of the stones, was an attractive theory for the Canadian Mission at one time. The idea was that it would have been a lot easier to float the giant stones on large rafts across the sea to the coast, rather than hauling them with great effort a considerable distance overland. But the experience of the Canadian team adrift in a boat with a broken motor was a useful reminder that sailing the Red Sea is not a casual matter. If we are to suggest people crossed it on a regular basis, then we must assume appropriate craft, and knowledge of tides and winds. Certainly this is plausible if one accepts the evidence from a late second millennium Egyptian tomb relief of sailors from Punt manoeuvring a huge raft by means of triangular sail and oar-rudder (Davies 1935: fig. 2; Kitchen 1993: fig. 35.7). But, in fact, no suitable stone source has been found so far on the various islands offshore from Yemen. However, basalt is reported in the Gulf of Zula, off the Eritrean coast on the opposite side of the Red Sea (Schneider 1983: 49).

Figure 21: Miniature horse head carved in bone Activity IV. Domestic settlement (sites HWN, HWA, and HWB) Traces of a domestic settlement of ephemeral structures, probably made from palm fronds is indicated mainly by the survival of hearths seemingly set some distance downwind from the flimsy and highly flammable structures. Some of the hearths have been excavated, but no other substantial features have been found. The excavations have exposed pottery, obsidian, grinding stones, and masses of fish bone. A calibrated radiocarbon date of 1360 BC or 1320-970 BC has been extracted from a charred doum palm fruit found in one of the ashy middens.

Re-used stone was used to create impressive, monumental buildings. There is no evidence that re-use of the megaliths was a destructive act. Because of the effort involved in re-using the former standing megaliths, one might argue that this represents a continuing reverence for the stone. Arguably, it would have been easier to make the buildings entirely out of mudbrick, rather than spending the energy on breaking up the megaliths. Two of the rectilinear structures were built with foundations and walls of stone, and internal subdivisions of mudbrick that perhaps served as bases for wooden roof columns. A third structure has no comprehensible ground plan preserved. Perhaps it was an open sanctuary. However, numerous blocks of dressed stone, now tumbled, as well as fragments of carved decoration, suggest at least an elaborate entrance to an enclosed compound.

Facet IVA. Hearths Hearths were created by placing large fragments of broken storage vessels in a pit. Facet IVB. Fire-cracked stone circles (sites HWN and WWW) This activity involved a fire being set beneath a block of rhyolite, which cracked in the process, leaving a circle of broken stone (Keall 1997: 16, fig. left). Pottery and grinding stones were also set deliberately onto the fire. The fragments are fire-blackened. Surface deflation of the soft ground around the circle has caused these features to be preserved as elevated heaps. Almost exclusively, these stone heaps are to be found out in the open, in isolation from the sherd scatters that represent the living settlement.

Besides being recycled for building use, stone was also employed in a cemetery to create circular tombs where pottery grave goods were unearthed. The vessel types represented in the cemetery duplicate those recorded for the domestic settlement, although the cemetery vessels were of inferior quality, having been produced purely for funerary purposes. As for the technology represented in the pottery vessels used in the domestic settlement, it is far superior to what we know from classic South Arabia. Although made by hand and slab-built, the pottery is well produced from good quality, oxidised red-firing clay. The technique of paring was commonly used to trim away excess clay and to make a thinner vessel wall. In many cases, only minimal effort was made to disguise the paring marks. In other instances, smoothing the rough paring marks led to a burnished effect. Burnished vessels were often decorated with punctate designs. The repertoire of shapes reflects pots designed for a variety of

Commentary on the Material Culture The distinctive elements of the main al-Midamman site record are megalithic stone use for commemorative monuments and re-use of stone for constructing monumental buildings and formal cemeteries. The unearthing of a deliberately buried cache of metal tools provides the crucial date of between c. 2400-1800 BC for the setting up of the megaliths. The date is based upon artefact typology and a comparison of the metallography of the tools, derived from instrumental analysis.11 As far 11

Keall 2000: 721-722; Giumlia-Mair et al. 1999; 2002.

50

EDWARD J. KEALL: POSSIBLE CONNECTIONS IN ANTIQUITY Carved stonework from a building facade (Figs 15-18), the occasional find of a three dimensional form modelled in baked clay, and a unique miniature bone pendant in the form of a horse’s head (Fig. 21) comprise the only figural art. Some of the pottery was painted with simple linear patterns in a monochrome red; ostrich eggs were painted in bichrome colours (Fig. 20). Loosely speaking, mostly based on typological analysis, this record falls roughly between c. 2400-900 BC. With regard to modelled pottery at al-Midamman, there is one bull’s head in relief from a pottery vessel (Fig. 11), and one human figurine (Fig. 12). One of the distinctive incense burners of pottery has alphabetic letters scratched onto the top of the vessel (Fig. 22). Each of these features is sufficient for some to point to a South Arabian connection. But we must be cautious in drawing this conclusion for the culture as a whole. Features like this are very rare within the corpus of finds, representing three out of 4000 recorded, diagnostic fragments. The fact that nondiagnostic sherds comprise an inventory of tens of thousands more vessel fragments only serves to emphasise the rarity of the ‘exotic’ elements.

food preparation and storage purposes. These features include heavily scored surfaces for grating and internal lugs for hanging pots over a fire. The mass of pottery stands out overwhelmingly as being ‘non-Near Eastern’ in terms of a ceramic assemblage. The domestic settlement itself is represented in the archaeological record mostly by open hearths, with ephemeral housing probably made out of palm branches. As described above, a radiocarbon date of towards the end of the second millennium gives us a vital reference point for the life of the settlement. Besides the inventory of red burnished pottery, the excavations have furnished ample evidence of other utilitarian artefacts associated with the life of the settlement, such as use of grindstones, an obsidian microlithic tool technology, and copper-alloy tools. Grindstones clearly played an important role in daily life, as they are found together with the pottery, as well as symbolically placed (as described above) on the heaps of the fire-cracked stone circles. As for aspects of personal adornment, there are indications of considerable stonebead use; and evidence for worked gold for jewellery. From small pieces of coloured stone recovered, partially faceted, there is ample evidence for the manufacture of beads from stone washed down with the floods. There is no definitive evidence for gold jewellery production on site; nor are the characteristics of the gold beads sufficiently distinctive to be culturally diagnostic (Fig. 19). However, it may be useful to record that the technique for bead production involved hammering gold sheet to form hollow ovoid shapes; thin, washer-like rings were perhaps stamped out with a punch, to serve as jewellery spacers.

Discussion Who the people of al-Midamman were, and how they actually sustained themselves is still an enigma. While there are some cultural affinities with what academic tradition has declared to be typical of the Sabaeans (implying something after the eighth century BC), there is just as much likelihood that the common expressions reflect elements that contributed to the development of classic South Arabian civilization, but were not products of it. The commonalities are too limited, and the

Figure 22: Symbols scratched in clay (before firing) of incense burner tops

51

TRADE AND TRAVEL IN THE RED SEA REGION that, since the 1950s, megalithic sites in Arabia have generally been attributed to the Chalcolithic-Early Bronze era.13 The most recent (2002) season at alMidamman has unequivocally demonstrated that while the practice of setting up menhirs may well have started before the end of the third millennium, if not before, their demolition did not occur until the end of the second millennium. However, there is no evidence at all of the so-called statue-menhirs either, which would have made a plausible contemporaneous link with those documented on the plateau that have been judged to reflect a Bronze Age elite in the interior of Yemen in the second millennium (Newton & Zarins 2000: 161). Conversely, one may usefully refer to the Mahal Teglinos excavations in the Gash Delta of Kassala Province, in southeastern Sudan, where there are reports of monolithic stelae erected, not as single grave head-stones, but as pillars placed randomly in cemetery areas in the early third-mid second millennium (Fattovich 1993: 440).

dissimilarities too great for this to be the case. We face the fact, then, that a previously unknown group of people entered the area and expended a great deal of energy and resources setting up giant monuments of stone somewhere around the turn of the third-second millennium, before disappearing again a thousand years later. Their tool technology did not develop out of the Arabian peninsula tradition; their pottery is superior in production quality, and more sophisticated in terms of specialised vessel function, than most of the contemporary pottery from the interior of Yemen.12 In short, there is no convincing evidence that these people were in any way colonists from the highlands of Arabia. One must conclude that the people already had a reverence for mountain stone when they chose to settle in al-Midamman, because they expended so much effort on getting it to their settlement. Whatever the explanation for the delivery of the stones, there is evidence that there were once innumerable giant stones imported into this landscape and set up as commemorative markers or in alignments. Most of them were re-used later during the second millennium BC for other purposes, so few of them are left in their original contexts. Yet in spite of the fact that the inhabitants appear to have been obsessed with stone, there are no inscriptions carved in stone; no sacrificial offering trays of stone; no stone incense burners; no three dimensional sculptures of either animals or humans, in stone. All of these would be appropriate for a culture linked to classic South Arabia in its traditional sense.

Pottery represents by far the most visible and tangible relic of the al-Midamman settlements. As indicated above, it reflects a fully developed ceramic tradition. We do not have evidence for a forerunner of the main repertoire of shapes, nor is there indication of a declining technical standard. It disappears from the immediate region with the abandonment of the settlement early in the first millennium BC.14 Some of the forms, methods of manufacture, and decorative techniques observed in the pottery can certainly be paralleled in the Horn of Africa. The punctate designs, often presented within triangular frames, are initially reminiscent of the punctate incised pottery from the Gash Delta, and dated to 1700-1400 BC (Fattovich 1980; 1997). Regardless of the al-Midamman finds, a hypothetical African-Arabian connection was explored extensively, and persistently, by Fattovich. He developed a plausible ‘Afro-Arabian cultural complex’ theory by pointing to generic similarities with the socalled wavy-line punctate pottery associated with CGroup peoples in the upper Nile area.15 But the similarities are superficial, and limited largely to decorative features. Most of the Nubian material is deliberately reduced in firing, to create a distinctive red and black look. This is entirely absent in the Tihāmah finds where the clay is fired to an oxidised all-over red. Such choices in firing pottery are highly significant when considering cultural continuity. Furthermore, according to Phillipson, a weakness of the Fattovich argument is that the Gash Delta pottery is not really ‘African’ either.16 The closest parallel for the al-Midamman corpus is to be found on the Yemeni side of the Red Sea, in the MaΜlayba/Sabr assemblages, which have been dated to

Notwithstanding the evidence of the newly discovered building decorations, it is striking that all of the utilitarian artefacts we find right across the site, in domestic, burial and ceremonial contexts alike, are identical for the entire time period of the settlement. It is true of the pottery, the obsidian, the metal, and the grindstones. The only differences are the different uses the objects were put to (and consequently the different types of object and quality involved). But the techniques of manufacture and methods of production are identical in all cases. We have no evidence that different people were involved in the two different phases of the site, so that one could explain in this way the arrival of new ideas along with newcomers to the area. However, when so many of the individual elements in the cultural record documented at al-Midamman do not conform exactly with those in the highlands or desert interior of Yemen, it is surely a tempting alternative to seek possible connections with the Horn of Africa. As in the case of much southern Red Sea research, it is hard to assign precise dates to specific cultural records, and especially difficult to attribute appropriate identities to social groups represented by cultural expressions. Most pertinent here is a reference to the traditional way

13

cf. Zarins 1992: 58-59; Vogt 1998: 124-126 Notwithstanding this observation, it can be argued that certain production characteristics present in the Bronze Age al-Midamman pottery (as well as at MaΉaΓΓat al-Kašawba) can be observed in the Islamic period pottery recorded in Zabīd and environs (Ciuk & Keall 1996: cf. pl.6 with pls.27 & 28). 15 Fattovich 1993: 439; 1997: 278; Munro-Hay 1993: 609. 16 Personal communication, Cambridge, May 2002. 14

12

For the relatively poor quality of South Arabian pottery, cf. Buffa & Vogt 2001: 443.

52

EDWARD J. KEALL: POSSIBLE CONNECTIONS IN ANTIQUITY Red Sea. This involves the location of the land of Punt — the territory that the New Kingdom Egyptians knew as the source of frankincense and other exotic products. The current academic consensus is that Punt lies in southeastern Sudan - northern Ethiopia - Eritrea (Kitchen 1993: 604). It includes the Gash Delta where the closest parallel, though not a precise match, to the Arabian peninsula cultural record is to be observed. One may argue that there is evidence that both sides of the Red Sea were loosely connected (in both ethnic and cultural terms). This certainly did not, however, involve mass migration or colonization. Perhaps it is more appropriate to talk of immigration, implying rather casual settlement (even though perhaps fairly large numbers can be involved at any given time). We should continue to see the land of Punt the way the Egyptians described it, for the Egyptian-Puntite interaction. But we should also think of these people as having cultural links further afield, on both sides of the Red Sea, and especially in the Tihāmah. We may see the people of the Tihāmah as related to the people of Punt, culturally speaking, though not necessarily as part of their polity.

somewhere between the thirteenth and ninth century BC.17 It should be repeated here that generally at the second millennium sites in the Tihāmah there is an absence of Sabaean characteristics. Apart from the main cache of copper-alloy tools, isolated examples of similar tool fragments have been found across the site. Typological comparisons from the Levant are used for dating the copper alloy tools (Giumlia Mair et al. 1999; 2002), not an absolute date derived from stratigraphic work. Square-sectioned, swallow-tailed adzes are an almost universal phenomenon of the Old World bronze ages, from Ireland to Egypt. Nevertheless, the square-sectioned points excavated at Sihi, with a date cited for beginning c. 2100 BC (Zarins & Badr 1986: 4849, fig. 64), and the double-riveted dagger blades from the Yemeni highlands attributed to c. 2300-2000 BC (Seipel (ed.) 1998: 141, cat. 76) seem particularly relevant. Also, the double-riveted dagger blade has been argued to be characteristic of second-first millennium Arabia, but not Mesopotamia (Newton & Zarins 2000: 161). Given the large numbers of bronze statues and plaques produced in the classic South Arabian period,18 copper working was not foreign in Yemen. We must admit that little is known about either how or where it was manufactured.

Having said that, we may turn briefly to the Dahlak archipelago off-shore from Eritrea. Erosion of the flat limestone bedrock of Dahlak Kabir island in the past resulted in the natural creation of numerous potholes that were engineered in Islamic times into an elaborate network of reservoirs that sustained a rich community in the eleventh – thirteenth centuries because of its ability to sell fresh water to ships navigating the Red Sea. The history of the island from Aksumite times to the Ottoman period is given in Insoll (2001). However, as evidence of a presence on the island before Aksumite times, the Canadian Mission has observed fragments of obsidian microliths in the area of the island where ancient, nonengineered water holes exist. Although non-diagnostic obsidian flakes are difficult to date, the general absence of pottery in the areas where the obsidian was observed suggests that it was derived from antiquity as well.

As far as the monumental building architecture is concerned (sites HWA, Building A, and BNF), the internal subdivisions are distinctive features. This kind of arrangement can be found in Burned Building V at Sabr, near Aden, dated to the end of the second or beginning of the first millennium BC. Vogt suggests this so-called hypostyle hall was a likely antecedent of the classic South Arabian temple (Buffa & Vogt 2001: 440). The third and most recent building at al-Midamman (HWB, Building B) remains for the moment a complete enigma. Its exact ground plan is by no means evident. However, it is clear that the shallowly carved decorations (Figs 1518) are precisely the same (in both technique and style) as those from the so-called Banāt ΚĀd temples in the al-Дazm area of the Jawf, which have been ascribed a date of the sixth-eighth centuries BC (Breton et al. 1990; Breton 1998: 215). But it should be emphasised that the Banāt ΚĀd decorations themselves are unique, and the origin of the style may actually lie outside of the Jawf, and their date may be considerably earlier than the eighth century. Audouin (1996: 142) suggests the second half of the second millennium BC on the basis of the artwork. Lundin reiterates this suggested date for epigraphic reasons (1997: 14).

Most of the elaborately inscribed head-stones in the medieval Islamic cemetery are fragments of felsite or basalt (Schneider 1983: 48-49). The stone is foreign to the island. Clearly a wealthy Muslim community may have opted to import the stone. But, conceivably, it was originally imported onto the island during the Bronze Age, and only much later re-used for the new funerary purposes. Although this may be circumstantial evidence at best, it nevertheless keeps alive the idea that there may have been contact between the Horn of Africa and Arabia in the Bronze Age.

The plausible second millennium BC date for the majority of the al-Midamman finds allows us to pursue another intriguing line of thought, as a way of explaining vaguely similar cultural expressions on both sides of the

References Audouin R. 1996. Étude du décor des temples des Banât ΚÂd. Pages 121-142 in C.J. Robin (ed.) with the collaboration of I. Gajda, Arabia Antiqua. Early Origins of South Arabian States. (Proceedings of the

17

Buffa & Vogt 2001: 439, 446; Vogt & Sedov 1998: 144-151. Witness the numerous catalogue entries of Yemeni made bronzes in the Queen of Sheba exhibition volume (Simpson 2002). 18

53

TRADE AND TRAVEL IN THE RED SEA REGION First International Conference on the Conservation and Exploitation of the Archaeological Heritage of the Arabian Peninsula, held in the Palazzo Brancaccio, Rome, by IsMEO on 28th-30th May, 1991). (Serie orientale Roma 70/1). Roma: Istituto Italiano per il Medio ed Estremo Oriente. Breton J-F. 1998. Der ΚAthtar-Tempel von as-SawdâΜ (dem antiken Nashshân). Pages 215-216 in W. Seipel (ed.), Jemen, Kunst und Archäologie im Land der Königin von Saba . Wien: Kunsthistorisches Museum. Breton J-F., Arramond J-C. & Robine G. 1990. Le Temple de Athtar d’As-Sawdâ. Sanaa: Mission archéologique française. Brunner U. 1998. Der grosse Damm von MaΜrib - ein techniches Weltwunder der Antike. Pages 63-67 in W. Seipel (ed.), Jemen, Kunst und Archäologie im Land der Königin von Saba. Wien: Kunsthistorisches Museum. ----------1999. Jemen. Vom Weihrauch zum Erdöl. Wein; Köln; Weimar: Böhlau. Buffa V. & Vogt B. 2001. Sabir - Cultural Identity between Saba and Africa. Pages 437-450 in R. Eichmann & H. Parzinger (eds), Migration und Kulturtransfer: Der Wandel vorderund zentralasiatischer Kulturen im Umbruch vom 2. zum 1. vorchristlichen Jahrtausend. (Kolloquien zur Vorund Frühgeschichte 6). (Akten des Internationalen Kolloquiums, Berlin, Nov. 23-26, 1999). Bonn: Habelt. Ciuk C. & Keall E.J. 1996. Zabid Project Pottery Manual 1995, Pre-Islamic and Islamic Ceramics from the Zabid area, North Yemen. (British Archaeological Reports, International Series 655). Oxford: BAR Publishing. Davies N. de G. 1935. Trading with the land of Punt. The Egyptian Expedition 1934-1935. The Work of the Graphic Branch of the Expedition. Bulletin of the Metropolitan Museum of Art 30/2: 46-49. Edens C. & Wilkinson T.J. 1998. Southwest Arabia during the Holocene: Recent Archaeological Developments. Journal of World Prehistory 12/1: 55119. Fattovich R. 1980. Materiali per lo studio della ceramica pre-aksumita etiopica. (Annali dell’ Istituto Orientale di Napoli 40/4, Supplemento 25). Napoli: Istituto Orientale di Napoli. ----------1993. The Gash Group of the Eastern Sudan: an outline. Pages 439-448 in L. Krzyzaniak, M. Kobusiewicz & J. Alexander (eds), Environmental Change and Human Culture in the Nile Basin and Northern Africa until the Second Millennium B.C. Poznań: Poznań Archaeological Museum. -----------1997. The contacts between Southern Arabia and the Horn of Africa in Late Prehistoric and Early Historical Times: a view from Africa. Pages 273-286 in A. Avanzini (ed.), Profumi d’Arabia: atti del convegno. (Saggi di storia antica 11). Roma: L’Erma di Bretschneider. Giumlia-Mair A., Keall E.J., Stock S. & Shugar A. 1999. Copper-based implements of a newly identified

culture in Yemen. Journal of Cultural Heritage 1: 3743. -----------2002. Investigation of a Copper-based Hoard from the Megalithic Site of al-Midamman, Yemen: an Interdisciplinary Approach. Journal of Archaeological Science 29/2: 195-209. Hehmeyer I. 1995. Physical evidence of engineered water systems in mediaeval Zabīd. Proceedings of the Seminar for Arabian Studies 25: 45-54. Hehmeyer I. & Keall E.J. 1993. Water and Land management in the Zabīd Hinterland. Al-ΚUΒūr alWuΒtā: Bulletin of Middle East Mediaevalists 5: 2527. -----------2000. Traditional Water Management Practices in Southern Arabia. Human Dimensions of Global Change: a Mandate for Anthropological Engagement. (Unpublished Conference paper. American Anthropological Association, 99th Annual Meeting, San Francisco, Nov. 15-19, 2000). Insoll T. 2001. Dahlak Kebir, Eritrea: From Aksumite to Ottoman. Adumatu 3: 39-50. Keall E.J. 1983. The dynamics of Zabid and its hinterland: the survey of a town on the Tihamah plain of North Yemen. World Archaeology 14: 378-392. -----------1997. Do you want to see the Stones? (Royal Ontario Museum). Rotunda 30/2: 12-19. -----------1998. Encountering megaliths on the Tihamah coastal plain of Yemen. Proceedings of the Seminar for Arabian Studies 28: 139-147. -----------1999a. Archäologie in der Tihamah. Die Forschungen der Kanadischen Archäologischen Mission des Royal Ontario Museum, Toronto, in Zabid und Umgebung. Jemen-Report 30/1: 27-32. -----------1999b. Les fouilles de la mission archéologique canadienne. Pages 19-23 in P. Bonnenfant (ed.), Zabîd, patrimonie mondial. (Saba: Arts - Littérature Histoire - Arabie méridionale 5-6). Bruxelles: UNESCO. -----------2000. Changing Settlement along the Red Sea Coast of Yemen in the Bronze Age. Pages 719-729 in P. Matthiae, A. Enea, L. Peyronel & F. Pinnock (eds), Proceedings of the First International Congress on the Archaeology of the Ancient Near East, Rome, May 18-23, 1998. Roma: Università La Sapienza. -----------2001. Canadian Archaeological Museum of the Royal Ontario Museum in Yemen. Al-Musnad 1/1: 96-90 (sic). Keall E.J. & Hehmeyer I. 1998. Sponsorship of a madrasa, reflecting the value of farmland in the urban economy of Zabīd, Yemen. Al-ΚUūΒr al-WuΒtā: Bulletin of Middle East Mediaevalists 10: 33-35, 47. Kitchen K.A. 1993. The land of Punt. Pages 587-608 in T. Shaw, P. Sinclair, B. Andah & A. Okpoko (eds), The archaeology of Africa: food, metals, and towns. (One World Archaeology 20). London/New York: Routledge. Lundin A.G. 1997. Der Ursprung des südarabischen Alphabets. (Staatliches Museum für Völkerkunde München). Mare Erythraeum 1: 9-18.

54

EDWARD J. KEALL: POSSIBLE CONNECTIONS IN ANTIQUITY Munro-Hay S. 1993. State development and urbanism in northern Ethiopia. Pages 609-621 in T. Shaw, P. Sinclair, B. Andah, & A. Okpoko (eds), The archaeology of Africa: food, metals, and towns. (One world archaeology 20). London/New York: Routledge. Newton L.S. & Zarins J. 2000. Aspects of Bronze Age art of southern Arabia: the pictorial landscape and its relation to economic and socio-political status. Arabian archaeology and epigraphy 11: 154-179. Rahimi D. 2001. Geometric microliths of Yemen: Arabian Precursors, African Connections. Parting the Red Sea: Holocene Interactions between Northeastern Africa and Arabia. (Unpublished Symposium paper. Society for American Archaeology, 66th Annual Meeting, April 18-22, 2001, New Orleans). Schneider M. 1983. Stèles funéraires musulmanes des Îsles Dahlak (Mer Rouge). (2 volumes). Cairo: Institut français d’archéologie orientale du Caire. Seipel W. (ed.) 1998. Jemen, Kunst und Archäologie im Land der Königin von Saba’. Wien: Kunsthistorisches Museum. Simpson St.J. (ed.) 2002. Queen of Sheba. Treasures from ancient Yemen. London: British Museum. Vogt B. 1998. Frühe Kulturen an der Küste des Roten Meeres und des Golfs von Aden. Pages 123-127 in W. Seipel (ed.), Jemen, Kunst und Archäologie im Land der Königin von Saba . Wien: Kunsthistorisches Museum. Vogt B. & Sedov A.V. 1998. Die Sabir-Kultur und die jemenitische Küstenebene in der 2.Hälfte des 2.Jahrtausends v. Chr. I. Pages 129-133 in W. Seipel (ed.), Jemen, Kunst und Archäologie im Land der Königin von Saba. Wien: Kunsthistorisches Museum. Wilkinson T.J. 1999. Settlement, soil erosion and terraced agriculture in highland Yemen: a preliminary statement. Proceedings of the Seminar for Arabian Studies 29: 183-191. Wilkinson T.J. & Edens C. 1999. Survey and Excavation in the Central Highlands of Yemen: Results of the Dhamār Survey Project, 1996 and 1998. Arabian archaeology and epigraphy 10:1-33. Zarins J. 1992. Archaeological and Chronological Problems within the Greater Southwest Asian Arid Zone, 8500-1850 B.C. Pages 42-62 in R.W. Ehrich (ed.), Chronologies in Old World Archaeology (2 volumes)/vol.1. Chicago; London: University of Chicago. Zarins J. & Al-Badr H. 1986. Archaeological Investigation in the Southern Tihama Plain II (including Sihi, 217-107 and Sharja 217-1720). 1405/1985. Atlal 10: 36-57.

55

56

Ancient interaction across the southern Red Sea: new suggestions for investigating cultural exchange and complex societies during the first millennium BC Matthew C. Curtis

northeastern Africa, including the Napatan state of the Nile valley. In the greater Asmara area these preAksumite communities are known as the ‘Ona’, a term used by local Tigrinya-speaking peoples meaning ‘ruins’.2

Introduction and background This paper briefly outlines key evidence for the existence of extensive ‘Ona’ village communities occupying the Asmara Plateau in the Hamasien region of highland Eritrea in the early to mid first millennium BC. This evidence is considered in relation to other developments concerning complex societies in the northern Horn of Africa and southern Red Sea region during the first millennium BC; and a new theoretical framework for examining relations between polities in the southern Red Sea area is suggested. In challenging the notion that South Arabian colonists and migration events were the prime movers in the development of sociopolitical complexity in the Horn of Africa, suggestions for developing alternative conceptualisations of interregional cultural exchange in the southern Red Sea are offered. In particular it is argued that archaeological and historical evidence suggests that during the early to mid first millennium BC some elite members of pre-Aksumite period highland communities of the northern Horn of Africa appropriated new forms of ceremonial material culture, iconography, religious architecture and inscriptional titles drawn from the wider southern Red Sea world in the effort to legitimise claims to political preeminence and expand involvement in interregional ideological, sociopolitical and economic networks.

The greater Asmara area and Hamasien region of central Eritrea are positioned at the crossroads of major culture areas of northeastern Africa and the Red Sea.3 (Fig. 23). The region is dominated by the Asmara Plateau, a relatively flat peneplain ranging from approximately 2200 to 2500 m above sea level, containing broad swales and shallow seasonal streams, and punctuated by small flat-topped hills. The Asmara Plateau has relatively fertile luvisols, receives 500-700 mm of rainfall annually and possesses gold deposits and copper mineralisation.4 The greater Asmara area’s position at a junction between the western lowlands of Eritrea, to the eastern Sudan and Nile Valley beyond, and to the larger Red Sea world may have been a key factor affecting its ancient political and economic development. In addition, its location in a region possessing environmental conditions favourable for permanent settlement and intensification of resources may have provided a natural focal point of economic and political power. The Ona communities of the preAksumite period, with their nascent urban forms and involvement within regional and interregional economic and sociopolitical networks and ideologies, may have been vital precursors to later state and imperial development during the Aksumite period.

Archaeological research carried out on the Asmara Plateau in the greater Asmara area, Eritrea, between 1998 and 2001 by the University of Asmara archaeological field schools program and the 1999-2000 Greater Asmara Regional Archaeological Survey Project (GARASP) has provided important new insights into the development of sociopolitical complexity in the northern Horn of Africa.1 Research results show that some of the earliest settled highland agropastoral communities known in the northern Horn of Africa were established in permanent villages and small towns in central Eritrea around Asmara between c. 800 BC and c. 400 BC, a time largely contemporaneous with the pre-Aksumite period of the Ethiopian/Eritrean highlands and the vaguely defined Ethio-Sabaean polity of ‘Daamat’ (DΚMT). In addition, it is a period that coincides with the development and expansion of the state of SabaΜ in South Arabia and overlaps temporally, in part, with other polities in

2

See Tringali 1965, 1967, 1969, 1979, 1981, 1987; Fattovich 1990; 1996, 1997a, 2000; Munro-Hay & Tringali 1993; Schmidt & Curtis 2001 for descriptions of the ancient Ona culture. There are various chronologies offered for the pre-Aksumite culture period; see Anfray 1964, 1967, 1968, 1990; Michels 1979, 1988, 1994; Fattovich 1988, 1990, 1997a, 1997b, 2000; Fattovich et al. 2000. Here pre-Aksumite refers to the complex societies that existed in the northern Horn of Africa during the first millennium BC (c. 900 BC to c. 300 BC) sharing basic elements of material culture, settlement patterns, economies, and sociopolitical organization. The pre-Aksumite culture period predates the proto-Aksumite (c. 300 BC to c. 100 BC) and Aksumite (c. 100 BC to c. AD 900) culture periods. It is recognized that the term preAksumite is wholly inadequate and we await the development of more precise culture historical terminology with future research in the northern Horn of Africa. 3 See Fattovich 1988, 1996, 1997a, 2000; Schmidt & Curtis 2001. 4 See Abul-Haggag 1961; FAO 1994; Nastasi 1994; Woldai et al. 2000 for various aspects of the Asmara Plateau’s physical environment.

1

See Schmidt 1999; Curtis 2001; Schmidt & Curtis 2001 for overviews of recent archaeological research in the Greater Asmara area, Eritrea.

57

TRADE AND TRAVEL IN THE RED SEA REGION

Figure 23: Map of the southern Red Sea region showing sites mentioned in the text. Only one large-scale systematic regional archaeological survey of pre-Aksumite settlements in the highlands of the northern Horn of Africa was conducted prior to 1998, providing, until very recently, the sole study of ancient settlement in relation to environmental features at a regional scale.5 Pre-Aksumite non-elite residential and production areas in the northern Horn of Africa have largely have been ignored by archaeologists, leaving unexplored a full range of specialised economic activities and adaptations to changing physical and social environments. Our view of pre-Aksumite society is coloured by a previous research bias toward specialised ceremonial and elite occupation sites such as Yeha and Hawlti-Melazo in the highlands of Tigray.6 Although

archaeological excavations in the 1990s in and around Aksum in northern Ethiopia have provided solid environmental and subsistence data, and have demonstrated a commitment to investigating systematically non-elite contexts, these excavations have focused mostly on Aksumite and proto-Aksumite sites and not the earlier pre-Aksumite occupations.7 Only very recently has a pre-Aksumite village settlement in the Aksum area been the focus of systematic excavation. The British Institute in Eastern Africa‘s excavations at the ‘D site‘ at Kidane Mehret represent the first thorough investigation of substantial non-elite pre-Aksumite 7

See Fattovich & Bard 1995, 1996, 1997a, 1997b; Phillipson & Reynolds 1996; Bard et al. 1997, 2000; Fattovich et al. 2000; Phillipson 1998, 2000 for descriptions of archaeological research conducted during the 1990s in the Aksum area concerning the proto-Aksumite and Aksumite periods. See Finneran 1998, 1999, 2000a, 2000b for research concerning Later Stone Age and early food-producing communities in the Aksum area.

5

Michels 1979, 1988, 1994. See Anfray 1963a, 1972a, 1972b, 1973 for descriptions of research at Yeha. See Drewes 1959; Leclant 1959; de Contenson 1961, 1962, 1963; Schneider 1961, 1973; Pirenne 1970 for descriptions concerning Hawlti -Melazo.

6

58

MATTHEW C. CURTIS: ANCIENT INTERACTION ACROSS THE SOUTHERN RED SEA While this African foundation of pre-Aksumite culture is stressed today by some archaeologists and historians working in the southern Red Sea region, the idea that South Arabian, particularly Sabaean culture, was the primary catalyst for state development and urbanism in the northern Horn of Africa is still very much evident in recent literature.12 An example can be seen in the following passage from an influential volume chapter concerning cultural connections between Yemen and the northern Horn of Africa in antiquity:

domestic contexts in the Aksum area dating to the first half of the first millennium BC (Phillipson 2000; see Phillips infra). Some consideration of pre-Aksumite settlement is evident in the Istituto Universitario Orientale/Boston University research in the area of Bieta Giyorgis at Aksum (Fattovich et al. 2000), but the overall focus at Bieta Giyorgis is on proto-Aksumite and early Aksumite periods. Outside the immediate area of Aksum, the excavation of pre-Aksumite period non-elite domestic occupation sites in northern Ethiopia is virtually nonexistent.

Archaeological evidence points to the fact that people from Southern Arabia arrived on the coast opposite no later than the middle of the first millennium, settled in large numbers in the hinterland, mingled with the indigenous population and founded several centres of habitation there. Finds from that early period and also from the following centuries, in which the areas of present day Eritrea and the north Ethiopian province of Tigre achieved a cultural peak, are completely South Arabian in character, i.e. they might just as well have been produced on Arabian soil (Raunig 1988: 412).

Prior to 1998, extant data on ancient site distribution in Eritrea came from a few ethnographic and archaeological reconnaissance projects conducted during the Italian occupation of Eritrea, the documentation of rock art sites conducted intermittently from the 1940s to 1960s, and from research carried out more than three decades ago by the French-sponsored Ethiopian Archaeological Institute in the Senafe area around the ancient site of Matara. Before the late 1990s, the limited excavation carried out in Eritrea had centered on the largest and most monumental of ancient settlements, such as the port city of Adulis on the Red Sea and the monumental urban center of Matara.8 Although Ona sites in the greater Asmara area were the focus of occasional investigation from the 1950s to the 1970s, the lack of standing monumental architecture and absence of written inscriptions seems to have ensured the obscurity of the ancient Ona sites in the archaeological literature prior to systematic investigation beginning in the late 1990s.9

The idea of the ‘sudden’ appearance by the middle of the first millennium BC of features such as writing and monumental stone architecture and sculpture is used as evidence to support the notion that the arrival of South Arabian peoples played a paramount role in the development of complex society in the pre-Aksumite period.13 The supposed ‘sudden’ appearance of South Arabian traits continues to pre-occupy archaeologists, yet our interpretive frameworks for investigating the incorporation of such traits have not moved beyond static migration models espousing exogenous innovation. Focusing on epigraphy, monumental architecture, and religious paraphernalia to the exclusion of other data creates a truncated perspective of interaction. This truncation results in a narrowing and channelling of critical examination so that the products of cultural exchange, such as influence and innovation and the resultant transformations they generate, are portrayed as emanating and proceeding from a single source and direction across the Red Sea. It is one thing simply to point out the existence of South Arabian elements in the material culture of ancient sites in the northern Horn of Africa, but it is altogether a different proposition to explain the processes by which these elements were incorporated into African communities. Although most recent interpretations concerning southern Red Sea

The first interpretations of early complex society in the northern Horn of Africa developed in the nineteenth and early twentieth centuries by historians, epigraphists and archaeologists envisaged a South Arabian colony ruling over local African populations.10 Later, in models developed by Francis Anfray and continuing with important modifications by Rodolfo Fattovich, preAksumite complex society is portrayed as an EthioSabaean polity where relatively small groups of South Arabian migrants were initially absorbed into an indigenous African population. In this view, South Arabians provided a written language and aspects of monumental architecture and ideology that were grafted onto a larger indigenous African culture. In particular, Fattovich has stressed the need to investigate the likely African substratum of pre-Aksumite communities.11 8

For early archaeological site documentation in Eritrea see Piva 1907; Dainelli & Marinelli 1912; Littmann et al. 1913. For rock art site descriptions and distribution in Eritrea conducted from the 1940s to 1960s see Franchini 1951, 1952, 1958; Ricci 1959; Graziosi 1964. For settlement site distribution in the Senafe area see Anfray 1963b. See Paribeni 1907; Sundström 1907; Anfray 1974 for excavations at Adulis. See Anfray 1963b, 1967, 1974; and Anfray & Annequin 1965 for Matara. 9 See Tringali 1965, 1973-77, 1980-81; Anfray 1970; Munro-Hay & Tringali 1993 for descriptions of initial documentation of Ona sites of the greater Asmara area. 10 Prominent examples include Glaser 1895; Conti Rossini 1906, 1928. 11 Both A. J. Drewes and R.Schneider challenged the ideas of ContiRossini and others that claimed that pre-Aksumite societies were largely

South Arabian transplants. As an alternative, Drewes and Schneider both offered a basic acculturative model of interaction and culture change in the northern Horn of Africa stressing South Arabian cultural influence through political and economic connections across the Red Sea. See Drewes 1956, 1959, 1962; Schneider 1976b for acculturative models. For descriptions of an ‘Ethio-Sabaean’ polity see Anfray 1964, 1967, 1968, 1990; Fattovich 1988, 1990, 1997a, 2000. See Fattovich 1978, 1996, 2000 for arguments for the African foundation of preAksumite society. 12 Relatively recent examples of this view can be found in Ricci 1984, 1994; Michels 1988, 1994; Raunig 1988. 13 See Phillipson 1998, pages 42-45 for a recent interpretation describing the ‘sudden’ appearance of South Arabian colonists and their writing, monumental stone architecture and sculpture.

59

TRADE AND TRAVEL IN THE RED SEA REGION interesting to note is that unlike specialised ceremonial and elite pre-Aksumite sites in northern Ethiopia, thus far the Ona communities of the Greater Asmara area and the pre-Aksumite community at Kidane Mehret have not produced material culture directly attributable to South Arabia (see Phillips, supra). There is no indication of goods manufactured in South Arabia, no inscriptions of South Arabian derivation, and no South Arabian-inspired statuary or iconography.

culture contact have moved away from a focus on the primacy of dominant foreign colonists in the development of sociopolitical complexity, researchers have yet to explore the dynamics of cultural transformation, and there exists a continued reliance on interpretations stressing exogenous change through immigrant-inspired innovation and passive indigenous assimilation. Recent research in the northern Horn of Africa

Thus, all available evidence suggests local origins for the Ona and Kidane Mehret pre-Aksumite communities and a continuity of occupation, material culture, and subsistence for at least four centuries. There is no evidence from pre-Aksumite domestic settlement sites that a Sabaean tradition of urbanism or technology diffused to the northern Horn of Africa in the ninth century BC. The pre-Aksumite settlements and contemporaneous Sabaean settlements of southern Arabia are best viewed as interacting but independent coeval communities. They seem to have developed within largely similar highland environments, but possess distinct material culture and settlement patterns. The presence of South Arabian-like elements of material culture, monumental architecture, and writing found at pre-Aksumite archaeological sites in the northern Horn of Africa ceremonial and elite sites such as Yeha, Addi Galamo, and Hawlti-Melazo seem to appear during the middle of the first millennium BC, after the establishment of the complex domestic settlements of the Ona sites and Kidane Mehret. Such elements are restricted to what seem to be specialised elite ceremonial centres such as Yeha, Addi Galamo and Hawlti-Melazo and are wholly absent in the surface and subsurface contexts of documented pre-Aksumite domestic sites.

Recent archaeological investigations of non-elite sectors of pre-Aksumite settlement provide a view of thoroughly endogenous origins and continuity in occupation for much of the first millennium BC. We now know from excavations and regional survey in the greater Asmara area of Eritrea and from excavations at the ‘D site’ of Kidane Mehret in the Aksum area of Ethiopia that by 800 BC people were living in fairly ‘large and complex settlements composed of permanent multi-room buildings constructed of stone’.14 These communities kept domesticated cattle, goats, and sheep and cultivated Near Eastern domesticates such as barley, wheat, linseed and lentils.15 Pre-Aksumite communities possessed ceramic and lithic traditions unique to the northern Horn of Africa. Well-crafted handmade vessels characterise the ceramics, including thin-walled types with red slip and burnish and large storage vessels. Ona flaked stone lithic assemblages include a wide and complex variety of scraper forms, perforators, points, backed blades, and microlithic types (Fig. 24). Ona communities possessed various copper-alloy objects, as well as many ground stone artefact types, ceramic and stone figurines, and a diversity of bead types, all of which seem to have been produced locally. The pre-Aksumite Ona communities produced thousands of abstract crescent-shaped stone figurines, termed ‘bulls’ heads’, which depict what seem to be the heads and horns of cattle. Given their abundance on Ona site surfaces and in archaeological deposits, these ‘bulls’ heads’ figurines may have played important functional and ideological roles in agropastoral Ona communities.

Call for a new perspective Despite over 6,000 extant Sabaean inscriptions documented on the southern Arabian peninsula (Breton 1999), there are no epigraphic data from South Arabian sites that mention, even in the most cursory manner, the existence of colonies or vassal polities in the northern Horn of Africa during the first millennium BC. PreAksumite inscriptions from the northern Horn of Africa provide limited detail of sociopolitical formations and record only fragmentary references to personal names and places. Lengthy textual description of any sort is absent. The inscriptions mention royal titles similar to those found in ancient South Arabia, list the names of probable deities closely similar to some in ancient South Arabian panthea, record possible clan or lineage names and provide possible names of ceremonial objects such as incense burners. Some inscriptions make reference to the polity of DΚMT (Daamat).17 The inscriptions from sites such as Addi Galamo (Caquot & Drewes 1955) and Enda

Regional survey in a 145 km² area around Asmara has thus far revealed over one hundred Ona village occupations ranging from less than one hectare to over fifteen hectares in size. Recent excavations and associated radiocarbon dates from sites such as Sembel, Sembel Kushet, Mai Hutsa, Ona Gudo, and others in the greater Asmara area demonstrate that Ona communities were fully sedentary and occupied continuously for several centuries.16 There is continuity in ceramic and lithic artefacts, architecture, and subsistence patterns from the ninth to the fourth centuries BC. What is 14 See Schmidt & Curtis 2001 for Ona sites of Eritrea. See Phillipson 2000 & Phillips (this volume) for Kidane Mehret ‘D site’, Aksum. 15 Boardman 1999, 2000; Cain 2000; Schmidt & Curtis 2001; D’Andrea 2002. 16 See Schmidt & Curtis 2001 for radiocarbon dates for Greater Asmara area Ona sites.

17

See Caquot & Drewes 1955; Pirenne 1956; Schneider 1961; Drewes 1962; Schneider 1973; Fattovich 1990, 2000.

60

MATTHEW C. CURTIS: ANCIENT INTERACTION ACROSS THE SOUTHERN RED SEA

Figure 24: Ona flaked stone lithics

61

TRADE AND TRAVEL IN THE RED SEA REGION Cherqos (Schneider 1961) mentioning Daamat and preAksumite rulers include the South Arabian royal/ religious titles of MKRB (mukarrib) and MLK (malik) and make reference to connections with the Sabaean polity, as evident in the phrase ‘MKRB of Daamat and SabaΜ’ (see de Contenson 1981: 353-354).18 While these inscriptions have been used to support the South Arabian colonist migration model for the origins of pre-Aksumite complex society, it is necessary to consider a couple of questions. Why must pre-Aksumite use of the terms MKRB and MLK in describing Daamat’s leaders and the linking of Daamat with SabaΜ necessarily imply a South Arabian mode of authority? Do such general terms, used within relatively unspecific contexts, provide definitive South Arabian signatures of identity on the pre-Aksumite landscape? Similar challenging questions have been posited indirectly before in a critique of the South Arabian colonist migration model (Isaac & Felder 1988), but no compelling alternative suggestion for an explanatory framework has been offered.

may have used Sabaean (Sabaic), ‘as a prestige language for inscriptional purposes, in somewhat the same way that the Nabataeans and Palmyrenes used Aramaic for their inscriptions, though they probably spoke Arabic themselves’ (Beeston 1988: 100). Beeston’s basic point is important and generally applicable to the context of the northern Horn of Africa in the first millennium BC. We need not assume that pre-Aksumite communities used a South Arabian language as the daily vernacular. Indeed, pre-Aksumite inscriptions are used in very limited and formalised ways and are largely restricted to elite political titles and religious terms. There is a general absence of epigraphic information concerning administrative function, trade, accounting or other more mundane, but essential, aspects of pre-Aksumite society. Writing in South Arabian-like script should not imply ipso facto wide-scale pre-Aksumite adoption of a South Arabian language, nor should it necessarily imply the existence of Sabaean migrations, colonisation and/or direct acculturation.

The author contends that we should consider whether the use of such terms might suggest the appropriation, manipulation, and elaboration of South Arabian symbols of authority by the pre-Aksumite elite to legitimate rule by stressing their relationship to the wider cultural and economic network of the southern Red Sea. The deliberate use of the term MKRB, and the reference to Daamat in conjunction with SabaΜ in the phrase ‘MKRB of Daamat and SabaΜ’ is, perhaps, not unlike that employed by other Red Sea polities, such as the South Arabian kingdom of Himyar. As Jean–François Breton has pointed out (Breton 1999: 178), the Himyarites, although removed both spatially and temporally from the Sabaean state, appropriated Sabaean titles for purposes of prestige and legitimisation. Pre-Aksumite use of South Arabian royal/religious titles need not necessarily suggest that pre-Aksumite leaders viewed SabaΜ as a patron state or political and cultural model. Rather, South Arabian titles may have been appropriated because their use served to further differentiate a leader from subjects and rival elites through connection to the exotic and to a perceived wider economic prosperity. The use of foreigninspired titles may have served to reinforce claims to esoteric knowledge of distant peoples, landscapes, technologies and sources of wealth.19 The use of the written Sabaean language by Himyarite political and religious leaders and by other peoples of the southern Arabian peninsula suggests a widespread recognition of the prestige associated with ancient Sabaean script. Alfred Beeston, for example, has stressed that Himyarites

As with the focus on a small number of fragmentary inscriptions, much attention has been paid to similarities in monumental architecture between South Arabian and pre-Aksumite ceremonial centres. In particular, the temple and large Grat BeΚal Gebri structure at the preAksumite site of Yeha, in Tigray, has been compared to South Arabian forms.20 The Yeha temple is a rectangular structure whose existing remains measure 18.5 by 15 m in area and more than 11 m in height. The building is constructed of ashlar masonry of large rectangular sandstone blocks fitted without mortar. The outer faces and corners of the structure’s walls are finely dressed. The walls sit on a seven-stepped podium base. A carved frieze of ibexes that is now incorporated into a more recent adjacent building was likely an element of the ancient temple. Denticulate plaques and South Arabian inscriptions found at the site may have also once decorated the walls. The monumental Grat BeΚal Gebri structure at Yeha possesses a number of large monolithic pillars of massive square-sectioned form. This structure seems directly associated with the temple, but its function is unclear. Although the overall plan of the temple structure is distinct in important ways from South Arabian examples, the temple’s stepped base, ashlar masonry and rectangular form are similar to temple structures in Yemen. In particular, Yeha’s monolithic pillars resemble those of monumental temple known as the Awwām temple near Mārib.21 Similar pillars have been documented at the pre-Aksumite sites of Kaskasè (Dainelli & Marinelli 1912) and Hawlti-Melazo (de Contenson 1963), suggesting that Yeha’s monumental architectural forms are not unique in the northern Horn of Africa.

18 For various discussions of the terms MKRB (mukarrib) and MLK (malik) see Ryckmans 1951; Audouin et al. 1988; Beeston 1988; Robin 1996; Breton 1999. In total, four rulers of Daamat and Saba’ are recorded in inscriptions from a number of sites in the northern Horn of Africa, including Addi Galamo (Caquot & Drewes 1955); Enda Cherqos (Schneider 1961); Melazo (Schneider 1978); Abuna Garima (Schneider 1973, Schneider 1976c); Matara (Schneider 1965; Drewes & Schneider 1967); Yeha and Kaskasè (Schneider 1976c). 19 See Pauketat 1994 & Peregrine 1999 for similar discussions concerning other world culture areas.

20 See Buxton 1970; de Contenson 1981; Anfray 1972a, 1972b, 1973, 1990; Fattovich 1990; Phillipson 1998. 21 See Phillips 1955; Bowen & Albright 1958; Doe 1971, 1983; Schmidt 1988.

62

MATTHEW C. CURTIS: ANCIENT INTERACTION ACROSS THE SOUTHERN RED SEA most common artefact classes, bear little affinity to South Arabian forms.23 Thus, rather than assuming simple preAksumite assimilation of Sabaean traditions, technology and innovations, we might view South Arabian elements of material culture as the products of pre-Aksumite engagement with various polities in the southern Red Sea area. Such a view suggests that the pre-Aksumites intentionally sought out and incorporated elements of a wider social and economic network that fitted within their own cultural sensibilities, traditions, and aesthetics. As with the use of South Arabian royal titles, incorporation of such elements may have been the result of calculated attempts by elites at legitimisation through appropriation of the symbols and material manifestations of wider socio-religious, political, and economic networks. Archaeologists have observed in many world areas and time periods that elements of initially exotic material culture associated with wider regional, interregional, or global networks or belief systems are crucial in legitimising claims to power.24 Often exotic items of material culture are perceived by populations as imbued with power arising from the items’ distant origins (spatially and sometimes temporally).25 The ability to acquire objects from distant realms enhances the charisma and reputation of the possessor. The appropriated objects help make known to others in society that the elites are linked to a wider network of high prestige socio-religious, political and economic interaction partners. The attraction of obtaining such objects necessitates establishing, cultivating, and maintaining relationships with elites of the distant realms, thus enhancing and deepening pre-existing forms of interregional interaction. In this view, pre-Aksumite incorporations of such elements may have taken place over many generations and need not be viewed as the products of a limited number of Sabaean colonisation or migration events. This is a view of pre-Aksumite active engagement and critical selection from a diverse repertoire of elements in the southern Red Sea cultural milieu, rather than passive incorporation from colonising populations. Such a perspective might be valuable in promoting the concept of pre-Aksumite agency and in creating a more dynamic modelling of cultural transformation in the first millennium BC.

Little is known about non-monumental pre-Aksumite domestic architecture. Domestic pre-Aksumite architectural components have been ignored largely by previous researchers in the northern Horn of Africa and only very recently have archaeologists examined the domestic architecture of pre-Aksumite communities.22 The virtual absence of archaeological investigation of non-monumental domestic architecture in both the northern Horn of Africa and the southern Arabian peninsula makes a systematic and detailed comparative perspective of the pre-Aksumite and South Arabian architectural traditions impossible. We simply do not have the comparative databases necessary to distinguish South Arabian from pre-Aksumite African forms and patterns at the scale of the individual household, community, or region. Indeed, there are important and profound similarities in architectural form and spatial configuration between temples of ancient Mārib and ancient Yeha. However, we must be cautious when identifying similarities in elite architecture, such as that of temples, not to assume that monumental forms represent the totality of pre-Aksumite architecture. Global archaeology has shown that temples and other religious physical structures are parts of religious systems that are often interregional in scale, products of larger inter-ethnic and macro-scale world-views. The non-elite, non-monumental architecture of the domestic setting may be less conditioned by an interregional world-view and more the product of adaptations and responses to localised, endogenous factors of settlement. We must, therefore, be very careful not to assume that the existence of similarities in architectural components like ashlar masonry and ibex friezes found on both pre-Aksumite and South Arabian ceremonial structures are testimony of South Arabian colonisation and/or large scale migration events. The focus on the monumental aspects of architecture and, until recently, almost complete neglect of research into domestic architecture and settlement construction has led some scholars to reach such limiting conclusions. The simple presence of monumental architecture with South Arabian affinities, or writing in South Arabian script does not provide immediate insight into the particular settings in which culture contact occurred nor the transformations within both preAksumite and South Arabian communities that probably resulted from such contact.

It has been proposed by some scholars (e.g. Schneider 1976b) that the first millennium BC saw the development of several competing intermediate-scale or chiefdomlevel polities that existed prior to the florescence of the Kingdom of DΚMT. If such polities existed, then the control over the ritual presentation and manipulation of

As Rodolfo Fattovich has pointed out (Fattovich 2000: 15), the presumed sources of the South Arabian-inspired artefact types come from various locales in the southern Arabian peninsula, including the realms of the Hadramawt, Aden, MaΚīn and Qatabān, in addition to that of SabaΜ. Furthermore, as has been noted above, possible South Arabian influence can be documented in only a small range of pre-Aksumite elite and ceremonial artefacts. Pre-Aksumite ceramics and stone tools, the

23

See Anfray 1966, 1967; Anfray & Annequin 1965; Fattovich 1980; Phillips 2000; Schmidt & Curtis 2001 for descriptions of pre-Aksumite ceramics. See Franchini 1953; Tringali 1969; Fattovich 1977a, 1988, 2000; L. Phillipson 2000a, 2000b; and Schmidt & Curtis 2001 for basic descriptions of pre-Aksumite lithic artefacts. 24 See Drennan 1991; Earle 1991; Feinman 1991; Gilman 1991; Steponaitis 1991; Schortman & Urban 1998; Smith 1998; Urban & Schortman 1999 for examples in various world culture areas. 25 See Helms 1979, 1988, 1992; Renfrew 1986; Schortman & Urban 1994, 1998 for examples.

22

See Phillipson & Reynolds 1996; Phillipson 2000; Schmidt & Curtis 2001 for basic descriptions of pre-Aksumite non-elite, domestic architectural features.

63

TRADE AND TRAVEL IN THE RED SEA REGION encounters. A shared belief system and its nomenclature may have facilitated the creation and maintenance of a pan-Red Sea interregional elite identity. Anthropologists and historians have long noted that trading partners from interacting societies are often afforded the titles and benefits of designation as kin.27 The use of common titles such as MKRB and MLK and formulations such as ‘King SRΜN of the tribe YGΜD, MKRB of Daamat and SabaΜ’ (see de Contenson 1981: 353) between South Arabian and northern Horn of Africa polities might have served to solidify inter-societal contacts between elites. In this view, shared script, titles, and elements of material culture and architecture would have helped foster and maintain a belief system that linked southern Red Sea elite into a pan-southern Red Sea interregional network. Pre-Aksumite elite who claimed the title MKRB may have helped maintain their status by writing in scripts derived from South Arabian types, worshipping in temples and sitting in thrones with iconography reminiscent of that found in South Arabia, and burning incense in honor of a god also found in South Arabian panthea. However, the benefit of such actions may have included more than just an effective means of legitimisation. Perhaps this allowed pre-Aksumite elites a more direct connection to their South Arabian peers and a better ability to emphasise and press upon visiting elites and dignitaries their common points of reference. Participants of this network might have felt linked together as part of a larger macro-culture bound by shared symbols of identity and mutually reinforcing privileged positions. Perhaps, only those individuals who could co-opt and display symbols of pan-Red Sea elite identity would have been given access to the economic and political benefits of this exclusive system.28

initially exotic and potentially controllable symbols and material manifestations of a wider pan-southern Red Sea socio-religious network may have been crucial in regional power contests possibly played out between preAksumite communities. In this scenario ceremonial material culture, iconography and religious architectural forms whose prototypes are first documented in South Arabia might have been utilised as authoritative resources by the elite of competing pre-Aksumite groups. Control over religious practices and associated symbols, sacred space in monumental architectural forms and artefacts might have been central in strengthening and reinforcing solidarity within the ruling or aspiring elite factions. A ritual community with its shared elite participation reinforced unity and amplified the factions’ separation from the larger pre-Aksumite population. Support for the idea that elite groups have controlled the performance of religious rites at large ceremonial centres is suggested by the observation that ritual material such as altars, religious iconography, inscriptions with references to deities and the monumental architectural forms of temple complexes are restricted in space to a limited number of sites such as Yeha and Hawlti-Melazo. A case roughly analogous to the pre-Aksumite situation might be found in complex societies of the Nubian Nile valley. Scholars long viewed the elites of Kerma and Napatan polities as being the products of wholesale assimilation of Egyptian culture. These Nubian societies were seen as Egyptianised and supposed proof was cited in their use of material culture and iconography, the origins and prototypes of which were found in Egypt. More recent scholarship suggests a very different situation, in which the use of Egyptian material culture and iconography is the result of a limited and very conscious borrowing by Nubian elites in order to emphasize their own power and authority in Nubian society. Such a view challenges directly previous Egyptocentric perspectives.26 An interpretive framework similar to that applied recently by Stuart Tyson Smith (1998) to model Nubian/Egyptian interaction might be well suited for exploring the dynamics of Southern Red Sea culture contact between the northern Horn of Africa and the southern Arabian peninsula during the first millennium BC.

It may be that the roots of a pan-southern Red Sea elite identity originate in an earlier (as of yet unknown) trade diaspora in which a specialised group linked by kin relations on both sides of the Red Sea emerged in interregional exchange. The transformation of a trade diaspora into full-blown ecumenical trade by the beginning of the first millennium BC might account for the appearance of shared symbols and material culture in South Arabian and pre-Aksumite elite and ceremonial sites by the latter half of that millennium.29 Are the inscriptions, titles, iconography, and monumental architectural elements found in common the products of a Red Sea ecumenical culture of exchange originating from long-term interaction? Could it be that what has been cited so often as the evidence of the physical presence of South Arabian colonists in the northern Horn of Africa (and the resulting sudden appearance of South Arabian

In addition to elites’ use of foreign titles, material culture and ideology for purposes of legitimisation, we should also consider how such elements might have served to create common terms of reference that helped facilitate interregional interaction between neighboring polities. The use of common political titles, such as MKRB and MLK, may have served to create common terms of reference that were important in interaction between rulers. Common titles may have helped ameliorate cultural differences in diplomatic and economic

27 See Cohen 1969, 1971; Freidel 1979, 1986; Curtin 1975, 1984; Shennan 1982, 1986; Schortman 1989 for examples. 28 See Donley-Reid 1990; Schortman & Urban 1998 for similar discussions in other world culture areas. 29 This is inspired by Phillip Curtin’s [Curtin 1984] discussion of trade diasporas and ecumenical trade zones in cross-cultural interactions. Also see Cohen 1969, 1971 for original descriptions of ‘trade diasporas’.

26 See Reisner 1923; Adams 1977 for examples of Egyptocentric perspectives concerning Nubia. See O’Connor 1993; Smith 1995, 1998 for challenges to Egyptocentric perspectives concerning Nubian societies.

64

MATTHEW C. CURTIS: ANCIENT INTERACTION ACROSS THE SOUTHERN RED SEA we have much data concerning external relations or specific pre-Aksumite roles in wider socio-religious, political and economic networks. However, aspects of elite material culture, inscriptions, titles, religious symbols and iconography suggest that pre-Aksumite elites may very well have actively sought out objects and ideology of wider networks as tools of legitimisation. This is a perspective of active pre-Aksumite engagement with external entities and execution of calculated agendas for the realisation of political and economic goals. It is a perspective very different from previous acculturative interpretations stressing foreign immigrants bestowing new material culture and belief systems and reception by passive pre-Aksumite populations.

features), may actually be the manifestation of a southern Red Sea ecumenical trade zone and a pan-Southern Red Sea elite identity? Discussion and conclusion In a number of articles published over the last decade, Rodolfo Fattovich has argued that continuous interaction throughout the Holocene between African and Arabian populations led to the emergence of distinctly AfroArabian cultures in the littoral areas of the southern Red Sea. Fattovich sees an Afro-Arabian cultural complex (also referred to as the Tihāmah Cultural Complex) in place by the second half of the second millennium BC.30 He supports this assertion through observation of similarities in material culture, particularly ceramics, from sites on both sides of the Red Sea, including Sihi, Wādī UrqΚ, and Sabr on the Arabian peninsula and Adulis in Eritrea. Based on evidence from early ceramic material at Matara and Yeha, Fattovich further suggests that this cultural complex expanded beyond coastal margins and included the highlands of central Eritrea and Tigray by the beginning of the first millennium BC. Fattovich’s hypothesis remains largely untested and the particular ways in which this presumed Afro-Arabian complex may have developed and affected the genesis of later southern Red Sea polities such as Daamat and SabaΜ remains unexplored. The ideas outlined above share a central common theme with Fattovich’s construct. It is a theme concerning the probable importance of long-term interaction between southern Red Sea communities; interaction in which both African and Asian archaeological and historical data are considered of equal significance. It is hoped that the ideas presented above might generate discussion on ways in which such themes might be conceptualised and tested through future archaeological and historical research in the southern Red Sea area. Such ideas might also provide avenues of inquiry for investigating what Christopher Ehret (Ehret 2002) has termed the ‘Commercial Revolution’ of the first millennium BC. The roles played by both South Arabian and pre-Aksumite polities in interregional exchange were probably integral to the economic, social, and political transformations witnessed in the larger Red Sea and Indian Ocean world during that millennium.

In order to make operable more dynamic models of ancient Red Sea interaction, we will require data on ancient political economies in the northern Horn of Africa and southern Arabian peninsula, particularly those relating to craft production, trade, redistribution and administration. Data concerning settlement patterns, site function and specialisation, rural/urban relations, agricultural traditions, technology and art and iconography, all of which are largely absent today, will be essential for developing models of cultural exchange in the first millennium BC. Collecting and interpreting such data will require a commitment to multi-scalar and interdisciplinary research projects. Investigating the multiple forms and variable ways in which interaction and cultural exchange are recorded in the archaeological and historical records will necessitate active, long-term collaboration between researchers working on both sides of the Red Sea. The October 2002 Red Sea Trade and Travel Conference and this volume of its proceedings might represent an important initial step towards that end. Acknowledgments This paper is dedicated to the memory of Michael Haile. Thanks to Peter Schmidt, Francine Stone, Paul Lunde and the Society for Arabian Studies. Special thanks to Jennifer Curtis. Much appreciation goes to Lalemba Tsehaie for the artefact drawings. Travel funds for workshop participation were provided by the University of Florida’s Department of Anthropology and College of Liberal Arts and Sciences. The author’s research in Eritrea during 1999-2000 was funded, in large part, through a student fellowship provided by the U.S. Fulbright Program. Research logistical support in Eritrea was provided by the University of Asmara, National Museum of Eritrea, and U.S. Embassy, Asmara. Many thanks to Yosief Libseqal, Wolde-ab Yisak, Asmerom Kidane, Rezene Russom, Alvaro Higueras, Jeheskel ‘Hezy’ Shoshani, Collette Christian, Seife Berhe, Cathy D’Andrea, Steven Brandt, the staff of the National Museum of Eritrea, the staff of the American Cultural Center in Asmara, and the graduates and current archaeology students of the University of Asmara.

Admittedly, a relatively limited variety and number of archaeological and historical data are yet available to support the theoretical framework suggested here. Such a framework is best viewed as a starting point toward testing hypotheses and building models concerning ancient Red Sea culture contact and the development of complex societies. We do not yet have enough information concerning the domestic economy, social structures, and indigenous belief systems of preAksumite communities to suggest the ways in which elites emerged and maintained power in the northern Horn of Africa during the first millennium BC. Nor do 30

Fattovich 1996, 1997a, 1997b, 2000.

65

TRADE AND TRAVEL IN THE RED SEA REGION of art and civilisation in Arabia Felix. Innsbruck: Pinguin. Boardman S. 1999. The Agricultural Foundation of the Aksumite Empire, Ethiopia: An Interim Report. Pages 137-147 in M. van der Veen (ed.), The Exploitation of Plant Resources in Ancient Africa. New York: Kluwer Academic/Plenum Publishers. ----------2000. Archaeobotany. Pages 363-368, 468-470 in D. W. Phillipson, Archaeology at Aksum, Ethiopia, 1993-7. (2 volumes)/vol. 2. (British Institute in Eastern Africa Memoir 17) (Reports of the Research Committee of the Society of Antiquaries of London 65). London: The British Institute in Eastern Africa and The Society of Antiquaries of London. Bowen R. LeBaron & Albright F. P. 1958. Archaeological Discoveries in South Arabia. (Publications of the American Foundation for the Study of Man 2). Baltimore: Johns Hopkins Press. Breton J-F. 1999. Arabia Felix from the Time of the Queen of Sheba; Eighth Century B.C. to First Century A.D. (transl. A. Lafarge). Notre Dame, Indiana: University of Notre Dame. Buxton D.R. 1970. The Abyssinians. London: Thames and Hudson. Cain C. 2000. Archaeozoology. Pages 369-372, in D. W. Phillipson, Archaeology at Aksum, Ethiopia, 1993-7. (2 volumes)/vol.2. (British Institute in Eastern Africa Memoir 17) (Reports of the Research Committee of the Society of Antiquaries of London 65). London: The British Institute in Eastern Africa and The Society of Antiquaries of London. Caquot A. & Drewes A. J. 1955. Les monuments recueillis à Maqallé (Tigré). Annales d’Ethiopie 1: 17-41. Cohen A. 1969. Custom and Politics in Urban Africa: A Study of Hausa Migrants in Yoruba Towns. Berkeley: University of California Press. ----------1971. Cultural Strategies in the Organization of Trading Diasporas. Pages 266-281 in C. Meillassoux (ed.), The Development of Indigenous trade and Markets in West Africa. London: Oxford University Press. ----------1961. Les fouilles à Haoulti-Melazo en 1958. Annales d’Éthiopie 4: 39-46, pls XXIII-XXXVI. ----------1962. Les monuments d’art sud-arabes découverts sur le site de Haoulti (Ethiopie) en 1959. Syria 39: 64-87. ----------1963. Les fouilles à Haoulti en 1959 – rapport préliminaire. Annales d’Éthiopie 5: 41-52, pls XXVILX. ----------1981. Pre-Axumite culture. Pages 341-361 in G. Mokhtar (ed.), UNESCO General History of Africa. (8 volumes)/vol.2: Ancient Civilizations of Africa. Berkeley: University of California Press. Conti Rossini C. 1906. Sugli Habašāt. Rendiconti della Reale Accademia dei Lincei 15: 39-59. ----------1928. Storia d’Etiopia: parte prima dalle origini all’avvento della dinastia salomonide. Bergamo: Istituto Italiano d’Arti Grafiche.

References Abul-Haggag Y. 1961. A Contribution to the Physiography of Northern Ethiopia. London: Athlone Press. Adams W. 1977. Nubia. Corridor to Africa. London: Alan Lane. Anfray F. 1963a. Une campagne de fouilles à Yĕhā (Février – Mars 1960). Annales d’ Éthiopie 5: 171192, pls CXIV-CLVII. ----------1963b. La première campagne de fouilles à Matara près de Sénafé (Novembre 1959 – Janvier 1960). Annales d’Éthiopie 5: 87-112, pls LXI-CXIII. ----------1964. Notre connaissance du passé éthiopienne d’après les travaux archéologiques recentes. Journal of Semitic Studies 9/1: 247-249. ----------1966. La poterie de Matarā. Rassegna di Studi Etiopici 22: 5-74. ----------1967. Matarā. Annales d’Éthiopie 7: 33-53. ----------1968. Aspects de l’archéologie éthiopienne. Journal of African History 9: 345-366. ----------1970. Notes archéologiques. Annales d’Éthiopie 8: 31-42. ----------1972a. Les fouilles de Yeha, Mai-Juin 1972. Documents pour servir à l’histoire des civilisations éthiopiennes 3: 57-63. ----------1972b. Les fouilles de Yeha. Annales d’Éthiopie 9: 45-56. ----------1973. Les fouilles de Yeha, Mai-Juin 1973. Abbay: Documents pour servir à l’histoire des civilisations éthiopiennes (CNRS) 4: 35-38. ----------1974. Deux villes axoumites: Adoulis et Matarā. Pages 745-765 in Atti del IV Congresso Internationale di Studi Etiopici- Roma, 10-15 aprile 1972. (2 volumes)/vol.1. (Accademia Nazionale dei Lincei. Problemi attuali di scienza e di cultura; quaderno n. 191). Roma: Academia Nazionale dei Lincei. ----------1990. Les anciens Ethiopiens, siècles d’histoire. Paris: Armand Colin. Anfray F. & Annequin G. 1965. Matarā: deuxième, troisième et quatrième campagnes de fouilles. Annales d’Éthiopie 6: 49-86. Audouin R., Breton J. & Robin C. 1988. Towns and temples – the emergence of South Arabian civilization. Pages 63-77 in W. Daum (ed.), Yemen: 3000 years of art and civilisation in Arabia Felix. Innsbruck: Pinguin. Bard K.A., Coltorti M., DiBlasi M.C., Dramis F. & Fattovich R. 2000. The Environmental History of Tigray (Northern Ethiopia) in the Middle and Late Holocene: A Preliminary Outline. African Archaeological Review 17/2: 65-86. Bard K.A., Fattovich R., Manzo A. & Perlingieri C. 1997. Archaeological Investigations at Bieta Giyorgis (Aksum), Ethiopia: 1993-1995 Field Seasons. Journal of Field Archaeology 24: 387-403. Beeston A.F.L. 1988. Pre-Islamic Yemeni Inscriptions. Pages 99-106 in W. Daum (ed.), Yemen: 3000 years

66

MATTHEW C. CURTIS: ANCIENT INTERACTION ACROSS THE SOUTHERN RED SEA Abbay: Documents pour servir à l’histoire de la civilisation éthiopienne (CNRS) 9: 25-30. ----------1980. Materiali per lo studio della ceramica preaksumita etiopica (Annali dell’ Istituto Universitario Orientale di Napoli 40/4: Supplemento 25). Napoli: Istituto Universitario Orientale. ----------1988. Remarks on the Later Prehistory and Early History of Northern Ethiopia. Pages 85-104 in T. Beyene (ed.), Proceedings of the Eighth International Conference of Ethiopian Studies (2 volumes)/vol. 1. Addis Ababa: Institute of Ethiopian Studies. ----------1990. Remarks on the Pre-Aksumite Period in Northern Ethiopia. Journal of Ethiopian Studies 23: 3-33. ----------1996. The Afro-Arabian circuit: interregional contacts between the Horn of Africa and Southern Arabia in the 3rd-2nd millennia B.C. Pages 395-402 in L. Krzyzaniak, K. Kroeper & M. Kobusiewicz (eds), Interregional Contacts in the Later Prehistory of Northeastern Africa. (Studies in African Archaeology 5). Poznan: Poznan Archaeological Museum. ----------1997a. The Near East and Eastern Africa: Their Interaction. Pages 479-484 in J. Vogel (ed.), Encyclopedia of Precolonial Africa: Archaeology, History, Languages, Cultures, and Environments. Walnut Creek, California; London: AltaMira Press. ----------1997b. Northeastern African States. Pages 484489 in J. Vogel (ed.), Encyclopedia of Precolonial Africa: Archaeology, History, Languages, Cultures, and Environments. Walnut Creek, California; London: AltaMira Press. ----------2000. Aksum and the Habashat: State and Ethnicity in Ancient Northern Ethiopia and Eritrea. (Working Papers in African Studies, No. 228). Boston, Massachusetts: African Studies Center, Boston University. Fattovich R. & Bard K.A. 1995. Scavi archeologici nella zona di Aksum. E. Ona Enda Aboi Zeugè e Ona Nagast (Bieta Giyorgis). Rassegna di Studi Etiopici 37: 5-35. ----------1996. Scavi archeologici nella zona di Aksum, F. Ona Nagast (Bieta Giyorgis). Rassegna di Studi Etiopici 38: 71-94. ----------1997a. The I.U.O./B.U. Excavations at Bieta Giyorgis (Aksum) in Tigray (Northern Ethiopia). Journal of Ethiopian Studies 30/1: 1-29. ----------1997b. Scavi archeologici nella zona di Aksum. G. Ona Enda Aboi Zeugè e Ona Nagast (Bieta Giyorgis). Rassegna di Studi Etiopici 39: 1-22. Fattovich R., Bard K.A., Petrassi L, & Pisano V. 2000. The Aksum Archaeological Area: A Preliminary Assessment. Napoli: Istituto Universitario Orientale. Feinman G. 1991. Demography, Surplus, and Inequality: Early Political Formations in Highland Mesoamerica. Pages 229-262 in T. K. Earle (ed.), Chiefdoms: Power, Economy, and Ideology. Cambridge: Cambridge University Press. Finneran N. 1998. Investigating the earliest foodproducing communities of the northern Ethiopian

Curtin P. 1975. Economic Change in Precolonial Africa: Senegambia in the Era of the Slave Trade. Madison, Wisconsin: University of Wisconsin Press. ----------1984. Cross-Cultural Trade in World History. Cambridge: Cambridge University Press. Curtis M. C. 2001. Recent Regional Archaeological Research in Eritrea: Investigating the Origins and Development of Early Complex Society in the Greater Asmara Area. Paper presented at the International Conference, Independent Eritrea: Lessons and Prospects, Asmara, Eritrea, July 22, 2001. D’Andrea A. C. 2002. Comments on Archaeobotanical Samples, Eritrea. Unpublished preliminary report submitted to University of Asmara archaeological field schools program and Greater Asmara Regional Archaeological Survey Project. Dainelli G. & Marinelli O. 1912. Risultati di un viaggio scientifico nella Colonia Eritrea. Firenze: Galletti e Cocci. Doe B. 1971. Southern Arabia. London: Thames and Hudson. ----------1983. Monuments of South Arabia. Cambridge/New York: Oleander Press. Donley-Reid L. 1990. A Structuring Structure: The Swahili House. Pages 114-126 in S. Kent (ed.), Domestic Architecture and the Use of Space. Cambridge: Cambridge University Press. Drennan R. 1991. Pre-Hispanic Chiefdom Trajectories in Mesoamerica, Central America, and Northern South America. Pages 263-287 in T. K. Earle (ed.) Chiefdoms: Power, Economy, and Ideology. Cambridge: Cambridge University Press. Drewes A. J. 1959. Les inscriptions de Melazo. Annales d’Ethiopie 3: 83-99. ----------1962. Inscriptions de l’Ethiopie Antique. Leiden: E. J. Brill. Drewes A. J. & Schneider R. 1967. Documents Epigraphiques de l’Ethiopie. Annales d’Ethiopie 7: 89-102. ----------1970. Documents Epigraphiques de L’Éthiopie – II. Annales d’Ethiopie 8: 57-72. ----------1976. Origine et development de l’ecriture éthiopienne jusqu’à l’époque des inscriptions royales d’Axoum. Annales d’Ethiopie 10: 95-107. Earle T. K. 1991. The Evolution of Chiefdoms. Pages 115 in T. K. Earle (ed.), Chiefdoms: Power, Economy, and Ideology. Cambridge: Cambridge University Press. Ehret C. 2002. The Civilizations of Africa: a History to 1800. Charlottesville, Virginia: The University Press of Virginia. Fattovich R. 1977a. Some data for the study of the cultural history in ancient Northern Ethiopia. Nyame Akuma 10: 6-18. ----------1977b. Pre-Aksumite Civilization of Northern Ethiopia: a Provisional Review. Proceedings of the Seminar for Arabian Studies 7: 73-78. ----------1978. Traces of a possible African component in the pre-Aksumite Culture of Northern Ethiopia. 67

TRADE AND TRAVEL IN THE RED SEA REGION Littmann E., Krencker D. & von Lüpke T. 1913. Deutsche Aksum-Expedition. (3 volumes in 4). Berlin: G. Reimer. Michels J. 1979. Axumite archaeology: an introductory essay, Pages 1-34 in Y. Kobishchanov (author), Axum. University Park: Pennsylvania State University Press. ----------1988. The Axsumite Kingdom: A Settlement Archaeology Perspective. Pages 173-183 in Proceedings of the Ninth International Conference of Ethiopian Studies. (Moscow, African Institute, USSR Academy of Sciences, 26-29 August 1986). Moscow: Nauka. ----------1994. Regional Political Organization in the Axum-Yeha Area during the Pre-Axsumite and Axsumite Eras. Pages 61-80 in C. Lepage (ed.), Études éthiopiennes. (Actes de la Xe conference internationale des études éthiopiennes, Paris, 24-28 août 1988). (3 volumes)/vol.1 Paris: Société Française pour les études éthiopiennes. Munro-Hay S. & Tringali G. 1993. The Ona sites of Asmara and Hamasien. Rassegna di Studi Etiopici 35: 135-170. Nastasi P. 1994. Notes Concerning Climatic and Floristic Regions of Eritrea. Roma: Istituto Italo-Africano. O’Connor D. 1993. Ancient Nubia: Egypt’s Rival in Africa. Philadelphia: The University Museum, University of Pennsylvania. Paribeni R. 1907. Ricerche nel luogo dell’antica Adulis. Monumenti Antichi, Reale Accademia dei Lincei 18/3: 437-572. Pauketat T.R. 1994. The Ascent of Chiefs: Cahokia and Mississippian Politics in Native North America. Tuscaloosa: University of Alabama Press. Peregrine P.N. 1991. Prehistoric Chiefdoms on the American Midcontinent: A World-System Based on Prestige-Goods. Pages 193-211 in C. Chase-Dunn & T. Hall (eds), Core/Periphery Relations in Precapitalist Worlds. Boulder, Colorado: Westview. ----------1999. Legitimation Crises in Prehistoric Worlds. Pages 37-52 in P. N. Kardulias (ed.), World-Systems Theory in Practice: Leadership, Production, and Exchange. Lanham: Rowman and Littlefield. Phillips J.S. 2000. Pottery and clay objects. Pages 303312 in D. W. Phillipson (ed.), Archaeology at Aksum, Ethiopia, 1993-7. (2 volumes)/vol.2. (Memoirs of the British Institute in Eastern Africa 17) (Report of the Research Committee of the Society of Antiquaries of London 65). London: The British Institute in Eastern Africa and the Society of Antiquaries of London. Phillips W. 1955. Qataban and Sheba. Exploring Ancient Kingdoms of the Biblical Spice Routes of Arabia. London: Victor Gollancz. Phillipson D.W. 1998. Ancient Ethiopia: Aksum, its antecedents and successors. London: British Museum Press. ----------2000. Archaeology at Aksum, Ethiopia, 1993-7. (2 volumes). (Memoirs of the British Institute in Eastern Africa 17) (Report of the Research Committee of the Society of Antiquaries of London

highlands: a case study from Aksum, Tigray. Nyame Akuma 49: 35-42. ----------1999. Post-Pleistocene Socio-Economic Developments in the Northern Ethiopian/Eritrean Highlands: A Case Study from Aksum Tigray. Unpublished Ph.D. thesis, University of Cambridge. ----------2000a. A New Perspective on the Late Stone Age of the Northern Ethiopian Highlands: Excavations at Anqqer Baahti, Aksum, Ethiopia 1996. Azania 35: 21-51. ----------2000b. Excavations at the Late Stone Age Site of Baahti Nebait, Aksum, Northern Ethiopia, 1997. Azania 35: 53-73. Food and Agricultural Organization of the United Nations (FAO). 1994. Eritrea: Agricultural sector review and project identification 2. Rome: FAO. Franchini V. 1951. Pitture rupestri a Sollum Ba’atti. Rassegna di Studi Etiopici 10: 121-123. ----------1952. Pitture rupestri nel Ba’atti Sollum nel Deghiem. Rassegna di Studi Etiopici 11: 42-48. ----------1953. Stazioni litiche in Eritrea. Bollettino, Istituto di Studi Etiopici, Asmara 1: 25-30. ----------1958. Altre pitture rupestri nell’Akkelè Guzay. Bollettino, Istituto di Studi Etiopici 2: 1-12, 47-51. Freidel D. A. 1979. Culture Areas and Interaction Spheres: Contrasting Approaches to the Emergence of Civilization in the Maya Lowlands. American Antiquity 44: 36-54. ---------------1986. Terminal Classic Lowland Maya: Successes, Failures, and Aftermaths. Pages 409-430 in J. Sabloff & E. W. Andrews (eds), Late Lowland Maya Civilization. Albuquerque: University of New Mexico Press. Gilman A. 1991. Trajectories Towards Social Complexity in the Later Prehistory of the Mediterranean. Pages 146-168 in T. K. Earle (ed.), Chiefdoms: Power, Economy, and Ideology. Cambridge: Cambridge University Press. Glaser E. 1895. Die Abesseiner in Arabien und Afrika. Munich: H. Lukaschik. Graziosi P. 1964. New Discoveries of Rock Paintings in Ethiopia. Antiquity 138: 91-99, 187-190. Helms M.W. 1979. Ancient Panama: Chiefs in Search of Power. Austin: University of Texas Press. ----------1988. Ulysses’s Sail: An Ethnographic Odyssey of Power, Knowledge, and Social Distance. Princeton: Princeton University Press. ----------1992. Long-Distance Contacts, Elite Aspirations, and the Age of Discovery in Cosmological Context. Pages 157-174 in E. M. Schortman & P. A. Urban (eds), Resources, Power, and Interregional Interaction. New York: Plenum Press. Isaac E. & Felder C. 1988. Reflections on the Origins of the Ethiopian Civilization. Pages 71-83 in T. Beyene (ed.), Proceedings of the Eighth International Conference of Ethiopian Studies. [1984, University of Addis Ababa]. (2 volumes/vol. 1). Addis Ababa: Institute of Ethiopian Studies. Leclant J. 1959. Haoulti-Melazo (1955-1956). Annales d’Ethiopie 3: 43-57. 68

MATTHEW C. CURTIS: ANCIENT INTERACTION ACROSS THE SOUTHERN RED SEA 3000 Years of Art and Civilization in Arabia Felix. Innsbruck: Pinguin. Schmidt P. R. 1999. Preliminary Report on the Results from the Training Excavations at the Sembel Site. [Unpublished report on file in the National Museum of Eritrea, Asmara]. Schmidt P.R. & Curtis M.C. 2001. Urban precursors in the Horn: early 1st-millennium BC communities in Eritrea. Antiquity 75: 849-859. Schneider R. 1961. Inscriptions d’Enda Cerqos. Annales d’Ethiopie 4: 61-65. ----------1965. Notes épigraphiques sur les découvertes de Matara. Annales d’Ethiopie 6: 88-92. ----------1973. Deux inscriptions sudarabiques du Tigré. Bibliotheca Orientalis 30: 385-389. ----------1976a. Documents épigraphiques de l’Ethiopie. V. Annales d’Ethiopie 10: 81-93. ----------1976b. Les débuts de l’histoire Ethiopienne. Abbay: Documents pour servir à l’histoire de la civilisation éthiopienne (CNRS) 7: 47-54. ----------1976c. L’inscription chrétienne d’Ezana en écriture sudarabe. Annales d’Ethiopie 10: 109-117. ----------1978. Documents épigraphiques de l’Ethiopie. VI. Annales d’Ethiopie 11: 129-132. Schortman E.M. 1989. Interregional Interaction in Prehistory: The Need for a New Perspective. American Antiquity 54: 52-65. Schortman E.M. & Urban P.A. (eds) 1992. Resources, Power, and Interregional Interaction. New York: Plenum Press. Schortman E.M. & Urban P.A. 1987. Modeling Interregional Interaction in Prehistory. Pages 37-95 in M. Schiffer (ed.), Advances in Archaeological Method and Theory. (11 volumes)/vol.11. Orlando, Florida: Academic Press. ----------1994. Living on the Edge: Core/Periphery Relations in Ancient Southeastern Mesoamerica. Current Anthropology 35: 401-430. ----------1998. Culture Contact Structure and Process. Pages 102-125 in J. G. Cusick (ed.), Studies in Culture Contact; Interaction, Culture Change, and Archaeology (Occasional Paper No. 25). Carbondale, Illinois: Center for Archaeological Investigations, Southern Illinois University. Shennan S.J. 1982. Exchange and Ranking: The Role of Amber in the Early Bronze Age of Europe. Pages 3345 in C. Renfrew & S. Shennan (eds), Ranking, Resource, and Exchange. Cambridge: Cambridge University Press. ----------1986. Interaction and Change in Third Millennium B.C. Western and Central Europe. Pages 137-148 in C. Renfrew & J. Cherry (eds), Peer Polity Interaction and Sociopolitical Change. Cambridge: Cambridge University Press. Smith S.T. 1995. Askut in Nubia: The Economics and Ideology of Egyptian Imperialism in the Second Millennium B.C. London: Kegan Paul International. ----------1998. Nubia and Egypt: Interaction, Acculturation, and Secondary State Formation from the Third to First Millennium B.C. Pages 256-287 in

65). London: The British Institute in Eastern Africa and the Society of Antiquaries of London. Phillipson D.W. & Reynolds A. 1996. B.I.E.A. Excavations at Aksum, Northern Ethiopia, 1995. Azania 31: 99-147. Phillipson L. 2000a. Lithics. Pages 352-363 in D. W. Phillipson (ed.), Archaeology at Aksum, Ethiopia, 1993-7. (2 volumes)/vol. 2. (Memoirs of the British Institute in Eastern Africa 17) (Report of the Research Committe of the Society of Antiquaries of London 65). London: The British Institute in Eastern Africa and the Society of Antiquaries of London. ----------2000b. Lithic Industries. Pages 449-453 in D. W. Phillipson (ed.), Archaeology at Aksum, Ethiopia, 1993-7. (2 volumes)/vol. 2. (Memoirs of the British Institute in Eastern Africa 17) (Report of the Research Committe of the Society of Antiquaries of London 65). London: The British Institute in Eastern Africa and the Society of Antiquaries of London. Pirenne J. 1956. Paléographie des Inscriptions SudArabes 1. Brussels: Académie Royale Flammande de Belgique. ----------1970. Haoulti, Gobochéla (Mélazo) et le site antique. Annales d’Ethiopie 8: 117-127. Piva A. 1907. Una civiltà scomparsa dell’Eritrea e gli scavi archeologici nella regione di Cheren. Nuova Antologia: 323-335. Raunig W. 1988. Yemen and Ethiopia-Ancient Cultural Links between two neighbouring Countries on the Red Sea. Pages 409-418 in W. Daum (ed.), Yemen: 3000 Years of Art and Civilisation in Arabia Felix. Innsbruck: Pinguin. Reisner G. 1923. Excavation at Kerma I-III, IV-V. Cambridge, Massachusetts: Harvard African Studies. Renfrew C. 1986. Varna and the Emergence of Wealth in Prehistoric Europe. Pages 141-168 in A. Appadurai (ed.), The Social Life of Things: Commodities in Cultural Perspective. Cambridge: Cambridge University Press. Ricci L. 1984. L’Expansion de l’Arabie Meridionale. Pages 249-257 in J. Chelhod et al. (eds), L’Arabie du Sud. Histoire et Civilisation. (3 volumes)/vol. 1: Le peuple yéménite et ses racines. Paris: Maisonneuve et Larose. ----------1994. On Both Sides of al-Mandab. Pages 409417 in H. G. Marcus (ed.), New Trends in Ethiopian Studies: Papers of the 12th International Conference of Ethiopian Studies, Michigan State University, 5-10 September 1994 (2 volumes)/vol.1. Lawrenceville, New Jersey: Red Sea Press. Robin C. 1996. Sheba dans les inscriptions d’Arabie du Sud. Supplement au Dictionnaire de la Bible. Paris: Letouzey. Ryckmans J. 1951. L’institution monarchique en Arabie meridionale avant L’Islam - MaΜîn et Saba. (Bibliothèque du Muséon 28). Louvain: Publications Universitaires. Schmidt J. 1988. Ancient South Arabian Sacred Buildings. Pages 78-98 in W. Daum (ed.), Yemen:

69

TRADE AND TRAVEL IN THE RED SEA REGION J. G. Cusick (ed.), Studies in Culture Contact; Interaction, Culture Change, and Archaeology. (Occasional Paper No. 25). Carbondale, Illinois: Center for Archaeological Investigations, Southern Illinois University. Steponaitis V. 1991. Contrasting Patterns of Mississippian Development. Pages 193-228 in T. K. Earle (ed.), Chiefdoms: Power, Economy, and Ideology. Cambridge: Cambridge University Press. Sundström R. 1907. “Letter to Dr. Enno Littmann” in E. Littmann, Preliminary Report of the Princeton University Expedition to Abyssinia. Zeitschrift für Assyriologie 20: 170-182. Tringali G. 1965. Cenni sulle “Ona” di Asmara e dintorni. Annales d’Ethiopie 6: 143-152. ----------1967. Necropoli di Curbacaiehat (Asmara). Journal of Ethiopian Studies 5: 109. ----------1969. Varietà di asce litiche in ‘ouna’ dell’altopiano eritreo. Journal of Ethiopian Studies 7: 119. ----------1979. Necropoli di Cascassè e oggetii sudarabici (?) dalla regione di Asmara (Eritrea). Rassegna di Studi Etiopici 26: 47-98. ----------1981. Note su ritrovamenti archeologici in Eritrea. Rassegna di Studi Etiopici 28: 99-113. ----------1987. Reperti antichi di scultura minore e di ornamenti dall’Eritrea e da Aksum. Rassegna di Studi Etiopici 31: 213-218. Urban P.A. & Schortman E.M. 1999. Thoughts on the Periphery: The Ideological Consequences of Core/Periphery Relations. Pages 125-152 in P.N. Kardulias (ed.), World-Systems Theory in Practice: Leadership, Production and Exchange. Lanham: Rowman and Littlefield. Woldai T., Schetselaar E. & Haile Mariam M. 2000. Structural Analysis of the Asmara Area, Eritrea: Using Remote Sensing and GIS. Pages 2-7 in Information for sustainable development: Proceedings of the 28th International Symposium on Remote Sensing of the Environment, March 27-31 2000 Cape Town, South Africa. Enschede: International Institute for Aerospace Survey and Earth Sciences.

70

The ‘pre-Aksumite’ state in northern Ethiopia and Eritrea reconsidered Rodolfo Fattovich (Zarins 1989, 1990, 1996). The ability of African peoples to navigate in the Red Sea at this time is supported by the occurrence of obsidian tools ascribable to the Wilton Technocomplex at Dahlak Kabir (Blanc 1955). There is equally early evidence for navigation at this time in the Gulf and the Mediterranean.

Introduction The history of northern Ethiopia and Eritrea was characterised by the development of an early state (the so-called ‘Kingdom of DΚMT’) with indisputable South Arabian (mainly Sabaean) features in the mid-first millennium BC (see e.g. Anfray 1990: 17-61). This specific culture historical period is conventionally named the ‘pre-Aksumite period’, and the material evidence of the early state is usually identified with the ‘preAksumite culture’.1

These contacts consolidated in the third to second millennia BC, as we can infer from a comparison between the ceramics from sites in eastern Sudan, Ethiopia, Djibouti and Yemen dating to c. 2500-1500 BC (Fattovich 1997a). In particular, the ceramics of the ‘Gash Group’ (c. 27001500/1400 BC) from Mahal Teglinos near Kassala, suggest that the western Eritrean-Sudanese lowlands were included in a network of contacts with the southern regions of the Horn of Africa and southern Arabia, as well as with eastern Sudan, Nubia and Egypt (Fattovich 1991, 1993).

The pre-Aksumite period was initially divided into two or possibly three phases, corresponding to the formation, consolidation and decline of the early state.2 There were two explanations for the origins of this state: a Sabaean colonisation of the Eritrean and Tigrean highlands in the mid- first millennium BC or a process of cultural contact beginning in the early first millennium BC.3

Beginning in the late third to mid-second millennia BC the people living along the coast of the southern Red Sea and the Gulf of Aden played an important part in the development of the Afro-Arabian circuit.5 Archaeological evidence from Sihi to the south of Jīzān in the Saudi Tihāmah and Adulis near the Gulf of Zula in the Eritrean Sahel suggests that these sites were part of a coastal cultural complex in the late third to mid-second millennia BC.6 This complex was characterised by bowls and pots with a rounded base and geometric decoration (mainly triangles along the rim and vertical lines on the body).

At present, the results of archaeological research in the southern Red Sea suggest that the development of the pre-Aksumite state was part of a long-term process of interaction between opposite regions of the Red Sea, and was related to the progressive inclusion of the southern Red Sea into the long distance trade between the Mediterranean and the Indian Ocean.4 Three phases can be distinguished: 1) the emergence and consolidation of an Afro-Arabian interchange in the fifth to early first millennia BC, 2) the development of the pre-Aksumite state in Eritrea and Tigray in the mid-first millennium BC, and 3) the decline of the pre-Aksumite state and the rise of a new polity, the kingdom of Aksum, in the late first millennium BC. In this paper this process will be outlined on the basis of fresh evidence.

Some similarities between ceramics from Sihi and Adulis, and Nubian C-Group and Kerma examples (late third to mid-second millennia BC), although not yet conclusive, may point to contacts with the Nile Valley,7 when the state of Kerma (c. 2500-1500 BC) controlled the trade between Egypt and the regions of the Horn of Africa, and the Upper Nile (see e.g. Bonnet ed. 1990). These contacts may also be supported by the occurrence of circular structures with vertical slabs at the edge, similar to Kerma graves, at Aqiq, on the coast to the south of the Barka Delta.8

The Afro-Arabian interchange circuit (fifth - early first millennia BC) A network of exchanges based on the circulation of obsidian emerged between the northern Horn of Africa (Eritrea, Tigray, and Eritrean-Sudanese border lowlands) and southern Arabia in the fifth millennium BC. In the second millennium BC, the obsidian circuit included the Eritrean Sahel, the southern Atbai Mountains and the Barka Delta, the Dahlak Islands, and the Arabian Tihāmah

5

See Fattovich 1997b, 1997a; Phillips 1998; Manzo 1999; Buffa & Vogt 2001. 6 See Zarins et al. 1981; Zarins & Zaharani 1985; Zarins & Al-Badr 1986; Paribeni 1908: 446-451. 7 For a comparison see Zarins & Zaharani 1985; Paribeni 1908; Gratien 1978. See also for a critical assessment of these similarities Buffa and Vogt (2001). 8 See Cremaschi et al. 1986; Fattovich 1995.

1

See Anfray 1964, 1967, 1968; Fattovich 1990. Anfray 1968; Fattovich 1980, 1990; Michels 1988, 1994. 3 Conti Rossini 1928; Drewes 1962; Anfray 1968; Schneider 1976a; Ricci 1984; Fattovich 1977, 1990. 4 See Fattovich 1996a, 1996b, 1997b, 1997a; Manzo 1998, 1999; Kitchen 2002. 2

71

TRADE AND TRAVEL IN THE RED SEA REGION Gash and Jabal Mokram Group specimens and might indicate contacts with the western lowlands in the midsecond millennium BC (see Tringali 1978: fig. 24). This is supported by the occurrence of typical jars of the Ona Culture in an assemblage of the Late Gash and/or Jabal Mokram Group at Ntanei, near Agordat, in the middle Barka valley.14 Some chipped stone bull-heads (Tringali 1978: figs 13-22.) may point to contacts with the Arabian peninsula, as bull-heads are a common motif in the rock art of central Arabia (see Cervicek 1978). Contacts with South Arabia are also suggested by the occurrence of vessels with a ring base in both regions. Finally, a clay artefact similar to the stone ‘bull-heads’ of the Ona Culture was collected at Sabr (see Buffa & Vogt 2001: fig. 3.3).

In turn, the role of coastal people as intermediaries between the Yemeni highlands and western EritreanSudanese lowlands may be supported by the occurrence of stelae similar to those of the Gash Group at Aqiq and elements of the Bronze Age Khawlan Culture at Sihi.9 Beginning in the mid-second millennium BC, a new cultural complex appeared along the south-western coast of Arabia (Sabr Culture, c. 2000-600 BC).10 Most likely, this culture emerged at MaΜlayba and Sabr in the region of Aden in the early second millennium BC, and progressively included the coastal regions of southwestern Arabia, as far as the Saudi Tihāmah, in the midsecond to early first millennia BC.11 The ceramics of the Sabr Culture were characterised by large jars and bowls with vertical or horizontal handles, tulip-shaped pots and bowls with a ring base. They were often decorated with burnished lines.

Moreover, a fine brown polished ware, including bowls decorated with geometric motifs like those on the pottery of Adulis and Sihi, as well as black-topped ware were found in the lowest strata (V-VIII) at Matara.15 Jars fragments similar to those from Sabr point to a date in the early first millenium BC for these strata.

Despite the scarce evidence, contacts and possibly exchanges between the Arabian and African regions of the southern Red Sea continued in the second millennium BC. This is suggested by fragments of bowls with everted rims and potsherds decorated with burnished linear motifs from the Terminal Gash Group (c. 1500-1400 BC) at Mahal Teglinos and at Wādī UrqΚ in the Yemeni Tihāmah.12 A few potsherds with geometric motifs, comparable to specimens of the Pan-Grave Culture (c. 1700-1500 BC) and/or Jabal Mokram Group (c. 14001000 BC) of the eastern Desert, have been also found at Sabr. They have been interpreted as imports from Africa (Buffa & Vogt 2001).

The pre-Aksumite state (c. 700/600-400 BC) In the first millennium BC, the interaction between the populations of the opposite shores of the southern Red Sea become more intense and culminated with the rise of a pre-Aksumite state (‘Kingdom of Daamat’, c. 700/600400 BC) in the highlands of Eritrea and central Tigray. The location of pre-Aksumite sites points to an expansion of the pre-Aksumite state along a restricted area from Qohaito in Eritrea to the Takaze River in Tigray. This expansion was marked by the foundation of ceremonial centres, such as Kaskasè, Addi Gerametem and Fikiya in Eritrea, and Yeha in Tigray. A town was located at Matara (Fattovich 1990).

In the late second to early first millennia BC, the people of the central Eritrean highlands were also part of a network of contacts with the coastal regions of the southern Red Sea as we can infer from the evidence of the Ona Culture in the Hamasien and the lower strata (VVIII) at Matara in the Akkele Guzay. The Ona Culture can be ascribed to a sedentary population settled in the region of Asmara in the late second to early first millennia BC.13 The sites include cemeteries and settlements which are sometimes very large in size. The ceramics include bowls and necked jars decorated with geometric motifs, similar to those of Kerma and Sihi, along the rim and/or the shoulder, as well as black-topped vessels (see Tringali 1978: fig. 23, 25; B.Gratien 1978: Pl. V-VII).

The archaeological evidence of this early phase is very scarce. Stratum IV at Matara and the occupation phase II at Yeha could be ascribed to it. The ceramics from these strata contain vessels that are comparable to those of the Sabr Culture.16 Although Sabaean was the official language of the kingdom, South Arabian cultural influence is only evident in the monumental architecture, inscriptions, art, votive incense altars, offering tables and bronze seals. Many of these elements can also be related to other South Arabian kingdoms and are not necessarily to be attributed to SabaΜ.17 Most of the ceramics have a local origin. Black-topped ware was a typical product of this period. Only very few vessels are South Arabian in style, and

The range of contacts of the Ona Culture is still uncertain. Some fragments of big storage pots with everted rims from Sembel-Cushet are similar to Late 9

See Fattovich 1989, 1995; Buffa & Vogt 2001. See Phillips 1998; Buffa & Vogt 2001. 11 Vogt 1997; Vogt & Sedov 1997, 1998, 2000; Buffa 2000. 12 Fattovich et al. 1988-89; Fattovich 1996a. For Wādī UrqΚ see Tosi 1985, 1986. This remark is based on a personal view of the ceramics and original illustrations from this site. The author wishes to thank Maurizio Tosi and Maurizio Cattani for providing these materials. 13 Tringali 1965, 1978, 1981; Munro-Hay & Tringali 1993; Schmidt & Curtis 2001; see also Fattovich 1988. 10

14

Personal observation at the Sudan National Museum, Khartoum; see also Arkell 1954. 15 See Anfray 1966; Fattovich 1980. 16 Anfray 1966; Fattovich 1972, 1976, 1980. 17 See Fattovich 1977, 1990; Anfray 1994; Robin & de Maigret 1998.

72

RODOLFO FATTOVICH: THE ‘PRE-AKSUMITE’ STATE IN NORTHERN ETHIOPIA AND ERITREA CONSIDERED foundation of a temple (Robin & de Maigret 1998: 797). This epigraphic evidence records four kings who ruled the SB (Sabaeans?) and ΚBR, suggesting a dual division of the people. The last two kings used the Sabaean title mukarrib. These kings worshipped South Arabian as well as indigenous gods. Moreover, the inscriptions suggest that Sabaeans were settled on the plateau in the mid-first millennium BC.21 The textual evidence however points to an indigenous origin for the pre-Aksumite state. The earliest monumental inscriptions record queens who were accorded very high status, and who were possibly equal to kings. Queens do not seem to have played such an important role in South Arabia, but high status queens are numerous in the ancient and traditional kingdoms of subSaharan Africa, such as in the Nubian kingdom of Kush.22

could be imported (Fattovich 1980; cf. Phillips this volume infra and Porter forthcoming). The main component of pre-Aksumite pottery is similar to the Ona Culture ware, suggesting that this was the local cultural background to the kingdom of Daamat. Black-topped ware, in particular, can be related to a late prehistoric tradition of the Nile Valley. The lithic tools, too, are similar to the late prehistoric ones of the Sudan and the Horn of Africa (Fattovich 1978, 1994). At present, we can tentatively distinguish two phases in the development of the pre-Aksumite state: 1) an initial penetration of the South Arabians, possibly coming from different regions of southern Arabia, into the Eritrean and Tigrean highlands; 2) the consolidation of a proper state as a consequence of the inclusion of the highlands within the area of direct political and economic influence of the Sabaeans.

The end of the pre-Aksumite state and the rise of Aksum

Rock inscriptions in the Akkele Guzay (central Eritrea) suggest that South Arabians were settled on the plateau in the early first millennium BC.18 These inscriptions, presently dated from the mid-ninth to the mid-sixth centuries BC, include South Arabian names, and other names which are not attested in South Arabia or in Ethiopia (Drewes 1962: 5-10). The location of these inscriptions suggests that the South Arabians used the valley of the Haddas river, connecting Adulis to Qohaito, to access the plateau, but so far no definite evidence for a South Arabian presence has been found at Adulis.19 In Arabia, the excavations conducted at al-Hāmid suggest that the valley of Wādī Sihām was a route from the highlands to the sea in the eighth to sixth centuries BC (Phillips 1997, 1998). This might point to a progressive shift from a northern maritime route, crossing the Red Sea from Jīzān to Massawa and the Gulf of Zula, in the early second millennium BC to a southern one from the Wādī Sihām region to the Danakil coast to the south of the Buri peninsula in the late second-early first millennia BC.

The end of the pre-Aksumite state is still obscure. At present, we can only suggest that this state declined or collapsed in Tigray in the fourth to third centuries BC, although most likely an Afro-Arabian urban (state?) society survived in the Akkele Guzzay (see Fattovich 1990). There is some slight evidence for warfare being the cause of the end of the pre-Aksumite state in Tigray as the monumental building, Grat BeΚal Gebri at Yeha, seems to have been destroyed by fire, and the Great Temple also shows some traces of fire (Anfray 1972, 1997). Two shrines at Hawlti and a small temple at Melazo, to the east of Aksum, may provide very intriguing evidence for the dating to this final phase of the pre-Aksumite Period in Tigray.23 The ceramics associated with these monuments, although very badly preserved, are similar to those from the top of the deposit at Grat BeΚal Gebri, and thus might be later than the monumental building (Fattovich 1980). The shrines at Hawlti and the small temple at Melazo are similar to Meroitic temples, and may suggest an inclusion (albeit temporary) of the region of Aksum into the Nubian area of influence (Fattovich 1990). These buildings were associated with well-carved artefacts in ‘Ethio-Sabaean’ and Sabaean style: a so-called ‘throne’, some fragments of another ‘throne’, and two female statues at Hawlti; a few small votive altars with Sabaean inscriptions at Melazo. Some pillars similar to the South Arabian examples were also arranged in a circle around the shrines at Hawlti.24

The use of the Sabaean monumental script and language in the royal inscriptions and the construction of large ceremonial centres at sites such as Kaskasè in Akkele Guzay and Yeha in Tigray characterised the second phase of the pre-Aksumite period.20 Most likely, the preAksumite state was consolidated in the seventh to sixth centuries BC, when the so-called ‘Great Temple’ was erected at Yeha over an earlier South Arabian shrine. This dating is supported by the strong similarity between the Great Temple and the Minaean temple at Baraqish, which is dated to 700-500 BC (Robin & de Maigret 1998). A 700 BC date is also supported palaeographically by some royal inscriptions from Yeha, which record the

In the absence of any exhaustive record of excavations at these sites, we can only speculate about the meaning of this association. Most likely, the thrones, statues and

18

Drewes 1962; Ricci 1959, 1960, 1994. Some Sabaean artefacts, which are exhibited in the Archaeological Museum in Asmara (see Ricci 1983), are said to come from Adulis, but they were initially imported from Yemen. 20 Anfray 1968, 1970; de Contenson 1981; Fattovich 1990. 19

21

See Drewes 1962; Schneider 1973, 1976b. See Schneider 1973, 1976a; Robin & de Maigret 1998. de Contenson 1962, 1963; Leclant 1959. 24 See de Contenson 1962, 1963; Pirenne 1967. 22 23

73

TRADE AND TRAVEL IN THE RED SEA REGION lowlands.26 There is also evidence of similarities in mortuary symbolism of proto-Aksumite Aksum and Upper Nubia. These include types of grave goods, basins for offerings in the mortuary cult, and possibly subsidiary burials associated with the tombs (Bard et al. 2002). This evidence suggests that in the fourth-third centuries BC the pre-Aksumite state was already in decline or most probably collapsed, at least in central Tigray.

small altars were not in their original location, but were moved from elsewhere (Pirenne 1970). This is supported by the circular arrangement of the pillars, which suggests that they were re-erected. In South Arabia pillars were always arranged in a straight line at the entry of temples. The similarity in plan between the shrines at Hawlti and the small temple 292 at Meroe, where a bronze head of Augustus was found (Welsby 1996: 189), might suggest that the thrones and statues were Meroitic war trophies. At present, there is no clear record in the Meroitic texts of any raid as far as the Tigrean highlands. The Annals of Harsiyotef (404-369 BC) and Nastasen (335-315 BC) demonstrate that in the fourth century BC the Meroites were active in the eastern Sudan and probably conducted frequent raids as far at the Eritrean-Sudanese lowlands.25 The Annals of Nastasen, in particular, record campaigns against many populations that cannot be conclusively located; the campaigns might have been conducted in the Tigrean highlands, and might have contributed to the decline and/or collapse of the pre-Axumite state in this region in the fourth century BC.

We have no corresponding factual evidence from Eritrea. It is possible that an urban society maintaining some preAksumite cultural traditions survived in Akkele Guzay until the first -second centuries AD, when the Aksumite kingdom progressively expanded to the east and included Eritrea and Yemen within the area of political control of Aksum. This could be supported by the apparent continuity of occupation from pre-Aksumite to Aksumite times at Matara (Anfray 1966, 1974). Conclusion •

Since 1993 the Joint Archaeological Expedition at Aksum of the University of Naples ‘L’Orientale’ and Boston University have conducted archaeological investigations on the hill of Bieta Giyorgis, to the northwest of Aksum. The excavations have demonstrated that a new polity emerged at Aksum in the fourth century BC (proto-Aksumite Period, c. 400-150 BC) (Fattovich & Bard 1994, 2001). The proto-Aksumite polity at Aksum was most probably part of a constellation of small-scale chiefdoms scattered in the region. Another chiefdom was probably located at Medaguè, 6 km to the south-west of Aksum, where a stelae-field with rough monoliths and proto-Aksumite ceramics similar to those from Bieta Giyorgis were recorded (de Contenson 1961).

•

The proto-Aksumite evidence shows only a few elements of continuity with the pre-Aksumite material culture. These consist of architectural ground plans and the morphology of the pottery. A building with long narrow rooms without any evidence for entrances or windows may be compared to similar buildings; in South Arabia; these structures formed the basement of the building and may have been used as storerooms (see e.g. Doe 1971, fig. 36). To date, the only evidence of continuity in the pottery consists of foot-ring bases. Proto-Aksumite mortuary evidence, on the other hand, points to a basic change in ideology between the pre-Aksumite and protoAksumite cultures. Pre-Aksumite public works were the cult temples of the gods whereas proto-Aksumite monumental works commemorated elite burials.

•

•

•

•

At present, the proto-Aksumite culture can be ascribed to an indigenous Tigrean tradition, possibly related to the late prehistoric cultural traditions of the Eritrean-Sudanese

25

26

See Eide et al. 1996: ii, 437-464, 471-501.

74

Any reconstruction and interpretation of the preAksumite Period in Eritrea and northern Ethiopia is still uncertain and mainly speculative because of the scarce archaeological evidence, the ambiguity of the epigraphic evidence, and the absence of a solid chronology. In the author’s opinion, however, the picture emerging from a preliminary and tentative review of all the evidence dating to the second to first millennium BC from the southern Red Sea is coherent and suggests the following: The emergence of the pre-Aksumite state in Eritrea and Tigray was the result of a long interaction between the populations of the opposite regions of the southern Red Sea. The coastal people of the Arabian Tihāmah and African Sahel had a crucial role in this interaction in the second millennium BC. This interaction was intense in the late second to early first millennia BC, when the African regions were included in the South Arabian area of influence and individuals or small groups of South Arabians settled on the plateau in Eritrea and mixed with the local people. An Afro-Arabian complex society arose in Eritrea in the early first millennium BC and progressively included Tigray into its area of economic and perhaps political influence. This complex society was exposed to a strong influence of the kingdom of SabaΚ and most likely became a satellite state of SabaΚ in the mid- first millennium BC. This state collapsed in Tigray, perhaps under Meroitic pressure, in the fourth-third centuries BC, but we cannot exclude the possibility that a preAksumite urban society survived in Eritrea until the early first millenium AD.

Fattovich et al. 1998; Bard et al. 2002.

RODOLFO FATTOVICH: THE ‘PRE-AKSUMITE’ STATE IN NORTHERN ETHIOPIA AND ERITREA CONSIDERED •

Buffa V. & Vogt B. 2001. Sabir - Cultural Identity between Saba and Africa. Pages 437-450 in R. Eichmann & H. Parzinger (eds), Migration und Kulturtransfer: Der Wandel vorderund zentralasiatischer Kulturen im Umbruch vom 2. zum 1. vorchristlischen Jahrtausend; Akten des Internationalen Kolloquiums Berlin, 23.bis 26. November 1999. (Eurasien-Abteilung [und] Orientabteilung des Deutschen Archäeologischen Instituts). Bonn: Habelt. Cervicek P. 1978. Some African Affinities of Arabian Rock Art. Rassegna di Studi Etiopici 27: 5-12. de Contenson H. 1961. Trouvailles fortuites aux environs d’Axoum (1957-1959) Annales d’Ethiopie 4: 15-23. ----------1962. Les monuments d’art sudarabe découverts sur le site de Haoulti (Ethiopie) en 1959. Syria 39: 6487. ----------1963. Les fouilles à Haoulti en 1959 – rapport préliminaire. Annales d’Ethiopie 5: 41-86. ----------1981. Pre-Aksumite Culture. Pages 341-361 in G. Mokhtar (ed.), UNESCO General History of Africa. (8 volumes)/vol.2: Ancient Civilizations of Africa. Berkeley: University of California Press. Conti Rossini C. 1928. Storia d’Etiopia; parte prima dalle origini all’avvento della dinastia salamonide. Bergamo: Istituto Italiano d’Arti Grafiche. Cremaschi M., D’Alessandro A., Fattovich R. & Piperno M. 1986. Gash Delta Archaeological Project: 1985 Field Season. Nyame Akuma 27: 45-48. Doe B. 1971. Southern Arabia. London: Thames and Hudson. Drewes A-J. 1962. Inscriptions de l’Ethiopie antique. Leiden: E. J. Brill. Eide T., Hägg T., Pierce R. & Török L. 1996. Fontes Historiae Nubiorum: textual sources for the history of the Middle Nile region between the eighth century B.C. and the sixth century A.D. (4 volumes)/vol 2: From the mid-fifth to the first century B.C. Bergen: Klassisk Institutt, Universitetet i Bergen. Fattovich R. 1972. Sondaggi stratigrafici - Yeha 1971. Annales d’Éthiopie 9: 65-86. ----------1976. Introduzione alla ceramica pre-aksumita di Grat BeΚal Guebri (Yeha). Annales d’Ethiopie 11: 105-128. ----------1977. Pre-Aksumite Civilization of Northern Ethiopia: a Provisional Review. Proceedings of the Seminar for Arabian Studies 7: 73-78. ----------1978. Traces of a possible African component in the pre-Aksumite Culture of Northern Ethiopia. Abbay: Documents pour servir à l’histoire de la civilisation éthiopienne (CNRS) 9: 25-30. ----------1980. Materiali per lo studio della ceramica preaksumita etiopica. (Annali dell’Istituto Universitario Orientale di Napoli 40/4: Supplemento 25). Napoli: Istituto Universitario Orientale. ----------1988. Remarks on the Late Prehistory and Early History of Northern Ethiopia. Pages 85-104 in T. Beyene (ed.), Proceedings of the Eighth International Conference of Ethiopian Studies. (2 volumes)/vol.1. Addis Ababa: Institute of Ethiopian Studies.

A new, indigenous, polity emerged in central Tigray and developed as the kingdom of Aksum in the late first millennium BC.

Finally, the kingdom of Aksum consolidated with the progressive incorporation of the Eritrean highlands in the early first millennium AD. References Anfray F. 1964. Notre connaissance du passé éthiopien d’après les travaux archéologiques recentes. Journal of Semitic Studies 9/1: 247-249. ----------1966. La poterie de Matarā. Rassegna di Studi Etiopici 22:5-74. ----------1967. Matarā. Annales d’Ethiopie 7: 33-53. ----------1968. Aspects de l’archéologie éthiopienne. Journal of African History 9: 345-366. ----------1972. Les fouilles de Yeha. Annales d’Ethiopie 9: 45-56. ----------1974. Deux villes axoumites: Adoulis et Matarā. Pages 745-765 in Atti del IV Congresso Internazionale di Studi Etiopici – Roma, 10-15 aprile 1972. (2 volumes)/vol.1. (Accademia Nazionale dei Lincei. Problemi attuali di scienza e di cultura; quaderno n. 191). Roma: Accademia Nazionale dei Lincei. ----------1990. Les anciens éthiopiens, siècles d’histoire. Paris: Armand Colin. ----------1994. Considérations sur quelques aspects archéologiques des relations de l’Ethiopie et de l’Arabie antiques. Pages 17-25 in Y. Beyene, R. Fattovich, P. Marrassini & A. Triulzi (eds), Etiopia e oltre. Studi in onore di Lanfranco Ricci. Napoli: Instituto Universitario Orientale. ----------1997. Yeha. Les ruines de Grat BeΚal Gebri Recherches archéologiques. Rassegna di Studi Etiopici 39: 5-23. Arkell A.J. 1954. Four Occupation sites at Agordat. Kush 2: 33-62. Bard K.A., Fattovich R., Manzo A. & Perlingieri C. 2002. Aksum Origins, Kassala and Upper Nubia: new evidence from Bieta Giyorgis (Aksum). Archéologie du Nil Moyen 9: 31-49. Blanc A-C. 1955. L’industrie sur obsidienne des Iles Dahlac (Mer Rouge). Pages 355-357 in L. Balout (ed.), Actes du Congrès Panafricain de Préhistoire, 2e session - Algier 1952. Paris:Art et Métiers Graphiques. Bonnet C. (ed.) 1990. Kerma, royaume de Nubie, l’antiquité africaine au temps des pharaons. Exposition organisée au musée d’art et d’histoire, Genève, 14 juin-25 novembre 1990. Genève: Mission archéologique de l’université de Genève au Soudan: Musée d’art et d’histoire. Buffa V. 2000. Malayba: una comunità di agricoltori dell’Età del Bronzo del II millennio a. Cr. Pages 7173 in A. de Maigret & A. Avanzini (eds), Yemen - Nel paese della Regina di Saba. Milano: Skira. 75

TRADE AND TRAVEL IN THE RED SEA REGION ----------1989. The stelae of Kassala: a new type of funerary monument in the Eastern Sudan. Archéologie du Nil Moyen 3: 55-70. ----------1990. Remarks on the Pre-Aksumite Period in Northern Ethiopia. Journal of Ethiopian Studies 23: 3-33. ----------1991. At the Periphery of the Empire: The Gash Delta (Eastern Sudan). Pages 40-47 in W. V. Davies (ed.), Egypt and Africa. Nubia from Prehistory to Islam. London: British Museum Press. ----------1993. The Gash Group of the Eastern Sudan: an outline. Pages 439-448 in L. Krzyzaniak, M. Kobusiewicz & J. Alexander (eds), Environmental Change and Human culture in the Nile Basin and Northern Africa until the Second Millennium BC. Poznan: Poznan Archaeological Museum. ----------1994. The Contribution of the Recent Fieldwork at Kassala (Eastern Sudan) to Ethiopian Archaeology. Pages 43-51 in C. Lepage (ed.), Études éthiopiennes. (Actes de la Xe Conference Internationale des Études Éthiopiennes, Paris, 24.-28. août 1988). (3 volumes)/vol.1. Paris: Société Française pour les études éthiopiennes. ----------1995. L’archeologia del Mar Rosso: problemi e prospettive. Note in margine alla recente pubblicazione di due siti costieri della Somalia settentrionale. Annali dell’Isituto Universitario Orientale di Napoli 55/2: 158176. ----------1996a. The Afro-Arabian circuit: interregional contacts between the Horn of Africa and Southern Arabia in the 3rd-2nd millennia BC. Pages 395-402 in L. Krzyzaniak, K. Kroeper & M. Kobusiewicz (eds), Interregional Contacts in the Later Prehistory of Northeastern Africa. (Studies in African Archaeology 5). Poznan: Poznan Archaeological Museum. ----------1996b. Punt: the archaeological perspective. Beiträge zur Sudanforschung 6: 15-29. ----------1997a. The Contacts between Southern Arabia and the Horn of Africa in Late Prehistoric and Early Historical times: A View from Africa. Pages 273-286 in A. Avanzini (ed.), Profumi d’Arabia: atti del convegno. (Saggi di storia antica 11) Roma: L’Erma di Bretschneider. ---------- 1997b. The Near East and Eastern Africa: Their Interaction. Pages 479-484 in J. Vogel (ed.), Encyclopedia of Precolonial Africa: Archaeology, History, Languages, Cultures, and Environments. Walnut Creek, California; London: Alta Mira Press. Fattovich R. & Bard K. A. 1994. The Origins of Aksum: A View from Ona Enda Aboi Zaguè (Tigray). Pages 16-25 in H.G. Marcus (ed.), New Trends in Ethiopian Studies: Ethiopia 94. (Papers of the 12th International Conference of Ethiopian Studies, Michigan State University, 5-10 September 1994). (2 volumes)/vol.1. Lawrenceville, New Jersey Red Sea Press. ----------2001. The Proto-Aksumite Period: An Overview. Annales d’Éthiopie 17: 3-24. Fattovich R., Manzo A. & Bard K.A. 1998. Meroe and Aksum: New Elements of Comparison. Archéologie du Nil Moyen 8: 43-53.

Fattovich R., Sadr K. & Vitagliano S. 1988-1989. Society and Territory in the Gash Delta (Kassala, Eastern Sudan), 3000 B.C.-A.D. 300/400. Origini 14: 329-357. Gratien B. 1978. Les Cultures Kerma. Essai de Classification. (Publications de l’Université de Lille III). Villeneuve-d’Ascq: Université de Lille. Kitchen K.A. 2002. Egypt, Middle Nile, Red Sea and Arabia. Pages 383-401 in S. Cleuziou, M. Tosi & J. Zarins (eds), Essays on the Late Prehistory of the Arabian Peninsula.(Serie orientale Roma 93). Roma: Instituto Italiano per l’Africa e l’Oriente. Leclant J. 1959. Haoulti-Melazo (1955-1956). Annales d’Ethiopie 3: 43-57. Manzo A. 1998. The Dynamics of External Contacts of Northern Ethiopia and Eritrea from Proto-Historical to Aksumite Times, Late 2nd Millennium BC - Late 1st Millennium AD. Pages 35-52 in Äthiopien und seine nachbarn/Ethiopia and its neighbours.(3. wissenschaftliche Tagung des Orbis Aethiopicus in Gniew/Polen, 25-29 September 1997). Frankfurt: Orbis Aethiopicus. ---------- 1999. Echanges et contacts le long du Nil et de la Mer Rouge dans l’époque protohistorique (IIIe et IIe millénaires avant J-C.). Une synthèse préliminaire. (Cambridge Monographs in African Archaeology 48/British Archaeological Reports, International Series 782). Oxford: BAR Publishing. Michels J.W. 1988. The Aksumite Kingdom: a Settlement Archaeological Perspective. Pages 173183 in Proceedings of the Ninth International Conference of Ethiopian Studies. (Moscow, African Institute, USSR Academy of Sciences, 26-29 August 1986). Moscow: Nauka. ---------- 1994. Regional Political Organization in the Aksum-Yeha Area during the Pre-Aksumite and Aksumite Eras. Pages 61-80 in C. Lepage (ed.), Etudes éthiopiennes. (Actes de la Xe conference Internationale des études éthiopiennes, Paris, 24.-28. août 1988). (3 volumes)/vol. 1. Paris: Société Française pour les études éthiopiennes. Munro-Hay S. & Tringali G. 1993. The Ona sites of Asmara and Hamasien. Rassegna di Studi Etiopici 35:135-170. Paribeni R. 1908. Ricerche sul luogo dell’antica Adulis. Monumenti Antichi, Reale Accademia dei Lincei 18/3: 437-572. Phillips C. 1997. Al-Hamid: A Route to the Red Sea? Pages 287-294 in A. Avanzini (ed.), Profumi d’Arabia: atti del convegno. (Saggi di storia antica 11). Rome: L’Erma di Bretschneider. ---------- 1998. The Tihamah c. 5000 to 500 BC. Proceedings of the Seminar for Arabian Studies 28: 233-237. Pirenne J. 1967. Haoulti et ses monuments, nouvelle interpretation. Annales d’Ethiopie 7: 125-133. ---------- 1970. Haoulti, Gobochela (Melazo) et le site antique. Annales d’Ethiopie 8: 117-127. Porter, A. (forthcoming) Amphora trade between South Arabia and East Africa in the first millennium BC: a

76

RODOLFO FATTOVICH: THE ‘PRE-AKSUMITE’ STATE IN NORTHERN ETHIOPIA AND ERITREA CONSIDERED Pages 64-69 in A. de Maigret & A. Avanzini (eds), Yemen - Nel paese della Regina di Saba. Milano: Skira. Welsby D. 1996. The Kingdom of Kush: the Napatan and Meroitic empires. London: British Museum Press. Zarins J. 1989. Ancient Egypt and the Red Sea Trade: the case for obsidian in the Predynastic and Archaic Periods. Pages 339-363 in A. Leonard & B.B. Williams (eds), Essays in Ancient Civilization Presented to Helene Kantor. Chicago: Oriental Institute of the University of Chicago. ---------- 1990. Obsidian and the Red Sea trade: prehistoric aspects. Pages 507-541 in M. Taddei & P. Callieri (eds), South Asian Archaeology 1987. Roma: Istituto Italiano per il Medio ed Estremo Oriente. ---------- 1996. Obsidian in the Larger Context of Predynastic/Archaic Egyptian Red Sea Trade. Pages 89-106 in J. Reade (ed.), The Indian Ocean in Antiquity. London: Kegan Paul International, in association with the British Museum. Zarins J. & Al-Badr H. 1986. Archaeological Investigations in the Southern Tihama Plain, II. Atlal 10: 36-57. Zarins J., Al-Jawarad Murad A. & Al-Yish K. S. 1981. The Comprehensive Archaeological Survey, a: The Second Preliminary Report on the Southwestern Province. Atlal 5: 9-37. Zarins J. & Zaharani A. 1985. Recent Archaeological Investigations in the southern Tihama Plain. 1404/1984. Atlal 9: 65-107.

re-examination of the evidence. Proceedings for the Seminar for Arabian Studies 33. Ricci L. 1959. Iscrizioni rupestri dell’Eritrea, I. Rassegna di Studi Etiopici 15: 59-95. ---------- 1960. Iscrizioni rupestri dell’Eritrea, II. Rassegna di Studi Etiopici 16: 77-119. ---------- 1983. Museo Archeologico di Asmara. Itinerario descrittivo. Rome: Istituto Italiano per l’Africa. ---------- 1984. L’expansion de l’Arabie Méridionale. Pages 249-257 in J. Chelhod et al. (eds), L’Arabie du sud. Histoire et Civilisation. (3 volumes)/vol. 1: Le peuple yéménite et ses racines. Paris: G.–P. Maisonneuve et Larose. ----------- 1994. On both sides of al-Mandab. Pages 409417 in H. G. Marcus (ed.), New Trends in Ethiopian Studies: Papers of the 12th International Conference of Ethiopian Studies. (Michigan State University, 510 September 1994). (2 volumes)/vol.1. Lawrenceville, New Jersey: Red Sea Press. Robin C.J. & de Maigret A. 1998. Le grand temple de Yéha (Tigray, Éthiopie), après la première campagne de fouilles de la Mission Française (1998). Académie des Inscriptions et Belles-Lettres, Compte Rendus des séances (Juillet-Octobre 1998): 737-798. Schmidt P.R. & Curtis M. 2001. Urban precursors in the Horn: early 1st-millennium BC communities in Eritrea. Antiquity 75: 349-359. Schneider R. 1973. Deux inscriptions sudarabiques du Tigré. Bibliotheca Orientalis 30: 385-389. ---------- 1976a. Le debuts de l’histoire éthiopienne. Abbay: Documents pour servir à l’histoire de la civilisation éthiopienne (CNRS) 7: 47-54. ---------- 1976b. Documentes epigraphiques de l’Ethiopie. V. Annales d’Ethiopie 10: 81-93. Tosi M. 1985. Tihama Coastal Archaeological Survey, Preliminary Report 1985. East and West 35/4: 363369. ---------- 1986. Tihama Archaeological Survey 1986: a Short Note. East and West 37/4: 400-414. Tringali G. 1965. Cenni sulle “Ona” di Asmara e dintorni. Annales d’Ethiopie 6: 143-152. ---------- 1978. Necropoli di Cascassè ed oggetti sudarabici (?) dalla regione di Asmara (Eritrea). Rassegna di Studi Etiopici 26: 47-98. ---------- 1981. Note su ritrovamenti archeologici in Eritrea. Rassegna di Studi Etiopici 28: 99-113. Vogt B. 1997. Sabr - une ville de la fin du IIe millénaire dans l’arrière-pays d’Aden. Pages 47-48 in C. Robin & B. Vogt (eds), Yémen - au pays de la reine de Saba. Paris: Flammarion. Vogt B. & Sedov A. 1997. La culture de Sabr, sur la côte yémenite. Pages 42-46 in C. Robin & B. Vogt (eds), Yémen - au pays de la reine de Saba. Paris: Flammarion. ---------- 1998. The Sabir culture and coastal Yemen during the second millennium BC - the present state of discussion. Proceedings of the Seminar for Arabian Studies 28: 261-270. --------- 2000. La cultura di Sabir e la piana costiera yemenita nella seconda metà del II millennio a. Cr. 77

78

Pre-Aksumite Aksum and its neighbours Jacke Phillips the kingdom must have allowed their dispersal, for sites with similar Sabaean features are found from the African Red Sea coast to west of Aksum, more than 250 km inland, although the majority of sites are found inland (See Anfray 1990: fig. 16; cf. Fattovich this volume).

As early as the Early Stone Age, cultural development in northern Ethiopia (here including modern Eritrea) (Fig. 25) generally follows the development of the rest of north-east Africa (Asamerew, Finneran & Phillips 2003). Enough features are either common to all or demonstrably interrelated within the region in toto that interaction and an ultimately common tradition are undeniable (Manzo 1999, with further bibliography). Within this ongoing development, and in northern Ethiopia sometime after c. 800 BC, extremely strong foreign features began to manifest themselves in northern Ethiopia, within a politically cohesive culture rising to ‘civilisation’ that is known to us as the DΚMT or ‘Daamat’ kingdom. Multiple sites in this kingdom of asyet indeterminate extent exhibit for the first time features astonishingly parallel to those that had appeared demonstrably earlier in the Sabaean civilisation, located in what is now Yemen.

Discussion of this influence, of course, especially emphasises contact at the élite level, and its political and mercantile overtones. The author will examine how all this contact that we can readily recognise in the archaeological and textual record affected those ‘ordinary’ elements of pre-Aksumite society who lived and worked at Aksum, and who would be less likely to be influenced and affected by outside forces. Examination indicates that the ‘ordinary’ pre-Aksumite individual was far less profoundly influenced by the outside world than were those in the higher echelons of society. This upper/lower hierarchal division can be quite marked when one culture impacts itself upon another so suddenly and blatantly. Direct impact as immediate and profound as we see here can be – but rarely is – recognised as a ‘veneer’ below which the distinct elements of the indigenous culture remain little if at all changed. It is a basic concept, but it may be for this reason that we historians and archaeologists tend to overlook the point

These features, to name but some of the most prominent visual parallels, include ashlar stone buildings of dressed masonry constructed without the use of mortar, largescale stone statuary in human form, and carved stone inscriptions in the same script and, apparently, language as in Yemen. An underlying political unification within

Figure 25: Map of northern Ethiopia and Eritrea showing the location of Aksum (‘D site’)

79

TRADE AND TRAVEL IN THE RED SEA REGION produced the vast majority of the evidence for ‘EthioSabaean’ contact and the strongly Sabaean influence to be recognised in their material remains – the dressed stone temples, the statuary, the inscriptions – together with other imports from Egypt and Nubia (Phillips 1995) that testify to the long-range external contacts of the DΚMT kingdom in other directions. And it is on these imports and influences from the Sabaean kingdom across the Red Sea that the ‘Ethio-Sabaean’ concept is based. The author does not dispute this concept, but some limitations need to be imposed upon our understanding of its permutations.

that foreign goods and influence may have affected only the smallest percentage of any given ancient population. In a general overview on the relationship of the Ethiopian area with its neighbours, the author has previously commented upon the material relations and influence of the Sabaean culture on that of Ethiopia in the first millennium: …..these influences do not seem to have been allpervasive, however, but the phenomenon is similar to the pattern readily recognised in the succession of ancient Nubian cultures that adopted a veneer of ‘Egyptian’ features at the élite levels. (Phillips 1997: 442)

Our excavations at what we have termed the ‘D site’ have revealed substantial stone-built architecture of well-built but undressed walls between half and more than one metre thick, having large stone ‘rubble’-filled interior, in clearly domestic habitation use. We uncovered well-laid pebble floors of an external courtyard and some large and open interior spaces, and sherd-laid hearths (Phillipson & Phillips 2000: 271-273, 277; Phillips, Reynolds & Spandl 2000: 280-283, 295-298). Our zooarchaeologist, Chester Cain, notes that the population relied heavily for meat on domesticates, mainly cattle, sheep and goat, rather than hunting wild game (Cain 2000: 369, fig. 324). They also grew a variety of cereals, to be further discussed below. The pottery associated with these levels is extremely well made, high-fired and technically of superior quality (Phillips 2000: 303-312, figs 266-271). They are the products of considerable knowledge and technical development by the potters who produced them (and their predecessors). In one area of the site, in what must have been some kind of storage facility, we recovered a large quantity of storage vessels, some almost a metre in height and yet less than a centimetre in thickness. Clearly the population living here at this time was both sedentary and sophisticated, with a high level of social organisation and material goods, understandable in a society that also produced the finer architecture of higher status buildings already known in the DΚMT kingdom.

Recent excavations at Aksum itself have amply demonstrated this comment. The recently published excavations at ‘D site’ by the British Institute in Eastern Africa (1994-1996) (Phillipson D.W. 2000), have begun to illuminate the living conditions of the ‘ordinary’ indigenous population of Aksum in pre-Aksumite times. These phases date to about the eighth – fifth centuries BC according to our 14C dates from seeds and wood samples in clearly stratified deposits (Phillipson D.W. 2000, ii: 267-379, 505 fig. 425 phases 1-3B, contra Phillipson D.W.1998: 48). These results do not compare well with previous interpretations of Sabaean influence exerted on the indigenous culture in the north Ethiopian highlands during this period. We did not anticipate uncovering pre-Aksumite levels at Aksum, as none had been reported in any earlier excavations here (Munro-Hay 1989), and it had seemed that no habitation at Aksum existed this early. The Periplus of the Erythraean Sea, our earliest reference to Aksum as a place, dates to some 500-800 years later.1 Actual pre-Aksumite habitation elsewhere within DΚMT has barely been brought to light through excavation and thus this aspect of the period has been very little studied. Nonetheless, contemporary sites have been excavated in the immediate region and within the DΚMT kingdom, such as the temples at Yeha and Hawlti, the élite graves at Yeha that may be (in part) a royal cemetery and, further east, the élite structures at Matara, to name only the most important sites. These and other sites have been published only in preliminary form, and thus are not easily discussed in any detail.2 Yet these sites have

However, what we did not find was evidence for many imported goods or even for much foreign-influenced locally produced goods. The ‘D site’, less than 20 km from the temple at Hawlti and less than 50 km from the temple at Yeha (two sites very strongly influenced by Sabaean culture) produced virtually no evidence for anything other than its own indigenous cultural material. What little can be cited nonetheless indicates that the levels in which they were recovered at Aksum generally are contemporary with the period of strong Sabaean influence at the nearby sites, or at least no earlier. Future

1 The Periplus of the Erythraean Sea is a c. 40-70 AD mercantile handbook that mentions, ‘την µητροπολιν τον ’Αξωµιτην’ (‘the metropolis of the Axomites’). The Periplus goes on to comment that ‘into it is brought all the ivory from beyond the Nile through what is called Kyeneion (Κυηνειον), and from there down to Adulis‘ (Periplus/Casson ed. 1989: 52-53, 106-108; on the dating, see 6-7). 2 Broad discussions of the period and its Sabaean influence can be found in Anfray 1990: 65-71; Munro-Hay 1991: 61-66; Phillipson 1998: 41-49, all with more specific references. Note the differences in both interpretation and dating of the evidence but that all agree strong influence from across the Red Sea is clearly evident for the preAksumite period. The author specifically highlights Yeha and Hawlti in this paper because they are located so near to Aksum and indicate

abundant evidence of Sabaean influence. Stone inscriptions also have been recovered in the immediate Aksum area, at Abba Pantaleon at what likely was a pre-Christian temple (?) before its conversion to a church (Munro-Hay 1991: 64, fig. 9, 209) and at Seglamien, slightly south-west of Aksum in an unclear context (Ricci & Fattovich 1984-86: pl. 1). Other sites also indicate Sabaean features, but are more distant from Aksum.

80

JACKE PHILLIPS: PRE-AKSUMITE AKSUM AND ITS NEIGHBOURS 343, fig. 342). Could it be that there is no tradition of metalworking at this period here, beyond the élite level and Sabaean-influenced sites? Metalworking certainly was practiced by the Sabaeans, as well as by the Egyptians and the Nubians, and abundant bronze artefacts – tools as well as jewellery items representative of the DΚMT civilisation – were recovered both at Yeha and at Hawlti. Therefore the knowledge of working the bronze itself existed immediately nearby, although perhaps not the knowledge of actually extracting the metal with which to work. The limited use of bronze as a material only for jewellery usually suggests that its potential had not yet been realised, or that it was either generally unavailable or not required. This may have been the case at Aksum, as discussed further below.

research or fieldwork, either at Aksum or elsewhere, may circumscribe the wide date range cited here. We have no large stone sculpture or other representations of humans or animals at the ‘D site’, and only one ‘inscription’, a single ‘alpha’ sign applied to the shoulder of a large storage jar found together with many others in what probably was a storage area (Fig. 26). Yet both inscriptions and large-scale statuary carved in stone are known at Hawlti and Yeha at this same period. Nor are there any buildings in ashlar masonry, with cut and fitted blocks, like the temple at Yeha, although the scale and quality of stone construction the inhabitants did use at Aksum is quite impressive. We have no glass, no faience, and no gold,3 despite their being plentiful both in Egypt and Nubia at this period, and found in pre-Aksumite graves at Yeha and at Sabaean sites across the Red Sea.4

In short, the innovations that characterise the ‘EthioSabaean’ period and the DΚMT kingdom, as well as certain materials and goods one would expect in a region exhibiting those characteristics, are virtually absent at a major and sophisticated habitation site not 50 km from its apparent epicentre at Yeha where they are found in large quantity. What else did, or did not, arrive at preAksumite ‘D site’? There is considerable evidence for the use of stone tools. The vast majority of chipped stone tools recovered are made of obsidian, a resource found nowhere near Aksum. It is, however, found in abundance at outcrops east of Mekelle, over 150 km south-east of Yeha where ‘EthioSabaean‘ influences are strongly evident. These outcrops are, in fact, the nearest source to Aksum (Zarins 1990). The few tools not made of obsidian at ‘D site’ are produced from locally abundant rock types, such as chalcedony, quartz and chert (Phillipson L.2000, i: 357, fig. 311). Yet its inhabitants preferred to fashion their tools from obsidian at this time, and deliberately obtained the stone from some distance away to do so. They were therefore not averse to using imported material when this was both obtainable and found to be more useful than alternatives locally available. And yet they did not obtain or use metal tools, which may have been a deliberate preference for stone tools on their part or, alternatively, availability of the metal might have been restricted only to the élite. Although not in the archaeological record, we also can assume that the salt fields of the Danakil depression near these Mekelle obsidian outcrops also would have been exploited at this time. Both obsidian and salt most probably had been exploited for millennia, and would have been in common use by pre-Aksumite times. Obsidian is not found at Late Stone Age sites in the immediate Aksum area but is found even further west in the Shire region, so must have come through Aksum from the Mekelle area.5

Figure 26: Sherd from the ‘D site’ at Aksum with a single ‘alpha’ sign Nothing is made of iron, although Yeha has quite a number and variety of objects in this material. Very few objects in copper or copper-alloy metals were recovered. Those few that were found are limited to small items of jewellery rather than tools (Phillipson D.W. 2000, II:

5

It is likely, then, that the Late Stone Age culture in the Aksum region would also have had obsidian, and its lack in excavation is simply due to chance. See Asamerew, Finneran & Phillips 2003 for prehistoric Shire.

3

The one ‘copper alloy’ sheet fragment recovered did show traces of gilding, see Phillipson 2000, ii: 343. 4 See Anfray 1963: passim; Simpson (ed.) 2002: 186-191 and passim.

81

TRADE AND TRAVEL IN THE RED SEA REGION further east at Matara (Fattovich 1980: pls XXXIII.5-6, XXXIV.2). Rocker decoration is a basic type, however, and its origin cannot be identified with any certainty. PreAksumite pottery, in any event, generally lies within another long-standing interaction with neighbouring cultures, as has long been understood. We also recovered a small number of very distinctive body sherds in a bright orange-coloured fabric having a ‘soapy’ feel and virtually no inclusions that can only have been imported. The author was unable to identify this fabric’s origin, but suspects that it came from somewhere further east, perhaps the Arabian peninsula. The sherds mostly represent thick-walled closed vessels having a narrow neck and elongated oval body, but one is an open bowl (Phillips 2000: fig. 270: f-h) (Fig. 28).

The evidence of food types grown and consumed by the ‘D site’ inhabitants point to a variety of crops. The crops identified by our archaeobotanist, Sheila Boardman in stratified pre-Aksumite levels at ‘D site’ include hulled two-row barley (Hordeum vulgare), emmer wheat (Triticum dicoccon), and an indeterminate wheat (T.sp.), possibly including bread wheat (T. aestivum), as well as linseed (Linium usitatissumum). Others are in more problematic contexts (Boardman 2000: 363-365, fig. 322). Barley and emmer wheat, as well as linseed, are part of the so-called ‘Near Eastern crop package’, and are Near Eastern imports to this part of Africa. The most clearly attested crops therefore are imported, not indigenous cereals, but some at least were cultivated in Ethiopia long before this period. Barley, for example, is attested in the aceramic Late Stone Age level at Baahti Nebait just outside Aksum, excavated by Niall Finneran in 1996 (Finneran 2000: 66-68). But whether all or any of these crops were recent or old arrivals by pre-Aksumite times is not yet known, therefore, they can have little to contribute to any discussion of pre-Aksumite foreign contacts and relations. It is important to emphasise that Ethiopian interaction with the Arabian peninsula and the Near East must be seen as already one of long standing by pre-Aksumite times. Direct Sabaean cultural influence, however, is a separate discussion.

The only other possible indication of any influence from across the Red Sea is the solitary lower body and base of a tall splay-footed closed vessel (Phillips 2000: fig. 267.d; Fig. 29) that may be a local version of the socalled ‘Yeha alabastron‘ well-known from its appearance at both Yeha and Matara (only) in Ethiopia/Eritrea, and also widespread at over 20 sites in the southern Arabian peninsula.6 Thin-section analysis has revealed that its composition is entirely distinct from examined vessels from Yeha, Matara and numerous pre-Islamic sites in Yemen (A. Porter, pers. comm. 28/08/03; Porter forthcoming), but it also stands out amongst the local Aksum fabrics as a whole. If indeed this vessel is an Aksum-made version of the form, it is only very loosely so, and would imply that the potter had no real idea what he was attempting to emulate.

Two open vessel (bowl?) sherds were recovered with ‘thumbnail’ rocker decoration (Fig. 27), which have not been attested at Aksum so far. Their fabric indicates that both are pre-Aksumite products; the fabric differs little if at all from others found at ‘D site’, but the unusual decoration suggests that the pottery is from a location other than Aksum. The design has not been reported at Yeha or Hawlti either, but similar decoration is found

But, what of the other directions? Briefly, Nubian imports also have been recovered in pre-Aksumite

Figure 27: Sherds with ‘thumbnail’ rocker decoration from the ‘D site’ at Aksum

6

The type is most common in the region of Yemen, where it is variously known as ‘amphora’, ‘type 4100’, and even ‘torpedo-shaped jar’. Those examined at Yeha and Matara were locally or regionally produced (Porter 2001; 2002:141, cat. no. 186; forthcoming).

82

JACKE PHILLIPS: PRE-AKSUMITE AKSUM AND ITS NEIGHBOURS

Figure 28: Potterywith orange fabric from the ‘D site’ at Aksum

Figure 29: Lower body and base of a splay-footed vessel (Figs. 25-29 are courtesy of Prof. D. W. Phillipson and the Aksum Archaeological Research Project). been identified at Mai Adrasha and other sites in some quantity (contra Phillips 2003).

contexts at Yeha, Hawlti and elsewhere in DΚMT, and they almost certainly must have passed through Aksum to their ultimate destination further east (Phillips 1995: 5-6; 1997: 440-445). The shortest and most logical route from the Nile Valley to the Red Sea for some distance both north and south emerges directly west of Aksum through the modern Ethiopian province of Shire, as the author has argued elsewhere (Phillips 1997: 440-445; 2002). This area west of Aksum, however, is only just beginning to be investigated, and in a recent survey further into Shire, east of Endasellassie, pre-Aksumite material has now

One import in particular brought from this western direction must be mentioned here. Little evidence is found in the pre-Aksumite levels for the vast quantities of ivory that the Periplus some five to eight centuries later tells us are brought through Aksum to the Red Sea coast at that time. Two pieces of ivory were recovered in the ‘D site’ excavations; both pieces are hippopotamus canines (Phillipson D.W. 2000, ii: 345; Cain 2000: 369, 83

TRADE AND TRAVEL IN THE RED SEA REGION fig. 324) rather than elephant tusk that later would be the majority of the ivory brought to the coast. Both fragments are unworked, but they at least tell us that ivory did find its way to Aksum (and, presumably, beyond it to the coast) long before the Periplus was written. Hippopotami are (and were) found no closer to Aksum than Lake Tana more than 250 km to the south, and in the Nile that is more than twice that distance to the west, and from beyond which the Periplus tells us ivory came through Aksum.

Eastern Africa Memoir 17) (Reports of the Research Committee of the Society of Antiquaries of London 65). London: British Institute in Eastern Africa; The Society of Antiquaries. Cain C.R. 2000. Archaeozoology. Pages 369 - 379 in D.W. Phillipson, Archaeology at Aksum, Ethiopia, 1993-97. (2 volumes)/vol.2. (British Institute in Eastern Africa Memoir 17) (Reports of the Research Committee of the Society of Antiquaries of London 65). London: British Institute in Eastern Africa; The Society of Antiquaries. Fattovich R. 1980. Materiali per lo studio della ceramica pre-aksumita etiopica. Annali dell’ Istituto Universitario Orientale di Napoli, Supplemento 25. Naples: Istituto Universitario Orientale di Napoli. Finneran N. 2000. Excavations at the LSA site of Baahti Nebait, Aksum, Northern Ethiopia, 1997. Azania 35: 53-73. Manzo A. 1999. Échanges et contacts le long du Nil et de la Mer Rouge dans l’époque protohistorique (IIIe et IIe millénaires avant J.-C.). Une synthèse préliminaire (Cambridge Monographs in African Archaeology 48) (British Archaeological Reports, International Series 782). Oxford: BAR Publishing. Munro-Hay S.C. 1989. Excavations at Aksum. An account of research at the ancient Ethiopian capital directed in 1972-74 by the late Dr Neville Chittick. (Memoirs of the British Institute in Eastern Africa 10). London: British Institute in Eastern Africa. ----------1991. Aksum: An African Civilisation of Late Antiquity. Edinburgh: Edinburgh University Press. Periplus/Casson, L. (ed. and transl.). 1989. The Periplus Maris Erythraei. Text with Introduction, Translation, and Commentary. Princeton: Princeton University Press Phillips J.S. 1995. Egyptian and Nubian material from Ethiopia and Eritrea. Sudan Archaeological Research Society Newsletter 9: 2-10. ----------1997. Punt and Aksum: Egypt and the Horn of Africa. Journal of African History 38: 423-457. ----------2000. Pottery and clay objects. Pages 303-337 in D.W. Phillipson, Archaeology at Aksum, Ethiopia, 1993-97. (2 volumes)/vol.2. (British Institute in Eastern Africa Memoir 17) (Reports of the Research Committee of the Society of Antiquaries of London 65). London: British Institute in Eastern Africa; The Society of Antiquaries. ----------2002. Egypt, Nubia and Ethiopia. Pages 434-442 in Z. Hawass (ed.), Egyptology at the Dawn of the Twenty-First Century: Proceedings of the Eighth International Congress of Egyptology, Cairo, 2000. (3 volumes)/vol.2: History and Religion. Cairo: American University in Cairo Press. ----------2003. Looking Forwards by Looking Backwards: West of Aksum. Pages 69-72 in P.J. Mitchell, A. Haour & J.H. Hobart (eds), Researching Africa’s Past: New Contributions from British Archaeologists (Oxford: Oxford University School of Archaeology Monograph 57). Phillips J., Reynolds A. & Spandl K. 2000. Structures

To sum up then, pre-Aksumite ‘D site’ at Aksum provides generally negative evidence for the ‘strong South Arabian (mainly Sabaean) influence’ that appeared in the kingdom of DΚMT in the first half of the first millennium BC. These are the pseudo-‘Yeha alabastron’ vessel base, the single applied letter on a storage pot that may or may not have been produced at Aksum, the sherds of foreign orange fabric and perhaps the introduction of copper alloy metal (but only as finished jewellery). At least three of these four pieces of evidence would represent only secondary influences, likely from Yeha or Hawlti which possessed the alabastra, writing and metallurgy, rather than directly from across the Red Sea. The only other possible contemporary evidence from further east, yet not from beyond the Red Sea, is the pair of pottery sherds with ‘thumbnail’ decoration and comparanda at Matara. We might also add the use of imported obsidian for tools and the ‘Near Eastern package’ crops, both known in the Aksum area long before the pre-Aksumite period. By the very nature of the evidence available to us we have little choice but to generalise, but the percentage of any given population in any given ancient culture exposed to any foreign influence at all was extremely limited. The South Arabian influence evident at élite DΚMT sites and contexts does not penetrate down to the lifestyle of the ordinary pre-Aksumite population elsewhere, even in their immediate vicinity. This ‘influence’ over the kingdom of DΚMT is a phenomenon of limited overlying ‘veneer’, apparently ‘political’ rather than ‘cultural’ in force and nature, recognisable and distinguishable only amongst certain elements within an otherwise wholly indigenous culture. References Anfray F. 1963. Une campagne de fouilles à Yĕhā (Février-Mars 1960). Annales d’Éthiopie 5: 171-233. ----------1990. Les anciens Éthiopiens: siècles d’histoire. Paris: Armand Colin. Asamerew D., Finneran F. & Phillips J.S. 2003. The prehistoric settlement of the Shire Region, Western Tigray, Ethiopia. Some preliminary observations. Nyame Akuma 59: 26-33. Boardman S. 2000. Archaeobotany. Pages 363 - 365 in D.W. Phillipson, Archaeology at Aksum, Ethiopia, 1993-97. (2 volumes)/vol.2. (British Institute in 84

JACKE PHILLIPS: PRE-AKSUMITE AKSUM AND ITS NEIGHBOURS and stratigraphy. Pages 280 - 301 in D.W. Phillipson, Archaeology at Aksum, Ethiopia, 1993-97. (2 volumes)/vol.2. (British Institute in Eastern Africa Memoir 17) (Reports of the Research Committee of the Society of Antiquaries of London 65). London: British Institute in Eastern Africa; The Society of Antiquaries. Phillipson D.W. 1998. Ancient Ethiopia. Aksum: Its Antecedents and Successors. London: British Museum Press. ----------2000. Archaeology at Aksum, Ethiopia, 1993-97 (2 volumes) (British Institute in Eastern Africa Memoir 17) (Reports of the Research Committee of the Society of Antiquaries of London 65). London: British Institute in Eastern Africa; The Society of Antiquaries. Phillipson D.W. & Phillips J. 2000. The site and its sequence. Pages 267 – 280 in D.W. Phillipson, Archaeology at Aksum, Ethiopia, 1993-97 (2 volumes)/vol.2. (British Institute in Eastern Africa Memoir 17) (Reports of the Research Committee of the Society of Antiquaries of London 65). London: British Institute in Eastern Africa; The Society of Antiquaries. Phillipson L. 2000. Lithics. Pages 352-363 in D.W. Phillipson, Archaeology at Aksum, Ethiopia, 1993-97 (2 volumes)/vol.2. (British Institute in Eastern Africa Memoir 17) (Reports of the Research Committee of the Society of Antiquaries of London 65). London: British Institute in Eastern Africa; The Society of Antiquaries. Porter A. 2001. Society Grant Report, Bulletin of the Society for Arabian Studies 6: 7-10. ----------2002. Ceramic jar with everted rim. Page 141 in St. J. Simpson (ed.) Queen of Sheba: Treasures from Ancient Yemen. London: The British Museum Press. ----------(forthcoming) Amphora trade between South Arabia and East Africa in the first millennium BC: a re-examination of the evidence. Proceedings for the Seminar for Arabian Studies 33. Ricci L. & Fattovich R. 1984-86. Scavi archeologici nella zona di Aksùm a Seglamièn. Rassegna di Studi Etiopici 30: 117-169. Simpson St.J. (ed.) 2002. The Queen of Sheba: Treasures from Ancient Yemen. London: British Museum Press. Zarins J. 1990. Obsidian and the Red Sea Trade Prehistoric Aspects. Pages 507-541 in M. Taddei & P. Callieri (eds), South Asian Archaeology 1987. Rome: Istituto Italiano per il Medio ed Estremo Oriente.

85

86

Adulis to Aksum: charting the course of antiquity’s most important trade route in East Africa Walter Raunig Pliny the Elder informs us (Natural History 12.41.83) that Rome paid one hundred million sesterces annually for luxury goods from Arabia, India and other parts of the east. It is difficult to convert this enormous sum into pounds, dollars or euros, but we know that in Pliny’s time a quarter or half litre cup of best-quality wine cost one sesterce (Overbeck and Klose 1986).

that is, towards the open sea, on the right are a number of other islands, small and sandy, called Alalaiu; these furnish the tortoise shell that is brought to the port of trade by the Ichthyophagoi. (Periplus/Casson 1989: 5152)

The route from Adulis to Aksum must have been the most important in East Africa in ancient times. Adulis was the most important town on the coast and Aksum was the centre of a rather important empire, starting about the time of Christ and lasting until the eighth or ninth century (Fig. 30).

What sorts of exotic goods were brought from the east to Rome and other centres of the empire? And where did these articles come from? Wealthy Roman citizens bought a wide variety of precious stones, coral, pearls, ivory, skins, tortoise shell, silk, spices like cinnamon, pepper, frankincense, myrrh, ebony and exotic animals for the games. These and other luxury goods were brought from the Levant, the Gulf, India, China and North and East Africa, to say nothing of the amber, furs, skins etc. from northern Europe and Siberia.

What was the course of the route from Adulis up into the mountainous country of what is now Ethiopia and Eritrea? The Periplus refers to a location named Koloe, a city ‘that is the first trading post for ivory’ and says ‘from Adulis it is a journey of three days to Koloe’. On their marches through the highlands of northern Ethiopia and Eritrea, Russel (1860), Bent (1893), Schoeller (1894), Littmann (1906) and others visited the plateau of Qohaito and discovered many ancient ruins (Fig. 31). They had no doubt that modern Qohaito, today consisting of several villages of the Saho people, was ancient Koloe. The distance between the coast and the plateau is 32 km, and this fitted with the text of the Periplus. Nevertheless, some specialists, including Kobishchanov and Scholz, attempted to identify Koloe with the more southerly sites of Toconda or Matara.

Everyone interested in the ancient history of the Near East, particularly in the ancient history of the Mare Erythraeum, a term that in classical times included the Red Sea, the Arabian Gulf and the Indian Ocean, knows the Periplus of the Erythraean Sea. Written by a Greek merchant from Egypt in the second half of the first century AD, this guidebook was written for sailors and merchants travelling in the Red Sea and Indian Ocean. The Periplus is the fundamental classical account of this region. In § 4, the anonymous author informs us: About 3000 stades beyond Ptolemais Theron is a legally limited port of trade, Adulis. It is on a deep bay extending due south, in front of which lies an island called Oreine [‘hilly’] that is situated about 200 stades from the innermost part of the bay towards the open sea and, on both sides, lies parallel to the coast; here at the present time arriving vessels moor because of raids from the mainland. Formerly they used to moor at the very outermost part of the bay at the island, called Diodoros Island, right by this part of the coast; there is a ford crossing to it by which the Barbaroi dwelling roundabout used to overrun the island. On this part of the coast, opposite Oreine, 20 stades in from the sea is Adulis, a fair-sized village. From Adulis it is a journey of three days to Koloe, an inland city that is the first trading post for ivory, and from there another five days to the metropolis itself, which is called Axomites [Axum]: into it is brought all the ivory from beyond the Nile through what is called Kyeneion, and from there down to Adulis. The mass of elephants and rhinoceroses that are slaughtered all inhabit the upland regions, although on rate occasions they are also seen along the shore around Adulis itself. In front of the port of trade,

Beginning in 1995, Professor Dieter Wenig from Humboldt University in Berlin began his campaign on the Qohaito plateau, in cooperation with the Kunsthistorisches Museum in Vienna and the National Museum of Eritrea. His team found some 900 sites in Qohaito (Fig. 32), and these ruins: ...cover the plateau, making it one of the most extensive archaeological areas in sub-Saharan Africa. Most of the ruins originate from the Axumite period, but potsherds from the pre-Axumite period were also found. Stone tools from the middle Stone Age and from the Neolithic period were found in abundant quantities, as well as Early Islamic open prayer places. This place on the plateau may have been a kind of garden city... for the upper classes of Adulis, which contained hundreds of houses, estates and religious buildings. Between them were, as today, extended areas for cultivation. There can be no doubt that the site must have been one of the most important ones of the Axumite empire. But the supposition that the ancient town of Koloe is to be located here.... has to be considered very cautiously,

87

TRADE AND TRAVEL IN THE RED SEA REGION

Figure 30: Ruins of ancient Adulis; Gulf of Zula in Eritrea.

Figure 31: Temple-ruin on the Qohaito plateau.

88

WALTER RAUNIG: ADULIS TO AKSUM

Figure 32: Ancient water reservoir on the Qohaito plateau.

report that they reach Zula/Foro within three days, but fighters from the Eritrean Peoples’ Liberation Front (EPLA) made it within two days’ (Wenig 1997: 20).

because Qohaito was apparently not an urban centre like Toconda or Matara (Wenig 1997: 20-21).

There are two possible routes from Adulis on the coast to Koloe/Qohaito, either through Wādī Haddas, or through Wādī Komalie. This latter route has been well known to Europeans since the days of General Napier. His military expedition against the Emperor of Ethiopia, Theodoros [Tewodros] II (1855-1868) began at Adulis in 1868 and followed Wādī Komalie up into the mountains. Gerhard Rohlfs, also in Eritrea at the time, was forced by the military to take the harder route, through Wādī Haddas. About eighty years later it was recommended better to take the route up into the mountains through the Wādī Haddas (see British military map of 1947)) (Fig. 33).

Thus, we are able to confirm that the important ancient route from the Red Sea coast to the first Aksumite inland emporium of Koloe passed through Wādī Haddas or through Wādī Komaile, either to modern Qohaito or to another place nearby, such as the ancient site of Hishmale, which like Qohaito is near the modern town of Adi Keyh. Hishmale is also a three days’ journey from the sea, and in it are to be found many ruins of large buildings, perhaps store-houses, but as yet nothing has been excavated. Beginning from Qohaito or from Hishmale a caravan thus could reach Aksum via Taconda-Kaskasè-Matara and Yeha within five days, just as the Periplus states.

In 1997, climbing up to the Qohaito plateau coming from Adi Keyh on the Haddas valley side, we met two ‘honey men’ from the coast, who were bringing their honey to the modern market place of Adi Keyh near Qohaito (Fig. 34). They told us they had been travelling for three days. We were able to confirm that this journey took three days both in Foro, near Adulis and from Professor Wenig: ‘The plateau of Qohaito could be reached from Adulis plain either via Haddas Valley west of the plateau or via Komaile Valley east of the plateau’ (cf. Duncanson 1947: 161, map). At least three very steep descents lead from the plateau to the valleys. People using the eastern-most

Koloe, referred to in the Periplus as an important transfer or reloading point, can be identified with one of the two localities mentioned above, either Hishmale or Qohaito. The importance of this area in East Africa to the history of the east-west trade in the ancient world is still underestimated. Future research is bound to bring considerable changes to our present state of knowledge.

89

TRADE AND TRAVEL IN THE RED SEA REGION

Figure 33: Map of routes to Qohaito plateau.

Figure 34: Ascending the plateau of Qohaito

90

WALTER RAUNIG: ADULIS TO AKSUM

Conference of Ethiopian Studies, 20-25 July 2003. (Hamburg University, Asia-Africa Institute). Hamburg: Asia Africa Institute. Rohlfs, G. 1869. Im Auftrage Sr. Majestät des Königs von Preussen mit dem Englischen Expeditionscorps in Abessinien. Bremen: J. Kühtmann. Rüppell, E. 1838-40. Reisen in Abessinien (2 volumes). Frankfurt/Main: Gedruckt auf Kosten des Verfassers und in Commission bei Siegmund Schmerber. Russel, S.J.M. 1884. Une mission en Abyssinie et dans le Mer Rouge, 23 octobre 1859 – 7 mai, 1860. Paris: E. Plon, Nourrit et Cie. Schoeller, M. 1895. Mitteilungen über meine Reise in der Colonia Eritrea (Nord Abyssinien). Berlin: Norddeutsche Buchdruckerei & Verlagsanstalt. Scholz, P. 1984. Auf den Sputen der äthiopischen vergangenheit zwischen dem Niltal und Arabia Felix. Antike Welt 3: 2-34. Sidebotham, S.E. 1986. Roman Economic Policy in the Erythra Thalassa 30 B.C - A.D. 217. Leiden: E.J. Brill. War Office (ed.). 1946. Asmara. Map 1:500.000. Wenig, S. 1997. German Fieldwork in Eritrea. Nyame Akuma. 48: 20-21.

References Bent, J.T. 1893. The Sacred City of the Ethiopians. London: Longmans, Green and Co. Duncanson, D.J. 1947. Girmaten: a new archaeological site in Eritrea. Antiquity 21:158-161. Eigner, D. 1999. German Archaeological Mission to Eritrea, Season 1996, The Ruin "Littmann no 8" at Qohaito. Mare Erythraeum 3: 41-56. Guida dell’Africa Orientale Italiana, 1938. Milano: Consociazione turistica italiana. Hahn, W.1994. Münz- und geldgeschichtliche Objekte Spätantike Zeitzeugen: Die axumitischen Münzen als Demonstration einer Handelsmacht. Pages 21-53 in W. Hahn (ed.) Äthiopien, Kunsthandwerk und Münzen aus österreichischen Sammlungen. Linz: Oberösterreichische Landesmuseum. Holland, T.J. & Hozier, H. 1870. Record of the Expedition to Abyssinia (2 volumes). London: H.M.S.O. Kobishchanov, Y.M. (ed. J.W. Michels, transl. T. Kapitanoff) 1979. Axum. University Park/London: The Pennsylvania State University Press. Littmann, E. 1913. Reisebericht der Expedition, Topographie und Geschichte Aksums. Volume 1 of E. Littmann, D. Krencker, Th. von Lüpke, & R. Zahn (eds). Deutsche Aksum Expedition. (3 volumes in 4). Berlin: G. Reimer Munro-Hay, S. 1989. The British Museum excavations at Adulis, 1868. The Antiquaries Journal 69 (1): 43-52. Overbeck, B. & D.A.O. Klose 1986. Antike im Münzbild. Munich. Paribeni, R. 1908. Ricerche nel luogo dell’antica Adulis. Monumenti Antichi, Reale Accademia dei Lincei 18/3: 437-572. Periplusi/Casson, L. (ed. and transl.) 1989. The Periplus Maris Erythraei. Text with Introduction, Translation, and Commentary. Princeton: Princeton University Press. Petermann, A. 1861. Der nördlichste Theil von Abessinien und die Land-schaften BOCOS und HABAB, 1:900.000. Page 298 in A. Petermanns Mitteilungen. Gotha: Justus Perthes. Phillipson, D.W. 1998. Ancient Ethiopia. Aksum: Its Antecedents and Successors. London: British Museum Press. Pliny/H. Rackham (ed. and transl.) 1969. Natural History. (10 volumes). (Loeb series). London: Heinemann/Cambridge, Massachusetts: Harvard University Press. Raunig, W. 1967. Orienthandel im Altertum. Führer durch das Museum für Völkerkunde und Schweizerische Museum für Volkskunde Basel. Basel: Museum für Völkerkünde und Schwizerische Museum für Volkskunde. ---------1971. Bernstein, Weihrauch, Seide - Waren und Wege der antiken Welt. Wien/München: Schroll. ---------2004. Wo verlief Nordostafrikas wichtigster antiker Handelsweg? Papers of the 15th International 91

92

The Egypto-Graeco-Romans and Panchaea/Azania: sailing in the Erythraean Sea Felix Chami during the Hellenistic period goes back far into the first half of the first millennium BC.’

Introduction This paper is another attempt to elucidate how the Egypto-Graeco-Romans had extended trade links to the coast of East Africa and probably to the interior of East Africa. This is probably the first time that a variety of archaeological evidence recently recovered from the coast and islands of Tanzania has been collated with existing Graeco-Roman documents. Previous efforts have relied mostly on historical data and a few unprovenanced objects such as coins1 and a few excavated finds.2

With this definition the Graeco-Roman period thus can be divided into two phases. The first phase is the one in which East Africa is linked to the economy of the Mediterranean world via the Red Sea. The emerging historical and archaeological data reveal the existence of contact between the regions of the Mediterranean/Red Sea and East Africa beginning from about the early third century BC. This is the period when the Greeks had taken over the control of the eastern Mediterranean and Ptolemaic rule had been installed in Egypt. Greek sailors voyaged in the Red Sea and had probably reached East Africa. The second phase of Graeco-Roman economic links with East Africa was during that of Roman dominance of the Mediterranean beginning with the reign of Caesar Augustus in 27 BC. Due to the pre-eminent power of the Arabs in the Red Sea and in the northern Indian Ocean in the last century BC, late Ptolemaic and early Roman Egypt may have linked with eastern Africa along the Nile via Meroe (for conspectus, cf. Chami 1999a). It was in the time of Caesar Augustus, more so after the BC/AD changeover, that Graeco-Roman sailors and traders gained access to the Indian Ocean through the Red Sea. Arab power was crushed and Aden, the town controlling the trade from the Indian Ocean to the Red Sea, was destroyed.3

The focus of this paper is on how the Red Sea was used to facilitate contact between communities of the Mediterranean Sea and those of the Red Sea and East Africa, or Panchaea/Azania. However, other evidence may also be used to show that similar contacts existed between East Africa, the Middle East and south-east Asia (Chami 2002a, 2002b). The present work describes new archaeological finds of 2000-2002 from the Tanzanian coast. The work also utilises a more confident interpretation of early Graeco-Roman documents hitherto regarded as the equivalent of fairy stories. The Egypto-Roman period: nomenclature It is important to define what is termed in this paper the Egypto-Graeco-Roman period. The Egyptian period of this paper is the Third Intermediate Period dating roughly between 1000 and 500 BC (Spencer 1997). During this time the Kushites ruled Egypt and Pharaoh Necho sent an expedition to circumnavigate Africa via East Africa. In this period the Egyptian traditional trading partner, Punt, is mentioned only once, in about 600 BC (Kitchen 1993). It will be shown later that Egyptian trade goods of the Third Intermediate Period were reaching East Africa.

The extension of Roman economic influence to the Indian Ocean and beyond, to East Africa and the Far East, created an unprecedented world economic order. The new era is known to have lasted three centuries; Roman power over the Red Sea and the Indian Ocean then passed to the Aksumites and Sasanians respectively.4

The Random House College Dictionary defines the concept of Graeco-Roman as that ‘of or having both Greek and Roman characteristics’. Such characteristics occurred in the period when the Romans took over the empire of the Greeks in the second part of the last century BC, fusing together Greek and Roman influences: this process continued until about the end of the third century AD. The use of the concept ‘GraecoRoman’ in this work extends further back to cover the period following the death of Alexander the Great in 323 BC until the end of the third century AD. However, Sales (1996: 253) has pointed out that ‘the history of trade

Egypto-Graeco-Roman documents and East Africa

1

3

2

The earliest Graeco-Roman accounts of eastern Africa are those of Euhemerus (c. 300 BC) as narrated by Diodorus Siculus (History 5.41-46) and Iambulus (c. 300-200 BC) also preserved in Diodorus Siculus (History 2.55-60). Recent scholars, like their ancient counterparts, have doubted the authenticity of Euhemerus and Iambulus’ report, regarding both as fantasy (Cary & Warmington 1963: 241).

Chittick 1980; Sheriff 1981; Periplus/Casson 1989; Horton 1996 Chami & Msemwa 1997; Chami 1999a, 1999b.

4

93

Begley & de Puma (eds) 1991; Cinimo (ed.) 1994. Whitehouse & Williamson 1973; Compareti 2002.

TRADE AND TRAVEL IN THE RED SEA REGION

Iambulus’ voyage and visit to Panchaea, c. 300 BC Eudoxus’ accidental visit to Panchaea, c. 115 BC Eudoxus’ first attempt to circumnavigate Africa from west-south-east, c. 110 BC Necho’s successful expedition to circumnavigate Africa from eastsouth-west 610-594 BC Figure 35: Pre-Roman expeditions to Africa south of the Sahara. the capital. Panara is located on one of the islands. Farming was the principal occupation of the people. The land is reported to have possessed rich mines of gold, silver, copper, tin, and iron.

Euhemerus referred to islands off the East African coast as Panchaea and he noted that the way to reach them was to sail from Arabia southward for a number of days (Diodorus Siculus History 6.1.4). That Panchaea also referred to the general area of eastern Africa is corroborated by the description of what is found in that land, including animals such as elephants, leopards, and gazelles. Euhemerus also describes a mountain and several towns of Panchaea, including Panara, probably

Diodorus recounts the voyage of Iambulus from somewhere on the Horn of Africa/Red Sea to islands off the coast of East Africa, at the equator, sometime in the third century BC (Fig. 35). Iambulus describes an 94

FELIX CHAMI: THE EGYPTO-GRAECO-ROMANS AND PANCHAEA/AZANIA archipelago of seven islands whose inhabitants shared a similar culture. They are literate and practice a communal, but quasi-Neolithic mode of production. The people of the islands venerated the sun and stars and hence the ‘the islands of the Sun.’ Another important aspect of Iambulus’ report is the evidence he provides for sailing times. His voyage took four months from the Horn of Africa and when he left the islands, after a stay of seven years, it took him another four months to reach the north-west coast of India. The report suggests that the three regions of the Red Sea, East Africa and India were in contact.

Zangion/Zanchion and in the Persian/Arabic term Zanj/Zanji used to denote the East African coast and its inhabitants and still preserved as the first element in the word Zanzibar. This phonetic correlation cannot be fortuitous. The identification of Panchaea with East Africa is supported by the fact that Euhemerus describes sailing south from Arabia which would definitely land one in East African islands; Iambulus sailed from the Horn of Africa south to the East African islands at the equator and Eudoxus was driven by monsoons south from the northern Indian Ocean to land at a locality in East Africa.

Another important aspect of Iambulus report, also mentioned by Euhemerus, is that of farming. The reed described by Iambulus as similar to ‘white vetch’ its seeds steeped in warm water until they have become about the size of a ‘pigeon’s egg’ and then crushed and rubbed skilfully and moulded into loaves which are baked and eaten (Diodorus Siculus History 2.57.2) is clearly a reference to rice. We also have evidence of cotton and coconut cultivation.

The description of Panchaea by those who reached it is anecdotal. Apart from Iambulus’ seven islands and the description of the inhabitants and their belief system, no geographical detail of East Africa is provided. Probably some additional information could be derived from Strabo, who also compiled data from his predecessors (Fig. 36). Hipparchus (second century BC) and Eratosthenes (third century BC), described the topography and climate and some cultures of the deep interior of East Africa (Strabo Geography 17 passim). These included well-populated highlands and people in some parts of the interior who ate locusts (Strabo Geography 16.4.12); the people of the Great Lakes region still do. According to Pliny some inhabitants smeared themselves with ochre, a practice well attested today ethnographically and archaeologically for the Rift Valley people (Natural History 6.35.190).

In about 115 BC another Greek voyager, Eudoxus of Cyzicus, is said to have reached the coast of East Africa after being driven there by monsoon winds when returning to the Red Sea from a mission in India (Strabo Geography 2. 4-5). It was this accidental voyage that led him to believe that Africa could be circumnavigated. His discovery there of a fragment from the wreckage of a typical Mediterranean vessel convinced him that sailors based on the Atlantic coast of Spain at Gades (today’s Cadiz), had reached East Africa by sea from the south (see Fig. 35).

Another important aspect of the Red Sea-East Africa relationship is the trade in spices, particularly cinnamon and cassia, reported at this time to have been reaching the Mediterranean and the Red Sea worlds from eastern Africa. Indeed Iambulus’ mission was in search of spices. Panchaea was reported to have traded spices with Arabia, another reason for suggesting that the country was East Africa rather than the Red Sea coast of Africa. Strabo speculated that the origin of the cinnamon and cassia was near the Nile sudd (Strabo Geography 16.4.14). However, Pliny corrected him, suggesting that it was the East African ‘cave dwellers’, related to the people of the Red Sea, who brought the spices from a place far off in the Ocean, which would point to south-east Asia (Natural History 12.42.85) (Fig. 36). As will be shown later, the Late Stone Age/Neolithic people on the islands of Tanzania lived in caves.

It has been mentioned that classical as well as modern scholars were/are sceptical of Greek knowledge of East Africa/Panchaea, referring to it as fantasy. It should be noted here, however, that Greeks may well have learned in Egypt of the existence of an island in the Indian Ocean called by the Egyptians Pa-anch, and this may well be the origin of the Greek Panchaia. Both Pa-anch and Panchaea are phonetically quite close to the Egyptian Pwnt, which is conventionally read as ‘Punt’ Kitchen 1993: 587). The Egyptian pronunciation of Pwnt was probably something like Pwanet, which is reminiscent of the modern East African word pwani literally meaning ‘coast’.

The second phase in the Graeco-Roman sources is that of the early centuries AD, comprising the celebrated Periplus Maris Erythraei and Ptolemy’s Geographia. These texts have been thoroughly discussed by other scholars.6 Only a few points need to be mentioned here. The Periplus, written in about AD 40-70, describes Azania as extending from a place south of what is now

Although scholars have always placed Punt and Panchaea in the Red Sea regions of Ethiopia and the Sudan,5 a new approach suggests that Punt was a large region extending south probably as far as East Africa. Furthermore, the Egyptian Punt is probably the same land known as Paanch/Panchaea. It has been shown elsewhere (Chami 2002b) that nchi which means in Kiswahili ‘country/territory/land’, and appears as an element in both words, is also preserved in Pliny‘s 5

6

See translations of the Periplus by Schoff (1912), Huntingford (1980), and Casson (1989) and discussion by Datoo (1970), Kirwan (1986), Horton (1990), Chami (1999a, 1999b).

Kitchen 1993; Huntingford 1980; Cary & Warmington 1963.

95

TRADE AND TRAVEL IN THE RED SEA REGION

Figure 36: Roman expeditions and knowledge of eastern Africa. Ptolemy’s Geographia, written in the second century AD, adds very little to the Periplus, apart from offering latitudes and longitudes to the settlements mentioned. The Geographia suggests the growth of more settlements and the development of Rhapta into a metropolis (Ptolemy Geographia 4.7). The coast of what is now Mozambique and the offshore islands were then known, because it is here where Ptolemy places his Menuthias.

Somalia to the general area of the Tanzanian coast (Fig. 36). An island called Menuthias, probably to be identified as Zanzibar, is mentioned and an emporium called Rhapta, probably located near the Rufiji River. The trade route connecting parts of the Red Sea to Azania and India is described and it is also reported that the Arabs of the Red Sea coast controlled trade at Rhapta, arguing that Azania was under their suzerainty.

96

FELIX CHAMI: THE EGYPTO-GRAECO-ROMANS AND PANCHAEA/AZANIA Africa and the site could have been serving the trade route to India.

Probably one of the most significant contributions made by Ptolemy’s Geographia is the inclusion of data about the deep interior. Much of this data concerning Azania derived from a certain Diogenes, who was driven to the East African coast returning from a voyage to India and visited Lake Nyanza (Victoria), and knew about the Ruwenzori range of mountains, ‘The Mountains of the Moon’, the Uganda Nile and a mountain with three peaks, probably Kilimanjaro (Fig. 36).7 Imports to East Africa included iron objects, wine and grain whereas exports included ivory, rhinoceros horn, tortoise shells and what was either nautilus shells or coconut oil, depending on whether one follows the maunuscript reading nauplios, or the restored reading nargilios (see Periplus/Schoff ed. 1912: 99).

The third discovery of material evidence linking the Mediterranean world with Azania, for the first time incontrovertible, was the report by this author of the Roman beads from the Rufiji delta region dating back to the early centuries AD (see Chami 1999b). The most important of the beads collected is a segmented gold/silver-in-glass bead made on Rhodes between 100 BC and 300 AD. If the beads reached the coast of East Africa before the first century AD, it is probable that they would have come through the Nile Valley route via Meroe, where similar beads are found (Welsby 1996). If they reached Azania after the BC/AD changeover, it is likely that they followed the Red Sea route, as the Indian Ocean was not yet opened to Roman traders earlier, and such beads have now been recovered from Berenike on the Red Sea coast of Egypt in a context dating back to the early centuries AD (Sidebotham’s illustrations in the Red Sea Workshop).

Archaeological evidence It was noted earlier that Graeco-Roman documents of any period were of little use up to 1995 because archaeological evidence to corroborate them had not been recovered in East Africa. According to Sales (1996: 253) East Africa, Madagascar and South Africa lay outside the scope of classical studies because ‘these areas are completely absent from written sources and have not provided any archaeological information for these centuries.’

The fourth item of evidence, also incontrovertible, consists of archaeological finds of both Roman and Indian material deposited in a limestone cave in Zanzibar, and also reported by this author.10 The artefacts include Roman red ware pottery similar to that reported by Smith and Wright (1988) from Ras Hafun (this and other finds have been illustrated in Chami 2002a). In association were green and blue glazed and non-glazed Persian ware and Indian beads, including one of carnelian.

It has already been pointed out that while the earlier Greek reports were not believed, the Periplus and Ptolemy’s reports were approved because they were strongly supported by archaeological evidence, particularly from India and Sri Lanka. Moreover, Roman historians corroborated the presence of Roman traders in the Indian Ocean, reaching as far as south-east Asia.8 The main thrust of this paper is to provide archaeological evidence spanning the period of the Egypto-GraecoRoman era, showing that there were settlements on the coast of East Africa linked by trade to the Red Sea and beyond. The earliest material evidence reported which suggests the existence of such settlements and trade links is that of Chittick (1966) who summarises reports of finds of various coins of rulers of Ptolemaic Egypt, the Roman Empire and other parts of the Near East from the last centuries BC. Horton (1996) provides a more exhaustive discussion of the coins of the classical period collected from eastern and southern Africa. As none of these coins were found in archaeological contexts the data are unreliable. Chittick has provided further material evidence for trade links between the Red Sea and the Indian Ocean from Ras Hafun.9 Here, finds of Roman and Partho-Sasanian pottery in an archaeological context dating to between the first century BC and third century AD were recovered. The problem with these finds, however, was that they were not directly linked to East

TRADITION

SITE

Neolithic Neolithic

Machaga/Zanzibar Machaga/Zanzibar

LAB. NUMBER Ua-14648 GrA-14155

Neolithic

Machaga/Zanzibar

GrA-14158

Neolithic Neolithic

Ukunju/Mafia Ukunju/Mafia

Pta-8522 Pta-8527

14

C DATE

135-45 BC 813-791BC 2847-2568 BC 800-400BC 800-400BC

Table 1: Recent 14C dates from Neolithic Zanzibar and Mafia caves, off the coast of Tanzania. A 14C sample from associated charcoal gives a date in the last century BC (see Table 1, row 1 of dates). Since then, evidence of many trade goods has been recovered from the island of Mafia (see Table 2). The preliminary report of the archaeological work yielding these finds is found elsewhere (Chami 2000) and a report of the early Graeco-Roman beads has been published elsewhere (Chami 2002a). A more detailed report covering the excavations on the Mafia Island is forthcoming.11

7

For conspectus, cf. Huntingford 1980; Chami 1999a. Begley & de Puma (eds) 1991; Cinimo (ed.) 1994. 9 Chittick 1980; Smith & Wright 1988. 8

10 11

97

Chami & Wafula 1999; Chami 2001. Chami 2004.

TRADE AND TRAVEL IN THE RED SEA REGION

POTTERY TYPE

CULTURAL PERIOD

RANGE OF DATES

Harrapa

10th –8th century BC

South Indian Megalithic

8th –4th century BC,

Chocolate ware

Early History

4th century-BC – 4th century AD

Coarse black ware

Early Historic

4th century BC – 4th century AD

Early Historic

4th century BC – 4th century AD

Early Historic

4th century BC- 4th century AD

Coarse Black Ware beaded rim

Early Historic

4th century BC – 4th century AD

Incised decoration

Early Historic

4th century BC– 4th century AD

Dull red ware/small ritual vessel

Early Historic

4th century BC – 4th century AD

Perforated lamp cover

Early Historic

4th century BC–4th century AD

Basket impressed design

Arikamedu

1st century BC–1st century AD

Black burnished ware & coarse black ware

Chalcolithic

Second millennium BC

Red painted

Gupta period

4th century – 6th century AD

Red ware/slightly polished

Kushan-Gupta transition

3rd –5th century AD

Buddhist monks’ vessel Dull Red Ware

small

ritual

Table 2: Imported ceramics from India and Egypt excavated from Mafia cave, off the coast of Tanzania.

Figure 37: Harappa and South Indian Megalithic 10th 4th century BC found in Mafia cave excavations.

Figure 38: Marl clay pottery from the Nile Valley excavated from Mafia cave excavations (10th-5th century BC).

The limestone caves and shelter preserving these materials is found on the small island of Mafia called Juani. The cultural layer, containing Late Stone Age/Neolithic materials, is now dated by 14C samples from associated large conch shells offering dates of between 800-400 BC (Table 1, row 4 and 5 of dates).

Imported goods include those of Egyptian, GraecoRoman and Indian origin. The potsherds of Mediterranean origin, probably Egyptian, are of great interest because they can also provide a relative date. According to Nordström, who examined the shards, four are of marl clay from the Upper Nile region. Such pottery is known to have been traded by Egyptians to the Nubian 98

FELIX CHAMI: THE EGYPTO-GRAECO-ROMANS AND PANCHAEA/AZANIA Dubin (1987: 60-61) as Hellenistic or Roman period, dating from between 300 BC to AD 400 (for illustration see Chami 2002a). St John Simpson (pers. comm.) suggested that the ‘eye’ bead could be pre-Islamic and there is evidence for both Roman and Partho-Sasanian material from the Horn of Africa (e.g. Aksum and Ras Hafun).

Kingdom of Napata in about the eighth and seventh centuries BC (Kendal 1997: 162). Spencer has discussed the nature of this kind of pottery, which was made from calcareous marl clays: ...regional differences in ceramic production are particularly clear during the third Intermediate period (c. 1070-650 BC), when the pottery of southern Egypt differs considerably from that being made in the Delta. The chief cause of this variation is the relative absence of marl clay products in the north, since the principal sources of these clays lay in the desert regions between Memphis and Esna. (Spencer 1997: 66)

Lastly, and of equal importance, is the find from the same Mafia island of a red ware, possibly of Roman origin, associated with local potsherds of the Early Iron Working period dated to the third century AD (Chami 1999c). The imported ceramic has a partial remain of a stamp which may be difficult to decipher. According to Greene (1992) and Brodribb (1987) it is only the Romans who stamped ceramics, including pottery, tiles and bricks. However, St John Simpson notes that Sasanian stamped pottery exists (pers.comm.).

Arnold has also shown that this type of pottery stopped being popular after 650 BC.12 His study would suggest that the marl clay pottery found in East Africa could have arrived there not later than the seventh century and hence supports the earlier 14C dates for the occupation of the Mafia site.

Discussion Prior to 1990 only historical documents existed to show that the Red Sea region had provided a vital economic link between East Africa and Mediterranean communities. The communities of the Red Sea regions involved in this link included Nabataeans, Zoskales and Homerites/Sabaeans (Casson 1989: 44). Although early Greek documents were known to modern scholars, no one seriously believed them and they were therefore not discussed in the historiography of Eastern Africa. Places such as Punt/Pwane and Panchaea, mentioned in the early literature, were therefore placed in the Red Sea region.13 Although the documentary sources of the Roman period were accepted by most scholars, the tendency was to situate Azania and Rhapta more to the north of the eastern African coast.14 As Trigger (1989: 130) has suggested the main reason for this attitude was racial bias in assuming that only Hamites/Cushites could have founded trading towns or developed elements of cultural sophistication (cf. Sanders 1969). However, those who were convinced that the core of Azania was on the Tanzanian coast had a problem substantiating that position due to the lack of archaeological evidence. As suggested in Chittick’s publications, the problem was how to substantiate the documents in the absence of archaeological records. In his attempt to locate Rhapta, Ptolemy’s metropolis of Azania, Chittick (1982: 58) suggested before his death in 1984: ‘It remains possible that all traces of Rhapta have been washed away or buried.’

Figure 39: Stamped ceramic found in association with early centuries AD ceramics from Mafia Island, Tanzania. Another important finding is that of red painted ware identified by Sunil Gupta (pers. comm.) as being of Indian origin and dating from the Early Iron Working period. According to Allchin (1995: 141) the iron using culture of India can be dated back to the eighth century BC. Vessels of red painted ware are also known to have been used in the Red Sea area in the late centuries BC and early centuries AD (Curtis/Sinclair pers. comm.). Another class of pottery found in the island of Mafia dating to the same early period is black ware, also identified by Gupta as belonging to the Early Iron Working period of India. More than twenty-five potsherds recovered from the island of Mafia have been identified as of Indian origin (see Table 2).

The archaeological finds presented here, collected during 1990 by this author and his colleagues now incontrovertibly prove that the Graeco-Roman documents preserved only a scanty, patchy picture of what happened

Another category of find linking the Red Sea and East Africa is that of beads now identified as Hellenistic/Phoenician or just Graeco-Roman. These include one ‘eye’ bead and one mosaic bead described by

13 12

14

Arnold 1997: 67; also cf. Hope 2001.

99

Kitchen 1993; Phillips 1997; Cary & Warmington 1963. Allen 1983; Horton 1990; Ehret 1998.

TRADE AND TRAVEL IN THE RED SEA REGION (1980) showed that the people of Ras Hafun, where he found trade goods of the last centuries BC, had also used stone tools like those of East Africa (also cf. Chami & Wafula 1999).

in Panchaea or Azania, now part of East Africa. This author has already attempted to redraw the ancient cultural and economic system that embraced the ancient Near East, Nile, Red Sea and Mediterranean.15 The archaeological data provided above add to what has previously been reported and hence strengthen the interpretation of the historical records.

It would seem that by the time of the Periplus and Ptolemy, in the early centuries AD, the people of the coast of East Africa had begun to smelt iron. If this was the case, as the archaeological finds imply (Chami 1999b, 1999c), then the East African population of the period of the Periplus were already producing iron objects. The question is therefore why, according to the Periplus, the main imports of Azania were iron tools? Various explanations, including that of exchanging iron bloom for fine and finished tools from the Red Sea or the Mediterranean regions, have been suggested.17

The discovery of glass and semi-precious stone beads in association with marine shells at sites in the Rift Valley of Kenya and Tanzania dating back to 800 BC would suggest the antiquity of trade networks going back to the time of the Egyptian Punt. Similarly, the find at Tell Asmar, ancient Eshnunna, of a pendant of gum copal originating from a location between Zanzibar and Madagascar and dating to between 2500-2400 BC would offer further evidence of such trade links in the early days of the rise of Egyptian and Mesopotamian civilizations (Meyer et al. 1991). It should be noted that, putting aside the stress in this paper on trade links, the actual date for the earliest occupation layer in the Zanzibar cave is of the early part of the third millennium BC (see Table 1, row 3 of dates). Chicken bones of Asiatic origin were also recovered from this context (see Chami 2001, 2002a).

We now know, from the archaeological evidence reported above, that the Periplus does not provide the complete list of what was imported and exported. Iron bloom could have been a major item of trade as the Roman Empire required a considerable amount of iron for use in shipbuilding and weaponry. In fact the Romans stimulated the production of iron throughout their empire and its periphery (Tylecote 1976), and the increased production of iron in Azania, as revealed by archaeological finds in eastern Africa, could have been stimulated by such demand. Euhemerus noted that Panchaea produced iron probably as ore rather than bloom (Diodorus Siculus 5.46.4).

In view of what the author has discussed here and elsewhere, only a few points need to be reiterated. First, East Africa had a thriving population with trade centres from the last centuries BC or even earlier, during the Puntite period. The zooarchaeological data suggest that the people of the Late Stone Age domesticated animals like chickens, dogs and cats. They also domesticated plants such as rice, cotton, coconut and other crops of African and Asiatic origin such as banana. They entered into contacts with other cultures to the west, north and east. It is likely that the Late Stone Age people of the coast of East Africa occupied caves as dwellings or just for sanctuary. They used pottery and worked small quartz pebbles to make microlithic tools and blades (see details in Chami 2004). In addition to their domesticates, they hunted wild animals and fished to supplement their diet. The pottery of these people and its comparative analysis have been illustrated elsewhere.16

Another possibility is that local iron production did not meet the demand for iron tools required by the new agricultural communities of Azania, now opened to the world market by the growth of Panara, Rhapta and other large settlements. Importation of iron tools may have been necessary to cater for the higher demand in the markets of Azania. Another controversial topic is whether coconut oil was exported from Azania to the Red Sea markets and beyond as stated in the early translations of the Periplus (Schoff 1912: 99). Casson (1980, 1989) has argued that it was not coconut oil referred to in the Periplus, but nautilus shells. Recent archaeological finds, however, suggest that other south-east Asian domesticates such as the chicken and banana had already reached the coast of East Africa during the first millennium BC, so it is not improbable that the coconut could also have been acclimatised in East Africa at the same time (Chami 2001).

It would seem that these are the people identified by Pliny as the cave dwellers involved in the transfer of spices from south-east Asia to the Red Sea region. Pliny, the Periplus, and the earlier accounts of Iambulus and Euhemerus preserved by Diodorus Siculus have shown that the people of East Africa had cultural and trade relationship with the people of the Red Sea and elsewhere. Euhemerus says there were Greeks settled in Panchaea. Iambulus says that the people of the Red Sea coast sent him to the islands of East Africa as a sacrifice to the people described by the South Arabians as ‘fair and good’. The Periplus reports that Azania was traditionally under the suzerainty of South Arabian rulers. Chittick 15 16

Another controversy, which can now be examined in view of the new archaeological finds, is whether it was East Africa which provided cinnamon and cassia to the Red Sea and the Mediterranean regions as stated by Strabo (Geography 11.5.7) and Pliny (Natural History 12.42.87-88). Miller (1969:153-172) suggested that it was the people of East Africa or Austronesians who

Chami 1999a, b; Chami 2002a, b. Chami 2001; Chami & Kwekason 2003.

17

100

Chami1999a; Mapunda 2002.

FELIX CHAMI: THE EGYPTO-GRAECO-ROMANS AND PANCHAEA/AZANIA (Chami 1999b). Future research in this general region may widen the knowledge we now possess of Rhapta and Azania.

obtained the spices from south-east Asia and trafficked them via the Red Sea and Nile Valley routes to the Mediterranean region. On the other hand, Casson (1984) has argued that it was the people of the Somali coast who received the Austronesian spices and passed them to the interior and to the Red Sea.

The last point for discussion is the possible interior trade link between the coast of East Africa and the Nile valley. It was noted earlier that Strabo suggested that cinnamon and cassia reached the Mediterranean region via the Nile sudd, south of Meroe. Pliny claimed that they originated from the coast of East Africa. Miller (1969) has argued that there was an interior route from the coast of East Africa to the Nile Valley. This idea is supported by the present author (Chami 1999a), who has argued that the route may have gone through central Tanzania to the Great Lakes region.

There is evidence on the Tanzanian coast of trade goods and domesticates of South Asian/Indian origin as well as objects from the Red Sea/Mediterranean regions, dating back to the first millennium BC. This would suggest that either East Africa or Somalia or both played a major role in trafficking the spices. The archaeological evidence for most of these trade goods has been found in caves. This corroborates Pliny’s statement, correcting Strabo, that it was the cave dwellers of East Africa who brought spices from far out in the ocean and passed them on to the Red Sea.

The point is that when the Mediterranean states failed to obtain south-east Asian products through the Red Sea, due to the growing power of the Arabs or perhaps the Persians, other routes were explored. According to Kosmas Indicopleustes (Freeman-Grenville 1975), in the sixth century AD the Aksumites used an interior route to trade with East Africa in order to avoid the Sasanians, who at that time dominated the waters of the northern Indian Ocean. In the late fifteenth century, the Portuguese had to circumnavigate Africa from the Atlantic Ocean via the Cape to get spices from India and south-east Asia in order to avoid the Muslim blockade of the eastern Mediterranean and the Red Sea.

The actual location of the Azanian capital, Rhapta, remains unknown. However, archaeological indicators reported above suggest that it was located on the coast of Tanzania, in the region of the Rufiji River and Mafia Island. It is in this region where the concentration of Panchaea/Azanian period settlements has been discovered. If the island of Menuthias mentioned in the Periplus was Zanzibar, a short voyage south would land one in the Rufiji region.18 Ptolemy locates Rhapta at 8º S, which is the exact latitude of the Rufiji Delta and Mafia Island. The metropolis was on the mainland about one degree west of the coast near a large river and a bay with the same name. While the river should be regarded as the modern Rufiji River, the bay should definitely be identified with the calm waters between the island of Mafia and the Rufiji area. The peninsula east of Rhapta would have been the northern tip of Mafia Island. The southern part of the bay is protected from the deep sea by numerous deltaic small islets separated from Mafia Island by shallow and narrow channels. To the north the bay is open to the sea and any sailor entering the waters from that direction would feel as if he were entering a bay. Even today the residents identify these waters as a bay, referring to it as a ‘female sea’, as opposed to the more violent open sea on the other side of the island of Mafia.

It is likely that before the rise of Roman power in the Indian Ocean an interior route via Meroe, and perhaps even circumnavigation of Africa, had been used to reach the Indian Ocean and obtain goods from south-east Asia. This is supported by the report in Strabo that spices reached the north via the Nile sudd and Eudoxus’ report that seafarers from Gades in southern Spain had circumnavigated Africa from the Atlantic Ocean, via the Cape. This could have occurred during the period of Persian dominance of the eastern Mediterranean and the Red Sea. Those that deny that either of these routes was possible should consider the late fifteenth century and later Portuguese voyages round the Cape of Good Hope to India and the journey of Speke and Grant in the nineteenth century from Zanzibar via the Great Lakes region to Egypt. A sliver of evidence for a Roman route to East Africa via Meroe is the Emperor Nero’s dispatch, in the first century AD, of an expedition to the source of Nile. One reason for this expedition was to find out if the overland trade route was still functioning. The mission reported that the route had shifted to the Red Sea (Welsby 1996). The implication of this report is that goods traded along this route must have originated from the coast of East Africa, and hence could easily be re-routed through the Red Sea. Additional circumstantial evidence is provided by Diogenes’ journey from the coast of East Africa/Rhapta to the source of the Nile (Huntingford 1980). Diogenes must have been following a trade route to the deep

The doyen of East African coastal archaeology, Nevelle Chittick, was quite convinced that Rhapta was in the area of the Rufiji River. As noted earlier, his suggestion (1982) that the remains of the metropolis are now buried in the ocean, near the mouth of the river, where he surveyed, may not be acceptable in view of the information provided by Ptolemy. Since Ptolemy located Rhapta 1° inland from the coast, it would seem that the settlement was about 40 km or further in the interior. In this region, larger ancient settlements, some with imports from the Mediterranean world, have been recovered 18

Datoo 1970; Kirwan 1986; Periplus/Casson 1989.

101

TRADE AND TRAVEL IN THE RED SEA REGION interior. Although his mission could have been exploratory, like those of the nineteenth century African explorers, another objective could also have been to ascertain whether the trade route continued to the Nile Valley. In view of the new archaeological data from the coast of Tanzania one can now start reassessing the shift of trade routes from time immemorial in response to ancient power politics. It is when sub-Saharan Africa is placed in the context of ‘classical’ history with more positive involvement that a better picture of these trade routes will be made clear.

References Allen J. de V. 1983. Swahili origins. London: James Curry. Arnold D. 1997. Introduction to Egyptian pottery. Pages 62-73 in I. Freestone & D. Gaimster (eds), Pottery in the Making: World Ceramic Traditions. London: British Museum Press. Begley V. & de Puma R. (eds) 1991. Rome and India: The Ancient Sea Trade. Madison, Wisconsin: University of Wisconsin Press. Brodribb G. 1987. Roman Brick and Tile. Gloucester: Alan Sutton. Cary M. & Warmington E. 1963. The Ancient Explorers. London: Penguin. Casson L. 1980. Periplus Maris Erythraei: Three notes on the text. Classical Quarterly 30: 495-497. ------------1984. Ancient Trade and Society. Detroit, Michigan: Wayne State University Press. Chami F. 1999a. Graeco-Roman trade link and the Bantu migration theory. Anthropos 94/1-3: 205-215. ------------1999b. Roman beads from the Rufiji Delta, Tanzania, first incontrovertible link with Periplus. Current Anthropology 40/2: 237-241. ------------1999c. The Early Iron Age on Mafia Island and its relationship with the mainland. Azania 34: 1-11. ------------2000. Further archaeological research on Mafia Island. Azania 35: 208-214. ------------2001. Chicken bones from a Neolithic limestone cave site in Zanzibar. Pages 84-97 in F. Chami, G. Pwiti, & C. Radimilahy (eds), People, Contacts and the Environment in the African Past. Dar-es-Salaam: Dar-es-Salaam University Press. ------------2002a. People and contacts in the ancient western Indian Ocean seaboard or Azania. Man and Environment 27/1: 33-44. ------------2002b. East Africa and the Middle East relationship from the last millennium BC to about 1500 AD. Journal de Africanistes 72/2: 21-38. ------------forthcoming 2004. The archaeology of Mafia Island, Tanzania. In Chami, F., Pwiti, G. & Radimilahy, C. (eds), African Archaeology Network: Reports and Views. (Studies in the African Past 4). Dar-es-Salaam: University Press. Chami F. & Kwekason A. 2003. Neolithic pottery traditions from the islands, the coast and the interior of East Africa. African Archaeological Review 20/2: 65-80. Chami F. & Msemwa P. 1997. A new look at culture and trade on the Azania coast. Current Anthropology 38/4: 673-677. Chami F. & Wafula G. 1999. Zanzibar in the Neolithic and Roman time. Mvita: Bulletin of the Regional Centre for the Study of Archaeology in Eastern and Southern Africa, Kenya 8: 1-4. Chittick N. 1966. Six early coins from near Tanga. Azania 1: 156-157. -------------1980. Pre-Islamic trade and ports of the Horn. Pages 364-366 in R.E. Leakey & B.A. Ogot (eds), Proceedings of the 8th Panafrican Congress of

Conclusion This paper has attempted to put East Africa or Panchaea/Azania in the context of the Egypto-GraecoRoman world via the Red Sea. This has been achieved using the most recent archaeological finds from the Tanzanian coast. The data has been collated with GraecoRoman reports in order to offer a clear reconstruction of relations between the two regions. It has been shown that the coast of East Africa had Late Stone Age/Neolithic communities, probably established about 3000 BC. These communities, apart from hunting wild animals and fishing, also domesticated animals such as chickens, dogs and cats. It is likely that plants such as the coconut and banana were also domesticated. It is within this Neolithic period that these East African communities began trading with the north and east, reaching the Mediterranean region and south-east Asia. The Indian Ocean islands, for example, Zanzibar, were occupied by such Neolithic people well into Hellenistic times, when Euhemerus mentions that the island had a capital called Panara. A new era opens with the adoption of iron technology in the early centuries AD or slightly before. The communities of East Africa grew in size, traded with Rome and Arabia and were more heavily cultivated and settled. The population traded with the Romans via the Red Sea, and their capital, identified by the Periplus and Ptolemy, was Rhapta. The culture and the economy of these people spread to the deep interior and as far south as southern Africa. The core of these communities may have been in the Rufiji-Mafia region where many sites of that period are found, some with remains of imported Mediterranean trade goods. It has been noted that at different periods of East African history, trade routes changed, according to power politics. Northerners sometimes reached East Africa via the interior and at other times by way of the Indian Ocean or the Red Sea. However, as trade routes are difficult to illustrate archaeologically, this is not easy to prove conclusively.

102

FELIX CHAMI: THE EGYPTO-GRAECO-ROMANS AND PANCHAEA/AZANIA second series, volume 151). London: The Hakluyt Society. Periplus/W. Schoff (ed. and transl.) 1912. The Periplus of the Erythreaan Sea. New York: Longmans, Green & Co. Phillips J. 1997. Punt and Aksum: Egypt and the Horn of Africa. Journal of African History 38: 423-457. Pliny/H. Rackam (ed. and transl.).1961. Natural History. (10 volumes) (Loeb series). London: William Heinemann/Cambridge, Massachusetts: Harvard University Press [Reprint]. Sales J. 1996. Archaemenid and Hellenistic trade in the Indian Ocean. Pages 251-268 in J. Reade (ed.), The Indian Ocean in Antiquity. London: Keagan Paul. Sanders E. 1969. The Hamitic hypothesis: its origin and functions in the time perspective. Journal of African History 10/4:521-532. Sheriff A. 1981. The East African coast and its role in maritime trade. Pages 551-567 in G. Mokhtar (ed.), UNESCO General History of Africa. (8 volumes)/vol. 2: Ancient Civilization of Africa. Berkeley: University of California Press. Smith M. & Wright H. 1988. The ceramics from Ras Hafun in Somalia: Notes on a Classical maritime site. Azania: 23:115-142. Spencer A. 1997. Dynastic Egyptian pottery. Pages 62-73 in I. Freestone & D. Gaimster (eds), Pottery in the Making: World Ceramic Traditions. London: British Museum Press. Strabo/Jones H. (ed. and transl.). 1966. The Geography of Strabo. (8 volumes). (Loeb series). London: William Heinemann/Cambridge, Massachusetts: Harvard University Press. [Reprint]. Trigger B. 1989. A History of Archaeological Thought. Cambridge: Cambridge University Press. Tylecote R. 1976. A History of Metallurgy. London: The Metals Society. Vogel J. & Visser E. 1981. Pretoria radiocarbon dates II. Radiocarbon 23/1: 43-80. Welsby D. 1966. The Kingdom of Kush. London: British Museum Press. Whitehouse D. & Williamson A. 1973. Sasanian maritime trade. Iran 11: 29-49.

Prehistory and Quarternary Studies (Nairobi, 5 to 10 September 1977). Nairobi: International Louis Leakey Memorial Institute for African Prehistory. -------------1982. Reconnaissance in coastal Tanzania. Nyame Akuma 20: 57-58. Cinimo R.M. (ed.) 1994. Ancient Rome and India. New Delhi: Munshiram Manoharlal. Compareti M. 2002. The Sasanians in Africa. Transoxiana 4: www.transoxiana.com Datoo B.A. 1970. Rhapta: the location and importance of East Africa’s first port. Azania 5: 65-75. Diodorus Siculus/C. Oldfather (transl.) 1961. History. (12 volumes). (Loeb series). London: William Heinemann/Cambridge, Massachusetts: Harvard University Press. [Reprint]. Dubin L. 1987. The History of Beads from 30,000 B.C. to the Present. London: Thames and Hudson. Ehret C. 1998. An African Classical Age. Eastern and Southern Africa in World History, 1000 B.C. to A.D. 400. Oxford: James Curry. Freeman-Grenville G. 1975. The East African Coast. Selected Documents from the First to the Earlier Nineteenth century. Oxford: Clarendon Press. Greene K. 1992. Roman Pottery. London: British Museum Press. Hope C. 2001. Egyptian pottery. (Shire Egyptology 5). (2nd edition). Princes Risborough: Shire Horton M. 1990. The Periplus and East Africa. Azania 25: 95-99. ------------1996. Early maritime trade and settlement along the coasts of eastern Africa. Pages 439-460 in J. Reade (ed.), The Indian Ocean in antiquity. London: Kegan Paul International in association with the British Museum. Kendall T. c. 1997. Kings of the sacred mountains: Napata and the Kushite twenty five dynasty of Egypt. Pages 161-228 in W. Dietrich (ed.), Sudan: Ancient Kingdoms of the Nile. Paris: Flammarion. Kirwan L.P. 1986. Rhapta, metropolis of Azania. Azania 21: 99-114. Kitchen K. 1993. The land of Punt. Pages 587-608 in T. Shaw, P. Sinclair, B. Andah, & A. Opoko (eds), The Archaeology of Africa. Food, metals, and towns. (One world archaeology 20). London/New York: Routledge. Mapunda B. 2002. Iron metallurgy along the Tanzanian coast. Pages 76-88 in F. Chami & G. Pwiti (eds), Southern Africa and the Swahili World. (Studies in the African Past 2.). Dar-es-Salaam: University Press. Meyer C., Todd J. & Beck C. 1991. From Zanzibar to Zagros: A copal pendant from Eshnunna. Journal of Near Eastern Studies 50: 289-98. Miller J. I. 1969. The Spice Trade of the Roman Empire, 29 B.C. to A.D. 641. Oxford: Clarendon Press. Periplus/L. Casson (ed. and transl.) 1989. The Periplus Maris Erythraei. Text with Introduction, Translation, and Commentary. Princeton: Princeton University Press. Periplus/G. Huntingford (ed. and transl.) 1980. The Periplus of the Erythreaan Sea. (The Hakluyt Society, 103

104

Reflections of ethnicity in the Red Sea commerce in antiquity: evidence of trade goods, languages and religions from the excavations at Berenike Steven E. Sidebotham Berenike (at 23° 54.62’ N/35° 28.42’ E) was a PtolemaicRoman emporium on the Red Sea coast for about eight hundred years, from the third century BC until the sixth century AD. Located about 825 km south of Suez and about 260 km due east of Aswan, Berenike was the premier emporium at the northern end of the Red Sea in Ptolemaic and Roman times. It was also the southernmost major Red Sea port in Ptolemaic and Roman Egypt (Figs. 40-44).

population patterns and their variations are evident from studying the plethora of artifacts, floral and faunal remains, the diverse linguistic record (evidence of twelve different languages/scripts: Hieroglyphs, Demotic, Greek, Latin, Palmyrene, Hebrew, Aramaic, Coptic, Tamil-Brahmi, a Prakrit-Sanskrit hybrid and two unknown) and the religious preferences of the city’s inhabitants. Interestingly, the Berenike harbours, like those elsewhere along the Red Sea coast, were subject to extensive silting as occasional heavy rains and the resulting flash floods carried huge amounts of water-borne sediments into them (Harrell 1996: 103-105). Instead of dredging periodically as needed, which seems to have been within their technological competence, judging by evidence from Quseir al-Qadim (Whitcomb 1979: 37), the inhabitants of Berenike simply moved their settlement ever eastwards towards the receding harbours. This phenomenon must be considered when investigating the different periods of activity and the changing ethnic composition of Berenike’s population. This silting process resulted in the earlier Ptolemaic settlement lying farther west while the Roman ones extended farther north-east, south-east and east.

Founded by Ptolemy II Philadelphus in about 275 BC and named after his mother (Pliny, Natural History 6.33.168), it was part of a larger project of Red Sea port, canal and trans-desert road construction (Bagnall et al. 1996: 320). Berenike’s final demise before the mid-sixth century AD was due, undoubtedly, to multiple causessilting of the harbour, competition from South Arabian and Aksumite middlemen and, possibly, the effects of a plague.1 Excavations conducted by the University of Delaware (co-director S.E. Sidebotham) and Leiden University (co-director W.Z. Wendrich) have recovered no evidence, thus far, for earlier Pharaonic or later Islamic occupation. The remains of the central part of the town cover an area of about seven hectares bounded on the east by the Red Sea and to the north and south by shallow wadis. The southern wadi ends in a shallow bay, once part of an ancient harbour; the northern/northeastern harbour has completely silted up. The central concentration of buildings in the town, that sits atop a natural mound comprising partly extinct coral reefs, measures about 250 m north-south x 320 m east-west. Scattered structures and other indications of activity reveal that the periphery of the city extended much farther towards the north and west for a total area of about 500 m north-south x 700 m east-west.

There were three major periods of peak activity at Berenike: early and middle Ptolemaic, early Roman and late Roman. It is from studying these that we have some insight into the ethnic diversity of the port’s inhabitants. Little is known thus far about the late Ptolemaic, middle Roman or the final - late fifth and sixth century AD occupation of the settlement. Early and Middle Ptolemaic Period Faunal evidence, predominantly, suggests that the population at Berenike in this period was Greek/Macedonian, Egyptian/Greco-Egyptian from the Nile valley and peoples from the Eastern Desert (van Neer & Ervynck 1999: 338). While there is evidence of contact between Berenike and regions beyond Egypt, there is, at this point, no evidence that peoples from beyond Egypt dwelt at Berenike in Ptolemaic times. Excavation of Ptolemaic remains has uncovered substantial data from the third-second centuries BC in the western-most part of the city and it is here that excavations have found most of the evidence thus far recovered from Berenike for Ptolemaic activity. Some

Surface surveying, geological explorations and excavations by the Delaware-Leiden team at Berenike between 1994-20012 have investigated in detail only about 2% of the site horizontally, much less vertically, but this limited fieldwork documents the rich and diverse ethnic and socio-economic backgrounds of the people who made Berenike their home. Analyses of excavated remains have also begun to record how the ethnic composition changed over the long life of the port. Ethnic 1 Procopius, History of the Wars 2.22.1-23.21; 2.24.8, 12; Keys 1999: 17-24; Sidebotham & Wendrich 2001-2002: 48. 2 Sidebotham & Wendrich eds, 1995, 1996, 1998a, 1998b, 1999, 2000, 2001-2002; 2002.

105

TRADE AND TRAVEL IN THE RED SEA REGION

Figure 40: Map showing the location of Berenike

Figure 41: Map showing routes and sites between the Nile and the Red Sea

106

STEVEN E. SIDEBOTHAM: REFLECTIONS OF ETHNICITY IN THE RED SEA COMMERCE IN ANTIQUITY

Figure 42: Berenike and surroundings.

107

TRADE AND TRAVEL IN THE RED SEA REGION Used as ballast, the basalt chunks would have been unloaded – and discarded – along with the cargoes to allow for refurbishment and repair of the ships. Larger basalt pieces may well have been carried off for use elsewhere on site as grinding and building stones. Silted up at some point, Roman structures later appeared in this area, thereby providing some terminus ante quem dates for the putative Ptolemaic port’s demise. Future excavations in this area aim to confirm this tentative harbour identification.

late Ptolemaic material has also been found here, as well as in a trench north of the Serapis temple (Fig. 44, n. 7) and beneath the early Roman rubbish dump at the northern end of the site (Fig. 43). This western area was an industrial zone where metals (lead sheeting, iron and copper alloy nails, tacks and fittings) and bricks were made3 (Fig. 44, nn. 4-5). The eventual identification and excavation of the Ptolemaic commercial and residential areas should provide a more complete idea of the diversity and size of the city’s population in this early period.

Excavations have provided little data indicating how large Berenike was in the Ptolemaic period, though it seems to have extended from at least the Ptolemaic industrial zone, noted above, to beneath the early Roman rubbish dump in the north and over to the region of the Serapis temple in the east; an approximate area of 175 m north-south x 425 m east-west. Nevertheless, there is insufficient data to allow estimation of population size in the early and middle Ptolemaic periods.

There is, generally, ample archaeological evidence from Berenike of Ptolemaic contacts with other areas of the desert (Nubia), Nile valley (Edfu and Alexandria), Mediterranean (Rhodes and Phoenicia) and, much less so, with the Red Sea (a single elephant tooth, some ivory). Evidence from a papyrus (not found at Berenike) records the trade in live elephants – and the hazards this commerce faced – coming up from coastal Sudan and Eritrea;4 Agatharchides also discusses this commerce (1989: 141 and n. 3). Graffiti depicting these pachyderms, undoubtedly from this trade, survive along the roads linking Berenike to the Nile.5 Written sources confirm Berenike’s contacts with other areas of the Red Sea in the early and middle Ptolemaic period though one would expect more archaeological data pertaining to the elephant trade that seems to have been one of the main raisons d’être of the port in early Ptolemaic times.

The excavation of carefully made copper-alloy objects, faience beads, small-scale sculpture, an ostracon in Demotic (probably from the late Ptolemaic period and found beneath the early Roman rubbish dump) dealing with wine and ostrich eggshell does suggest that at least some segment of the population was moderately well– off. The ethnic backgrounds of these well-to-do inhabitants are, however, unknown. There is very little evidence, aside from the ostracon written in Demotic dealing with wine (noted above) and a little pottery from the Ptolemaic industrial area and from a trench immediately north of the Serapis temple, for late Ptolemaic activity at Berenike. Thus, nothing can be said about the composition or size of the population at that time.

Epigraphic evidence from India indicating diplomatic contacts with the Mauryan empire of Asoka (cf. Sidebotham 1986: 7) and ample epigraphic and archaeological data attesting contacts between southern Arabia (Sidebotham 1986: 1-12) and Egypt and with Egyptian possessions in the Aegean in the Ptolemaic period (Bagnall 1976) are, at present, not evident in the archaeological record at Berenike. The archaeological evidence from Berenike itself (an elephant’s tooth, ivory and some pottery) of contact with other areas of the ancient world outside Egypt at this time is, clearly, very limited. There is none for the presence of groups other than those noted above residing at Berenike. There must have been; but at this point in the limited excavations of Ptolemaic areas, the evidence is simply lacking that this was the case.

Early Roman Period It is in this era, and in the late Roman period, that excavations have unearthed the most data bearing on socio-economic and ethnic diversity at Berenike. This information derives from extensive written documents found at the site dating primarily, but not exclusively, from the first century AD, from large quantities of floral and faunal remains, from textiles, pottery, coins and other organic artifacts. There is modest evidence of religious preferences in the city at this time.

Surface surveying may have located the Ptolemaic harbour. Immediately east and south of the Ptolemaic industrial area a large semicircular berm with ‘arms’ protruding from the concave side suggests the outlines of an artificial port facility (Figs 43 & 44, n. 3). The discovery of quantities of vesicular basalt – not found in Egypt, but imported from more southerly parts of the Red Sea - in this area may lend support to this identification.

The written documents derive predominantly, though not solely, from the early Roman rubbish dump in the northern part of the site (Fig. 43). Hundreds of JulioClaudian and Flavian period (30s-early 70s AD) ostraca and dozens of papyri shed light on official and personal activities of the city’s residents (Bagnall et al. 2000; forthcoming). The official documents tend to appear on the ostraca while the private ones are written on papyri (Bagnall et al.2000; forthcoming). The former are mainly passes granting permission for goods to go aboard ships.

3

Sidebotham 1998: 106-107; Sidebotham & Wendrich eds 2001-2002: 25-26. 4 Wilcken 1912: 533-535; cf. Scullard 1974: 126-133. 5 cf. Bernand 1972: 44-46, no. 9 bis and pl. 54, nos 1-2 at el-Kanaïs; Casson 1993; Sidebotham & Zitterkopf 1996: 376, fig. 21-20 at Abraq.

108

STEVEN E. SIDEBOTHAM: REFLECTIONS OF ETHNICITY IN THE RED SEA COMMERCE IN ANTIQUITY

Figure 43: Site plan, Berenike

Figure 44: Site view with places of interest numbered 1-15.

109

TRADE AND TRAVEL IN THE RED SEA REGION from Wādī Lahma in the northwest to Shenshef in the southwest (Fig. 42) protected land approaches to the city (Sidebotham & Wendrich eds 2001-2002: 38, n. 9). At least three of these (two in Wādī Kalalat and one at Siket: see Fig. 42) provided potable water to Berenike (Sidebotham & Wendrich eds 2001-2002: 38-39) and two of these preserve epigraphic evidence of the military’s role in that endeavour.6 The military’s discovery, acquisition, protection and, one must assume, control of transport of potable water to the city attest its critical role in the life of this important port. As no potable water could have been obtained from wells dug close to Berenike – water from these would have been salty or too brackish for any human and most animal consumption – the residents were dependent upon the military to supply this necessary commodity. The administrative and technical mechanisms involved in the transport of water from these outlying hydreumata to the city are unknown. One ostracon (O. Berenike II, 191 from the first century AD rubbishdump north of town) may refer to a leather water bag (R.S. Bagnall pers comm.), suggesting, perhaps, that some kind of permit was required to ship/receive water. The bag may indicate that pack animals transported the water; as no aqueducts/water channels have been found the conclusion that pack animals were used gains some credence.

In some instances the names of the ships are recorded. The bulk of these ostraca are from the Berenike customs house through which trade goods and objects consumed by ships’ crews passed. The personal names of dozens of men appear on these ostraca; some of the names occur repeatedly thereby indicating their importance in the customs house and its bureaucratic processes. These names are Egyptian, Greek, Roman and a few Semitic (one, perhaps, Palmyrene); there were also mixed GreekEgyptian personal names. Surprisingly few official titles appear in these documents and those that do denote military ranks. Undoubtedly, these officials and those dealing with them interacted frequently; the officials’ titles were well known and these, therefore, do not appear on the ostraca. The bulk of these ostraca are written in Greek. One ostracon found in the early Roman rubbish dump is written in a script that has yet to be identified (cf. Sidebotham & Wendrich eds 2001-2002: 28, fig. 9). The papyri number a few dozen and include a wellpreserved bill of sale for a white male donkey for 160 drachmai. The buyer was clearly illiterate or barely literate as his signature on the papyrus suggests; evidently, a professional scribe drafted this document. Fragments of land registers indicate propertied classes (Bagnall et al. forthcoming). A letter from a mother complaining that her son, who was in Arabia, had not written her for some time could, of course, stem from any time in history!

Pottery, especially from Italy, but also from elsewhere in the western and eastern Mediterranean, indicates widespread contacts with those regions. Undoubtedly, merchants and their representatives from some of those areas (attested elsewhere by name in graffiti in the Eastern Desert on the road between Berenike and the Nile) (Fig. 41) (cf. De Romanis 1996: 203-217) must also have resided at Berenike at least temporarily.

Two inscriptions that are duplicates of one another, found in a late Roman building (in the area on Fig. 44, n. 9), indicate that some women played prominent public roles in Berenike. A woman named Philotera offered the two stones, dedicated to Jupiter during the reign of Nero (5468 AD) (Sidebotham & Wendrich eds 2001-2002: 32; Bagnall et al. forthcoming).

Tantalisingly, a few sherds of the eggshell thin and delicately painted Nabataean fine ware have turned up both as surface finds and from excavated contexts (Hayes 1995: 38; 1996: 150). There is no way to determine if this indicates that Nabataeans visited or resided at Berenike – a possibility since their presence is epigraphically attested quite extensively elsewhere in the Eastern Desert7 – or merely denotes contacts with that caravan kingdom.

A few texts also come from a trench near the Serapis temple (Fig. 44, n. 7) and date to the early Roman period. Several nineteenth century ‘clearing’ operations of the temple to this syncretised deity (a combination of Osiris and Apis) revealed Hieroglyphic texts on temple walls that attest its operation from early Roman times until at least the joint reigns of Marcus Aurelius and Lucius Verus (161-169 AD) or in the last years of Aurelius’ reign (died 180 AD) (Meredith 1957: 61, 69). The temple may well pre- and post-date the epigraphic evidence preserved therein.

An abundance of Indian-made pottery, both fine table wares and cooking and coarse wares found in early levels throughout the excavations,8 and dated from the late first century BC into the first century AD, plus the find of a Tamil-Brahmi graffito from the first century AD scratched on an Italian amphora fragment (Mahadevan 1996), probably indicate the presence of people from

Faunal remains indicate a predilection for pork and chicken in early Roman times. Mediterranean peoples outside Egypt consumed pork, which was also popular with the Roman army (Groenman-van Waateringe 1997: 263). The ostraca noted above confirm Roman military presence at Berenike at this time.

6 Bagnall 2000; Bagnall et al. 2001; Sidebotham & Wendrich eds 20012002: 38-39; Sidebotham forthcoming. 7 Clermont-Ganneau 1919; Littmann & Meredith 1953; Littmann & Meredith 1954; Jones et al. 1988; Toll 1994; Briquel-Chatonnet & Nehmé 1998. 8 Begley & Tomber 1999; Tomber & Begley 2000; Tomber 2000.

Several forts in the environs of Berenike also provide archaeological and epigraphic evidence for the military’s prominence in the area. In early Roman times ten forts 110

STEVEN E. SIDEBOTHAM: REFLECTIONS OF ETHNICITY IN THE RED SEA COMMERCE IN ANTIQUITY Surprisingly little has been excavated from the second century at Berenike apart from some pottery, a few coins and an inscription – found in a later Roman context (in the area on Fig 44, n. 13) - from the time of Trajan dedicated by a secretary and translator to Isis.17 Earlier travellers recorded Hieroglyphic texts perhaps from the period of Marcus Aurelius and Lucius Verus (cited above) (Meredith 1957: 61, 69) in the Serapis temple. The Delaware-Leiden team also recovered fragments of one or more inscriptions in Greek from the reign of Trajan in a large hydreuma in Wādī Kalalat (Fig. 42), c. 8.5 km west of Berenike; that site certainly provided drinking water to the city (Bagnall 2000).

southern India residing at Berenike. The graffito mentions a Chera chieftain named Korra. Interestingly little, except possibly some pottery, has been identified from southern Arabia. A small corpus of handmade wares, mainly from early Roman levels, is similar to fabrics from South Arabia and this may well be the source for the Berenike sherds; further study must be undertaken, however, to confirm this (R.S. Tomber pers. comm.). The ancient literary sources are replete with references to the export of Roman merchandise to southern Arabia and importation of quantities of frankincense and myrrh from there to the Roman empire undoubtedly via the Red Sea coast of Egypt as well as by overland caravan across the Arabian peninsula.9 An Abrus precatorius seed from India was found in the excavations.10 In India today these are used to weigh gold and are often mixed with incense. The recovery of this seed thus may be indirect evidence of the presence of (South Arabian or Somali) incense on site (R.T.J.Cappers pers. comm).

From the late second/early third century AD come two inscriptions from the same location (Fig. 44, n. 6). One carved in Greek and dedicated by Marcus Aurelius Mokimos, a Palmyrene archer serving as an auxiliary in the Roman army, to the Roman Imperial cult of Caracalla and Julia Domna dated 8 September 215. It was on a base atop which excavations recovered portions of a nearly life-sized bronze statue of an unknown female deity holding either a snake or cornucopia in her left hand (Sidebotham et al. 1996: 230, 234, fig12-5). Another text, a bilingual in Palmyrene and Greek (dates c. 180212 AD), mentions several Roman officials including Aemilius Celer (prefect of Mount Berenike), the ala Heracliana (probably a dromedary rather than a cavalry unit) and a chiliarch (military tribune) named Valerius Germanion. It also records that the artist Berechei (a Semitic and probably Palmyrere name) made a statue of the greatest god Hierobol/Yarhibol.18 A statue base and a bronze hand may be all that is left of Berechei’s efforts.19 Long before the excavation of this cult centre, mounted Palmyrene archers were thought to have patrolled the road between Berenike and the Nile emporium of Coptos (Sidebotham 1986: 66); the discovery of a ‘Palmyrene’ shrine in the city confirms this. Its presence is not that surprising, though the relative lavishness of the dedications is noteworthy.

The discovery of large quantities of teak wood,11 a few scraps of bamboo from the early Roman rubbish dump (Vermeeren 2000: 318, 321, 340) and sandalwood (Vermeeren 2000: 318, 321, 329) indicate contacts with South Asia and the Indian Ocean basin. Recovery of some Indian sailcloth (Wild & Wild 2000: 266-269; 2001) and other textiles (Wild & Wild 2000: 271-273), rice (which may be from Palestine and not India) (Cappers 1998: 305-306), sorghum,12 coconut,13 peppercorns,14 agate cameo blanks (P. Francis, Jr. pers. comm.) and beads (Francis 2000: 221-222) also highlights extensive trade contacts with South Asia (especially, but not exclusively, India and Sri Lanka) at this time. These discoveries reinforce the ceramic evidence and the graffito mentioned above for the presence of Indian and other South Asian merchants and sailors at Berenike. The finds of a gold and pearl earring,15 nicely carved semi-precious stones from finger rings, marble (from Proconessus in Asia Minor) floor and wall revetment,16 fragments of embroidery, plush carpets, furniture coverings or tapestries (Wild & Wild: 2000: 254, 258) and even escargot (Helix pomatia) from the Mediterranean (van Neer & Ervynck 1999: 339), indicate a wealthy element in the city at this time. Unfortunately, there is no way to determine from these finds nor is there any context which would indicate what the ethnicities of these wealthy Berenikeans might have been.

Also found in this small multi-purpose shrine (Fig. 44, n. 6) was a stone statue of a sphinx,20 a small stone head of Harpocrates,21 a small stone altar (Sidebotham 1999: 75, pl. 2-41 and 76, pl. 2-42), and 100 wooden bowls with burnt offerings inside (Sidebotham 1999: 70-74). From the early Roman rubbish (Fig. 43) comes an exquisite statuette of Aphrodite.22 All this rupestral art was carved from local gypsum/anhydrite (Harrell forthcoming) suggesting the presence of an indigenous sculpture school at Berenike.

9

cf. Sidebotham 1986: 13-20; Young 2001: 34-36. Cappers 1999: 302; Cappers 2000: 307. 11 Vermeeren 1998: 343; Vermeeren 1999: 311, 319; Vermeeren 2000: 335, 340-342. 12 Cappers 1996: 327-329, 333-335; Cappers 1999: 302, 305. 13 Cappers 1998: 313-317; Cappers 1999: 300, 303, 305. 14 Cappers 1998: 311-313; Sidebotham & Wendrich eds 2001-2002: 30. 15 Sidebotham & Wendrich eds 1998a: 90; Francis 2000: 223; Sidebotham & Wendrich eds 2002: 30. 16 Harrell 1996: 111; Harrell 1998: 142; Harrell forthcoming. 10

17 Sidebotham & Wendrich eds 2001-2002: 32; Bagnall et al. forthcoming. 18 Dijkstra & Verhoogt 1999; Sidebotham & Wendrich eds 1998: 93-94. 19 Sidebotham & Wendrich eds 1998: 94-95, fig. 17; Sidebotham 2000: 52-61. 20 Sidebotham & Wendrich eds 1998: 93; Sidebotham 2000: 63-64. 21 Sidebotham & Wendrich eds 1998: 93; Sidebotham 2000: 60. 22 Sidebotham & Wendrich eds 2001-2002: 27; Sidebotham & Wendrich eds 2002: 28-29.

111

TRADE AND TRAVEL IN THE RED SEA REGION A thriving business and residential area in Berenike between the Serapis temple and the church (Fig. 44, nn. 9 and 13) reinforces the view that the city was prosperous from the middle of the fourth and into the fifth century AD. Multi-storeyed structures appear in abundance: ground floors dedicated to commercial activities of great value (small single entrances, door locks, tiny and delicate scales and weights) while upper storey/ies (all but one have staircases) were used for domesticresidential purposes. Many ground floor walls contain niches complete with extant wooden shelves (Sidebotham & Wendrich eds 2001-2002: 39-40).

Clearly, people from various socio-economic backgrounds, from many parts of the Mediterranean world, the Near East, the Red Sea and Indian Ocean lived and worked at Berenike in the early Roman period through the early third century AD. We have, however, relatively little information on the population in the second and early third centuries (aside from the inscriptions noted) and virtually none for the bulk of the third or first part of the fourth centuries AD. Late Roman Period Beginning sometime in the middle of the fourth century AD through part of the fifth, Berenike experienced a renaissance. There was a large building boom along the eastern edge of the city and in the city center (Fig. 44, nn. 8-14). There is a fair amount known about this period as most excavations conducted thus far (aside from the Ptolemaic industrial and early Roman rubbish dump areas) have dealt with the latest buildings on site. Most are from this period of the renaissance, viz. second half of the fourth through fifth centuries AD. The architectural, textual, floral, faunal, bead and numismatic records provide abundant data on commercial contacts, the degree of prosperity and the diverse and changing ethnic composition of the population in this period.

Excavations in some parts of the city have revealed substantial quantities of bones: camel, goat, sheep and by-products together with a fine hand-made pottery [Eastern Desert Ware (EDW)] that point to a population group from the desert/Upper Egypt/Lower Nubia – perhaps the Blemmyes noted in ancient sources (e.g. Olympiodorus and Procopius) – residing in the port at that time. Other areas of town have numerous marine remains (fish bones and shells) suggesting a completely different group (perhaps the Ichthyophagoi: ‘fish eaters’). In any case, the almost total absence of pork bones and nearly complete lack of pottery and coins from the western Mediterranean suggest a contraction of contacts with and, concomitantly, perhaps absence at Berenike, of people from those regions. The archaeological record indicates that trade networks in the late Roman period with the Mediterranean were mainly with the eastern portion.

Religious cults in late Roman times included a mystery religion (inside a small structure) complete with temple pools, lamps and torches (Fig. 44, n. 8) and a broken sculptural relief reused as a seat (Sidebotham & Wendrich eds 2001-2002: 28). This enclosed, most likely barrel vaulted, structure housed, perhaps, a cult to a syncretised Isis. The sanctuary noted above with the imperial cult (Fig. 44, n. 6) seems to have continued in use in the late fourth and fifth centuries though whether dedicated to the same deities as in the late second/early third century AD is not known. Another building, possibly a temple, on the northeastern edge of the city also flourished at this time (Fig. 44, n. 10). Ostraca recording ‘Jewish delicacies’, (Bagnall et al. 2000: 66, n. 99 and 70, n. 109) and a fifth century graffito in Hebrew (Schmitz: 2000: 183-186) suggest the presence of Jews on site though no synagogue has yet been identified. A large Christian ecclesiastical structure at the eastern end of Berenike (Fig. 44, n. 11) complete with church and living areas (lamps, bronze cross)23 was larger than the other contemporary pagan structures suggesting that while Christianity was not the only religion practiced in fifth century Berenike, it appears to have been one of the more prominent. Interestingly, the only non-Roman coins found at Berenike (one from Aksum of the last preChristian king, Aphilas, dating from the late third to first third of the fourth century, and the other of the Indian monarch Rudrasena III dating from about 362 AD)24 both came from excavations in the church.

Yet, contacts with South Asia and Aksum were extensive at this time. Over half of all beads found in late Roman levels are from South Asia (southern India and/or Sri Lanka) (Francis 2000: 222-223). Resist dyed cotton textiles from India have been found in fifth century refuse (Wild & Wild 2000: 264-273) and bamboo appears in the edifice housing the mystery cult noted above (Fig. 44, n. 8) (Vermeeren 2000: 325, 328, 340). A single Job’s tear (Coix lacryma-jobi L) (Cappers 1996: 331-332), an Indian grass seed, had been deliberately pierced and used as a bead. Quantities of Aksumite pottery indicate increased importance of that Red Sea kingdom in the trade (Tomber forthcoming). Economically, a high status late Roman cemetery at the northwestern edge of town (Fig. 44, n. 2) and continued importation of peppercorns, teak wood, textiles, sorghum, rice, coconuts, Mediterranean escargot and even bamboo, etc. (all noted above in early Roman levels, too) indicate a degree of wealth. Decapitated corpses unceremoniously dumped in the abandoned Ptolemaic industrial area (Fig. 44, nn. 4-5) suggest the other end of the economic spectrum. Conclusion

23

Sidebotham & Wendrich eds 2001-2002: 32-35; Sidebotham & Wendrich eds 2002: 29, 31. 24 Sidebotham & Wendrich eds 2001-2002: 41; Sidebotham & Wendrich eds 2002: 30, figs 9-10.

Clearly, the population of Berenike varied in size and composition (both ethnically and from the socio112

STEVEN E. SIDEBOTHAM: REFLECTIONS OF ETHNICITY IN THE RED SEA COMMERCE IN ANTIQUITY Bagnall R.S., Manning J.G., Sidebotham S.E. & Zitterkopf R.E. 1996. A Ptolemaic Inscription from Bir ΚIayyan. Chronique d’Égypte 71: 317-330. Begley V. & Tomber R.S. 1999. Indian Pottery Sherds. Pages 161-181 in S.E. Sidebotham & W.Z. Wendrich (eds), Berenike 1997. Report of the 1997 Excavations at Berenike and the Survey of the Egyptian Eastern Desert, including Excavations at Shenshef. (CNWS Publications, special series n. 4). Leiden: Centre of Non-Western Studies. Bernand A. 1972. Le Paneion d’El-Kanaïs: Les Inscriptions grecques. Leiden: E.J. Brill. Briquel-Chatonnet L. & Nehmé L. 1998. Graffiti nabatéens d’al-Muwayah et de BiΜr al-Hammámát (Égypte). Semitica 47: 81-88. Cappers R.T.J. 1996. Archaeobotanical Remains. Pages 319-336 in S.E. Sidebotham & W.Z. Wendrich (eds), Berenike 1995. Preliminary Report of the 1995 Excavations at Berenike (Egyptian Red Sea Cost) and the Survey of the Eastern Desert. (CNWS Publications, special series n. 2). Leiden: Centre of Non-Western Studies. ----------1998. Archaeobotanical Remains. Pages 289-330 in S.E. Sidebotham & W.Z. Wendrich (eds), Berenike 1996. Report of the 1996 Excavations at Berenike (Egyptian Red Sea Coast) and the Survey of the Eastern Desert (CNWS Publications, special series n. 3). Leiden: Centre of Non-Western Studies. ----------1999. The Archaeobotanical Remains. Pages 299-305 in S.E. Sidebotham & W.Z. Wendrich (eds), Berenike 1997. Report of the 1997 Excavations at Berenike and the Survey of the Egyptian Eastern Desert, including Excavations at Shenshef (CNWS Publications, special series n. 4). Leiden: Centre of Non-Western Studies. ----------2000. Archeobotanical Remains. Pages 305-310 in S.E. Sidebotham & W.Z. Wendrich, (eds), Berenike 1998. Report of the 1998 Excavations at Berenike and the Survey of the Egyptian Eastern Desert, including Excavations in Wadi Kalalat. (CNWS Publications, special series n. 5). Leiden: Centre of Non-Western Studies. Casson L. 1993. Ptolemy II and the Hunting of African Elephants. Transactions of the American Philological Association 123: 247-260. Clermont-Ganneau Ch. 1919. Les Nabatéens en Égypte. Revue de L’Histoire de Religions 80: 1-29. De Romanis F. 1996. Cassia, Cinnamomo, Ossidiana. Uomini e Merci tra Oceano indiano e mediterraneo (Saggi di Storia antica 9). Rome: “L’Erma” di Bretschneider. Dijkstra M. & Verhoogt A.M.F.W. 1999. The GreekPalmyrene Inscription. Pages 207-218 in S.E. Sidebotham & W.Z. Wendrich (eds), Berenike 1997. Report of the 1997 Excavations at Berenike and the Survey of the Egyptian Eastern Desert, including Excavations at Shenshef. (CNWS Publications, special series n.4). Leiden: Centre of Non-Western Studies.

economic point of view) during the approximately 800year history of the port. Regarding size, the ancient authors indicate, and excavations complement and supplement, the fact that the early Roman period was the zenith of activity at Berenike. The second greatest era was the late Roman and the third was the early to middle Ptolemaic. It may well be that future excavations will alter the relative rankings of these three periods. Future fieldwork should also reveal something of the epochs that are relatively unknown at the present time: the late Ptolemaic and middle Roman. Archaeological evidence at present indicates that the ethnic composition of the population also seems to have been most varied in the early Roman period, less so in the late Roman and least so in the Ptolemaic era. In the late Roman period the population seems to have been, as it had in the early Ptolemaic, more Egyptian and desert oriented in its composition and less ‘cosmopolitan’. The reasons for the ultimate demise of Berenike (silting of the harbour, stiff commercial competition from ‘nonRoman subject’ entrepreneurs and shippers, and/or a plague) should also eventually become clearer as excavations continue to unearth this city’s secrets. References Agatharchides of Cnidus/S.M. Burstein (ed. and transl.) 1989. On the Erythraean Sea. (The Hakluyt Society, second series n. 172). London: The Hakluyt Society. Bagnall R.S. 1976. The Administration of the Ptolemaic Possessions Outside Egypt. Leiden: E.J. Brill. ----------2000. Inscriptions from Wadi Kalalat. Pages 403-412 in S.E. Sidebotham & W.Z. Wendich (eds), Berenike 1998. Report of the 1998 Excavations at Berenike and the Survey of the Egyptian Eastern Desert, including Excavations in Wadi Kalalat.(CNWS Publications, special series n. 5). Leiden: Centre of Non-Western Studies. Bagnall R.S., Bülow-Jacobsen A. & Cuvigny H. 2001. Security and water on the Eastern Desert roads: the prefect Iulius Ursus and the construction of praesidia under Vespasian. Journal of Roman Archaeology 14: 325-333. Bagnall R.S., Helms C.C. & Verhoogt A.M.F.W. 2000. Documents from Berenike. Volume I. Greek Ostraka from the 1996-1998 Seasons (Papyrologica Bruxellensia 31). Brussels: Fondation Égyptologique Reine Élisabeth. Bagnall R.S., Helms C.C. & Verhoogt A.M.F.W. (forthcoming). The Inscriptions, Papyri and Ostraka. In S.E. Sidebotham & W.Z. Wendrich (eds), Berenike 1999-2000. The 1999 and 2000 Seasons of Excavations at Berenike (Egyptian Red Sea Coast) and Survey of the Eastern Desert, including Excavations at Wadi Kalalat and Siket. Los Angeles: Cotsen Institute of Archaeology.

113

TRADE AND TRAVEL IN THE RED SEA REGION Berenike 1995. Preliminary Report of the 1995 Excavations at Berenike (Egyptian Red Sea Coast) and the Survey of the Eastern Desert (CNWS Publications, special series n.1). Leiden: Centre of Non-Western Studies. Meredith D. 1957. Berenice Troglodytica. Journal of Egyptian Archaeology 43: 56-70. Pliny/H. Rackham (ed. and transl.) 1969. Natural History. (10 volumes). (Loeb series). Cambridge, Massachusetts: Harvard University Press; London: Heinemann. Procopius/H.B.Dewing (ed. and transl.) 1914. History of the Wars. (5 volumes). (Loeb series). Cambridge, Massachusetts: Harvard University Press; London: Heinemann. Schmitz P.C. 2000. Semitic Graffiti. Pages 183-189 in S.E. Sidebotham & W.Z. Wendrich (eds), Berenike 1998. Report of the 1998 Excavations at Berenike and the Survey of the Egyptian Eastern Desert, including Excavations in Wadi Kalalat (CNWS Publications, special series n. 5). Leiden: Centre of Non-Western Studies. Scullard H.H. 1974. The Elephant in the Greek and Roman World. Ithaca, New York: Cornell University Press. Sidebotham S.E. 1986. Roman Economic Policy in the Erythra Thalassa 30 B.C. – A.D. 217. (Mnemosyne Supplement 91) Leiden: E.J. Brill. ----------1998. The Excavations. Pages 11-120 in S.E. Sidebotham & W.Z. Wendrich (eds), Berenike 1996. Report of the 1996 Excavations at Berenike (Egyptian Red Sea Coast) and the Survey of the Eastern Desert. (CNWS Publications, special series n. 3). Leiden: Centre of Non-Western Studies. ----------1999. The Excavations. Pages 3-94 in S.E. Sidebotham & W.Z. Wendrich (eds), Berenike 1997. Report of the 1997 Excavations at Berenike and the Survey of the Egyptian Eastern Desert, including Excavations at Shenshef. (CNWS Publications, special series n. 4). Leiden: Centre of Non-Western Studies. ----------2000. Excavations. Pages 3-147 in S.E. Sidebotham & W.Z. Wendrich (eds), Berenike 1998. Report of the 1998 Excavations at Berenike and the Survey of the Egyptian Eastern Desert, including Excavations in Wadi Kalalat. (CNWS Publications, special series n. 5). Leiden: Centre of Non-Western Studies. ----------2003. Ptolemaic and Roman Water Resources and their Management in the Eastern Desert of Egypt. Pages 87-116 in M. Liverani (ed.), Arid Lands in Roman Times. Papers from the International Conference, Rome, July 9th-10th, 2001. (Arid Zone Archaeology Monographs 4). Firenze: all’Insegna del Giglio. Sidebotham S.E. & Zitterkopf R.E. 1996. Survey of the Hinterland. Pages 357-409 in S.E. Sidebotham & W.Z. Wendrich (eds), Berenike 1995. Preliminary Report of the 1995 Excavations at Berenike (Egyptian Red Sea Coast) and the Survey of the Eastern Desert.

Francis P., Jr. † 2000. Human Ornaments. Pages 211-225 in S.E. Sidebotham & W.Z. Wendrich (eds), Berenike 1998. Report of the 1998 Excavations at Berenike and the Survey of the Egyptian Eastern Desert, including Excavations in Wadi Kalalat. (CNWS Publications, special series n. 5). Leiden: Centre of Non-Western Studies. Groenman-van Waateringe W. 1997. Classical authors and the diet of Roman soldiers: true or false? Pages 261-265 in W. Groenman-van Waateringe, B.L. van Beek, W.J.H. Willems & S.L. Wynia (eds), Roman Frontier Studies 1995. Proceedings of the XVIth International Congress of Roman Frontier Studies (Oxbow Monograph 91). Oxford: Oxbow. Harrell J.A. 1996. Geology. Pages 99-126 in S.E. Sidebotham & W.Z. Wendrich (eds), Berenike 1995. Preliminary Report of the 1995 Excavations at Berenike (Egyptian Red Sea Coast) and the Survey of the Eastern Desert.(CNWS Publications, special series n. 1). Leiden: Centre of Non-Western Studies. ----------1998. Geology. Pages 121-148 in S.E. Sidebotham & W.Z. Wendrich (eds), Berenike 1996. Report of the 1996 Excavations at Berenike (Egyptian Red Sea Coast) and the Survey of the Eastern Desert. (CNWS Publications, special series n. 3). Leiden: Centre of Non-Western Studies. ----------(forthcoming). The Geology. In S.E. Sidebotham & W.Z. Wendrich (eds), Berenike 1999-2000. Report of the 1999 and 2000 Excavations at Berenike and Survey of the Eastern Desert, including Excavations in Wadi Kalalat and Siket. Los Angeles: Cotsen Institute of Archaeology. Hayes J.W. 1995. Summary of Pottery and Glass Finds. Pages 33-40 in S.E. Sidebotham & W.Z. Wendrich (eds), Berenike 1994. Preliminary Report of the 1994 Excavations at Berenike (Egyptian Red Sea Coast) and the Survey of the Eastern Desert (CNWS Publications, special series n. 1). Leiden: Centre of Non-Western Studies. ----------1996. The Pottery. Pages 147-178 in S.E. Sidebotham & W.Z. Wendrich (eds), Berenike 1995. Preliminary Report of the 1995 Excavations at Berenike (Egyptian Red Sea Coast) and the Survey of the Eastern Desert (CNWS Publications, special series n. 2). Leiden: Centre of Non-Western Studies. Jones R.N., Hammond, P., Johnson, D. & Fiema Z. 1988. A Second Nabataean Inscription from Tell eshShuqafiyah, Egypt. Bulletin of the American Schools of Oriental Research 269: 47-57. Keys D. 1999. Catastrophe. An Investigation into the Origins of the Modern World. New York: Ballantine Books. Littmann E. & Meredith D. 1953. Nabataean Inscriptions from Egypt. Bulletin of the Schools of Oriental and African Studies, University of London 15: 1-28. ----------1954. Nabataean Inscriptions from Egypt – II. Bulletin of the Schools of Oriental and African Studies, University of London 16: 211-246. Mahadevan I. 1996. Tamil-Brāhmi Graffito. Pages 205208 in S.E. Sidebotham & W.Z. Wendrich (eds), 114

STEVEN E. SIDEBOTHAM: REFLECTIONS OF ETHNICITY IN THE RED SEA COMMERCE IN ANTIQUITY Van Neer W. & Ervynck A.M.H. 1999. The Faunal Remains. Pages 325-348 in S.E. Sidebotham & W.Z. Wendrich (eds), Berenike 1997. Report of the 1997 Excavations at Berenike and the Survey of the Egyptian Eastern Desert, including Excavations at Shenshef. (CNWS Publications, special series n. 4). Leiden: Centre of Non-Western Studies. Vermeeren C.E. 1998. Wood and Charcoal. Pages 331348 in S.E. Sidebotham & W.Z. Wendrich (eds), Berenike 1996. Report of the 1996 Excavations at Berenike (Egyptian Red Sea Coast) and the Survey of the Eastern Desert. (CNWS Publications, special series n. 3). Leiden: Centre of Non-Western Studies. ----------1999. Wood and Charcoal. Pages 307-324 in S.E. Sidebotham & W.Z. Wendrich (eds), Berenike 1997. Report of the 1997 Excavations at Berenike and the Survey of the Egyptian Eastern Desert, including Excavations at Shenshef. (CNWS Publications, special series n. 4). Leiden: Centre of Non-Western Studies. ----------2000. Wood and Charcoal. Pages 311-342 in S.E. Sidebotham & W.Z. Wendrich (eds), Berenike 1998. Report of the 1998 Excavations at Berenike and the Survey of the Egyptian Eastern Desert, including Excavations in Wadi Kalalat. (CNWS Publications, special series n. 5). Leiden: Centre of Non-Western Studies. Whitcomb D.S. 1979. Trench Summaries. Pages 1165 in D.S. Whitcomb & J.H. Johnson (eds), Quseir al-Qadim 1978. Preliminary Report. Cairo/Princeton: American Research Centre in Egypt. Wilcken U. 1912. Grundzüge und Chrestomathie der Papyruskunde. (Erste Band, Historische Teil, Zweiter Hälfte. Chrestomathie). Leipzig-Berlin: Teubner. Wild J.P. & Wild F.C. 2000. Textiles. Pages 251-274 in S.E. Sidebotham & W.Z. Wendrich (eds), Berenike 1998. Report of the 1998 Excavations at Berenike and the Survey of the Egyptian Eastern Desert, including Excavations in Wadi Kalalat. (CNWS Publications, special series n. 5). Leiden: Centre of Non-Western Studies. ----------2001. Sails from the Roman port at Berenike, Egypt. The International Journal of Nautical Archaeology 30.2: 211-220. Young G.K. 2001. Rome’s Eastern Trade. International commerce and imperial policy, 31 BC-AD 305. London-New York: Routledge.

(CNWS Publications, special series n. 1). Leiden: Centre of Non-Western Studies. Sidebotham S.E., Wendrich W.Z. & Hense A.M. 1996. Statuary and Cult Objects. Pages 229-243 in S.E. Sidebotham & W.Z. Wendrich (eds), Berenike 1995. Preliminary Report of the 1995 Excavations at Berenike (Egyptian Red Sea Coast) and the Survey of the Eastern Desert.(CNWS Publications, special series n. 1). Leiden: Centre of Non-Western Studies. Sidebotham S.E. & Wendrich W.Z. (eds), 1995. Berenike 1994. Preliminary Report of the 1994 Excavations at Berenike (Egyptian Red Sea Coast) and the Survey of the Eastern Desert. (CNWS Publications, special series n. 1). Leiden: Centre of Non-Western Studies. ----------1996. Berenike 1995. Preliminary Report of the 1995 Excavations at Berenike (Egyptian Red Sea Coast) and the Survey of the Eastern Desert. (CNWS Publications, special series n. 2). Leiden: Centre of Non-Western Studies. ----------1998a. Berenike: Archaeological fieldwork at a Ptolemaic-Roman port on the Red Sea coast of Egypt: 1994-1998. Sahara 10: 85-96. ----------1998b. Berenike 1996. Report of the 1996 Excavations at Berenike (Egyptian Red Sea Coast) and the Survey of the Eastern Desert. (CNWS Publications, special series n. 3). Leiden: Centre of Non-Western Studies. ----------1999. Berenike 1997. Report of the 1997 Excavations at Berenike and the Survey of the Egyptian Eastern Desert, including Excavations at Shenshef. (CNWS Publications, special series n. 4). Leiden: Centre of Non-Western Studies. ----------2000. Berenike 1998. Report of the 1998 Excavations at Berenike and the Survey of the Egyptian Eastern Desert, including Excavations in Wadi Kalalat.(CNWS Publications, special series n. 5). Leiden: Centre of Non-Western Studies. ----------2001-2002. Berenike. Archaeological fieldwork at a Ptolemaic-Roman port on the Red Sea coast of Egypt 1999-2001. Sahara 13: 23-50. ----------2001. Berenike: A Ptolemaic-Roman port on the ancient maritime spice and incense route. Minerva 13/3 (May/June): 28-31. Toll C. 1994. Two Nabataean Ostraca from Egypt. Bulletin de l’Institut français de Archéologie Orientale 94: 381-382. Tomber R.S. 2000. Indo-Roman trade: the ceramic evidence from Egypt. Antiquity 74: 624-631. ----------(forthcoming). Aksumite Sherds from Berenike 1996-2000. In S.E. Sidebotham & W.Z. Wendrich (eds), Berenike 1999-2000. Los Angeles: Cotsen Institute of Archaeology. Tomber R.S. & Begley V. 2000. Indian Pottery Sherds. Pages 149-167 in S.E. Sidebotham & W.Z. Wendrich (eds), Berenike 1998. Report of the 1998 Excavations at Berenike and the Survey of the Egyptian Eastern Desert, including Excavations in Wadi Kalalat. CNWS Publications, special series n. 5). Leiden: Centre of Non-Western Studies.

115

116

Gold dinars and silver dirhams in the Red Sea trade: the evidence of the Quseir documents Li Guo The bulk of the hundreds of Arabic documents uncovered in the ancient port Quseir during the 1980s consists of letters, notes and accounts that shed rare light on commercial activities and economic life in this Red Sea community in the thirteenth century (Guo 1999; 2001). How to pay for the goods and services was naturally an integral part, and perhaps the focal point, of the whole process. It is commonly held that medieval Near Eastern commerce was primarily a credit economy in that various forms of credit, such as the famous Ήawālah system (transfer of credit, bill of exchange), the advanced pledges of property for debts (al-rahn), and the exchange of goods, were widely used in business practices and commercial transactions (Goitein 1999: 197-200; 240245).1 However, at the end of the day, payment still had to be made and discussions about money would have been inevitable.

Mecca. Various currencies must have changed hands along the way. In one particular case (RN 970a), the issue of ‘Egyptian dinars’ versus ‘Meccan dinars’ became a bone of contention and was discussed in some detail. In another letter (RN 1003c/1004d), the question of exchange rates between the gold dinars and silver dirhams circulated in different towns was a major issue. This article examines these two texts within the historical context of the Red Sea trade in Upper Egypt during the Ayyubid period. The first letter (RN 970a), published below for the first time,2 was sent by two clients, one named al-MuΞaffar and the other Abū Bakr, to Shaykh Abū Mufarrij, the owner of the warehouse known as the ‘Sheikh’s house’ (figs. 45, 46). The text as a whole reveals a complex web of business networking and social circles around the ‘Sheikh’s house’. From the content of the letter we gain considerable insights into how a commission was made and carried out; we learn that Abū Mufarrij’s duties, as an agent and middleman, included re-selling the goods his warehouse received from the senders (lines 9-11, recto), picking up his clients’ money from a third party (lines 17-19, recto), and shipping back some new items (‘to be wrapped separately’) to al-MuΞaffar and Abū Bakr (lines 19-20, recto). Cash flow seems to be a major concern: the if-you-don’t-pay-on-time-you-won’t-getthe-goodies warning is raised twice (lines 13-15, recto; lines 4-5, verso). Many issues are at stake here; but what concern us for the present purpose are the strongly worded instructions on how to handle the money:

In this respect, the Quseir documents suggest that monetary payments were made in various ways. A usual method seems to have been advance payment. One memo (RN 1070), for example, makes this point plain and simple. ‘As soon as you send me five dirhams,’ the author of the note promises, ‘I will rent [pack animals to carry] the crops (ukrī Κalā al-Υhallah) [to you].’ Another letter (RN 1003c/1004d) makes a similar point by insisting on receiving the advance payment as a precondition for dispatching the goods. On the other hand, the documents also reveal some more complicated payment arrangements insofar as multilateral partnership and network played a central role in the Red Sea shipping business at the port of Quseir, where a broker might receive the shipment on behalf of his client, re-sell the commodities, collect the money, buy new goods at the request of the client, and then pack and ship them back (Guo 2004).

Don’t accept, on our behalf, the payments unless [they are paid in] Egyptian gold dinars (dahab misrī). Cash the money (intaqid-hu) carefully: no Meccan dinars (dahab makkī). Egyptian dinars only. (lines 17-19, recto)

Then, there is the question of exchange rates. The archaeological and textual evidence indicates that the port of Quseir was essentially a transit stop on the longdistance itineraries that linked Egypt to the Red Sea and Indian Ocean trade routes; the question of various currencies and other related matters thus apparently became a major concern among the merchants and brokers. One of the most prominent sites excavated so far, the so-called ‘Sheikh’s house’ in the central area of the town, is believed to have been involved in supplying provisions for the Egyptian pilgrims on their way from Cairo and other parts of the country via the Nile, overland across Upper Egypt to the Red Sea and on to

Now what is the significance of this? What exactly are the differences between ‘Egyptian‘ money and Meccan‘ money in terms of value, purchasing power, and credibility? Why was the latter vehemently rejected in Quseir? This particular indication of preferences for, and prejudices against, different types of dinars then circulating in the Red Sea commercial network has raised questions the answers to which are speculative, given the current state of our knowledge. S.D. Goitein, in his study 2

1

Other texts discussed in this article are published in Guo 2004, chapter 5.

Also cf. Udovitch 1972, especially 77-86, 261.

117

TRADE AND TRAVEL IN THE RED SEA REGION still being used, as is seen, for example, in another letter (RN 969). Here, the sender, apparently a merchant specialised in textiles and clothing, demands that the recipient (whose identity is unknown; perhaps a member of the Abū Mufarrij clan?) sell ‘the turbans’ on his behalf. He then switches gear and informs the recipient:

of the Cairo Geniza papers, has held that the ‘Egyptian’ dinars had a higher value in comparison with the ‘foreign’ currency (Goitein 1999: 234, 237, 239-240, 359-360).3 Since the Cairo Geniza was roughly contemporary to the period when Quseir as a Red Sea port was active (Whitcomb & Johnson 1979; 1982a; 1982b), and given the overall monetary situation in Ayyubid Egypt, we would assume that the ‘Egyptian’ dinars circulating at Quseir came from the same source and usually had a higher value vis-à-vis that of the ‘foreign’ currencies. As for the ‘Meccan’ dinars, or Hijazi dinars, one theory has it that they were perhaps being struck at new mints supplied by a relatively new source of gold in East Africa (Zimbabwe, Kilwa, etc.) that was perhaps less ‘pure’ and therefore of lower value than that of the miΒrī, or Egyptian dinar; miΒrī, according to Goitein, specifically means ‘minted in Fustat’ (Goitein 1999: 234).4 This is, of course, just one interpretation.5

NuΚmān has sold coat-tails (al-dayl) for twenty-six dirhams, and he still has eight [coat-tails?] left [unsold], by God Almighty! As for me, all [the items] in my [storehouse] are sold out (inqadā), and we need [more] children’s clothes. We have pure gold (dahab aΒfar), which we will use [to pay for them], by God’s blessing.

It is unclear whether by ‘pure gold,’ literally ‘yellowish gold,’ he meant gold proper, or golden dinars; but the latter is perhaps more likely, as our next text (RN 1003c/1004d) sheds more light on the matter. In this lengthy letter to Shaykh Najīb, a senior associate of Shaykh Abū Mufarrij, the urgent demand for cash to pay for the rental pack animals to carry the merchandise is raised repeatedly:

Another interpretation of the preference for the Egyptian over the Hijazi dinar lies in the notion of uniform investments in a joint partnership. This concept had been stipulated by Muslim legal writers, especially the ShāfīΚīs, in the early Islamic centuries and is confirmed in practice by later documents, such as the Cairo Geniza papers. (Udovitch 1970: 32-33, 177-180).6 The insistence on, or the preference for, a similar type of coinage as capital in joint investments was very commonplace in business practices. The uniform coinage also ought to be the most desirable and current in the marketplaces. In addition, the writers of the Cairo Geniza letters made it clear that they wished to receive ‘exactly the type of money one paid to the issuing banker’, to avoid losses over time due to the devaluation of the currency (Goitein 1999: 242). Whatever the case, there were certainly enough incentives to convince Shaykh Abū Mufarrij and his clients to stick to ‘Egyptian dinars’.

I need nothing but cash (al-darāhim), because I intend to use them to rent [pack animals]. . . . Do not pay me golden dinars (lā taΒrifu lī dahab); change them (ištarihā, literally ‘buy them’) to silver dirhams (darāhim). The exchange rate (al-Βarf) in Qina and Qus is thirtyseven [dirhams per dinar], and [if it is] the Yūsufī (? or Tawfīqī?)8 [dirham], then it is nineteen and a quarter dirhams [per dinar]. O God, O God! Send me the cash in silver dirhams [only].

Here, the exchange rate of gold dinars (dahab) versus silver dirhams, with its probable local variants (Yūsufī? Tawfīqī?),9 is in keeping with Goitein’s observation of a Cairo Geniza ‘writer’s assumption that the provincial town has an exchange rate by itself’ (Goitein 1999: 378). The letter shows that the nearby towns Qus and Qina did in fact have their own exchange rate, which was perhaps different from that of Quseir, and Cairo, where the ‘standard’ rate was 1: 35-40, with 1: 34 hitting the ‘bottom’ (Goitein 1999: 390-391).10 The rate of 1: 19 certainly strikes one as off the scale, but without a clearer notion of the currency in question (Yūsufī? Tawfīqī?), the question is perhaps better left open until a more accurate reading of the term can be achieved and the term identified. Speaking of money exchange, the activities of the so-called al-Βayrafī, ‘money changer,’ are also recorded in some Quseir texts. Gold and silver, after all, were themselves commodities, to be freely traded everywhere in Islamic Near East (Goitein 1999: 229-66), and the documents examined above add to that history with some regional details.

One more intriguing thing about the reference to dahab, ‘gold,’ in the aforesaid text is the fact that in later Ayyubid times, gold had become scarce (Goitein 1999: 234);7 the currency dealt with most commonly in Quseir was, as everywhere else, silver dirhams. Although this question may best be left for specialists to debate, the documents do seem to indicate that gold currency was 3

On exchange rate of the ‘good old Misri dinar’ versus Tripoli, Lebanon, Damascus, and other ‘Mediterranean‘ dinars in the twelfth century; another case of the devaluation of the Maghribi dirhams in relation to Egyptian ones is discussed in Goitein 1999: 390. 4 The idea was discussed with John Sutton on October 7, 2002. For details about the gold mining industry in Kilwa, see Sutton 1997; 2001. I thank Dr. Sutton for sending me the articles and sharing with me his expertise on the subject matter. 5 A historical survey of the currency in pre-modern Mecca is found in Mortel 1989. Although Mortel’s article deals mainly with the Mamluk period, some data may prove to be relevant to the late Ayyubid and perhaps early Mamluk times under discussion. 6 Based on the MabsūΓ by the Дanafi al-Sarākhsī (d. 1090) and the Minhāj al-Γālibīn by the ShāfiΚī al-Nawāwī (d. 1277). 7 For the monetary policy under the Ayyubid sultan al-Kāmil after 622/1225 when copper fulūs attained the status of official currency in addition to silver dirham, see Allouche 1994: 68-69; Rabie 1972: 182.

8

The word is unpointed in the manuscript; I have been unable to reach a satisfactory reading. The usual practice was to name a coin after the ruler; I have been unable to identify any ruler with the name, or title, of Yūsuf or Tawfīq. 10 For the time period in question, the sources usually refer to the Kāmilī dirham, that is, after the Ayyubid sultan al-Mālik al-Kāmil (d. 635/1238), as the official currency circulated in Cairo. 9

118

LI GUO: GOLD DINARS AND SILVER DIRHAMS IN THE RED SEA TRADE

Figure 45: RN 970a, recto.

Figure 46: RN 970b, verso.

Photos courtesy of Janet Johnson and Donald Whitcomb, the Oriental Institute, Chicago.

119

TRADE AND TRAVEL IN THE RED SEA REGION Appendix: The Text11 Description: RN 970a, light brown paper, badly torn in the upper part, large holes, 8.2 x 24.6 cm, recto twenty lines, verso eight lines on both ends of the paper, in naive but clear hand, with generously supplied diacritics, black ink. Edition:

11

[
conjectural additions by the editor } deletions by the editor

120

LI GUO: GOLD DINARS AND SILVER DIRHAMS IN THE RED SEA TRADE

Translation:

Commentary:

recto:

recto:

1-3. In the name of God, the Merciful, the Compassionate. His servants al-Muzaffar and Abū Bakr 4. [write to inform] Shaykh Abū Mufarrij—may God prolong 5. his life and make his strength and happiness everlasting!— 6. [. . . . . . . . that] I have dispatched to you 7. [ . . . . . . . . . . . . . . in] small loads 8. [ . . . . . . . a total of] thirty; [and that] every thing shipped to him (i.e.to you) is of good quality. 9. By God, by God! as soon as the aforesaid [shipment] arrives, 10. don’t let it sit idle, not a second! 11. Sell it, as God so feeds and provides. 12. We would like to rent a large cargo ship 13. and dispatch it to [you] the Master. [But] we will not send it out 14. unless you would help us out on account of the aforesaid [shipment]. By God, 15. by God! All you need to do is send me the down payments, 16. to be accompanied by KhiΡr the young porter. 17. Don’t accept, on our behalf, the payments unless [they are paid in] Egyptian gold [dinars]. 18. Cash the money carefully: no Meccan 19. gold [dinars]. Egyptian dinars only. Wrap 20. everything separately and hand it all to KhiΡr.

6. The first part of the line is damaged by worm-eaten holes. In light of the formulaic phraseology, the missing part should likely be wa-siwā dalika, a phrase to end the greeting segment and introduce the main content. 8. The first part of the line is missing and the reference to the suffix ‘him’ (ilay-hi) is perhaps an unidentified third party. 9. WuΒūl la-ka is odd (wuΒūl-ka is better), but the second l is visible in the manuscript. 13. NaΓlubu nukrī . . . nusayyiru, since all the verbs are unpointed, there remains a possibility that the reading could be naΓlubu tukrī . . . nusayyiru, i.e., we request that you, Abu Mufarrij, rent a cargo ship and dispatch it to the Master, a third party. 18. Intaqid, ‘cash the money,’ could also mean ‘examine it,’ as in, ‘detect defects in it!’ which also fits the context here. verso: 2-3. The mawlā here is obviously Abu Mufarrij. Note the shift of the subject from the third person (iΉtāja) to second (taktubu). 5. mil [Μ] al-taman, ‘paid in full’; the statement implies that the money mentioned on the recto was perhaps some sort of down payments, and that the ‘full’ amount was due after the delivery had been completed.

verso Bibliography 1. By God, by God! Any thing you want, let [me] know. 2. Whatever [you,] the Master need, 3. write me a memo [and send it] through the porters; 4. I will ship off your orders. Upon the delivery 5. of the crops as ordered hereby, you should send us the full payment. 6. Peace be upon you. God’s mercy and blessings.

Allouche A. 1994. Mamluk Economics: A study and translation of al-Maqrizi’s Ighathah. Salt Lake City: University of Utah Press. Goitein S.D. 1999. A Mediterranean Society: The Jewish Communities of the World as Portrayed in the Documents of the Cairo Geniza. (5 volumes). (First paperback edition. Berkeley, Los Angeles). Berkeley; London: University of California Press. Guo L. 1999. Arabic Documents from the Red Sea Port of Quseir in the Seventh/Thirteenth Century, Part 1: Business Letters. Journal of Near Eastern Studies 58/3: 161-190. ---------2001. Arabic Documents from the Red Sea Port of Quseir in the Seventh/Thirteenth Century, Part 2: Shipping Notes and Account Records. Journal of Near Eastern Studies 60/2: 81-116.

address a. The service was commissioned [ . . . . . . . . . . ] b. Shaykh Abū Mufarrij to [ . . . . . . . . . . . . . . . . ]

121

TRADE AND TRAVEL IN THE RED SEA REGION ---------2004. Commerce, Culture, and Community in a Red Sea Port in the Thirteenth Century: The Arabic Documents from Quseir. Leiden: E.J. Brill. Mortel R. 1989. Prices in Mecca during the Mamluk Period. Journal of the Economic and Social History of the Orient 32: 299-302. Rabie H. 1972. The Financial System of Egypt, A.H. 564741/A.D. 1169-1341. London: Oxford University Press. Sutton J.E.G. 2001. Kilwa in the early fourteenth century: Gold trade, monumental architecture and Sunni conformity at the southern extremity of dar al-Islam. Pages 425-439 in B.S. Amoretti (ed.), Islam in East Africa: New Sources (Archives, Manuscripts and Written Historical Sources. Oral History. Archaeology). Rome: Herder. ---------1997. The African Lords of the Intercontinental Gold Trade before the Black Death: al-Hasan bin Sulaiman of Kilwa and Mansa Musa of Mali. The Antiquaries Journal 77: 221-242. Udovitch A. 1970. Partnership and Profit in Medieval Islam. Princeton: Princeton University Press. Whitcomb D.S. & Johnson J.H. (eds) 1979. Quseir alQadim 1978: Preliminary Report. (American Research Center in Egypt Reports 1). Cairo/Princeton: American Research Center in Egypt. -----------1982a. Quseir al-Qadim 1980: Preliminary Report. (American Research Center in Egypt Reports 7). Malibu, California: Undena. -----------1982b. 1982 Season of Excavations at Quseir alQadim. American Research Center in Egypt Newsletter 120: 24-30.

122

The merchants’ diet: food remains from Roman and medieval Quseir al-Qadim Marijke Van der Veen •

Introduction Recent excavations of the ancient port at Quseir alQadim on the Red Sea coast of Egypt have revealed botanical evidence for the Roman and medieval spice trade between Egypt and India, as well as a wealth of information (both botanical and faunal remains) for the diet of the inhabitants of this port. Spices from the Far East have long represented a significant and lucrative commodity. The profits of the pepper trade, as well as cloves, cinnamon, cardamom and nutmeg, made Venice a key player in world trade during the fifteenth and sixteenth centuries, yet little is known about the earlier history of this trade, other than from records in classical and early Islamic sources. However, we know that the first introduction of pepper into the Mediterranean world dates to the Roman period, when ports on the Red Sea coast of Egypt like Berenike and Myos Hormos (now called Quseir al-Qadim) were used to transship spices and other goods obtained from India.1 Quseir also flourished during the medieval period when the merchants known as the ‘Kārimī’ used Quseir (as well as ΚAydhāb further south) as a port from which they organised their trade with southern Arabia and India (Fischel 1958). Documentary sources refer to the Kārimī as ‘the merchants of pepper and spices’, but we know that they also dealt with other commodities, such as wheat, rice, sugar, silk, textiles and wood (Fischel 1958). Both the Roman and medieval sources, given their nature, are largely silent, however, on the inhabitants and living conditions at the port.

• •

the reconstruction of the food supply to the inhabitants of this port which, located in the coastal desert of the Red Sea, was dependent for all its supplies on food brought in from the Nile valley (in antiquity a journey of c. seven days by donkey and camel) or further afield; the recovery and identification of plants which may represent the trade with India, and possibly with Arabia, Yemen, China and the Moluccas; the exploration of changes through time in either the trade or the food supply, and of the interrelationship between diet and culture.

The analysis of the food remains is in an early stage, and this paper will set out the potential of the data sets obtained, rather than offer definitive answers. Quseir al-Qadim The site is located some 8 km north of the present town of Quseir, on the Red Sea coast of Egypt (Figs 47 and 48).

During the spring of 1999 a five-year project of survey and excavations was initiated at Myos Hormos/Quseir alQadim (Peacock et al. 1999, 2000, 2001, 2002) with permission of the Supreme Council for Antiquities of Egypt and with support from the Peder Sager Wallenberg Charitable Trust. The project aims to explore the origin and development of the port, establish the street plan and layout of the settlement, and identify the location of the harbour. Moreover, it will investigate the economy of the town and establish what goods were traded. The results will complement those from the current excavations at Berenike.2 During the excavations food remains have been recovered in abundance, and their study has three main objectives:

The Roman and Islamic remains at Quseir al-Qadim (Old Quseir) have been known for a long time and were previously identified with the ancient port of Leucos Limen.3 Recent research has demonstrated that the remains do, in fact, belong to Myos Hormos, known from classical sources as one of the two major ports for the sea trade with India during the Roman period.4 Excavations carried out between 1978 and 1982 by an American team under the aegis of the American Research Center in Egypt (Whitcomb & Johnson 1979: 1982) established that the Roman occupation of the site was dated to the mid-first and second centuries AD, while the Islamic occupation spanned the eleventh to fifteenth centuries AD. Archaeobotanical work carried out by Wetterstrom (1982) established that the deposits at this site were rich in plant remains, and she recorded the first evidence for the diet of the inhabitants, as well as two imports from India, pepper and coconut, both dated to the Islamic occupation. Wattenmaker (1979, 1982) published preliminary reports on the faunal remains and recorded a heavy reliance on marine resources (especially fish), and the importance of sheep and goat, with camel, cattle and birds playing a supplementary role and pig present in the

1

3

2

4

Begley & De Puma 1991; Casson 1989; Dalby 2000; Miller 1969. e.g. Cappers 1999a, 1999b; Sidebotham & Wendrich eds 1998; Van Neer 1997; Van Neer & Lentacker 1996.

123

e.g. Bowman 1990; Whitcomb & Johnson 1979, 1982. The other being Berenike; Peacock 1993; Bülow-Jacobsen et al. 1994.

TRADE AND TRAVEL IN THE RED SEA REGION

Figure 47: Quseir and the Indian Ocean. Roman but not the Islamic period. Full reports on the excavations or the archaeobotanical and faunal work were never published.

Preliminary Results Samples of plant macrofossils, animal and fish bones have been collected from all well-dated deposits. The majority of the plant foods (grains, seeds, fruit stones, nut shells, herbs and spices) are preserved in desiccated form, although some carbonized remains are also present. Preservation is generally excellent, although some of the remains (faunal and botanical) are affected by salt damage. To date, four seasons of excavation have taken place and a vast data set has been recovered. Both the botanical and faunal assemblages measure well over 20,000 specimens each,5 and much remains to be done; the results given below are thus preliminary.

The coastal desert of the Red Sea forms the immediate environment of the port. Plant growth here is extremely restricted (annual rainfall is c. 4 mm) and consists of a limited number of desert plants, mostly xerophytes (plants able to grow in very dry conditions) and halophytes (plants adapted to saline conditions). Agriculture is not feasible in the area today, nor was it in the Roman or medieval periods; the region is frequented today only by nomads as it would have been in the past. The present town of Quseir relies on most of its water and food being brought in from the Nile Valley, with fresh fruit and vegetables coming from places such as IsmaΜiliya on the Suez Canal, and the Nile Delta. Similarly, the ancient port must have relied on the import of food, probably primarily from the Nile valley, but which foods were they? The extreme aridity of the Red Sea coastal plain has ensured the excellent preservation of food remains and kitchen refuse (seeds, fruits, nuts, bones of domestic animals as well as fish) discarded by the inhabitants during the Roman (early first to early third centuries AD) and medieval periods (Ayyubid and Mamluk periods, late eleventh to fifteenth centuries AD). For the first time we can now start to address questions such as: what did they eat, how varied was their diet and did they have access to the exotic foods that were transshipped at the port?

‘Local’ plant foods Hulled barley (Hordeum vulgare) and hard wheat (Triticum durum) are the two principal grain crops recovered. Grains and chaff fragments (primarily rachis segments) are present in virtually every sample (Roman and medieval). These two cereals represent the main staple crops in both periods, barley having been grown in Egypt since the pre-dynastic period, while hard wheat replaced emmer wheat as the principal wheat crop during the Ptolemaic period, contra Watson (1983 and repeated in Insoll 1999) who claimed that hard or durum wheat

5

Van der Veen 1999, 2000, 2001; Hamilton Dyer 1999, 2000, 2001, 2002.

124

MARIJKE VAN DER VEEN: THE MERCHANTS’ DIET

Figure 48: The location of Quseir al-Qadim and other sites mentioned in the text. the herbs coriander (Coriandrum sativum) and fennel (Foeniculum vulgare). Nuts such as hazelnuts (Corylus avellana; Fig. 50) and walnuts (Juglans regia) are found in both periods; pine kernels (Pinus pinea) more in the Roman period and pistachio (Pistacia vera) more in the medieval period (Fig. 51).

was an Islamic introduction. Spikelets of emmer wheat (Triticum dicoccum) have been found in one medieval deposit, while sorghum (Sorghum bicolor) was found in several medieval deposits (but in restricted areas only). Sorghum originates from Africa south of the Sahara. The timing of its domestication and widespread cultivation in Africa is still a matter of some debate (Rowley-Conwy et al. 1998, 1999), but it may have become a fodder crop of some importance in Egypt during the early Islamic period. A number of pulse crops have been found: lentils (Lens culinaris), fava beans (Vicia faba, var. minor), chickpeas (Cicer arietinum), and termis beans (Lupinus albus). Lentils are particularly common in the Roman period and fava beans in the medieval period. Fruits found in both periods include dates (Phoenix dactylifera), watermelon (Citrullus lanatus), grapes (Vitis vinifera) and cucumber (Cucumis sativus). Olives (Olea europea) and sebesten (Cordia myxa) are more common in the Roman period, carob (Ceratonea siliqua) in the medieval period. Vegetables include artichoke (Cynara cf. scolymus; Fig. 49), although only found in deposits dating to the Roman period), onion (Allium cepa), garlic (Allium sativa), and

Figure 49: Artichoke bracts (mid second century AD). 125

TRADE AND TRAVEL IN THE RED SEA REGION

Figure 50: Hazelnut shells (Mamluk period). Figure 52: Parrot fish (Mamluk period; photo Sheila Hamilton-Dyer and Wilfried Van Rengen).

Figure 53: Chicken foot and feathers (Roman period; photo Sheila Hamilton-Dyer).

Figure 51: Pistachio shells (Mamluk period). Animal protein

‘Foreign’ foods

Marine fish and the meat of sheep (Ovis aries) and goat (Capra hiscus) were the three most important sources of animal protein. Bones of these species were abundantly present in the refuse deposits of both periods. Most of the fish comes from the Red Sea: parrot fish (Scaridae; Fig. 52), emperors (Lethrinidae), groupers (Serranidae), jacks (Carangidae), trigger fish (Balistidae) and seabreams (Sparidae), but some bones of a small Nile river fish have also been recovered. After sheep and goat, camel (Camelus dromedarius) was important too as a source of meat, while cattle (Bos sp.), donkey (Equus asinus), and horse (Equus caballus) must have made much smaller contributions. Remains of chicken (Gallus domesticus) were also found (egg shell, bones, feathers and even a foot; Fig. 53), as were marine shells (in large quantities), including giant clam (Tridacna) and oysters (Saccostrea sp.). The bones of pig (Sus scrofa) are quite frequent in the Roman deposits, but absent in the medieval (Islamic) period.

Several imports have already been recovered. Pepper (Piper nigrum) has been found in both Roman and medieval deposits; the earliest finds so far date to the early first century AD (Fig. 54). Pepper originates in south-west India. The plant is a climbing vine that can grow to more than four metres. The small fruits are borne in long hanging spikes. Black pepper is made from unripe peppercorns that have been dried in the sun which causes the black and wrinkled surface structure (Vaughan & Geissler 1998) whereas white pepper is made by removing the outer covering of the ripe fruits. The peppercorns at Quseir represent black pepper. Peppercorns formed an important and profitable article of commerce in Rome‘s sea trade with India and special spice warehouses (horrea piperataria) were constructed and maintained in Rome. Pepper was used chiefly as a culinary spice and quickly became an essential part of the everyday life of respectable households in Rome (Warmington 1928: 180-183). It maintained this status 126

MARIJKE VAN DER VEEN: THE MERCHANTS’ DIET Coconut (Cocos nucifera) is probably native to south-east Asia (Simmonds 1976). Like pepper and rice, it has been found in deposits dating to the early first century AD, as well as in second century and medieval layers. The remains of coconut consist of fragments of the epicarp and fibrous husk (Fig. 56) and of the shell (Fig. 57). In all cases the endosperm, the white coconut ‘meat’, had been removed. The coconut fibre, ‘coir’, was and is widely used in the production of ropes and matting.

throughout the medieval and post-medieval periods and represented one of the main sources of wealth for the Kārimī merchants (Fischel 1958).

Figure 56: Epicarp and fibrous husk of coconut imported from India (mid second century AD). Figure 54: Peppercorns imported from India (early first century AD). Rice (Oryza sativa) has been found in small quantities in both Roman and medieval deposits (Fig. 55). It almost certainly originates in China and/or south-east Asia and was first domesticated there. During the Roman period it was one of the food plants imported from India. It still needs to be established at what point in time it became a crop cultivated in Egypt, but documentary evidence suggests it was grown in the Fayyum by Mamluk times (Canard 1959).

Figure 57: Shell of coconut imported from India (mid second century AD). Finally, aubergine (Solanum melongena) has been found in medieval contexts. It is a native of South Asia and was first cultivated in India (Vaughan and Geissler 1998). The remains of aubergine at Quseir do not consist of the fruit, or the seeds, but the calyx (the outer whorl of united sepals; Fig. 58). These are small (c. 30 mm wide), suggesting a small egg-sized fruit (hence possibly the other common name: egg-plant). Like rice, it remains to be established exactly when the status of this crop changed from import to locally grown crop in Egypt. Figure 55: Rice kernels imported from India (early first century AD).

127

TRADE AND TRAVEL IN THE RED SEA REGION Islamic periods are known to be periods of major agricultural innovation, with the introduction of both new species and new agricultural techniques such as water management systems. However, to what extent did the products of such innovations reach a site like Quseir alQadim? What new information can the food remains from this site add to what we know from historical sources? How important are summer crops, which allow an additional season of agricultural production? The introduction of summer crops is widely attributed to the early Islamic period (e.g. Watson 1983), but both Rowley-Conwy (1989) and Samuel (2001) have published archaeobotanical evidence suggesting that summer cultivation of some field crops started earlier, and may have been introduced gradually, rather than as part of the unification of the Near East and North Africa by Islam. Other questions concern the diet in the two periods. Are the observed differences economically or culturally determined? Does the Mamluk assemblage contain a greater range of Far Eastern [south Asian?] exotics than the Roman assemblage (i.e. did the Mamluks obtain foods from further afield)? Can the food remains from this site contribute to our understanding of the functioning of the state in these two periods? The functioning of the port must have been heavily dependent on the central organisation of the state and that of the nearest urban centre, Coptos in the Roman period and Qus in the Mamluk period. Judging by these preliminary results, we can look forward to the answers to at least some of these questions.

Figure 58: Aubergine calyces (Mamluk period). Discussion These first results reveal a fascinating glimpse of the merchant’s diet at Quseir al-Qadim. While much work remains to be done, we can already answer the first few questions posed. In terms of what they ate and how varied their diet was, we can say that in both Roman and medieval periods the inhabitants of the port ate a wide variety of different foods. A number of staples provided the much needed carbohydrates (wheat, barley, dates, pulses), and they had several sources of protein (fish, meat, pulses and nuts). Fruits and vegetables supplied additional vitamins and minerals.

Acknowledgements

In both periods the foods comprised both day-to-day staples, as well as foods eaten for enjoyment and at celebrations, e.g. meat, nuts, exotics, certain fruits. The presence of expensive food items, such as nuts, and the imported pepper, coconut and rice, point to a considerable degree of wealth of at least some of the inhabitants. The occurrence of pepper, coconut and rice suggests that these imports were not just trans-shipped at the port, but were also available for consumption in Quseir itself (a similar pattern has been found in Berenike; Cappers 1999a). Detailed analysis of the chronological and spatial distribution of these exotic and expensive items may help identify social differences among the inhabitants and help assess differences in status of the inhabitants between the two periods. There is already some evidence for chronological differences, such as pig present in the Roman but not medieval (Islamic) period, and changes in the predominance of certain fruits and nuts, though caution is necessary at this stage as only part of the assemblages has yet been analysed.

I would like to thank Professor Wilfried Van Rengen for his help with the photography, Sheila Hamilton-Dyer for the use of Figs 52 and 53, Chris Hart for editing the photographs for publication, and Debbie Miles-Williams for Figs 47 and 48. References Begley V. & De Puma, R.D. 1991. Rome and India. The Ancient Sea Trade. Madison, Wisconsin: University of Wisconsin Press. Bowman A. 1990. Egypt after the Pharaohs. Oxford: Oxford University Press. Bülow-Jacobsen A., Cuvigny H. & Fournet J.L. 1994. The identification of Myos Hormos: new papyrological evidence. Bulletin de l’Institut Français d’Archéologie Orientale 94: 27-42. Canard M. 1959. Le riz dans le Proche Orient aux premiers siècles de l’Islam. Arabica 6: 113-131. Cappers R.T.J. 1999a. Trade and subsistence at the Roman port of Berenike, Red Sea coast of Egypt. Pages 185-197 in M. Van der Veen (ed.), The Exploitation of Plant Resources in Ancient Africa (Proceedings of the 2nd International Workshop on Archaeobotany in Northern Africa, held June 23-25,

Once the full data set has been identified and quantified, we can address a range of further, wider issues concerning agricultural changes and the interrelationship between food and culture. Both the Roman and early 128

MARIJKE VAN DER VEEN: THE MERCHANTS’ DIET Periplus/L. Casson (transl.) 1989. The Periplus Maris Erythraei. Text with Introduction, Translation, and Commentary. Princeton: Princeton University Press. Rowley-Conwy P. 1989. Nubia AD 0-550 and the “Islamic” agricultural revolution: preliminary botanical evidence from Qasr Ibrim, Egyptian Nubia. Archéologie du Nil Moyen 3: 131-138. Rowley-Conwy P., Deakin W. & Shaw C. 1998. Ancient DNA from archaeological sorghum (Sorghum bicolor) from Qasr Ibrim, Nubia: implications for domestication and evolution and a review of the archaeological evidence. Sahara 9: 23-34. Rowley-Conwy P., Deakin W. & Shaw C.H. 1999. Ancient DNA from Quasr Ibrim, Egyptian Nubia. Pages 55-61 in M. Van der Veen (ed.), The Exploitation of Plant Resources in Ancient Africa (Proceedings of the 2nd International Workshop on Archaeobotany in Northern Africa, held June 23-25, 1997, in Leicester, United Kingdom). New York; London: Kluwer Academic; Plenum Publishers. Samuel D. 2001. Medieval irrigation agriculture in the Syrian Middle Euphrates valley: an archaeobotanical case study. Pages 347-438 in S. Berthier (ed.), Peuplement Rural et Aménagements Hydroagricoles dans la Moyenne Vallée de l’Euphrate Fin VIIe-XIXe Siècle. Damas: Institut Français de Damas. Sidebotham S.E. & Wendrich W.Z. (eds) 1998. Berenike 1996. Report of the 1996 Excavations at Berenike (Egyptian Red Sea Coast) and the Survey of the Eastern Desert. (CNWS Publications, special series, n. 3). Leiden: Centre of Non-Western Studies. Simmonds N.W. 1976. Evolution of Crop Plants. London: Longman. Van Neer W. 1997. Archaeological data on the food provisioning of Roman settlements in the Eastern Desert of Egypt. Archaeozoologia 9, 137-154. Van Neer W. & Lentacker, A. 1996. The faunal remains. Pages 337-355 in S.E. Sidebotham & W.Z. Wendrich (eds). Berenike 1995. Preliminary Report of the 1995 Excavations at Berenike (Egyptian Red Sea Coast) and the Survey of the Eastern Desert (CNWS Publications, special series, n. 2). Leiden: Centre of Non-Western Studies. Vaughan J.G. & Geissler, C.A. 1998. The New Oxford Book of Food Plants. Oxford: Oxford University Press. Van der Veen M. 1999. The plant remains: food supply, diet and trade. Pages pp. 40-43 in D. Peacock, L. Blue, N. Bradford & S. Moser (eds), Myos Hormos Quseir al-Qadim: a Roman and Islamic Port/Trade Site on the Red Sea Coast of Egypt. (Interim Report 1999). Southampton: University of Southampton. -----------2000. The plant remains: Roman imports from India. Pages 59-60 in D. Peacock, L. Blue, N. Bradford, & S. Moser (eds), Myos Hormos - Quseir al-Qadim: a Roman and Islamic Port Site on the Red Sea Coast of Egypt. (Interim Report 2000). Southampton: University of Southampton. -----------2001. The plant remains: diet and fuel in Roman Myos Hormos. Pages 59-61 in D. Peacock, L. Blue,

1997, in Leicester, United Kingdom). New York; London: Kluwer Academic; Plenum Publishers. -----------1999b. Archaeobotanical evidence of Roman trade with India. Pages 51-69 in H.P. Ray (ed.), Archaeology of Seafaring: the Indian Ocean in the ancient period (Indian Council of Historical Research 1). Delhi: Pragati Publications. Dalby A. 2000. Dangerous Tastes: the Story of Spices. London: British Museum. Fischel, W. 1958. The Spice Trade in Mamluk Egypt. Journal of the Social and Economic History of the Orient, 1: 157-174. Hamilton-Dyer S. 1999. Faunal Remains. Pages 38-39 in D. Peacock, L. Blue, N. Bradford & S. Moser (eds), Myos Hormos - Quseir al-Qadim: a Roman and Islamic Port/Trade Site on the Red Sea Coast of Egypt. (Interim Report 1999). Southampton: University of Southampton. -----------2000. Faunal remains: update. Page 57 in D. Peacock, L. Blue, N. Bradford, & S. Moser (eds), Myos Hormos - Quseir al-Qadim: a Roman and Islamic Port Site on the Red Sea Coast of Egypt. (Interim Report 2000). Southampton: University of Southampton. -----------2001. Faunal remains. Page 57 in D. Peacock, L. Blue, N. Bradford, & S. Moser (eds), Myos Hormos - Quseir al-Qadim: a Roman and Islamic Port Site on the Red Sea Coast of Egypt. (Interim Report 2001). Southampton: University of Southampton. -----------2002. Faunal remains: update. Pages 67-71 in D. Peacock, L. Blue, N. Bradford, & S. Moser (eds), Myos Hormos - Quseir al-Qadim: a Roman and Islamic Port Site on the Red Sea Coast of Egypt. (Interim Report 2002). Southampton: University of Southampton. Insoll T. 1999. The Archaeology of Islam. Oxford: Blackwell. Miller J.I. 1969. The Spice Trade of the Roman Empire, 29 B.C. to A.D. 641. Oxford: Clarendon Press. Peacock D.P.S. 1993. The site of Myos Hormos: a view from space. Journal of Roman Archaeology 6: 226232. Peacock D., Blue L., Bradford N. & Moser S. (eds) 1999. Myos Hormos - Quseir al-Qadim: a Roman and Islamic Port/Trade Site on the Red Sea Coast of Egypt. (Interim Report 1999). Southampton: University of Southampton. -----------2000. Myos Hormos - Quseir al-Qadim: a Roman and Islamic Port Site on the Red Sea Coast of Egypt. (Interim Report 2000). Southampton: University of Southampton. -----------2001. Myos Hormos - Quseir al-Qadim: a Roman and Islamic Port Site on the Red Sea Coast of Egypt. (Interim Report 2001). Southampton: University of Southampton. -----------2002. Myos Hormos - Quseir al-Qadim: a Roman and Islamic Port Site on the Red Sea Coast of Egypt. (Interim Report 2002). Southampton: University of Southampton. 129

TRADE AND TRAVEL IN THE RED SEA REGION N. Bradford, & S. Moser (eds), Myos Hormos Quseir al-Qadim: a Roman and Islamic Port Site on the Red Sea Coast of Egypt. (Interim Report 2001). Southampton: University of Southampton. Warmington, E.H. 1928. The Commerce between the Roman Empire and India. Cambridge: Cambridge University Press [Reprint, 1995, New Delhi: Munshiram Manoharial Publishers]. Watson A.M. 1983. Agricultural Innovation in the Early Islamic World. Cambridge: Cambridge University Press. Wattenmaker P. 1979. Flora and fauna. Pages 250-252 in D.S. Whitcomb & J.H. Johnson (eds), Quseir alQadim 1978. Preliminary Report. (American Research Center in Egypt Reports 1). Cairo/Princeton: American Research Center in Egypt. -----------1982. Fauna. Pages 347-353 in D.S. Whitcomb & J.H. Johnson (eds), Quseir al-Qadim 1980. Preliminary Report. (American Research Center in Egypt Reports 7). Malibu, California: Undena. Wetterstrom W. 1982. Plant remains. Pages 355-377 in D.S. Whitcomb & J. H. Johnson (eds), Quseir alQadim 1980. Preliminary Report. (American Research Center in Egypt Reports 7). Malibu, California: Undena. Whitcomb D.S. & Johnson J.H. (eds) 1979. Quseir alQadim 1978. Preliminary Report. (American Research Center in Egypt Reports 1). Cairo/Princeton: American Research Center in Egypt. -----------1982. Quseir al-Qadim 1980. Preliminary Report. (American Research Center in Egypt Reports 7). Malibu, California: Undena.

130

‘What the devil are you doing here?’ Arabic sources for the arrival of the Portuguese in the Red Sea and Indian Ocean Paul Lunde did you get here?’ Yet contemporary Arabic sources for the most part share the Tunisian’s lack of curiosity as to how the Portuguese succeeded in reaching India by setting sail from Portugal.

Introduction ‘What the devil are you doing here?’ These were the first intelligible words the convict João Nunes heard when he set foot in Calicut on May 16th, 1498.1 Vasco da Gama had sent him ashore in one of the native craft that rowed out to investigate the three unusual ships that had suddenly appeared in the roadstead. It was Portuguese policy to employ expendable crewmembers for potentially dangerous first contacts, and indeed their ships normally carried a number of convicted criminals for just this purpose. Realising from his outlandish garb, the rig of his ship and his incomprehensible speech that their visitor must have come from somewhere unusual, the locals took him to two widely travelled Muslim merchants residing in Calicut. These men were from Tunis, and spoke, in addition of course to Arabic, Castilian and Genoese. One of them, called Monçaide in our source, later accompanied Vasco da Gama back to Portugal.2 The presence of Maghribis in Calicut is no more to be wondered at than their linguistic ability; the extraordinary mobility of Muslim merchants, religious scholars and condottiere is well documented.

This is all the more surprising in view of the extensive geographical literature of Islamic lands, a literature that includes several well-known speculative theories on the possibility of circumnavigating Africa. As early as 1029, al-Birūnī, struck by a story he read in the tenth century historian al-MasΚūdī, of the finding of a carved plank from a characteristic Indian Ocean vessel floating in the waters of the eastern Mediterranean, had postulated, as an explanation, that the waters of the Atlantic and the Indian Ocean joined south of Africa.3 Hence, Africa was theoretically circumnavigable. A sketch map accompanying one of al-Birūnī’s works actually shows this graphically.4 But by the later middle ages, such speculations in learned circles were a thing of the past, largely thanks to the retrograde ideas of the twelfth-century geographer alIdrīsī, who followed Ptolemy when he should have known better, and extended the East African coast almost all the way to China, placing Sofala directly opposite the South China coast. He produced his geography at the very time that the thriving ports on the East African coast were actively trading with Arabia, India, south-east Asia and even China and were being visited by Arabic speaking merchants and mariners who must have known perfectly well from the position of the stars where they were. The authority of Ptolemy, reinforced by al-Idrīsī, was such that it overrode, for the learned elite, both the speculations of men like al-Birūnī and the practical experience of Indian Ocean and Red Sea mariners like Ahmad ibn Mājid, which in any case only began to be committed to writing at the very time the Portuguese were engaged in their great endeavour of pushing down the coast of west Africa towards the Cape of Good Hope.

It is clear from the conversation between Nunes and Monçaide recorded in the Roteiro that the Tunisian’s surprise was due not to Vasco da Gama’s achievement in reaching Calicut by sailing around Africa, but to the fact that the vessels that had just arrived in the port of Calicut were Portuguese rather than Spanish, French or Venetian – all at the time more powerful states than Portugal. Monçaide and his companion seem to have evinced no curiosity at all as to how the Portuguese had managed to sail from Portugal to Calicut. The modern reader of the Roteiro cannot help feeling that Monçaide asked the wrong question. Surely his question should have been not ‘what brought you here?’ but ‘how 1

Since the eighteenth century it has been the western habit to seek historical watersheds and define historical periods by discoveries or technological innovations. Islamic scholars have instead continued to emphasise continuity

The conversation between Nunes and the Muslim merchants of Calicut is given in Velho 1987: 54-55. A more literal translations of the Portuguese phrase Al diabro que te doo! Quén te traxo acá would be ‘The Devil take you! What brought you here! (my emphasis). For an interesting account of this first contact see Subrahmanyam, 1997: 129130. 2 The Roteiro also refers to this personage as ‘Bontaibo’, which Subrahmanyam (1997: 129) suggests conceals the name Ibn Tayyib. The name ‘Monçaide’ may be a composite of a Romance possessive and the Arabic sayyid (cf ‘Mio Cid’). Monçaide is mentioned a number of times by Camões in Os Lusíadas (VII.26; 28; 46; 67; 68; IX. 5; 15) as having loyally served Vasco da Gama.

3

This story seems to echo Eudoxus’ account of finding a plank from a ship from Cadiz off the East African coast and coming to the same conclusion (Strabo 2.3. 4-5). 4 For a reproduction of this map, see Encyclopedia of Islam (New edition) iv: 1080.

131

TRADE AND TRAVEL IN THE RED SEA REGION Portuguese ships are infesting the Red Sea, attacking India merchantmen. They have taken so many cargoes that there is a severe shortage of turban sashes and muslins from India in the markets of Cairo. And now Ibn Iyās gives his explanation of the Portuguese presence in the Indian Ocean; it is worth quoting in full:

with the past. The watershed of Islamic history is the revelation of the Qur’ān, which provides the only ‘before and after’ of significance. As late as the eighteenth century, a sophisticated scholar like the Moroccan alZayyānī, widely travelled in both Europe and the East, could write a summary of world geography without mentioning the discovery of America or the discovery of the sea route to India. His geographical notions are little different than those of a scholar living in tenth century Baghdad (al-Zayyānī 1967: passim). There are nevertheless three contemporary Arabic sources which do attempt to account for the presence of the Portuguese in the Red Sea and Indian Ocean, and from them we can sense something of the popular reaction to this remarkable event.

The reason for this disaster is that the Franks have cleverly succeeded in breaching the barrier constructed by Alexander, the son of Philip the Greek. They have made a breach in a mountain which separates the Sea of China [baΉr al-Βīn] from the Mediterranean. The Franks have spent years enlarging the breach so that their ships could pass into the Red Sea. This has been the major cause of the present war (Ibn Iyās 1955-60, i: 106).

It is hard to believe that better information than this was not available in higher court circles in Cairo in 1506. Pietro Martire d’Anghiera himself, first and best chronicler of the discovery of the New World, had been in Cairo on a diplomatic mission in 1502, as we know from his Legatio babilonica, and must surely have spoken of the recent discoveries. The Venetians, minutely informed of the latest Spanish and Portuguese discoveries, had been urging the Mamluk Sultan since at least 1500 to take steps against the Portuguese, even suggesting the building of a canal linking the Mediterranean to the Red Sea – rather seriously overestimating the resources of the Mamluk state. There were a number of people in Cairo – Andalusians fresh from Granada, corsairs, renegados, ambassadors of European powers – who were fully informed of the recent Spanish and Portuguese discoveries. Yet Ibn Iyās chose to recount what can only be described as the gossip of the bazaars.

The Chronicle of Ibn Iyās When Vasco da Gama reached Calicut in 1498, the major power in the Islamic world was the Mamluk dynasty of Egypt. Its days were numbered and it was in decline, but it was still the most powerful and highly organized state in the East. The Mamluks controlled Syria and Palestine as well as Egypt; they were the protectors of the Holy Cities of Arabia, Mecca and Medina, with all the political prestige attendant on this claim. The Mamluk rulers received embassies from European and Indian princes; Venetians, Genoese and Catalans had their funduqs at Alexandria. Cairo was thus ideally placed to be a clearing-house for information relating to Spanish and Portuguese discoveries. We possess a detailed chronicle of the last years of the Mamluk regime written by Ibn Iyās (1448-1524), son of a prominent official closely related to a number of highly placed Mamluks. Ibn Iyās studied under the greatest scholar of the age, al-Suyūtī, a prolific polymath of enormous erudition.

This is itself revealing, both of the cast of mind of the author and the tendency of the time to link innovations to pre-existing frames of reference, in this case the wellknown QurΜānic verses in which Alexander/Dhū alQarnayn is enjoined by God to build a barrier against the hordes of Gog Magog [jūj wa majūj], to stop them ‘working corruption in the earth’ (Q. 18: 90-95). At the end of time, they will be loosed upon the world (Q. 21: 95). But in addition to these QurΜānic verses, Ibn Iyās was influenced by the widespread popular belief, ultimately deriving from Arabic elaborations of the second century AD Greek Alexander Romance, that Alexander built a barrier [sadd] across the strait of Gibraltar, between Cadiz and Ceuta, making of the Mediterranean a land-locked sea.5 An identical story is told of his building a similar barrier between the Horn of Africa and Yemen, enclosing the Red Sea.6 The Portuguese broke through Alexander’s barrier, then tunnelled through a mountain and entered the Sea of China, the baΉr al-Βīn. Although this could possibly be interpreted as a vague and garbled reference to the westward voyages of Columbus and Cabral, it should be

Thanks to his chronicle it is possible to follow in some detail Mamluk military preparations to confront the Portuguese in the Indian Ocean. He tells us of the construction of a fleet at Suez, the mustering of troops and provisions, the selection of the two commanders, Amīr Дusayn al-Kurdī and Salmān RaΜīs, both frequently mentioned in the Portuguese chronicles. All of these details are extremely interesting and throw a great deal of light on the strengths and weaknesses of the Mamluk regime and its perception of the Portuguese threat. But in only one passage does Ibn Iyās attempt to explain the Portuguese presence in the Indian Ocean and the Red Sea. It occurs in an entry made in 912/1506. He first tells us that the Mamluk army had successfully defeated the rebellious governor of YanbuΚ and entered Mecca to perform the rites of the pilgrimage. The Portuguese – always referred to simply as ‘Franks’ – are continuing their attacks on the coast. Amīr Дusayn, commander in chief of the Mamluk forces, has begun fortifying Jiddah. Two ships are ready to depart for Aden. Twenty

5 6

132

See al-Himyari 1984: 308-311, s.v. al-sadd. See al-Himyari 1984: 509, s.v. lablaba.

PAUL LUNDE: ‘WHAT THE DEVIL ARE YOU DOING HERE?’ disease, incidentally, incapacitated a large number of troops destined to fight the Portuguese in the Red Sea and India.

borne in mind that Balboa did not sight the Pacific until 1513, seven years after this entry was written and that Magellan did not set out on his circumnavigation of the globe until 1521. Ibn Iyās nevertheless appears to suggest that the Portuguese reached the Red Sea by circumnavigating the globe, rather than sailing around the continent of Africa.

Under the following year Ibn Iyās records the fall of Granada to their Catholic Majesties, eight years after the event (Ibn Iyās 1945: 489). The traditional trading relationship with the Christian world, symbolized by suppliant Christian merchants seeking the balsam of Matariyah, was broken. It is improbable that Ibn Iyās associated syphilis with the New World, but he calls it the Frankish disease, so in a sense it too constituted a form of European aggression. Finally, the kingdom of Granada, last bastion of Islam on the Iberian peninsula, had fallen to the armies of the reconquista.

The breaking of the barrier in this story clearly has an eschatological significance for Ibn Iyās. Hints of this can be found in an earlier entry, that for the year 905/1500. In that year the balsam trees at Matariyah, just outside Cairo, inexplicably withered away. The Copts believed that the Messiah had planted them during the sojurn of Mary in Egypt, and indeed the Coptic Festival of Balsam commemorated the day of her arrival, May 19th. For centuries European merchants had come to procure the miraculous balm of these trees, considered a universal panacea. The little balsam garden at Matariyah in ΚAyn Shams, the ancient Heliopolis, symbolised the traditional commercial relationship between Islam and Christendom, a relationship changed forever by the voyages of Vasco da Gama and Cabral. This trade had been conducted on Muslim terms. Here was a commodity of legendary rarity, to which miraculous curative powers were ascribed. Production was severely limited. Tending the balsam trees and the careful collection and processing of the balsam, were carried out by Coptic monks, under the aegis of the Islamic state. All stages in production were accompanied by great ceremony and attended by state officials.7 The balsam was sold to favoured trading partners, such as the Venetians, whose presence in Islamic ports was governed by treaty.

So, as the tenth century of the hijrah began, the Franks were loosed upon the world like Gog Magog, no longer pent up in the prison of their Mediterranean by Alexander’s barrier, but sailing the Atlantic, the Indian Ocean and the Red Sea itself. Their coming struck at the lifeblood of the Mamluk state, which derived a great part of its revenues from taxes levied on the India trade. The Mamluk sultan Qānsawh al-Ghawrī, reacted by levying ever increasing taxes on an already burdened population. Qānsawh al-Ghawrī was killed at the battle of Marj Dābiq outside Aleppo in 1516, having made the illconsidered decision to oppose the Ottoman armies of Selim the Grim. The following year, Selim invaded Egypt and massacred the Mamluk cavalry on the plain near Matariyah, where the balsam trees had once flourished. The Kitāb al-Sufālīyyah of Ahmad ibn Mājid

Suddenly the balsam trees withered and died, ‘as if they had never been’. ‘This was one of the most precious specialities of Egypt,’ says Ibn Iyās, ‘which thus disappeared during the first years of the tenth century’ (Ibn Iyās 1945: 483-484).

The next text we are going to examine is of a very different character. The author is AΉmad ibn Mājid, the pre-eminent Red Sea and Indian Ocean navigator of his time. He wrote some thirty works on navigational and related matters, in both prose and verse and was a firstrate practical navigator as well, with a lifetime’s experience sailing the southern reaches of the Red Sea and the Indian Ocean.

Immediately after the passage recording the withering away of the balsam trees, Ibn Iyās describes the epidemic of syphilis which had struck Egypt, an epidemic the doctors were unable to control (Ibn Iyās 1945: 484). The

Towards the end of his life AΉmad ibn Mājid composed a rajaz poem on the route from Sofala to India. This is the Kitāb al-Sufālīyyah, preserved in a unique manuscript in Leningrad and published in facsimile along with a Russian translation by Shumovsky in 1957. The latest date mentioned in the poem is 1500, but internal evidence suggests that it was written about 1510 or possibly a few years later. If it is indeed by Ahmad ibn Mājid, it must have been written when he was very old, for he was born sometime between 1432 and 1437.8

7

There is a good description of the balsam garden of Matariyah in Simeon 1960: 57-58. The author was in Matariyah in 1323. Arnold von Harff visited Matariyah in 1497, and ascribes the destruction of the balsam trees to the rebellion of Qansuwayh against al-Nāsir Muhammad (1496-1498). He says Qansuwayh’s troops ‘pulled up the little bushes on which the balsam grew, and broke down the water wheels with which they used to water the garden, and took the oxen which drove the water wheels, so that they told me, as indeed I saw with my own eyes, than no balsam would grow there for the next ten years.’ (Harff 1946: 104-105). It is significant that Ibn Iyās says nothing of this, preferring the story of a mysterious withering away of the balsam trees. In fact, he returns to the subject of the balsam garden in an entry under 914/1509 (Ibn Iyās 1955-60, i: 145), in which he states that the balsam garden ceased to be cultivated at the beginning of 900/1494, that is at the beginning of the tenth hijri century. In this second account, he goes on to say that in 914/1509 Qānsawh al-Ghawrī succeeded in growing new balsam trees from seed, and that this was one of the major achievements of his reign. For the Arabic text of this entry, see Ibn Iyās 1960: 149.

8

Ibrahim Khoury, the editor of the Kitāb al-Sufālīyyah, regards those lines dealing with the coming of the Portuguese as later additions to Ibn Mājid’s text. He was led to this belief by the fact that the colophon of the manuscript states that it contains 701 lines, when in fact it contains 807. In his edition, he separates 106 lines from the text and relegates them to an appendix (Khoury 1987-88: 373-378, ll. 1-106. For a full

133

TRADE AND TRAVEL IN THE RED SEA REGION

The Kitāb al-Sufālīyyah is the only work attributed to Ahmed ibn Mājid in which the Portuguese are mentioned. Unfortunately, the poem is often very obscure and the author’s drift is not easy to follow. The passage that concerns us is that devoted to the coming of the Portuguese. First he mentions the gold mines of Sofala, and says that he has been told on good authority that they have recently been taken over by unbelievers from the west. He then says that the Nile is divided into three branches: the Nile of Nubia (al-nūbah), the Nile of Kuwāma, obviously the Rio Cuama, the major estuary of the Zambezi, and finally the Nile of Egypt. The ‘people of the west’ (ahl al-Υarb) have come seeking the gold of Nubia and succeeded in taking it all. Now the Portuguese rule the whole coast of Africa and all the islands, all the way from the Strait of Gibraltar to Madagascar (alqumr). They sailed through the Strait of Gibraltar, heading south for twenty watches, until they found the Eternal Isles (al-kālidāt). The Fortunate Isles are to the east (sic) of the Eternal Isles (i.e. the Canaries), and the Portuguese rule them all. In the year 906/1500 they came to Calicut, bought and sold, and suborned the Samudri Raja (al-sāmirī). They have greatly damaged Islam and the people are terrified of them. They have prevented pilgrims going from Calicut to Mecca. These are the same people who have conquered Andalucia and parts of Morocco. North of the Franks are a number of large islands (juzr kabīr) which they control. Going east (sic) from that land, inclining to the south, they have reached the land of China.9

One of the strangest events of the beginning of the 10th century was the coming of the cursed Portuguese, a branch of the accursed Franks, to the lands of India. Some of them sailed through the Strait of Ceuta (zuqāq sabta) into the ocean and set off into the Darkness. They passed by a place near the mountains of alqamr/al-qumr, that is ‘The White Mountains’ and they are the source of the River Nile. They thus arrived in the East. They passed by a place near the shore, in a strait, on one side of which was a mountain and on the other the Sea of Darkness. It was a place of huge waves, through which their ships could not sail and were broken up. All of them remained like that, none of them able to escape, being destroyed in that place, none of them able to reach the Indian Ocean, until a single grab (Υurāb) from among them reached India. They continued gathering information about this sea until a very clever person named Ahmed ibn Majid showed the chief of the Franks, who was called al-malindi (admiral) the way, having first been plyed with drink and made drunk. He told them: ‘Don’t go near the coast on that side, head out to sea, then turn back, so the waves don’t get you’. When they followed his instructions, many of the ships were saved from being wrecked and their numbers increased in the Indian Ocean…’ (alNahrawālī 1967: 18-19).

The word al-qumr is the usual name for Madagascar in Arabic geographical literature. Here it seems to be confused with the Mountains of the Moon, jibāl al-qamr, which since ancient times have been considered to lie near the sources of the Nile. But al-Nahrawālī appears to locate the mountains of al-qumr/qamr far to the south of Abyssinia, for the ‘strait’ mentioned in the text is surely the Cape of Good Hope. The similarity between the words qumr, referring to Madagascar, and qamr, ‘moon’, must have led him to believe, just as Ibn Mājid states in the Sufālīyyah, that the Zambezi was a branch of the Nile.

The statement about the Nile being divided into three branches probably indicates that he thought the Portuguese arrived in Sofala by sailing first down the ‘Nile of Nubia’, then down the Zambezi, to issue forth into the Indian Ocean near Sofala. The ‘large islands’ north of the Franks may be a reference to the New World, and the statement that they reached China by going east (he surely means west) is reminiscent of Ibn Iyās’ story.

Although it is clear that Ibn Iyās, Ahmad ibn Mājid and QuΓb al-Dīn al-Nahrawālī had only the vaguest notions of how the Portuguese arrived in the Indian Ocean, their accounts do show that the question was asked. The answers, all of which are imbued with legendary and folklore motifs, are obviously derived from popular rather than learned sources. As late as 1580, the anonymous author of the Turkish Tarikh-i Hind-i Gharbi could write:

The Barq al-Yamānī of QuΓb al-Dīn al-Nahrawālī The identification of Ahmad ibn Mājid with the pilot who guided Vasco da Gama from Malindi to Calicut was first made by Gabriel Ferrand in 1922, based on a passage in the Barq al-Yamānī fī l-fatΉ al-ΚUtmānī by QuΓb al-Dīn al-Nahrawālī (Ferrand 1924: 391-484). As a young man QuΓb al-Dīn studied in Cairo under teachers who had been pupils of the famous al-Suyūtī, Ibn Iyās‘ professor. The Barq al-Yamānī, which he completed in 1573, is a history of Mamluk and Ottoman military interventions in the Arabian Peninsula. As a boy, QuΓb al-Dīn witnessed Amīr Дusayn and Salmān RaΜīs mustering their troops in Mecca. The second chapter of the first part of the work contains a short account of the coming of the Portuguese:

[In 900/1494] the disaster-bearing ships of the evildoing Portuguese entered the Atlantic passing secretly through the Strait and voyaging along the western shores, then into the south. They turned east and passed behind the Mountains of the Moon (jibāl al-qamr), which are the sources of the Nile. Their coming caused great fear and apprehension in these places (Goodrich 1990: 85).

Here the date of the arrival of the Portuguese in the Indian Ocean is pushed back to 1494, in order to coincide with the beginning of the 10th century of the hijrah; they pass through the Strait of Gibraltar secretly – an echo of Ibn Iyās‘ story of the secret breaching of Alexander‘s

discussion of this problem, which does not affect my argument, see the French introduction to his edition. 9 Khoury 1987-88: 373-378; Shumovsky 1957, facsimile ff. 92v-93v.

134

PAUL LUNDE: ‘WHAT THE DEVIL ARE YOU DOING HERE?’ ----------/G. Wiet (transl.) 1955-60. Journal d’un bourgeois du Caire. (2 volumes). Paris: Armand Colin. ----------/M. Mostafa (ed.). 1960. Die Chronik des Ibn Ijâs. Badā ī al-Zuhūr wa Waqā ī al-Duhūr. (Bibliotheca Islamica 5d). Volume 4 (906-921/15011515). Wiesbaden: Franz Steiner. Ibn Mājid, Shihāb al-Dīn AΉmad/Khoury, Ibrahim (ed.). 1987-1988. Les poèmes nautiques d’Ahmad ibn Magid. 3ème partie. Les poèmes à rime variable: alΚAragiz. Tex arabe établi avec introduction et analyse en française. Bulletin d’Études Orientales 39-40:190424. al-Nahrawālī, Qutb al-Dīn Muhammad/Hamad al-Jāsir (ed.) 1967. al-Barq al-Yamānī fī al-Fath al-’Utmānī (Nusūs wa AbΉāt JuΥrāfīyyah wa Ta’rīkīyyah Κan Jazīrat al-ΚArab 6). Riyadh: Dar al-Yamāmah Simeonis, Symon/M. Esposito ed. 1960. Itinerarium Symonis Semeonis ab Hybernia ad Terram Sanctam. (Scriptores Latini Hiberniae 4). Dublin: Dublin Institute for Advanced Studies. Shumovsky, T.A. 1957. Tri neizvestnye lotsii Akhmada ibn Madzhida, Arabskogo lotsmana Vasko da-Gamy, v unikal’noi rukopisi Instituta vostokovedeniia AN SSSR. Moscow: Izd-vo Akademii nauk SSSR. Strabo/H.L. Jones (ed. and transl.). 1966. The Geography of Strabo. (8 volumes). (Loeb series). Cambridge, Massachusetts: Harvard University Press; London: Heinemann. Subrahmanyam, S. 1997. The Career and Legend of Vasco da Gama. Cambridge: Cambridge University Press Tibbetts, G.R. (ed. and transl.) 1971. Arab Navigation in the Indian Ocean before the Coming of the Portuguese, being a Translation of Kitāb al-fawāΜid fī uΒūl al-bahr wa ‘l-qawāΜid of Ahmad b. Mājid alNajdī. London: Royal Asiatic Society. [Reprint 1981]. Velho, Á./N. Águas (ed.) 1987 Roteiro da primeira viagem de Vasco da Gama. Mem Martins, Portugal: Publicações Europa-América. al-Zayyāni, Abu al-Qāsim/Abd al-Rahmān al-Filālī (ed.) 1967. al-Tarjumānāt al-kubrā. Rabat: Wizārat alanbā.

barrier; and their coming is somehow connected with the Mountains of the Moon and the sources of the Nile, as in Ahmad ibn Mājid and Qutb al-Dīn. Yet the author of the Tarikh-i Hind-i Gharbi had full access to European accounts; the bulk of his book, a history of the discovery of the New World, is based almost entirely on the texts collected by Ramusio in his Navigazioni e Viaggi. The linking of the coming of the Portuguese to that perennial geographical mystery, the source and true course of the Nile, reflects a historical reality. The Portuguese were not merely seeking the sea route to India, they were searching for Prester John, the legendary ruler of a Christian kingdom in the east. The appearance of an Ethiopian mission in Lisbon in 1481 had revealed the existence of a Christian kingdom in Africa. Where else could it be but at the headwaters of the Nile, one of the four rivers of Paradise? As Portuguese navigators pushed down the west coast of Africa in the fifteenth century, they excitedly identified each of the major rivers they encountered as tributaries of the Nile, assuming that by following them to their headwaters they would encounter Prester John. The three Arabic texts we have examined may well have been influenced by tales of this search. When Monçaide asked João Nunes in Calicut what he was seeking, he replied ‘Christians and spices’ – in that order. The linking of the appearance of the Portuguese in the Red Sea and Indian Ocean to the mystery of the course of the Nile, reinforced by the eschatological overtones of the breaching of Alexander‘s barrier, show how conscious our three Muslim authors were of the significance of the event, which they saw not in economic terms but as the herald of a new and disquieting future. References d’Anghiera, Pietro Martire, 1966. Legatio Babylonica. De orbe novo decades octo. Opus epistolarum. Nachdruck [der Ausg. von] 1516 [und] 1530. Graz: Akademische Druck und Verlagsanstalt. Ferrand, G, 1922. Le pilote arabe de Vasco da Gama et les instructions nautiques des arabes au XVe siècle. Annales de Geographie 7: 391-484. Goodrich, T.D. 1990. The Ottoman Turks and the New World. A Study of Tarih-i Hind-i Gharbi and Sixteenth-Century Americana. Wiesbaden: Otto Harrasowitz. Harff, Arnold von./M. Letts (ed. and transl.) 1946. The Pilgrimage of Arnold von Harff. (The Hakluyt Society, second series 94). London: The Hakluyt Society. al-Himyārī, Ibn ΚAbd al-MunΚim/Ihsān ΚAbbas (ed.), 1984. al-Rawd al-miΚΓār fī kabar al-aqΓār. Beirut: Librairie du Liban. Ibn Iyās/G. Wiet (transl.) 1945. Histoire des Mamlouks Circassiens. (Textes et traductions d’auteurs orientaux 6). Volume 2 (AH 872-906). Cairo: Institute Française d’archéologie orientale. 135

136

Mamluk and Ottoman activity in Yemen in the sixteenth century: coastal security and commercial significance Clive Smith following:

Stock taking in 1512 In the Musée du Louvre collection there is a painting, attributed to the Venetian school, of Qānsawh al-Ghawrī (1501-1517), the Mamluk sultan of Egypt, receiving in his maydān at Damascus a Venetian delegation led by Domenico Trevisan whose purpose was to renegotiate relations of friendship and commerce. The painting depicts the sultan, seated on a low platform to the right, and receiving the splendidly dressed Domenico heading his delegation of five, who are soberly dressed. On the platform with the sultan are two of his senior officials: to the right of the ambassador facing the sultan is the interpreter, the Amīr Younis, a Venetian renegade. The sultan wears a turban folded at the front into six high ‘horns’. In other portraits he wears such a turban but with only two accentuated ‘horns’. In the maydān can be seen Mamluks and others of the sultan’s court, together with merchants and a number of animals. Beyond the walls of the maydān are several buildings, with typical Mamluk domes and minarets; and the walls are hung with heraldic shields.1

He was a prince who inspired respect, of sombre bearing, imposing in processions, pleasing to look upon. Had it not been for his injustices, the frequency of his confiscations against the masses, his insatiable love of riches, he would have been one of the better rulers over Egypt (Petry 1993: 119).

Under normal conditions Venice enjoyed close commercial relations with Egypt and in fact depended on her for much of her wealth. Embassies from Venice were not infrequent, and since the turn of the century, had often come to discuss a subject that had vastly affected the commercial health of both their states: the arrival of the Portuguese in the Indian Ocean in 1498, and their virtual monopoly of the Indian trade in spices and textiles, with consequent effects on the finances of both their states. To highlight their plight, we can offer some anecdotal evidence. In 1498, the Venetians did not possess the funds to buy all the spices available in Alexandria, whereas in 1502 their ships were unable to find enough spices to load. Again, the official Venetian fleet of eight to thirteen ships, which had been coming to Alexandria twice a year (apart from other trading vessels) and returning laden with spices, was reduced, after the Portuguese discovery of the Cape route, to three ships, only to be seen in Egyptian ports once every two years. A colossal reversal of fortune! (Sālim 1974: 55-56).

The Mamluk sultan, Qānsawh al-Ghawrī, was one of five colourful personalities through whom we can perhaps illustrate Yemen’s importance for Mamluk and Turk during the sixteenth century. Of the others, one was Portuguese, another a Turk from Lesbos, and the remaining two Ottoman pashas, high in the sultan’s service. Qānsawh al-Ghawrī was now nearly seventy and had been sultan for eleven years. He was to rule for another four and to die, in 1516, under his horses’ hooves while fighting the Ottoman Turks. A rather hostile cameo portrait from his obituary in the work of the contemporary Egyptian historian, Ibn Iyās, contains the

The Venetians had been rigorous in arguing their case with the Mamluk sultan. They informed him as to their intelligence reports from Lisbon; they asked him to tell the Indian states to cease their dealings with the Portuguese; and they urged him to reduce his taxes on spices that did arrive. For as might be expected, there was an enormous difference between the cost of spices in Calicut and Alexandria. Indeed, we understand that the Portuguese were able to achieve profits of 500% on the loads they brought from India to Lisbon; and as early as 1503 they had opened a depot in Antwerp (Sālim 1974: 54-57).

1

In a contemporary account of the visit it is stated that the delegation was received by the sultan on no less than seven occasions, two of which took place in the maydān. Although the painting is not completely true to either of these visits and indeed appears to incorporate aspects of many of the visits, the most appropriate occasion would appear to have been the second in the maydān, on 20 June, Ascension Day. However, the Venetian consul from Damascus, who had been accused by the sultan of treasonable correspondence with the Persian court, was not depicted! It would appear that the painting had been at least sketched in detail by somebody on the spot for working up later. In reaching these conclusions I was helped by Gilles Peter (Schefer 1884: 184-185, 194-196). The call on the sultan by the delegation is also described by the Egyptian historian, Ibn Iyās, who states that the purpose of the visit was to negotiate the reopening of the Church of the Resurrection in Jerusalem which had been closed on the sultan’s orders (Ibn Iyās 1931, iv: 259).

The arrival of the Portuguese transformed the dynamics of the Indian Ocean. They exploited local differences to establish their presence, harassed and sank local craft with their warships and generally broke up the longstanding Indian trade with the South Arabian ports, notably, of course, with Aden. Aden, commanding the 137

TRADE AND TRAVEL IN THE RED SEA REGION entrance to the Red Sea, was the greatest entrepôt of goods from India and beyond, as well as Africa, the taxes from which brought a healthy return to the Yemeni ruling dynasty, the Tahirids, whose reigning sultan had already controlled Yemen for some twenty-two years. From Aden the goods would be transshipped to Jiddah and Suez. It was to the Mamluks, as the greatest Muslim power within reach and reputedly with firearms, not previously seen in India, that the sultan of Gujerat and other Muslim rulers on the Indian littoral looked for help against the aggression of the Portuguese (Serjeant 1974: 7; Sālim 1974: 60-61).

preparation had been made against the expected arrival of Portuguese squadrons in the Red Sea.

Initial Mamluk efforts against the Portuguese

Albuquerque’s ambition was to found a great Portuguese empire in the east. He did not merely mean to protect trade on shore but to dominate native rulers and force them to acknowledge Portugal as suzerain. The island of Socotra had been seized in 1507, and in 1510 Goa was taken and made the seat of government. Not without reason was he regarded as the first European colonialist in the east, although he was to the end distrusted by the Portuguese King Manoel and, unlike his predecessor and successors, he was not given the title of viceroy.5

1513: Entry of the Portuguese into the Red Sea Meanwhile, the advance of the Portuguese had been remorseless. With the arrival of Afonso de Albuquerque, our second colourful personality, and his appointment as governor of the Portuguese settlements in western India in 1509, the network of forts was tightened in the east and round the southern Arabian coasts (Serjeant 1974:1516).

The Mamluks were in fact ill equipped to help. They were beset by political and strategic problems of their own2 and their traditional strength lay in their cavalry and not in their seamanship or use of firearms, although, to his credit, al-Ghawrī had by now established a cannon foundry and organised a unit armed with handguns. Despite his internal concerns, however, al-Ghawrī had a fleet of some twelve ships constructed at Suez, with Дusayn, a brutal and ambitious Kurd (Bacqué-Grammont & Kroell 1988: 71-72, n. 5) as commander. Wood and boat builders had to be brought from the Mediterranean to construct the fleet at Suez; and the seamen were a motley of Moroccans, sons of Mamluks, Turkomans and Ethiopians.3 He instructed Дusayn, before the fleet left for India, to consolidate Mamluk power in the Red Sea and improve the fortifications at Jiddah and Suakin. The fortifications constructed by Дusayn at Jiddah, in their completed form, can be clearly seen in an illustration from a Portuguese account of the expedition of 1517 by Lopo Soares de Albegaria, which shrank from any attempt to breach its defences (Correa 1858-1864, i: 494).

Albuquerque was to realise that Aden, rather than Socotra, guarded the entrance to the Red Sea. In February 1513, with Dom Manoel’s authority, he set out from Goa to take the port and enter the Red Sea. He wished, among other things, to investigate reports of a new fleet being constructed by the Mamluk sultan in Cairo, to put an end to Mamluk power and to destroy the Muslim Holy Cities of Mecca and Medina. He also wished to make contact with Prester John in Ethiopia; an ongoing mission, close to the hearts of the Portuguese. His attack on Aden on Easter eve, the 26th of March, 1513, has been graphically described in his own reports and in the commentaries written by his son.6 Two images also illustrate the attack by the Portuguese on the wall with scaling ladders.7

Two sixteenth century Arabian geographical reports confirm that the town had been fortified on all sides save the sea, presumably against the marauding bedouin. The rampart walls to the landward side and six of the towers belong to the 1507 construction ordered by al-Ghawrī while the further towers and defences for the port, houses and cargo display date from his later consolidation. The towers were packed with artillery and muskets (Serjeant 1974: 160-162). In Aden Commander Дusayn was given expected assistance by the Yemeni sultan who had long been an ally of the Mamluks.4 Unfortunately the expedition of 1508, joined by a number of Indian ships without firearms, after some initial success at Diu and Chaul, was soundly defeated in India (Ross 1921: 548-549); nor were any of its successors more successful. But at least some

Albuquerque had decided to concentrate his assault on one of the towers or bastions, with the scaling ladders brought from Cochin, at least one of which was able to take six abreast. In the disordered attack the ladders kept breaking with the weight of the men but not before a group of them had climbed over and broken open one of the embrasures, to the height of a man. Aden’s governor, however, wisely eschewing battle outside the walls, succeeded in blocking the breech with earth, stones and a ‘large store of lighted straw’ which suffocated all who tried to enter from outside. Fighting lasted from dawn till midday but the Portuguese had to withdraw while those inside tried as best they could to escape. As they left, the Portuguese burnt the Arab merchant ships at anchor. At

2

5

Arising from prolonged economic decline, administrative ossification and the growing threat from Persia and Turkey (Holt 1978: 552-553). 3 Ibn al-Дusayn 1968, ii: 635; Ibn Iyās 1931, iv: 85; Bacqué-Grammont & Kroell 1988: 72, note 8. 4 Ibn al-Dayba 1977, ii: 211; Ibn al-Дusayn 1968, ii: 636.

There is a sequence of oil paintings of successive governors and viceroys on the upper floor of the Archaeological Museum in Old Goa. 6 Earle & Villiers 1990: 211-221; Albuquerque 1883, iv: 7-24. 7 Correa 1858-64, ii: 342 and a large contemporary woodcut engraving in the British Library, probably from Antwerp.

138

CLIVE SMITH: MAMLUK AND OTTOMAN ACTIVITY IN YEMEN IN THE SIXTEENTH CENTURY Indian Ocean. In this primary aim, the expedition, with its fleet of twenty ships and six thousand men ready, at long last, by the middle of 1515, was a complete failure and indeed never sailed beyond Aden. However, Salmān RaΜīs was to remain at the heart of Mamluk, and later Ottoman, activity in the Red Sea for the next twelve years.9

one stage, Albuquerque had shouted to his nephew, covered in blood: Nephew, do not excite yourself, for this orchard cannot produce any other fruit than this, and although these Moors now get the better of us, as we have broken our ladders, I trust in Our Lord that at some other time we may be revenged upon them (Albuquerque 1883, iv: 1718).

Unexpectedly, the expedition’s call for help from the Tahirid state in Yemen was rebuffed, with far reaching consequences. The Yemeni sultan’s adviser had counselled against such help on the grounds that a positive reply would merely encourage an endless series of demands for further aid.10 A Meccan historian was to observe that ‘a counsel of meanness is persuasive. Miserliness and parsimony are planted in nature’ (alNahrawālī 1967: 20). But by then the reduction in revenues from the Indian trade was causing a radical change to power politics within Yemen, for the ruling dynasty, the Tahirids, was dependent upon such revenue for the cost of the army, the civil service and expensive subsidies to the northern tribes (Porter 2002:183). They were now reduced to raising taxes to meet such needs, often in a heavy-handed way, thereby causing widespread unpopularity and prompting all sorts of groups to throw in their lot with the Mamluk forces against them. Among such groups was the emergent Zaydī Imām Sharaf al-Dīn who saw his opportunity for expansion.11

They never were. Albuquerque then headed for the Bab al-Mandab, the straits leading to the Red Sea. Deeply conscious that this was the first time the Portuguese had penetrated so far since their entry into the Indian Ocean fifteen years before, he ordered his ships to be dressed with their flags, and artillery to be fired amid general rejoicing (Albuquerque 1883, iv: 27). In fact Albuquerque achieved little but intelligence gathering in the Red Sea. He did reach the island of Kamarān, where there was water and which was an important staging post between Aden and Jiddah. There he was responsible for widespread depredation but contrary winds, disease, poor intelligence and the torrid heat of midsummer put paid to his plans as many of these elements did to the expedition that followed his death in December 1515. However, his very presence spurred on the Mamluks to action against the Portuguese, which would, perhaps inevitably, draw them into the internal affairs of Yemen.

Mamluk cannon were used to bombard al-Hudaydah; and in a battle nearby, firearms were used, according to the Yemeni chronicler, Ibn al-DaybaΚ, for the first time in the Arabian peninsula.12 Yemeni ports such as Jīzān, LuΉayyah and Mocha fell into their hands and the ancient city of Zabīd, capital of the Tihāmah, was seized. Presumably also, the anchorages near Zabīd at alMutaynah and al-BuqΚah were also taken, to which small boats would carry men and merchandise from the deepsea ships.

Enter the Turkish corsair, Salmān RaΜīs, and Mamluk involvement in Yemen Reports of the arrival of the Portuguese fleet off Aden had caused consternation and panic in Cairo. The sultan, al-Ghawrī, had been forced to seek help for the construction of a new fleet from the Ottoman sultan in Istanbul who, at his own expense, dispatched to Cairo materials and ship-builders together with a large amount of artillery and gunpowder which al-Ghawrī forwarded to Suez where there also appeared in 1514 some two thousand Turkish volunteers, among whom was a notable Mediterranean corsair from Mytilene, in Lesbos, called Salmān RaΜīs, who is our third figure.8 Mytilene was also the birthplace of his famous contemporary, Khayr al-Dīn Barbarossa, the redoubtable Turkish corsair and Grand Admiral of the Ottoman fleet, who ensured Turkish domination in the Mediterranean for over thirty years.

Foiled in their sustained attempt to take Aden, in which Salmān RaΜīs was wounded three times, the Mamluk commanders abandoned any thought of proceeding to India and returned to consolidate their position in Yemen, the island of Kamarān and Jiddah (al-Nahrawālī 1967: 23-24). In Yemen their forces were able to take the capital СanΚāΜ, near which the last Tahirid sultan and his brother were slain. They even got as far as laying siege to the highland fortress of Thulā where the Yemeni Imam, refusing to

Clearly a colourful and controversial character, Salmān RaΜīs caught the sultan’s eye to the extent that he was put in charge of arrangements for the construction of the fleet and then, with Дusayn al-Kurdī, made joint commander of the new expedition to oust the Portuguese from the

9

For a full account of the arrangements for the expedition and a comparison of the figures given by the sources for the men, ships and arms involved, see Bacqué-Grammont & Kroell 1988: 4-7. 10 al-Nahrawālī1967: 20; Bacqué-Grammont & Kroell 1988: 8. 11 al-Nahrawālī 1967: 20-21; Ibn al-DaybaΚ 1977, ii: 225; Stookey 1978: 131. 12 Cannon was used by the Mamluks near al-Hudaydah (Ibn al-DaybaΚ 1977, ii: 224; al-Nahrawālī1967: 21). The first battle between Tahirid and Mamluk forces, when firearms were used, took place al-Mazhaf in Wādī Mawr in May 1516 (al- ZaylaΚ ī 1992: 202).

8

Ibn al-DaybaΚ 1977, ii: 226; Ibn Iyās 1931, iv: 466; Soucek, 1997: 135-136. On Salmān RaΜīs’ introduction to Egypt, see Rang and Denis 1837, ii: 13-14.

139

TRADE AND TRAVEL IN THE RED SEA REGION Ocean and further east (Lesure 1976: 151-154). Furthermore, he continued, Yemen was a country as rich as Egypt, without master, the whole of which was ripe to fall into Ottoman hands. Aden had no equal in the entire Portuguese east. The Indian trade and Yemen’s exports of red dye or madder (fuwwah or Rubia tinctorum) to India, for use in dyeing textiles (Porter 2002: 182 and n. 62), could provide immense revenues in gold for the Ottoman treasury. ‘Whoever controlled Yemen would dominate the Indies’ (Lesure 1976: 156-157).

arrange meeting or truce with the now successful Mamluks, had taken refuge.13 However, by now the Mamluk state in Egypt had fallen to the Ottoman Turks and the Mamluks in Yemen decided to declare for the Ottoman sultan, Selim I. The country was far from united, however, and remained a battleground between warring Turk, Mamluk and elements of opposing Yemeni regimes. The territory acquired by the Mamluks thus remained as an uneasy basis for later Ottoman expansion and there is some disagreement as to the aim of the Mamluk commanders in its acquisition. It would seem that it had been no part of the Mamluk sultan Qānsawh al-Ghawrī’s plan for Yemen to be seized, although the Yemeni chronicler, Ibn al-DaybaΚ, has dryly observed that the Ottoman sultan had no idea that the Mamluk sultan and Дusayn al-Kurdī intended to seize Yemen!14 It was only when appeals for help were rebuffed that aggressive action was taken, in the first place to ensure supplies.

The arrival of the Portuguese and their diversion of the Indian trade round the Cape of Good Hope had led to the demise of both Mamluk Egypt and Tahirid Yemen by depriving them of essential revenues. On the Ottomans now fell the responsibility for protecting the Holy Places and the ongoing contest with the Portuguese. They would have heard of the wealth of Yemen under two successive dynasties and news would have reached them of aggressive Mamluk activity in Yemen amid echoes of the riches of the falling Tahirid state.

However, when the parlous state of the Tahirid state became clear it is possible that the two commanders, Дusayn al-Kurdī and Salmān RaΜīs, saw the scope for their own private advancement and conquest. Of the two, it would appear that Дusayn remained loyal to the Mamluks whereas Salmān,15 from Jiddah, contrived his erstwhile partner’s death by drowning (according to one source only three days before the arrival of a Portuguese fleet before Jiddah) and came out in support of the Ottoman sultan. Barbarossa’s earlier successes as corsair in Algiers and Tunis, leading to his appointment as Grand Admiral of the Ottoman fleet, may have acted as sufficient spur to this compatriot corsair of his.

The Ottomans put their faith in Salmān RaΜīs, entrusting him with a new expeditionary force, which like its predecessor, dissipated its energies in Yemen. Salmān, in all fairness, seems to have devoted all his energy to their cause, albeit at times leaving chaos in his wake. He cleared Kamarān and the shores of pockets of Portuguese. He took on remaining Mamluk enclaves in Zabīd and the highlands and at one time declared himself governor. He caused and quelled rebellion among the bedouin near Mecca amid rising inflation and scandalised Mecca with his overbearing behaviour, even raising his standards in the Holy Sanctuary. He seized the revenue of Jiddah, for long shared between the Mamluk sultan and the Grand Sharif of Mecca.16

Salmān RaΜīs and the Ottomans In any event, Salmān RaΜīs remains a figure essential to our story. He was quick to claim to the Turks, recently arrived in Cairo, in August 1517, with some justice, that he had saved Jiddah from the Portuguese fleet and justified his own role in the Yemeni campaign. The Ottomans must not have been entirely convinced and sent him for a time to prison in Istanbul before allowing him to return to the Red Sea (Soucek 1997: 135-136).

Salmān schemed and plotted against his co-commander who eventually, in 1528, contrived Salmān’s assassination on an offshore Yemeni island while playing chess.17 What he left behind in Yemen was an uncertain Ottoman presence in Zabīd and elsewhere, prone to rebellion and uncertainty, in a country that remained a battleground for local rival and foreign intruder. The determination of the Ottomans

Later, in 1525, he urged on the Ottoman vizier in Cairo, sent to restore order in the rebellious province, the need for a further expedition against the Portuguese and the case for conquering Yemen. In Jiddah, he claimed, from the original Mamluk fleet, there were no less than eighteen ships, together with a large quantity of artillery and arms. With such a force it would be possible to destroy all the Portuguese settlements in the Indian

It was another ten years, in 1538, before the Ottomans had assembled an expedition capable of taking on the Portuguese in the Indian Ocean; and, if the expedition’s campaign there resulted in failure, at least they were able to consolidate their position in the coastlands of Yemen, albeit with great brutality, with a view to later expansion in the highlands. Such expansion was only gained by successive pashas, at tremendous cost in life and funds, against the determined resistance of the Yemeni Imam

13

Ibn al-Дusayn 1968, ii: 651-4; Stookey 1978: 131-132. Ibn al-DaybaΚ 1977, ii: 226. In an analysis of Egyptian motives in Yemen, L. O. Schuman comes down against the idea of a concerted plan to seize the country (Schuman 1960 : 70-71). 15 For an analysis of their deteriorating relationship, see BacquéGrammont, & Kroell 1988:12-20. 14

16

al-Nahrawālī 1967: 42-44; 47-52. These chapters are included in the translation into English by C. K. Smith (2002b). al-Nahrawālī 1967: 53. See note 61 of the article mentioned in the above footnote.

17

140

CLIVE SMITH: MAMLUK AND OTTOMAN ACTIVITY IN YEMEN IN THE SIXTEENTH CENTURY Ottoman practice of establishing some sort of control or influence over states neighbouring the targets of their arms (Smith 2002: 206). Furthermore, Suleyman Pasha had provided valuable endorsement of Salmān RaΜīs’ earlier report concerning the riches to be achieved from control of Yemen, and his allegation that the Yemeni Imām Sharaf al-Dīn had submitted to Ottoman authority clearly added a further incentive to Ottoman intervention in the central highlands.18

Sharaf al-Dīn and his son, MuΓahhar, whose rebellion, in 1567, succeeded in reducing the Ottoman presence to an enclave surrounding Zabīd in the Tihāmah. Even Aden was wrested from Ottoman hands. And it was only with great trouble that the Ottomans were able to restore their rule in the province. Two campaigns during this period throw some light on continuing Ottoman attitudes towards the Yemen. The campaigns were led by two Ottoman pashas with much in common, Suleyman Pasha and Sinān Pasha. Both were seasoned Ottoman campaigners of ministerial rank and ex-governors of Egypt. Both were mature in age (indeed Suleyman, a eunuch, was said by one presumably hostile source to be over eighty and so fat that he needed four men to lift him) and had been captured and inducted into the corps of Mamluks when young. And most appositely, both had intrigued to be dispatched with such a force to Yemen.

There was a marked, if erratic, revival of the Red Sea spice trade following the Ottoman capture of Aden, and by the middle of the sixteenth century, it had retrieved its earlier scale. Between 1560-4, the Venetian purchases alone, at 12,000 quintals [1 quintal = c. 100 kg], were as high as before Vasco de Gama‘s second visit to the Indian Ocean in 1502 (Braudel 1972, i: 547). A modus vivendi in the highlands between the advancing Turks and the Yemeni Zaydī led by the Imām Sharaf alDīn‘s son, MuΓahhar, was not established until 1552, after some forty battles between the leaders involved (Ibn Husayn 1968, ii: 712). It allowed the Zaydī leader a virtually autonomous province in the highlands and lasted some sixteen years until his rebellion against the Turks in 1568.

Suleyman Pasha’s campaign, 1538-9 In fact, just before Suleyman Pasha’s expedition in 1538, the Portuguese had made overtures for peace with the Ottomans (Smith 2002: 243-271). The sultan wished for pepper in return for wheat: the Portuguese wished for the dismantling of the Mamluk fleet which they knew was in an advanced stage of preparation, as well as free access to the Red Sea. The sultan offered access to Aden but an end to Portuguese attempts to reach Zabīd, Suakin and Jiddah. Nothing came of them (Inalcik 1999, i: 327).

During this period the Turks were able to profit from the province and the ensuing taxes from spices, from about 1545, mark the revival of the Indian spice trade. Moreover, the period 1554-1567 witnessed an expansion in the direct spice trade with the sultan of Acheh, the most powerful ruler in Sumatra. Participation in the trade by the Achenese had in fact begun by the 1530’s. Ships, often Gujerati, reached Mocha and Jiddah and further north, from Acheh, and carried pepper, ginger, benzoin and cinnamon as well as gold, camphor, sandalwood, sulphur and silk (Inalcik 1999, i: 345-346). In August 1564, twenty-three ships unloaded at Jiddah; and in May 1565, 20,000 quintals of spices reached the same port. The Egyptian trade alone was worth between 30,000 and 40,000 quintals, a higher volume than reached Lisbon (Braudel 1972, i: 551, 553).

The vast armada contained some one hundred ships and 20,000 men under the command of Suleyman Pasha, who was ruthless and brutal in the way he conducted his campaign. The young governor of Aden, a relic of Tahirid power in Yemen, opened his gates in friendship, only later to be treacherously hanged from the yardarm of the pasha’s flagship, and Aden to be occupied by a strong Ottoman garrison. Suleyman alleged, in his report to the sultan that the governor had been in cahoots with the Portuguese. He also added that Aden was the key to the control of Yemen, with its vast swathe of territory, worth thirty sanjaqs, stretching to Mecca.

The Portuguese were unable to stop such traffic in Achenese, Ottoman, and perhaps mainly Gujerati ships, which was assisted by a network of Turkish merchant communities established from Gujerat to Acheh (Inalcik 1999: 346), although in the early 1560s they tried consistently to blockade the routes west and east of the Arabian Peninsula past the Bab al-Mandab and Hormuz. The fact that in 1561 fifty large ships (naos) escaped them during one month is witness to the extent to which the trade had recovered! (Boxer 1969: 418).

And after failure in India, Suleyman had the Mamluk governor of Zabīd killed and replaced by a loyal Ottoman, with responsibility for the coastal Tihāmah, including the ports of Jīzān and Mocha. He also took the opportunity of dispatching from Zabīd, as sop to his sultan in Istanbul, the salted and flayed head of the Portuguese leader, together with the noses and ears of 146 other Portuguese he had captured during his campaign in the east (Kerr 1811-1824, vi: 280). At least the remaining warring Tahirid and Mamluk elements in the Tihāmah had now been purged and the Yemeni coastal waters secured for Ottoman shipping. Such consolidation was a necessary precursor to any further expeditions to India and consistent with the

From 1549, spices acquired from tax revenues in Yemen were being sold in Cairo at a profit for the Ottoman state, and by 1556 a steady income was being received from 18

141

Blackburn 1980: 65; Stookey 1978: 136.

TRADE AND TRAVEL IN THE RED SEA REGION such taxes. The province began to be regarded as a lucrative source of revenue. One governor-general, RiΡwān Pasha, is said to have spent 50,000 gold ducats in bribes for the post. He is also reported to have been exempted taxation on the spices he shipped through Jiddah in 1565 for sale on the open market in Egypt which he had acquired in Yemen in lieu of salary, following a tradition which began for senior officers such as sanjaq begs as early as 1560 (Blackburn 1979: 131132, nn. 47 & 48). No wonder that the Ottomans were determined to help safeguard their interests in the Red Sea through the control of this recalcitrant province!

recaptured through attrition and re-conquest. We have no gunpowder left for the guns, and the roads have been cut and the supplies finished. There are few troops as they are scattered throughout the country, some of them to keep an eye on the area that has been captured, others have been sent to subdue the rebels and those who plunder the roads. Others have been martyred in God’s cause; yet others have fallen sick and died or remained unwell; and others still have fled and moved far away from us. All we have left now are about 1000 men, apart from the tribesmen who are under authority, the duΚāh and suchlike (al-Nahrawālī 1967: 418).

Sinān was forced to agree peace terms with MuΓahhar over the siege, allowing him to deal with the renewed and widespread rebellion which had broken out elsewhere in the country. One aspect of the campaign, which a reading of QuΓb al-Dīn’s history affords, is the help given by the IsmaΜīlī community to the Turks. This IsmaΜīlī community had been in Yemen for four hundred years but had been victimised by MuΓahhar’s virulent campaign against it and had been widely dispersed.

Such control as they were able to exercise in Yemen, however, came at a tremendous cost. In 1547, even before the largest campaigns were mounted, the Meccan author, QuΓb al-Dīn al-Nahrawālī, remarks: We have seen no foundry like Yemen for our soldiers. Each time we have sent an expeditionary force there, it has melted away like salt dissolved in water. Rarely does one come back. We reviewed the registers in the Egyptian chancery from the days of Ibrāhīm Pasha up to the present, and found that 80,000 soldiers were sent from Egypt to Yemen; those now remaining in Yemen amount to less than 7,000.19

The campaign had only been won by the Turks at great cost in men and funds. There is no doubt that Sinān Pasha used it to win credit with the sultan in Istanbul but the writing was on the wall. The cost of maintaining control of Yemen was too great for the diminishing returns involved and by 1636 the Ottomans were out of Yemen.

Sinān Pasha’s campaign, 1569-71 The second Ottoman pasha, Sinān Pasha, from his post as governor-general of Egypt, undoubtedly intrigued to replace the pasha appointed to put down the rebellion in Yemen by MuΓahhar, the Zaydī leader in Yemen. By July 1567, Ottoman control had been reduced to an enclave around Zabīd in the coastal plain; but it was not until March 1569 that his large expeditionary force reached Yemen.

References Albuquerque, A. de/Gray Birch, W. de (transl.). 1883. The Commentaries of the Great Afonso Dalboquerque, second viceroy of India. (4 volumes)/vol.4. (The Hakluyt Society, first series, vol. 69). London: The Hakluyt Society. Bacqué-Grammont, J-L. & Kroell, A. 1988. Mamlouks, Ottomans et Portugais en Mer Rouge. (Supplément aux Annales islamologiques; cahier no. 12). Le Caire: Institut Français d’Archéologie Orientale du Caire. Blackburn, J.R. 1979. The Collapse of Ottoman Authority in Yemen, 968/1560-976/1568. Die Welt des Islam, new series, 19:119-176. Blackburn, J.R. 1980. The Ottoman Penetration of Yemen: an annotated translation of Őzdemür Bey’s Fethnâme for the conquest of СanΚāΜ in Rajab, 954/August, 1547. Archivum Ottomanicum 6: 55-100. Boxer, C.R. 1969. A Note on Portuguese Reactions to the Revival of the Red Sea Spice Trade and the Rise of Atjeh, 1540-1600. Journal of South East Asian History 10:415-428. Braudel, F./Reynolds, S. (transl.) 1972. The Mediterranean and the Mediterranean World in the Age of Philip II. (2 volumes). New York; London: Harper & Row. Correa, G./Lima Felner, R.J.de (ed.) 1858-64. Lendas da India. (4 volumes). (Colleçao de monumentos ineditos para a historia das conquistas dos Portuguezes). Lisbon: Na Typographia da Academia Real das Sciencias.

QuΓb al-Dīn al-Nahrawālī, highlights the motives for the campaign which, in essence, remain true to those inherited from the Mamluks some fifty-two years earlier. The recapture of Aden had prior place. As the Porte instructed, ‘we have to restore it [Aden] to the control of Yemen... since he [the late sultan] had maintained the port of Aden as protection for the Holy Places against the heretical heathens [the Portuguese]’. And, early in the campaign, Aden was so restored, the admiral in charge being Qurt Öghlū of Mediterranean corsair fame. This is not the place for a detailed account of the campaign, save to indicate that it was a very close run thing indeed, as al-Nahrawālī, for all his pro-Turkish views, clearly shows. Of the situation during the long siege of Kawkabān, which proved the key to the campaign, he puts into the mouth of one of the Turkish commanders pleading to Sinān Pasha: It must be clear to your sharp mind that the fortress of Kawkabān cannot be taken by force: nor can it be 19

al-Nahrawālī 1967: 91-92 (as translated by Stookey 1978: 133-134).

142

CLIVE SMITH: MAMLUK AND OTTOMAN ACTIVITY IN YEMEN IN THE SIXTEENTH CENTURY Schefer, C. 1884. Le Voyage d’Outremer: (Égypte, Mont Sinay, Palestine) de Jean Thenaud; suivi de la Relation de l’ambassade de Domenico Trevisan auprès du Soudan d’Égypte, 1512. Paris: E. Leroux. Schuman, L.O.1960. A Political History of the Yemen at the Beginning of the XVI Century: Abu Makhrama’s account of the years 906-927 H. (1500-1521 A.D.). Groningen: V.R.B.Kleine. Serjeant, R.B. 1963. The Portuguese off the South Arabian Coast. Hadramī Chronicles. Oxford: Oxford University Press. [Reprint: Beirut: Librairie du Liban 1974.] Smith, C.K. 2002a. Suleyman Pasha’s Lost Opportunity in India. Studies on Arabia in honour of G. Rex Smith. Journal of Semitic Studies. Supplement 14: 243-271. ----------2002b. Lightning over Yemen: a history of the Ottoman campaign (1569-71): being a translation from the Arabic of part III of al-Barq al-Yamani fi alFath al-’Uthmani by Qutb al-Din al-Nahrawali alMakki as published by Hamad al-Jasir (Riyadh, 1967). (Library of Ottoman Studies 3). London: I.B.Tauris. -----------forthcoming. Indecisive Reactions to the Portuguese by Mamluk and Ottoman Turk. New Arabian Studies 6. Soucek, S. Selmān ReΜīs. 1997. Pages 135-136 in The Encyclopaedia of Islam (New Edition). ix. Leiden: E.J. Brill. Stookey, R.W. 1978. Yemen: the politics of the Yemen Arab Republic. (Westview Special Studies on the Middle East). Boulder, Colorado: Westview Press. al-ZaylāΚī, A.U. 1992. al-Awdā al-Siyāsīyyah wa-lAlāqāt al-Kārijīyyah li-Mintaqah Jāzān: al-Miklaf alSulaymānī fī al-ΚUΒūr al-Islāmiyyah al-WāΒitah. Riyadh: Matābi al-Farazdaq al-Tijāriyyah.

Earle, T.F. & Villiers, J. 1990. Albuquerque Caesar of the East. Selected texts by Afonso de Albuquerque and his son. Warminster: Aris & Phillips Holt, P.M. 1978. Kānsawh al-Ghawrī. Pages 552-553 in The Encyclopaedia of Islam (New Edition). iv. Leiden: E.J.Brill. Ibn al-DaybaΚ, ΚAbd al-RaΉman ibn ΚAlī/MuΉammad ibn ΚAli al-Akwa (ed.) 1977. Qurrat al-ΚUyūn bi-Akbār al-Yaman al-Maymūn (2 volumes)/vol.2. (Min TurāΕnā). Cairo: MaΓbaΚah al-Salafīyyah. Ibn al-Дusayn, YaΉya/Dr S.A. Ashūr (ed.) 1968. Гāyah al-Amānī fi AΫbār al-Qutr al-Yamānī (2 volumes) Cairo: Dar al-Kātib al-’Arabī. Ibn Iyās, Muhammad ibn Ahmad/P.Kahle & Muhammad Mustafa (eds.). 1931. Die Chronik des Ibn Ijâs. BadāΜī al-Zuhūr wa WaqāΜī Κal-Duhūr. Volume 4. (Bibliotheca Islamica 5).Istanbul: Staatsdruckerei [MaΓbaΚah al-Dawlah]. In Konmission bei F.A.Brockhaus, Leipzig. Inalcik, H. 1999. An Economic and Social History of the Ottoman Empire (2 volumes). Cambridge: Cambridge University Press. Kerr R. & Eden, F.A. 1811-1824. Particular Relation of the Expedition of Solyman Pacha from Suez to India against the Portuguese at Diu, written by a Venetian officer who was pressed into the Turkish Service on that occasion. Pages 257-286 in volume 6 [1824] of A General History and Collection of Voyages and Travels. (18 volumes). Edinburgh: William Blackwood; London: T Cadell. Lesure, M. 1976. Un document Ottoman de 1525 sur l’Inde Portugaise et les pays de la Mer Rouge. Mare Luso-Indicum 3:137-160. al-Nahrawālī, Qutb al-Din Muhammad/Hamad al-Jāsir (ed.) 1967. al-Barq al-Yamānī fī al-Fath al-ΚUtmānī (Nusūs wa Abhāt Jugrafīyyah wa TaΜrīkīyya Κan Jazīrat al-ΚArab 6). Riyadh: Dar al-Yamāmah. Petry, C.F. 1993. Twilight of Majesty: the reigns of the Mamlūk Sultans al-Ashrāf QaΜitbay and Qansūh alGhawri in Egypt. (Occasional papers, Middle East Center, Washington; n. 4). University of Seattle: University of Washington Press. Porter, V. 2002. The Ports of Yemen and the Indian Ocean Trade during the Tahirid Period (1454-1517), Studies on Arabia in honour of G. Rex Smith. Journal of Semitic Studies. Supplement 14:171-190. Rang, S. & Denis, F. 1837. Fondation de la Régence d’Alger. Histoire des Barberousse, chronique arabe du XVIe siècle, publiée sur un manuscrit de la Bibliothèque royale, avec un appendice et des notes. Expédition de Charles-Quint. Aperçu historique et statistique du port d’Alger etc. (2 Volumes). Paris: Angé. Ross, E.D. 1921. The Portuguese in India and Arabia between 1507 and 1517. Journal of the Royal Asiatic Society: 546-563. Sālim, al-Sayyid MusΓafā. 1974. al-Fath al- Utmānī alAwwal li-l-Yaman. (Second edition). Cairo: MaΚhad al-BuΉūt wa-l-Dirāsāt al- ΚArabīyyah.

143

144

Quseir Fort and the archaeology of the Дajj Charles LeQuesne

Introduction This, the ‘old town’, is overlooked by a low limestone bluff and it is on this that the Ottoman fort was built in the early 1570s (Fig. 61).

Recent archaeological and restoration work (1997-99) at Quseir Fort, part of the Antiquities Development Project of the American Research Center in Egypt,1 has opened a new page in the archaeology of the Egyptian Red Sea coast, looking for the first time in detail at the material culture of the past five hundred years in this region. The work has demonstrated not only the viability and potential significance of such studies for the period – in revealing links through various aspects of the material culture and architecture to other Red Sea sites and points beyond (Yemen in particular) – but it also produced further material for historical study in the form of around fifty letters dating largely to the late eighteenth century, many of which relate to the transport of wheat to the Hijaz. It provides a continuation, almost to the present, of the archaeological sequence at Quseir al-Qadim, c. 5 km to the north, which ends in the early sixteenth century. It is interesting to compare the patterns of economy and trade evident in the Islamic levels of the two sites. Equally, the evidence from Quseir Fort displays telling similarities and differences with the later levels at Qasr Ibrim, letters from which refer to the garrison at Quseir on more than one occasion (they were in the same military province in the sixteenth century).

Historical background There is no evidence for a pre-Ottoman town at Quseir. Indeed, the clear evidence for a Mamluk port at Quseir al-Qadim to the north makes this extremely unlikely. There is, however, some reason to believe that there may have been a Ptolemaic temple in the harbour area. This was first hinted at by Weigall’s discovery of hieroglyphic blocks built into houses at the southern end of the port as well as by subsequent observations.3 At least fifteen reused limestone and sandstone blocks from a monumental Graeco-Roman building have been identified in the walls of the fort, displaying joints for dovetail clamps and, in three cases, very worn traces of relief carvings and hieroglyphics. It seems highly likely that these come from a temple of Graeco-Roman and probably Ptolemaic date (Fig. 62). At present, though, there is no evidence to indicate any subsequent settlement at modern Quseir until the Ottoman period. Given the favourable position of the site in relation to the Hammamat road and the presence of a sheltered anchorage (albeit inferior to that at Roman Quseir al-Qadim), this seems surprising and it is to be expected that this understanding will change with future discoveries.

Site location Ottoman and modern Quseir is situated on the Red Sea coast at the eastern end of one of the key east-west routes connecting the sea with the Nile Valley at Qift (ancient Coptos): the Hammamat desert road (Fig. 59). Its long and colourful history as a route for trade, as well as for the transport of the ores and stones from the desert mines and quarries, is now well documented.2 The eastern end of the Hammamat route is formed by the Wādī Ambagi. The ancient road to Quseir al-Qadim cuts northwards over 5 km west of the coast towards the silted lagoon of the former Roman port. The later town, and its protecting fort, stand immediately north of the wide floodplain formed where the wadi reaches the sea. This outflow is also marked by a substantial embayment protected from the constant north winds by a marked promontory on which the key port buildings, the quarantine and the customs house, as well as most of the more substantial Ottoman merchants’ houses stand (Fig. 60).

The events of the early to mid-sixteenth century are, obviously, crucial to our understanding of the fort and the relationship of Mamluk Quseir al-Qadim to Ottoman Quseir. The obvious question is, why the change of site? The answer seems to bear largely on security issues. In 1541 a Portuguese fleet commanded by the future governor and viceroy of India, Don João de Castro, crept up the African shore of the Red Sea, apparently destroying each port it encountered. By de Castro‘s account this included Quseir, which he described as being ‘in a miserable condition’ (Kennedy Cooke 1933). In his Roteiro de Castro clearly refers to both old and new Quseir: Alcocer o velho and Alcocer o nouo.4 It is not entirely clear what the relative condition of the two was at this point but, if the harbour had moved south, which seems the most likely explanation for the creation of a new settlement, it may well be that most if not all of

1

3

The post-excavation work was partially funded, in addition, by a grant from the United States Agency for International Development. 2 See Prickett 1979; Sidebotham & Zitterkopf 1989.

Weigall 1909: 60-61, 80; Prickett 1979: 296; el-Zeini 1982: 399. de Castro 1833: 185. My thanks to Paul Lunde for pointing me towards this reference.

4

145

TRADE AND TRAVEL IN THE RED SEA REGION

Figure 59: Map of the northern Red Sea showing the location of Quseirand the Hammamat desert road.

146

CHARLES LEQUESNE: QUSEIR FORT AND THE ARCHAEOLOGY OF THE ДAJJ

Figure 60: Map of modern Quseir showing buildings in existence in 1931 and approximate boundaries of the town in 1799.

Figure 61: Early nineteenth century view of Quseir from the south painted by Robert Moresby. 147

TRADE AND TRAVEL IN THE RED SEA REGION

Figure 62: Part of Napoleonic bastion at the south corner of the fort incorporating blocks taken from a monumental Graeco-Roman building. the population had moved away from Quseir al-Qadim by this time. One possible historical context for this would be the Ottoman expansion into the Red Sea following their conquest of Egypt in 1517. A recently unearthed document from the Muhimme Defteri, the official Ottoman archives in Istanbul, provides further clarification of this issue as well as providing definitive evidence of the date and context of the fort’s foundation.5

from central government needed to be provided. Accordingly the firmān declares: As you asked for, at the mentioned place, build a firm fortress, being cautious to avoid wasting and squandering, place enough arms and guards, bring the runaways back to their places and make them settle.

This provides an interesting point of historical comparison with, for example, the extensive fortification of key sites in the Eastern Desert under the Romans and, by contrast, the apparent lack of such fortifications for the Upper Egyptian trade and pilgrimage routes (Qus/Edfu–ΚAydhāb) in the early medieval period.

The document is a firmān, from the sultan, Selim II, to his vizier in Egypt, ordering the construction of a fort at Quseir, written on September 19, 1571. It goes into some detail about the circumstances that led to this decision, which was made in response to a request from the vizier. It states that Quseir was:

Archaeological investigations 1997-98

…a big port with all necessaries, from (the time of) its conquest being prosperous and solid, but cannot be defended from disorders and wickednesses of mischiefmaking Arabs, who many times killed the notables and plundered all their properties.

The archaeological work at Quseir Fort began with a short preliminary season of trial excavations in the autumn of 1997 designed to assess what impact the restoration proposals might have on buried and standing archaeology as well as to establish whether the archaeology of the site extended backwards beyond the sixteenth century. The main period of survey, excavation and building study took place over ten weeks in the late winter and spring of 1998.

Combined with the evidence of João de Castro, it seems almost certain that this is a description of new Quseir. It is striking that the firmān refers to the threat, not of the Portuguese, but the perennial problem of bedouin raids. These had clearly reached such a pitch that the settlement had ceased to be viable: it either was abandoned permanently or substantial security measures resourced

A summary phasing of the site based on this work is given below. A fundamental characteristic of the fort that has profoundly influenced the form in which it survives

5

Istanbul, Basbakanlik Osmanli Arsivi, Muhimme Defteri No.12, Document 906. The author is deeply grateful to Dr Slobodan Ilic of Bilkent University, Ankara for finding, transcribing and translating this document.

148

Figure 63: Plan of the fortress showing location of archaeological trenches.

CHARLES LEQUESNE: QUSEIR FORT AND THE ARCHAEOLOGY OF THE ДAJJ

149

TRADE AND TRAVEL IN THE RED SEA REGION

Figure 64: North tower retaining a portion of the original sixteenth century masonry beside the curtain wall – note also the break in the Napoleonic ashlar refacing of the curtain wall short of the tower to accommodate a planned bastion. the Roman road stations in the Eastern Desert), to protect an existing watering hole.

and therefore the methods of working is the fact that its interior courtyard was completely cleared of archaeological material by Napoleonic troops during their major reconstruction of 1799. This material was then dumped behind mudbrick retaining walls set 3 m back from the inner face of the Ottoman defences to form contre-murs designed to resist cannon fire. The result of this, archaeologically speaking, is that the only preserved elements of pre-Napoleonic stratigraphy lie under the contre-murs. Most of the excavation trenches were, therefore, concentrated in this zone (Fig. 63). The main stages of the fort’s development were as follows:

Evidence of life led within the fort during its first century of use includes fragments of possible armour, including cotton quilting and possible chain mail. A bread oven, which seems to have been fired on repeated occasions, was found within one of the barrack rooms. Tobacco pipes were found in some of the earliest occupation levels (Fig. 65.1). Chinese porcelain cups and bowls with blue-and-white underglaze decoration were also fairly common. Among the coarse wares, grey-brown marl water jars similar to those known from Ottoman Red Sea wrecks were also well represented (Fig. 66).6

Late sixteenth-late seventeenth century

Late seventeenth – mid-eighteenth century

The fragmentary structural remains preserved beneath the later contre-murs show that the first phase of the fortress comprised substantial limestone fortifications, 75 m square, 1.5 m thick and c. 4.3 m high, with rounded towers at the corners, c. 6 m high (Fig. 64). Traces of an original internal gatehouse survive but subsequent reconstruction has obscured any further details of the gate (e.g. the presence of projecting flanking towers).

By the late seventeenth century, military discipline in the fortress seems to have been in decline. A series of midden-type deposits in a barrack room against the northwestern defences suggest a period of fairly casual settlement, perhaps similar to the semi-civilian villagetype occupation found inside the Nile forts at Qalat Sai, Qasr Ibrim and the Cairo citadel during the later seventeenth and eighteenth centuries.7 This was followed by a period of apparently total abandonment: walls were robbed (if stone) or fell down. The defences by the later eighteenth century were in a semi-ruinous state, patched liberally with mudbrick, as described by occasional travellers (Irwin 1780: 121). The artefactual material

The excavations were able to recover just enough structural evidence to suggest that the internal plan comprised a series of casemate rooms, 4.5-5 m wide x 78 m deep, set against the inside of the curtain wall. There probably was a well in the courtyard, although the quality of the groundwater in Quseir is notoriously bad. The fort was certainly not sited, like many classic hajj forts (and

6 7

150

Raban 1971; Haldane 1996; 1997. Alexander 1998a, 1988b; Lyster 1993: 46-47.

CHARLES LEQUESNE: QUSEIR FORT AND THE ARCHAEOLOGY OF THE ДAJJ

Figure 65.1-4: Tobacco pipes

Figure 66.1-7: Examples of sherds from grey-brown water-carriers (qulal). 151

TRADE AND TRAVEL IN THE RED SEA REGION

Figure 67: Late eighteenth century structures in Trench B overlying sixteenth century stone cross-wall. transport of wheat to the Hijaz was found which was dated to 1208/1798.

from these levels contained porcelain, cream slip vessels decorated with characteristic redpainted cross-hatching, and Hays-type green-glazed coffee cups (see Keall 1991: 83; 1992: 34) as well as glass bangles (perhaps indicating the presence of women within the fort) and intriguing black-burnished tobacco pipes that may have originated in the Tophane workshops of Istanbul (Figs 65.2-3).

Napoleonic occupation (May 1799- May 1801) As with Egypt as a whole, the Napoleonic occupation of Quseir was short but (literally) shattering in its impact and profound in its long-term consequences. There is, as one might expect, exceptionally good documentation for this period of the fort’s history, including a series of detailed maps and views of the area, plans of the existing state and proposed changes to the fort and officers’ letters. French infantry under General Belliard marched across the desert from the Nile and arrived in the town, unopposed, on May 30, 1799.

Late eighteenth century On top of a blanket of silty sand which covered the collapsed remains of the internal buildings of the sixteenth and seventeenth centuries, there was clear evidence for renewed informal occupation in the later eighteenth century. This is in accordance with the descriptions of Bruce and Irwin who describe the fort (in 1769 and 1777 respectively) as being in an advanced state of disrepair but still manned and armed, albeit with decrepit and outdated cannon.8

Belliard immediately realised that the defensive qualities of the fort were completely inadequate in the context of contemporary military technology and, in particular, the curtain walls provided minimal defence against potential cannon fire. Work began immediately on refacing the exterior of the curtain in stone, thickening it in places and in others building casemate barrack and store rooms (the first use of fired brick in the fort). Before this work was near completion, two British warships, HMS Daedalus and HMS Fox, arrived in the harbour. In three days of

A mudbrick structure was found built inside the northwestern side of the old curtain wall (Fig. 67). Animals, probably goats, had been kept between this building and the defences. In their straw bedding a letter relating to the 8

Bruce 1790: i, 194-95; Irwin 1780: 121.

152

CHARLES LEQUESNE: QUSEIR FORT AND THE ARCHAEOLOGY OF THE ДAJJ

Figure 68: View of seaward defences taken in 1960 with partially collapsed Napoleonic bastion around original Ottoman tower visible in the foreground and Napoleonic entrance redan in front of the south-eastern curtain wall in the background (courtesy of Gisela Kircher). near continuous bombardment, British cannon demolished most of the seaward defences of the fort, but, despite several attempts, were frustrated in their attempts to land troops.

the incised pattern (Fig. 65.4); and a considerable number of European-style wine and spirit-bottles.

Following this destructive interlude, the French garrison renewed its restoration (now closer to a reconstruction) of the fort. All internal structures were swept away and the debris dumped between the new inner retaining wall and the existing curtain to create a more realistic defence against cannon fire. The French military plans show their intentions to build thick bastions around the old circular stone corner towers, although only the southern one was ever built (cf. Fig. 64). Similarly, the main seaward entrance was now protected by the surviving massive redan (Fig. 68). The internal buildings of the fort were reconstructed against the interior of the new defences.

The name of MuΉammed ΚAli is still attached to the fort in the minds of modern Quseiris and his hand can clearly be seen in a number of aspects of the fort’s structure. It would appear that further extensive renovation was carried out during his rule, possibly during the Egyptian campaigns led by his sons against the Wahhabis in Arabia in 1811 and 1819. It is clear that the inner and outer arches of the main entrance were replaced at this time (Fig. 69), the curtain walls again partially refaced, at least some of the towers heightened and new barrack buildings constructed against the north-eastern defences. Owing to subsequent disturbance it was difficult to identify many in situ deposits of this period. It seems probable, given Quseir’s increasing strategic and economic decline from the middle of the nineteenth century, that it ceased to be maintained during this period. Certainly, the fortifications were in a state of partial collapse by the 1960s.

The nineteenth century

In artefactual terms, much of the material that seems to date from the period of the Napoleonic occupation, and perhaps also from the subsequent short-lived British occupation, had much in common with the previous levels: cross-painted domestic wares, Hays coffee cups, porcelain bowls and cups and grey-brown pitchers. There are also some distinctively new elements: the first occurrence of New World vegetables, including chilli and pumpkin; characteristic flared, red-burnished, Asyut-style pipe bowls, in one case with a face incorporated within

The Egyptian Ήajj Perhaps the greatest significance of the findings at the Fort is that they provide a starting point for assessing the

153

TRADE AND TRAVEL IN THE RED SEA REGION

Figure 69: The Napoleonic entrance redan with reconstructed arch from the reign of MuΉammad ΚAli. extensive but largely neglected material evidence for the Upper Egyptian Ήajj. There were essentially two pilgrimage routes from Egypt to the Hijaz. The northern, better-known land-route, originating in Cairo, crossed Sinai from Suez (al-Agroud) in the west, via Nakhl, to ΚAqabah in the east (Tamari 1982). The southern route, in the Mamluk and Ottoman periods at least, followed the Wādī Hammamat road from Qift to Quseir. Both routes were used by Egyptian and African pilgrims, although it would appear that the southern route, at least in the later periods, was the one favoured by Maghribis and West and Central Africans.9 This is supported by observations at western desert oases such as Dakhla, where evidence for the passage of African pilgrims has also been recognised (S. Akram pers.comm.).

outside the walls. In some examples in Sinai, where wells were absent, elaborate water management systems including dams and reservoirs were constructed to gather and conserve water from the heavy winter rains. It appears that the only primary Islamic period fortifications on the darb al-Ήajj of Middle Egypt were at its terminals: Qena and Quseir. Sixteenth century Ottoman forts were constructed at both of these locations. However, these do not appear to have been classic Ήajj forts: they were much larger and accommodated substantial permanent Janissary garrisons. The firmān ordering the construction of Quseir fort, issued in 1571 in Istanbul, is quite specific in stating that its purpose was to provide ‘defence from disorder makers’ and that it should accommodate a ‘sufficient number of fort guards and armour’. It would appear then that the existing infrastructure of the desert road, albeit antiquated, was considered sufficient to accommodate the needs of the pilgrim traffic. Intriguingly, there is evidence from eighteenth century travellers that a signalling system, using fire and smoke, connecting Quseir to the Nile was in operation in the late eighteenth century, as in the Roman period (Irwin 1780: 138). The continuity of patterns of use and activity on the desert routes from ancient to medieval times is striking.

By contrast with the northern route, there is apparently very little sign of the passage of Ήajj pilgrims at the wells of the Wādī Hammamat route. Studies of the road stations have only identified Roman remains. However, this may be because this is what was being looked for (by contrast with the northern route, where some preliminary survey has been carried out). It seems highly likely that the fortified wells were used and maintained at least in the Mamluk and Ottoman periods and many of the structures within and around them, upon closer examination, may well turn out to be Islamic in date.

The function of Quseir Fort Forts of the upper Egyptian darb al-Ήajj While it is clear that Qusier Fort was designed primarily as a military installation with a permanent Janissary garrison, its role in protecting pilgrims and the traffic of goods associated with their passage was essential. It is the latter which the foundation firmān emphasises in describing Quseir as being the port ‘from which the annual provision for the servants (officials) of the Holy Places is sent by vessels to Jiddah and Yanbu’. It is clear from both the Mamluk letters at Quseir el-Qadim and the later letters recovered from Quseir Fort that this was, in the minds of the authorities at least, the fundamental importance of Quseir – that it was the port of embarkation for the Upper Egyptian wheat that kept the

The architectural form most closely identified with the Ήajj is the so-called ‘Ήajj fort’. While this label is used to cover a broad range of installations, it is normally applied to the fortified wells, such as those found in the Sinai and on the Syrian and Hijazi darb al-Ήajj.10 As is clear from Ottoman-period accounts of the Ήajj, these installations were first and foremost designed to protect water sources and food supplies, with the pilgrims generally camping 9

One of the most important tomb-shrines of Quseir was that of alTakrūnī (i.e. al-Takrūrī), probably belonging to a pilgrim from Timbuktu (el-Zeini:1982: 404-405). 10 Jomier 1951; Tamari 1982; Petersen 1989.

154

CHARLES LEQUESNE: QUSEIR FORT AND THE ARCHAEOLOGY OF THE ДAJJ repeated. This same pattern can be observed at Quseir alQadim with its abandonment in the later Roman period and again at the end of the Mamluk period. This ephemeral quality - the reliance upon the presence of the political will to perpetuate their existence - is a characteristic of sedentary settlements of all periods along the desert shores of the Red Sea.

Holy Cities alive. Indeed the letters from the Fort, which have yet to be fully studied, suggest that it may have been used to store this wheat during the Ottoman period. Although the fort was primarily a military garrison, in contrast to the small Ήajj forts of Sinai, Syria and the Hijaz, it does display a number of common characteristics with them. Quseir, amongst its other economic roles, was the main port of departure for pilgrims from Upper Egypt and right across North Africa to the holy cities of Mecca and Medina. So, while accommodation for pilgrims was provided within the town, the fort provided crucial security from desert and sea raiders for the often wealthy travellers as well as some muscle to the endeavours of the central government to collect customs duties. A very similar fortification exists almost directly opposite Quseir at al-Dhubayr, near al-Wajh, on the Arabian coast (al-Mughannam et al. 1993: pl. 44b). This may indicate a concerted programme of refortification of the pilgrimage ports by the Ottomans in the early years of their administration of the region.

References Alexander J. 1998a. The Forgotten Garrisons of the Nile Valley. Pages 37-54 in Actes du IIème Congrès du Corpus d’Archéologie Ottomane. Zaghouane, Tunisia: Fondation Tamimi pour la Récherche Scientifique et l’Information. Alexander J. 1998b. The Ottoman Frontier Fortress of Qalat Sai (Sudan). Pages 31-36 in Actes du IIème Congrès du Corpus d’Archéologie Ottomane. Zaghouane, Tunisia: Fondation Tamimi pour la Récherche Scientifique et l’Information. Bruce J. 1790. Travels to Discover the Source of the Nile in the Years 1768, 1769, 1770, 1771, 1772 and 1773. (Five volumes). Dublin: William Sceater. de Castro, J./Nunes de Carvalho, A. (ed.). 1833. Roteiro, em que se contem a viagem que fizeram os Portuguezes no anno de 1541, partindo da Nobre Cidade de Goa atee Soe, que he no fim, e stremidade do Mar Roxo com o sitio, e pintura de todo o syno Arabico por Dom Joam de Castro. Paris: Baudry Haldane C.W. 1996. Sadana Island Shipwreck Excavation 1995. Institute of Nautical Archaeology Quarterly 22/3:3-9. Haldane C.W. 1997. Sadana Island Shipwreck Excavation 1996. Institute of Nautical Archaeology Quarterly 23/3:3-8. Irwin E. 1780. A Series of Adventures in the Course of a Voyage up the Red-Sea, on the coasts of Arabia and Egypt; and of a route through the Desarts of Thebais, hitherto unknown to the European traveller in the year 1777 in Letters to a lady. London: J.Dodsley. Jomier R.P.J. 1951. Un caravan-sérail sur la route des pélerins de la Mekke. Bulletin de la société d’études historiques et géographiques de l’Isthme de Suez 3: 33-56. Keall E.J. 1991. Drastic Changes in 16th Century Zabid. Proceedings of the Seminar for Arabian Studies 21:79-96. Keall E.J. 1992. Smokers’ Pipes and the Fine Pottery Tradition of Hays. Proceedings of the Seminar for Arabian Studies 22: 9-46. Kennedy Cooke B. 1933. The Red Sea coast in 1540. Sudan Notes and Records 16:151-59. Lyster W. 1993. The Citadel of Cairo. A History and a Guide. Cairo: Palm Press. Al-Mughannum M., al-Helwa S. & Mursi J. 1983. Catalogue of Stations on the Egyptian (coastal) and Syrian (inland) Pilgrim Routes. Atlal 7: 42-75. Petersen A. 1989. Early Ottoman Forts on the Darb elHajj. Levant 21: 97-118.

Conclusion The 1997-98 ARCE investigations at Quseir Fort have succeeded in delineating the broad outlines of the site’s history and development. It is clear that there was no preOttoman building on the site. The recently discovered firmān has not only provided a clear foundation date, but has also confirmed the nature and purpose of the fort. It was a response to the abandonment of (probably) Quseir al-Qadim in the face of constant bedouin (and possibly also Portuguese) harassment. The raison d’être of the port is also clearly stated to have been primarily supply of the Holy Cities. Other evidence, notably the letters found at the two sites, indicates that this refers primarily to Upper Egyptian wheat. Equally, it is clear that the fort was not a Ήajj station in the normal sense of the term. Clearly pilgrims did pass through Quseir in considerable numbers at some points in its history. The fort enabled this movement in the sense that the security that it provided kept the port open. However, it was not designed to cater for the needs of travellers: it did not even have a well with drinkable water. These observations also illustrate some fundamental truths about trade and settlement in the Eastern Desert. As anyone who has worked in this essentially hostile environment can attest, a high degree of organisation and infrastructure is required to support life. Without such systems, a sedentary lifestyle is impossible. In the words of the 1571 firmān, following the attacks of ‘mischiefmaking Arabs’ the town was ‘completely devastated, the port out of use, and that it would be impossible to bring back not even one person unless a fortress...will be established’. The stratigraphy of the fort strongly suggests that, with the weakening of central government in the later seventeenth century, this scenario was 155

TRADE AND TRAVEL IN THE RED SEA REGION Prickett M. 1979. Quseir Regional Survey. Pages 257352 in D.S. Whitcomb & J.H. Johnson (eds), Quseir al-Qadim 1978: Preliminary Report. (American Research Center in Egypt Reports 1). Cairo/Princeton: American Research Center in Egypt. Raban A. 1971. The Shipwreck off Sharm el-Sheikh. Archaeology 24/2:146-55. Sidebotham S.E. & Zitterkopf R.E. 1989. Stations and Towers on the Quseir-Nile Road. Journal of Roman Archaeology 75:155-90. Tamari S. 1982. Darb al-Hajj in Sinai, an historicalarchaeological study. Atti della Academia Nazionale dei Lincei. Memorie. Classe di Scienze morali, storiche e filologiche 8:10, fasc.4. Weigall A.E.P. 1909. Travels in the Upper Egyptian Deserts. Edinburgh: William Blackwood. el-Zeini H. 1982. Notes on Some of the Old Mosques in Quseir. Pages 397-406 in D.S. Whitcomb & J. Johnson, Quseir el-Qadim 1980: Preliminary Report. (American Research Centre in Egypt Reports 7). Malibu, California: Undena.

156

Les échanges commerciaux entre les rives africaine et arabe de l’espace mer Rouge golfe d’Aden aux seizième et dix-septième siècles Michel Tuchscherer Les échanges entre les deux rives bénéficiaient donc en premier lieu de conditions naturelles favorables, mais les facteurs humains étaient tout aussi importants. Hommes et cultures n’ont pas cessé de circuler entre les deux rives depuis l’Antiquité en s’enrichissant mutuellement. Ici un premier constat s’impose. Les langues parlées et utilisées autour de l’espace mer Rouge-golfe d’Aden appartiennent presque toutes au groupe des langues sémitiques. Ce qui n’a pu que faciliter l’émergence de l’une d’entre elles, l’arabe, comme lingua franca ou langue de communication au niveau régional. Ce phénomène était sans doute bien antérieure à la période qui nous intéresse, mais il se poursuivait et même s’accélérait aux seizième et dix-septième siècles. Cette progression était à l’évidence liée à l’expansion de l’islam. Si en Nubie l’islamisation était à peu près achevée, les conversions s’accéléraient dans et autour du Royaume de Sennar.1 Mais l’islamisation prenait surtout de l’ampleur dans le sud de l’Ethiopie où la plupart des groupes de Gallas adoptèrent la religion du prophète de l’Arabie.2 Cet islam partagé favorisait le déplacement et les rencontres des élites religieuses des deux rives. Celles-ci se rendaient bien sûr vers des centres prestigieux comme al-Azhar. Dans la grande université du Caire, les Djabart d’Ethiopie et les Arabes du Yémen disposaient d’un riwāq. Ces élites avaient aussi l’habitude de se rendre à La Mecque et à Médine, à Zabīd ou à Aden, à СanΚāΜ et dans les villes du Hadramawt. Les confréries religieuses, šādilīyyah et qādirīyyah en particulier, se développaient parallèlement, enserrant les rives africaine et arabique dans un réseau de plus en plus dense. Le pèlerinage à La Mecque déplaçait chaque année des groupes plus ou moins nombreux à partir du Yémen, de l’Ethiopie, du Sennar et de l’Egypte vers le Hédjaz. Ce phénomène n’était pas strictement limité aux communautés musulmanes. Chez les chrétiens aussi les contacts avec le monde arabe se renforçaient. On pense évidemment à l’église copte d’Ethiopie dirigée par un patriarche égyptien. Mais songeons aussi à cette caravane de chrétiens éthiopiens qui se rendait chaque année à Jérusalem et au Sinaï en passant par Suakin (Pirès 1944: 8) ou même par Djedda.

Pour étudier les échanges commerciaux entre les rives africaine et arabe de la mer Rouge au début de la période moderne (seizième/dix-septième siècle), il convient d’élargir le domaine spatial pour y inclure également le golfe d’Aden. En effet, cet ensemble constituait une unité économique articulée autour de deux systèmes d’échanges. L’un reposait sur l’axe qui reliait la Méditerranée à l’océan Indien. En prenant appui sur quelques cités portuaires (Shihr, Aden, Suakin/Sawākin, Djedda, al-Tur et Suez), il constituait l’une des grandes routes du commerce international au début du seizième siècle. Mais ce système était fortement menacé par les ambitions portugaises dans l’océan Indien. L’autre système englobait l’ensemble des mouillages et des ports situés de par et d’autre de la mer Rouge et du golfe d’Aden. Il regroupait un ensemble de réseaux relativement denses que générait une complémentarité économique forte entre les deux rives. L’arrivée des Ottomans durant le premier quart du seizième siècle, puis l’instauration progressive de leur hégémonie durant le quart suivant, ne bouleversa pas les équilibres anciens qu’avaient façonnés des facteurs solides à la fois physiques, sociaux et culturels. La présence ottomane eut d’abord pour conséquence de stimuler les activités. L’instauration d’un modus vivendi entre Ottomans et Portugais contribua grandement à la reprise du commerce des épices à travers l’ancienne route de la mer Rouge. De même la pax ottomanicum en mer Rouge facilita la circulation des hommes et des biens entre les deux rives. Des conditions générales favorables Ce qui frappe, c’est la densité des échanges entre les ports des deux rives, arabique depuis Shihr jusqu’à Muwaylih, et africaine depuis Berbera jusqu’à Suez. Cette intensité des relations était favorisée par un certain nombre de facteurs. Le premier, c’était évidemment la mer qui, plus qu’ailleurs, rapproche ici les deux rives. Les distances sont très courtes, de Moka par beau temps on distingue très nettement la rive d’en face, la distance en droiture n’excède nulle part les 300 kms. Si les traversées d’une rive à l’autre étaient courtes, elles étaient aussi faciles. Un régime de vents assez favorable rendait la traversée possible pratiquement tout au long de l’année. Elle ne durait généralement que quelques jours. Il n’en était pas de même pour les liaisons nord sud, beaucoup plus longues et plus aléatoires, en raison de vents souvents contraires, soufflant surtout du nord et rendant la remontée vers les parties septentrionales très difficile, en particulier dans la partie entre Djedda et Suez.

Ces contacts entre monde éthiopien et monde arabophone favorisèrent indéniablement la langue arabe. Ils contribuèrent aussi à renforcer la position des marchands arabes ou arabophones dans les réseaux commerciaux régionaux. Il s’agissait en premier lieu de marchands 1 2

157

Triaud 1995: 419-426; Holt 1994, 33-34. Trimingham 1965: 96-103; Abir 1975: 552.

TRADE AND TRAVEL IN THE RED SEA REGION des chevaux4, du beurre, du suif et même des volailles, ainsi que du miel et de la cire. Les produits de la chasse y tenaient une place non négligeable sous forme de peaux les plus variées, ou de cornes d’hippopotame (Van Donzel 1979: 101-104) sans doute destinées à la fabrication des manches ornant les Τambīyyah ou poignards des notables tribaux au Yémen. Les ports éthiopiens fournissaient aussi du bois, ainsi que les produits d’un artisanat encore rudimentaire tel que des paniers tressés, des cordages et des nattes. Durant la première moitié du seizième siècle, ZaylaΚ expédiait aussi du café vers l’Arabie (Serjeant 1963: 102-105), avant que le Yémen ne se lançât à son tour dans la production de la précieuse graine et y occupât une position de quasi monopole au dix-septième siècle (Tuchscherer 2001: 6989). D’après des chroniques yéménites, des céréales amenées d’Ethiopie permirent à plusieurs reprises, notamment en 1668-69, en 1676 et en 1678, de soulager les populations menacées de graves disettes en raison de calamités naturelles, notamment des sécheresses et des invasion de criquets. Les marchands ravitaillèrent non seulement les cités portuaires, mais aussi les villes et villages de l’intérieur avec des céréales éthiopiennes ou du riz apporté à la hâte depuis l’Inde (Ibn al-Qāsim 1996: 173, 299, 328).

musulmans venus surtout du Yémen, mais aussi du Hédjaz ou de l’Egypte. Mais au dix-septième siècle, et sans doute plus tôt, un nombre croissant de chrétiens maronites et arméniens venus de Syrie3 et donc arabophones participaient aux échanges commerciaux, souvent d’ailleurs dans des réseaux qui s’étendaient bien au-delà de la mer Rouge vers la Méditerranée et surtout vers l’Inde et l’Asie du Sud-Est. Pour ces raisons à la fois culturelles, religieuses et commerciales, l’arabe renforçait sa position dans la région pour s’ériger en une sorte de lingua franca de la mer Rouge et de ses territoires riverains. Des rives aux économies complémentaires Un autre facteur essentiel vint renforcer la cohérence de l’espace mer Rouge-golfe d’Aden. Il s’agissait de la complémentarité économique entre les deux rives. L’Arabie avait besoin de la rive africaine pour l’approvisionnement de ses villes. Sur ses côtes, les activités portuaires avaient contribué au développement de noyaux urbains importants, depuis Muwaylih au nord jusqu’à Shihr et Qishn au sud. Aden, à en croire des sources portugaises, aurait compté près de 50 000 habitants au début du seizième siècle. L’instauration de l’hégémonie ottomane autour de la mer Rouge durant le second quart du seizième siècle et la reprise du commerce de transit des épices vers la Méditerranée favorisèrent aussi le développement urbain en Arabie. Moka, érigée en base navale ottomane vers 1550, puis transformée en vaste emporium à l’entrée sud de la mer Rouge, connut une croissance particulièrement forte à la fin du seizième et au cours du dix-septième siècle (Brouwer 1997: 201-202). La Mecque et Médine, placées sous la protection du sultan d’Istanbul, accueillaient un nombre croissant de pèlerins (Faroqhi 1990: 63-64). Mais l’environnement le plus souvent désertique de ces cités posait évidemment le problème de leur approvisionnement. Seule la rive africaine de la mer Rouge, vaste zone de savanes et de forêts dans sa partie méridionale, immense oasis prodigieusement fertile dans la partie égyptienne de la vallée du Nil, était en mesure d’assurer la survie de ces villes. L’Egypte, par Suez et accessoirement par Quseir, envoyait d’importantes quantités de vivres vers celles du Hédjaz. Les sultans mamelouks avaient établi certaines des plus riches terres à blé de l’Egypte en waqfs en faveur des villes saintes d’Arabie. De la sorte, les céréales produites sur ces terres étaient acheminées chaque année par une importante flotte depuis Suez jusqu’à YanbuΚ et Djedda, les ports de Médine et de La Mecque (Tuchscherer 1994: 79-99). De même l’Ethiopie, à partir de Berbera, ZaylaΚ , Massawa, Suakin et de ports de moindre importance comme Baylūl, Asab ou Raheyta, contribuait à l’approvisionnement des villes de la rive opposée. Outre des céréales pour l’essentiel du sorgho, on acheminait sur de petits navires appelés Τalbah les produits les plus divers tirés de l’élevage tels que des boeufs, des moutons, des chèvres et

Si l’Arabie avait besoin de céréales et de bétail fournis par la rive opposée, elle s’y pourvoyait aussi en hommes. Pour bon nombre d’entre eux, la traversée de la mer Rouge se faisait en tant qu’esclaves dans les cales d’un boutre de marchands. En Arabie, ils fournissaient des troupes aux souverains et notables. Ceux-ci ne pouvaient espérer assoir leur autorité sur les tribus et préserver leur pouvoir au sein de leur propre famille qu’en s’appuyant sur des gardes personnelles dont la loyauté leur était totalement acquise. Les exemples abondent. Dans les années 1520, les Turco-mamelouks de Zabīd s’appuyaient sur une armée constituée surtout d’esclaves noirs (Nahrawallī 1967: 57). De même à La Mecque, les esclaves éthiopiens jouaient un rôle très important dans les querelles qui déchiraient les différents clans de chérifs entre 1672 et 1732. Certains groupes disposaient ainsi parfois de plusieurs centaines d’hommes. Ils venaient renforcer leurs rangs dans des combats qui se poursuivaient parfois jusqu’à l’intérieur du sanctuaire de La Mecque (Dahlān 1977: 99-178). Les imams qāsimides, qui avaient unifié le Yémen en chassant les Ottomans entre 1612 et 1638, s’étaient eux-aussi entourés de nombreux esclaves éthiopiens qu’ils utilisaient comme soldats (Serjeant 1983: 84b). Cependant, ils n’hésitaient pas à affranchir les plus doués d’entre eux pour leur confier de hautes fonctions dans l’administration, notamment à la tête des provinces. Les sultans kaΕīrī au Hadramawt s’appuyaient eux aussi sur des groupes d’esclaves africains (Serjeant 1963: 73). En outre, ces esclaves fournissaient aussi une main d’oeuvre appréciée

4

3

Pires 1944: 11-13; Van Donzel 1979: 52-60, 71-81, 101-104, 113; Brouwer & Kaplanian 1989: 88-99; Pitts 1949: 47.

Poncet 1949: 155; Van Donzel 1979: 13-38.

158

MICHEL TUCHSCHERER: LES ÉCHANGES COMMERCIAUX jonction entre les deux réseaux, les productions régionales s’écoulaient vers la Méditerranée ou l’océan Indien. Ainsi, les plateaux éthiopiens exportaient surtout de l’or en poudre, de l’ivoire, de la civette et des esclaves, tandis que les alentours des îles Farasan et Dahlak fournissaient des perles et de l’ambre gris. Les gommes récoltées sur les acacias du Sennar et du Darfur étaient appréciées aussi bien en Inde qu’autour de la Méditerranée. Il en était de même pour l’encens et la myrrhe produits sur les escarpements du nord de l’actuelle Somalie et sur les plateaux du Dhofar. Si au seizième siècle le Yémen exportait encore des chevaux (Serjeant 1963: 27), de l’opium, de la cornaline et de la garance vers l’Inde (Magalhaes 1969: 758, Inalcik 1994: 365), le café prit la relève au siècle suivant. Du Yémen, qui entre temps en était devenu le principal producteur, le café gagnait aussi bien les marchés de l’Empire ottoman, que ceux de la Perse (Matthee 1994: 5-15), de l’Inde des Moghols (Gopal 2001: 297-318), et bientôt ceux de l’Europe. Les comptoirs portuaires situés à la croisée des réseaux permettaient aussi de capter les marchandises nécessaires à la satisfaction des besoins locaux. Les textiles, tout particulièrement les cotonnades indiennes, venaient largement en tête (Van Santem 1982: 220). Les populations riveraines de la mer Rouge se contentaient cependant de marchandises communes (Das Gupta 1985: 412), laissant filer les qualités supérieures vers les clientèles plus riches du Caire (Brouwer & Kaplanian 1989: 199-200), d’Istanbul, de Venise ou de Marseille. En raison de la pauvreté en métaux de la région, étain, cuivre, fer, plomb et autres provenaient également de l’extérieur, c’est à dire de l’Europe, de l’Empire ottoman ou de l’Asie du Sud.6 Outre les métaux, les armes à feu ottomanes ou européennes étaient fort prisées autour de la mer Rouge.7 A la fin du seizième siècle et surtout au dix-septième siècle, Moka était devenu le premier port du Yémen et le principal emporium de la mer Rouge. A partir des ports éthiopiens (Van Donzel 1979: 52, 54-60) et arabes voisins, il drainait une part croissante des productions de la région en vue de leur exportation. De même, le port s’érigea en grand centre de redistribution dans toute la mer Rouge des marchandises venues principalement d’Inde. Cependant Moka ne parvint jamais à s’imposer comme relais unique. Des navires indiens continuaient à se rendre en droiture dans les autres ports, à ZaylaΚ, Massawa, Suakin, et surtout à Djedda. De même, Aden et surtout Shihr (Brouwer & Kaplanian 1989: 54-59) surent préserver leur fonction d’intermédiaire entre l’Inde, l’Afrique orientale et les ports de Berbera et ZaylaΚ.

dans les oasis du Hédjaz ou sur les champs en terrasses du Yémen. Des réseaux commerciaux hiérarchisés Si l’espace mer Rouge-golfe d’Aden offrait une forte cohérence, c’était aussi parce que ses réseaux et ses flux commerciaux étaient largement complémentaires et fortement hiérarchisés. Au sommet de cet ensemble, la route de transit nord sud assurait la liaison entre l’espace méditerranéen d’une part, celui de l’océan Indien d’autre part, c’est à dire l’Inde, l’Asie du Sud-Est et l’Afrique orientale. Tissus indiens, porcelaines d’Extrème-Orient, drogues diverses -entendons par là plantes tinctoriales et médicinales- ainsi que les épices s’échangeaient contre les produits venus de la Méditerranée: métaux divers ou produits de luxe tels que le corail maghrébin, l’ambre tiré de la Baltique, les draps et la verroterie de Venise, les fruits secs d’Anatolie ou de Syrie, etc. Ces échanges affichaient un déficit à peu près permanent aux dépens de la Méditerranée. Il ne pouvait être compensé que par un transfert de métaux précieux. Jusqu’à la fin du seizième siècle, il s’agissait surtout de monnaies d’or frappées dans l’Empire ottoman, notamment le dinār sulΓānī šarīfī du Caire ou dans une moindre mesure le ducat de Venise. A partir du début du siècle suivant l’argent d’Amérique, qui s’était imposé en Europe puis dans l’Empire ottoman à cette époque (Pamuk 1999: 112-158), finit aussi par gagner la mer Rouge et par se substituer très largement aux monnaies d’or (Tuchscherer 1999: 274-277). Là comme ailleurs, le grand commerce international, qui reliait entre elles les plus grandes places commerciales de la Méditerranée à celles de l’océan Indien, nécessitait des monnaies stables. Ce n’était plus le cas des frappes ottomanes d’argent qui, à partir des années 1580, étaient manipulées et dévaluées de plus en plus fréquemment par les autorités.5 C’est pourquoi, dès les premières années du dix-septième siècle, la piastre espagnole puis l’écu de Hollande s’imposèrent dans les ports de la mer Rouge comme monnaies fiduciaires. De là, ces pièces poursuivaient leur route vers l’Inde et la Chine. Mais le grand commerce international dans les ports de la mer Rouge ne représentait, en volume comme en valeur, sans doute qu’une partie réduite des activités marchandes de la région. Pourtant, c’est ce commerce qui a généralement retenu l’attention des observateurs, aussi bien à l’époque que de nos jours. C’est sans compter les activités qui se situaient à un échelon inférieur. Celles-ci s’effectuaient le long des routes maritimes multiples qui reliaient entre eux les ports des deux rives et par delà, les routes caravanières de l’Arabie à celles de l’Afrique. Ces routes constituaient un véritable réseau régional dont une des fonctions essentielles était de contribuer à l’approvisionnement des villes d’Arabie à partir de l’Afrique, comme nous l’avons déjà évoqué plus haut. Elles venaient aussi se greffer sur les grandes routes du commerce international de transit. A partir des points de

A la hiérarchie des réseaux répondait une hiérarchie non moins forte des acteurs et des techniques commerciales. Au sommet, le grand commerce de transit entre Méditerranée et océan Indien était pour l’essentiel entre les mains de négociants étrangers à la région, qui contrôlaient les activités en mer Rouge à partir de centres 6

5

7

Pamuk 1999: 131-148; Raymond 1973: 16-52.

159

Brouwer 2000: 30-47; Van Donzel 1979: 52, 101-104. Magalhaes 1969: 309; Cuoq 1981: 167, 215, 227.

TRADE AND TRAVEL IN THE RED SEA REGION part l’Egypte, des moyens rudimentaires de paiement dominaient encore partout. Voilà ce que notait Poncet en Nubie en 1698 :

extérieurs à la zone. Du Caire, les grands négociants maghrébins, turcs, égyptiens ou syriens entretenaient des agents dans les principaux ports de la mer Rouge. Ainsi, au début du dix-septième siècle, le riche négociant cairote IsmāΜīl Abū Кāqiyyah avait des représentants à Suez, Suakin, Djedda, La Mecque et Moka (Hanna 1991: 214-216). De même, les négociants des grandes places marchandes de l’Inde s’appuyaient sur un réseau d’agents dans la région. Au seizième siècle, ils venaient des multiples ports de la côte occidentale indienne. Au siècle suivant, le réseau de Surat était très largement dominant. Ces intermédiaires hindu ou gujaratiens, généralement appelés banians, constituaient de fortes communautés dans tous les ports du Yémen, en particulier à Moka (Brouwer 1997: 202, 206-7). Ils étaient aussi installés en nombre sur la côte africaine.8 Ils avaient même essaimé dans toutes les villes de l’intérieur du Yémen.9 Dans ces réseaux reliant la mer Rouge à l’extérieur, des techniques monétaires sophistiquées étaient pratique courante. Les négociants du Caire utilisaient assez largement le système des Ήawālah, une forme de lettres de changes ainsi que l’attestent les documents dans le registres des tribunaux du Caire de l’époque. Quant aux négociants indiens, ils recouraient systématiquement à la lettre de change (Ahmad Khan 2001: 329-331). Et chaque année en mars, au moment du nowruz ou nouvel an iranien, ils procédaient aux compensation entre leurs comptes (Das Gupta 1979: 69-70).

Silver is of no use in this country in the way of trade; all is done by exchange of commodities, as in the primitive times. With pepper, anissed, fenil, cloves, and chourga (which is wool dy’d blue), with spica of France, mahalab of Egypt, and other like things, passengers buy such provisions as they have occasion for (Poncet 1949: 98).

De même en Ethiopie, la poudre d’or au poids, le sel coupé en barres, le savon ou les toiles de lin servaient de monnaie en l’absence de toute pièce métallique (Van Donzel 1979: 18). Les échanges entre les deux rives, contrairement aux liaisons avec l’extérieur, étaient entre les mains de négociants d’origine surtout locale. Dans la partie septentrionale de la mer Rouge, entre l’Egypte et le Hédjaz, les réseaux étaient dominés par les négociants cairotes.11 Si les marchands du Hédjaz y tenaient une place non négligeable au seizième siècle (ΚAbd al-Rahīm 1985: 142-158), leur position semble avoir diminué par la suite. La caste dirigeante ottomane et mamelouke du Caire était aussi fortement impliquée dans ces activités.12 Ce réseau s’appuyait essentiellement sur la flotte stationnée à Suez. Elle comprenait une quarantaine de navires à la fin du dix-septième siècle, dont près de la moitié appartenaient à l’Etat ou aux divers waqfs (Tuchscherer 1997: 47). Dans la partie méridionale de la mer Rouge et dans le golfe d’Aden, les routes maritimes entre l’Arabie et l’Afrique étaient largement entre les mains d’armateurs et de négociants yéménites, établis principalement à Moka, al-LuΉayyah, al-Hudaydah, Aden et Shihr. Cependant les Somaliens, du moins selon un témoignage tardif de la fin du dix-huitièmee siècle, excluaient les navires arabes des ports de leur côte (Valentia 1811, ii: 354). Les Ethiopiens, et plus particulièrement les Chrétiens, ne dépassaient que rarement les ports de Massawa et de Suakin à partir de leurs plateaux. S’ils se rendaient à Moka ou à Djedda, ils se faisaient généralement passer pour musulmans et portaient les vêtements en conséquence (Van Donzel 1979: 59). Par contre, les marchands yéménites étaient présents dans différents réseaux à l’intérieur de l’Ethiopie où ils constituaient de petites communautés, à Gondar notamment (Van Donzel 1979: 9). Cependant l’insécurité quasi permanente, en particulier dans les basses terres côtières du fait de populations nomades non contrôlées, et sur les plateaux méridionaux en raison de la poussée des populations gallas, rendait le passage des caravanes depuis les ports jusqu’aux marchés de l’intérieur souvent périlleux. Au dix-septième siècle, les négociants musulmans de la principauté d’Awsa en pays Danakil servaient d’intermédiaires entre le port de ZaylaΚ et le

Le niveau inférieur de cette hiérarchie des activités était constitué par un réseau régional où les échanges reposaient très largement sur le troc. Là les instruments monétaires, si besoin en était, restaient rudimentaires. A Moka, ZaylaΚ, Baylūl, Massawa, Suakin ou Djedda, les marchands venus d’Ethiopie ou du Sennar échangeaient poudre d’or, esclaves, civette ou bétail contre des cotonnades indiennes, des métaux, des cauris des Maldives ou des épices.10 Dans les années 1530, les grains de poivre servaient couramment de moyen d’échange, non seulement au Hadramawt, mais aussi à La Mecque, notamment pour les pèlerins yéménites (Serjeant 1963: 77 fn 1). Au dix-septième siècle, les échanges à l’intérieur du Yémen étaient sans doute davantage monétarisés, grâce à l’économie du café qui drainait vers le pays d’importantes quantités de numéraire. Au Hédjaz, la présence des lieux saints et le pèlerinage avaient une fonction à peu près similaire, de sorte que les monnaies circulaient couramment, y compris parmi les nomades du désert. Elles provenaient en partie des donations en espèces appelées Βurrah que le sultan d’Istanbul et les autorités d’Egypte faisaient distribuer chaque année aux habitants des deux villes saintes ainsi qu’aux tribus bédouines dont l’aide était indispensable pour assurer la protection de la caravane du pèlerinage (Faroqhi 1990: 75-101). La situation était bien différente sur la rive africaine de la mer Rouge. Mise à 8

11

Van Donzel 1979: 134-6, 144; Das Gupta 2001: 390 fn 12. Ibn al-Qāsim 1996: 89, 231; Serjeant 1983: 76a, 83a, 110a. 10 Van Donzel 1979: 54-60, 71-81, 101-104, 130-135.

Raymond 1973: 107-156; Tuchscherer 1991: 321-356; ΚAbd al-MuΚtī 1999: 105-173. 12 Raymond 1973: 717-719; Hathaway 1997: 132-138.

9

160

MICHEL TUCHSCHERER: LES ÉCHANGES COMMERCIAUX pouvoirs de la région en rivalité avec les Ottomans. Autour de 1580, le négus Sartsa Dengel s’était allié avec son ancien ennemi l’émir de Harar pour tenter de faire de Baylūl le débouché maritime de son royaume, afin de se passer de l’intermédiaire des Ottomans à Massawa. La réaction fut brutale. Les Ottomans détruisirent entièrement Baylūl (Orhonlu 1974: 78-82). Quelques années plus tard en 1587, les Ethiopiens s’allièrent cette fois avec le souverain du Sennar pour tenter d’expulser les Ottomans de Suakin, mais sans plus de succès (Orhonlu 1974: 67-8). Puis après le retrait des Ottomans du Yémen en 1635, le souverain chrétien éthiopien Fāsiladas prit contact avec l’imam zaydite à deux reprises en 1642 et 1647 en vue de développer les relations commerciales entre Baylūl et Moka, à l’évidence pour éviter Massawa (Van Donzel 1979: 5-10). Le même souverain s’efforça aussi de développer ses relations commerciales à partir de l’Ethiopie en direction de l’Inde et de la Méditerranée. Pour cela, il fit appel à des marchands maronites et arméniens d’Alep (Van Donzel 1979: 19-38). Son successeur Yohannes poursuivit cette politique, non seulement en direction des Moghols, mais aussi en direction de l’Asie du sud en entrant en contact avec les Hollandais à Batavia (Van Donzel 1979: 39148).

royaume chrétien d’Ethiopie (Van Donzel 1979: 79). Par la suite, la conquête de ZaylaΚ par les troupes de l’imam zaydite al-Mahdī Muhammad en 1696 renforça indéniablement la position des marchands yéménites dans la partie méridionale de l’Ethiopie. Cette occupation montre combien les Etats de la région s’intéressaient au contrôle des flux commerciaux. Des activités commerciales au service des pouvoirs politiques Depuis l’antiquité les Etats, qui s’étaient successivement développés autour de la mer Rouge, prenaient appui sur les sociétés agraires et tribales des terres intérieures, aussi bien sur les plateaux éthiopiens que sur les hautes terres du Yémen, dans les oasis du Hédjaz ou dans la vallée du Nil en Egypte. Les souverains de ces Etats, même si en général ils n’accordaient qu’une attention limitée aux affaires de la mer, voyaient dans les ports une source non négligeable pour leurs revenus. Ils s’efforçaient donc par tous les moyens d’attirer les routes commerciales vers les cités placées sous leur contrôle et de soumettre les marchands à leur autorité. La vision des Ottomans au seizième siècle n’était guère différente, même si leur intérêt dans la région était initialement porté vers les lieux saints de l’islam. Il s’agissait de les protéger contre la menace portugaise avec l’espoir d’en tirer une légitimité accrue pour leur domination sur les différents peuples musulmans. Après l’échec de la dernière incursion portugaise d’envergure en mer Rouge en 1541, les Ottomans s’étaient assurés le contrôle de l’espace maritime. Mais contrairement à leur rivaux qui, à partir de la possession de quelques cités portuaires clé, visaient au contrôle des principaux réseaux commerciaux sur mer, les Ottomans gardaient une vision stratégique ancrée dans les profondeurs continentales. C’est pourquoi en 1547, ils s’emparèrent successivement de TaΚizz et de СanΚāΜ. Puis trois années plus tard, ils se lancèrent dans la conquête de l’Ethiopie (Orhonlu 1974: 1-42). Dès lors, comme du temps des Mamelouks et des Rasulides, les activités des marchands intéressaient les autorités d’abord en ce qu’elles contribuaient au financement de leur appareil politique et militaire à travers le prélèvement de diverses taxes. D’après le budget de la province ottomane du Yémen pour l’année 1008/1599-1600, les ports, et en premier lieu celui de Moka, assuraient à eux seuls 29% des revenus.13 De même pour le gouverneur ottoman de Massawa, les 15 000 filori ou pièces d’or tirés de ses douanes en 1581 (Orhonlu 1974: 100) représentaient sans nul doute une partie importante de son budget. Cet état de fait entrainait évidemment une vive compétition entre les gouvernants de la région. Dès 1575, puis à nouveau en 1590 le chérif de La Mecque se plaignit auprès du sultan d’Istanbul de la baisse de ses revenus dans le port de Djedda. Il accusait les autorités ottomanes du Yémen de contraindre les navires à décharger dans leurs ports au détriment de ceux du Hédjaz (Faroqhi 1990: 208-9). De même, le contrôle des ports était un enjeu décisif pour les 13

Conclusion L’hégémonie progressive des Ottomans en mer Rouge durant la première moitié du seizième siècle, puis leur retrait partiel au cours du dix-septième siècle ne modifièrent pas fondamentalement le système des échanges commerciaux dans l’espace mer Rouge-golfe d’Aden. Durant la seconde moitié du seizième siècle, leur présence contribua cependant à les stimuler. Les véritables changements dans les économies de la région se firent surtout sous l’effet de facteurs extérieurs à la zone. A ce titre, la présence portugaise dans l’océan Indien n’eut que des effets limités et temporaires sur les activités commerciales en mer Rouge. Au contraire, l’installation des Hollandais à Java et des Anglais en Inde au tout début du dix-septième siècle porta des coups bien plus sévères au commerce de transit entre Méditerranée et océan Indien. Mais là encore, les effets véritables tardèrent longtemps à se faire sentir. Ceci parce que depuis le dernier quart du seizième siècle, l’économie de la région était stimulée par le rapide développement de la production du café au Yémen. Dans l’Empire ottoman, en Iran et dans l’Empire moghol d’abord, puis en Europe l’usage du café se répandait dans des couches de plus en plus larges dans les sociétés. Le commerce du café à partir du Yémen vint fort opportunément plus que compenser le transit déclinant des épices. Il en résulta une véritable prospérité pour la région au dix-septième siècle. Le commerce du café drainait alors d’importantes quantités de numéraire vers la mer Rouge sous forme d’espèces européennes en argent. Contrairement au flux monétaire antérieur lié au transit des épices, l’argent cette fois-ci ne faisait pas que transiter par quelques cités portuaires avant de poursuivre sa route vers l’Inde et

Sahillioglu 1985: 294, 303; Inalcik 1994: 85, 538.

161

TRADE AND TRAVEL IN THE RED SEA REGION Century as recorded by the Dutch. Yemen Update 42: 30-50. Brouwer C.G. & Kaplanian A. 1989. Early SeventeenthCentury Yemen. Dutch Documents Relating to the Economic History of Southern Arabia, 1614-1630. Selected, translated into Arabic, introduced and annotated. Amsterdam: D’Fluyte Rarob. Cuoq J. 1981. L’islam en Ethiopie des origines au XVIe siècle. Paris: Nouvelles éditions latines. Dahlān A. 1977. Эulāsat al-kalām fī bayān ΚumarāΜ albalad al-Ήarām. al-Qāhirah: Maktabat al-Kullīyyāt alAzhārīyyah. Das Gupta A. 1985. Indian Merchants and the Western Indian Ocean: the Early Seventeenth Century. Modern Asian Studies 19/3: 481-499. [Reprint: Variorum 1994]. -----------1979. Indian Merchants and the Decline of Surat c. 1700-1740. Wiesbaden: Franz Steiner. -----------2001. The World of the Indian Ocean Merchant 1500-1800. Oxford: Oxford University Press. Faroqhi S. 1990. Herrscher über Mekka. Die Geschichte der Pilgerfahrt. München/Zürich: Artemis. Gopal S. 2001. Coffee Trade of Western India in the Seventeenth Century. Pages 297-318 in M. Tuchscherer (ed.), Le commerce du café avant l’ère des plantations coloniales: espaces, réseaux, sociétés (XVe-XIXe siècle). Le Caire: IFAO. Hanna N. 1991. IsmaΚīl Abū Кāqīyya et le commerce international au Caire 1585-1625. Pages 211-220 in D. Panzac (ed.), Les villes dans l’Empire ottoman: activités et sociétés. Paris: CNRS. Hathaway J. 1997. The Politics of Households in Ottoman Egypt. The Rise of the Qazdaglis. Cambridge: Cambridge University Press. Ibn al-Qāsim, YaΉya b. al-Дusayn/ΚAbd Allāh al-Дibšī (ed.) 1996. Yawmiyyāt СanΚāΜ fī-l-qarn al-Ήādī ašar, 1046-1099. Abu Dabi: Cultural Foundation Publications. Inalcik H. 1994. The Ottoman State: Economy and Society 1300-1600. Pages 9 - 410 in H. Inalcik & D. Quataert (eds), An economic and social history of the Ottoman Empire: 1300-1914. Cambridge: Cambridge University Press. Magalhaes Godinho V. 1969. L’économie de l’Empire portugais aux XVe et XVIe siècles. Paris: SEVPEN. Matthee R. 1994. Coffee in Safavid Iran: Commerce and Consumption. Journal of the Economic and Social History of the Orient 37/1: 1-32. Nahrawallī QuΓb al-Dīn M./Д. al-Jāsir (ed.) 1967. alBarq al-Yamānī fī-l-fatΉ al- Utmānī. al-RiyāΡ: Dār alYamāmah. Orhonlu G. 1974. Osmanli Imparatorlugu’nun Güney Siyaseti. Habes Eyaleti. Istanbul: Edebiyat Fakültesi Matbaasi. Pamuk S. 1999. A Monetary History of the Ottoman Empire. Cambridge: Cambridge University Press. Pires T./A. Cortesao (ed. and transl.)1944. The Suma Oriental of Tomé Pires. (2 volumes). (The Hakluyt Society, second series, volumes 89, 90). London: The Hakluyt Society.

l’Extrème-Orient. Il filait vers les multiples marchés situés profondément à l’intérieur des terres où des intermédiaires venaient se fournir en café directement auprès des paysans. Cette monétarisation d’une partie de l’économie yéménite eut deux conséquences. La première est que les paysans se mirent à acheter les cotonnades indiennes, marché dont bénéficièrent tout particulièrement les tisserands du Gujarat et les négociants de Surat. La seconde conséquence fut la consolidation de la dynastie des Qāsimides. Sans les revenus tirés du café, il n’est pas certain qu’elle aurait pu durablement financer l’achat d’esclaves et le soutien des tribus pour imposer son autorité après le retrait des Ottomans et unifier pour la première fois le Yémen depuis Najrān au nord jusqu’au Hadramawt à l’est. En Ethiopie aussi, après un seizième siècle marqué par des luttes à peu près incessantes entre chrétiens, musulmans, ottomans et tribus gallas, la stabilisation de l’émirat d’Awsa, la consolidation du royaume chrétien de Gondar et au-delà, l’extension du royaume Funj de Sennar stimulèrent le commerce et le développement de réseaux marchands. Les ambassades des empereurs chrétiens d’Ethiopie et des imams zaydites s’inscrivaient évidemment dans ce mouvement. Dans la moitié nord de la mer Rouge, si Djedda bénéficiait du transit du café, l’intérieur du Hédjaz profitait du passage annuel des deux grandes caravanes de pèlerins, celle du Caire et celle de Damas. Ce qui permit de resserrer fortement les liens économiques avec l’Egypte voisine. A la fin du dixseptième siècle, les pays riverains de l’ensemble mer Rouge-golfe d’Aden commençaient à être intégrés dans leurs profondeurs dans un système complexe d’échanges qui liaient entre elles non seulement les deux rives arabe et africaine, mais aussi celles-ci avec l’Inde et la Méditerranée References ΚAbd al-MuΚΓī, Husām Muhammad 1999. al-ΚAlaqāt almiΒrīyyah al-Ήijāziyyah fī-l-qarn al-Εāmin ašar. alQāhirah: al-Hay at al-misrīyyah li-l-kitāb. ΚAbd al-Rahīm, ΚAbd al-Rahmān 1985. al-Дijāziyyūn fī MiΒr fī-l-qarn al- āšir al-hijrī al-sādis ašar al-milādī. (2 volumes)/vol.1. RiyāΡ: al-Dāra. Abir M. 1975. Ethiopia and the Horn of Africa. Pages 537-577 in R. Gray (ed.), The Cambridge History of Africa, c. 1600-1790. (8 volumes)/vol. 4. Cambridge: Cambridge University Press. Ahmad Khan, Iftikhar 2001. Coffee Trade of the Red Sea. First Half of the Eighteenth Century. Pages 319-331 in M. Tuchscherer (ed.), Le commerce du café avant l’ère des plantations coloniales: espaces, réseaux, sociétés (XVe-XIXe siècle). Le Caire: IFAO. Brouwer C.G. 1997. Al-Mukha. Profile of a Yemeni Seaport as Sketched by Servants of the Dutch East India Company (VOC) 1614-1640. Amsterdam: D’Fluyte Rarob. -----------2000. The Supply of Minerals in the Port of alMukha during the first half of the Seventeenth 162

MICHEL TUCHSCHERER: LES ÉCHANGES COMMERCIAUX Pitts J. 1949. An Account of his Journey from Algiers to Mecca and Medina and back. Pages 1-50 in W. Foster (ed.), The Red Sea and Adjacent Countries at the Close of the Seventeenth Century as Described by Joseph Pitts, William Daniel and Charles Jacques Poncet. (The Hakluyt Society, second series, volume 100). London: The Hakluyt Society. Poncet Ch. 1949. A Narrative by Poncet of his Journey from Cairo into Abyssinia and back 1698-1701. Pages 89-165 in W. Foster (ed.), The Red Sea and Adjacent Countries at the Close of the Seventeenth Century as Described by Joseph Pitts, William Daniel and Charles Jacques Poncet. (The Hakluyt Society, second series, volume 100). London: The Hakluyt Society. Raymond A. 1973, Artisans et commerçants au Caire au XVIIIe siècle. Damas: Institut française de Damas. Sahillioglu H. 1985. Yemen’in 1599-1600 bütçesi. Pages 287-319 in Yusuf Hikmet Bayur’a Armaganlarar. Ankara: Türk Tarih Kurumu. Serjeant R.B. 1963. The Portuguese off the South Arabian Coast. Hadramī Chronicles. With Yemeni and European Accounts of Dutch Pirates off Mocha in the Seventeenth Century. Oxford: Clarendon Press. Triaud J.-L. 1995. L’expansion de l’islam en Afrique. Pages 397-429 in J-C. Garcin (ed.), Etats, sociétés et culture du monde musulman médiéval Xe-XVe siècle. Paris: Presses Universitaires de France. Trimingham J.S. 1965. Islam in Ethiopia. London: Frank Cass & Co. Tuchscherer M. 1991. Activités des Turcs dans le commerce de la mer Rouge au XVIIIe siècle. Pages 321-364 in D. Panzac (ed.), Les villes dans l’Empire ottoman: activités et sociétés. Paris: CNRS. -----------1994. Approvisionnement des villes saintes d’Arabie en blé d’Egypte d’après des documents ottomans des années 1670. Anatolia Moderna 5: 7999. -----------1999. Quelques réflexions sur les monnaies et la circulation monétaire en Egypte et en mer Rouge au XVIe et au début du XVIIe siècle. Annales islamologiques 33: 263-281. -----------2001. Commerce et production du café en mer Rouge au XVIe siècle. Pages 69-90 in M. Tuchscherer (ed.), Le commerce du café avant l’ère des plantations coloniales : espaces, réseaux, sociétés (XVe-XIXe siècle). Le Caire: IFAO. Valentia G. 1811. Voyages and Travels to India, Ceylon, the Red Sea, Abyssinia and Egypt in the Years 1802, 1803, 1804, 1805, and 1806. (3 volumes). London: Printed for F.C. and J. Rivington [Reprint 1994, Delhi: Manohar]. Van Donzel E. 1979. Foreign Relations of Ethiopia 1642-1700. Documents relating to the Journeys of Khodja Murad. Istanbul: Nederlands HistorischArchaeologisch Instituut. Van Santen H. 1982. De Verenigde Oost-Indische Compagnie in Gujarat. Leiden: Academisch Proefschrift.

163

164

Luxury wares in the Red Sea: The Sadana Island shipwreck Cheryl Ward international commercial links even as Ottoman power waned. Wealthy merchants specialising in coffee and the products of the East obtained significant political and social power as a result (Raymond 1974; Hanna 1998; Tuchscherer 2001) although the impact of European and American competition on traditional trade routes and products and a rise in taxation began to have severe consequences by the end of the 1700s.

The most ancient story of a shipwreck in the entire world, the early Twelfth Dynasty ‘Story of a Shipwrecked Sailor’, is set in the Red Sea, and hints of exotic contacts and exchange between the countries bordering the Red Sea have tantalized scholars of all eras. International travel began early in the Red Sea. More than five thousand years ago simple craft braved its waters to bring obsidian, a black volcanic stone valued for its sharpness, from the Arabian peninsula to Egypt (Zarins 1989). About the same time, travellers threaded their way through wadis in the rocky Eastern desert, stopping at rest stops like the one at Wādī Hammamat to inscribe the walls with images of boats, animals and descriptions of the task that brought them there (Winkler 1938: 36-39).

Strict commercial and cultural zones existed. Egyptian ships and merchants generally did not sail south of Jiddah in the later eighteenth century, in part because Yemen and Arabia controlled the central and southern Red Sea, and in part because sailing south of Jiddah required specialised knowledge of a very different wind regime (Facey, supra). Merchants from throughout the Muslim world lived in the port cities of the Red Sea, but Yemeni, Hijazi and Indian Ocean merchants were not at all common in Cairo (Raymond 2002: 51-52), despite the proximity of those regions to Suez and the importance of Egypt to their livelihoods.

Egypt’s pharaohs sent fleets into the Red Sea to visit copper and turquoise mines in Sinai and to sail much farther south, eventually approaching the Bab al-Mandab, to the fabled and nearly mythical land of Punt, where dancing dwarfs, giraffe tails, huge gold rings, and incense sacred to the gods could be obtained for mere trinkets (Kitchen, supra; Fattovitch, supra). Millennia later, Roman ships left Egyptian ports such as Berenike for Indian cities, sailing with the monsoon winds and returning with heady cargoes of aromatic resins and spices, elephant ivory and silks from the Far East (Periplus/Casson ed. 1989). Mamluk merchants used fine glass and imported Chinese and Near Eastern ceramics in medieval Quseir (Whitcomb 1980, 1982), and the Red Sea became a virtual Ottoman lake after the Turks took Cairo in 1517.

The focus of Red Sea trade for Egypt was on the route between Suez and Jiddah. Suez, still a small town of about fifteen hundred inhabitants at this point, was only a twenty-six hour walk and ninety-mile camel ride from Cairo. It had broad sea roads, food, water, and facilities to support both trade and pilgrim travel, all documented for Europeans by Carsten Niebuhr in the early 1760s (Niebuhr 1792, i: 175-177). Although western European ships were rare in Suez until the late 1700s, Indian ships (markab hindī) are recorded already in the mid-1600s (Raymond 2002: 52). Most vessels anchored at Suez were owned and operated by Egyptians, and many of these were engaged in the supply of Mecca (Tuschsherer 1994), transport of pilgrims, and the luxury trade made possible by acquiring goods at Jiddah. Despite all this traffic based in Egypt, only a few ships have been scientifically documented (Raban 1971, 1990; Haldane 1996b). Excavation of the Sadana Island shipwreck by the Institute of Nautical Archaeology in cooperation with Egypt’s Supreme Council of Antiquities suggests that it was part of this trading network. The exotic cargo, along with the ship‘s construction, provides a detailed look at part of the Muslim world and its integration into other trade networks of the early modern period.

By the later eighteenth century, international markets expanded traditional regional trading routes to systems that regularly spanned half the globe. In the Red Sea, Ottoman control limited direct European access to ports and maintained a seasonally timed sailing schedule for large Egyptian-owned and operated ships (Raymond 1974: i, 108; Pearson 1994: 146-153). These ships brought exotic wares into central Egyptian markets which served as redistribution points for the Mediterranean, particularly the Mediterranean provinces of the Ottoman Empire. Egypt’s foreign commerce was divided into two spheres in the Ottoman years: African/Oriental and Mediterranean. The former provided the Egyptians with luxuries they acquired at a deficit, but the re-export of most of these luxury goods to western Mediterranean countries, to the Balkans and other Ottoman lands, and to the Maghrib brought balance, if not a surplus, to the national accounts (Panzac 1992). The markets for Red Sea imports in Cairo allowed Egypt to maintain its

The Sadana Island shipwreck Shortly after 1764, a large ship sank near an eroded section of fringing reef known today as Sadana Island (Fig. 70). Capable of carrying more than 900 tons of cargo, this vessel belonged to the largest class of ships 165

TRADE AND TRAVEL IN THE RED SEA REGION recorded, its cargo and the few personal items recovered, contribute to the identification of this ship as Egyptian, operated by Muslims and occupied in the northbound luxury trade, much as the vessels André Raymond described in his 1974 study of historical documents.

then operating in the Red Sea. Sea travel, despite the known risks of sailing along a coastline almost continuously lined by coral reefs, proved to be much cheaper and safer than land caravans (Raymond 1974, i: 123-126). Although European ships had been sailing to Suez sporadically since the sixteenth century, they rarely operated north of Jiddah during the mid-eighteenth century. European ships brought Chinese export porcelain designed for the Middle Eastern market from India or China to trade at Mocha and Jiddah for coffee, and Muslim ships took it along the next leg north in the Red Sea to Suez (Brouwer 1991, 1992; Raby 1986; Ward 2000a).

The Sadana Island shipwreck site, discovered by a group of recreational divers in 1991, had been significantly disturbed in the area forward of midships, but the remainder was largely intact (Ward 2000b). The damage caused by the diving group was quite extensive and had resulted in large deposits of recently broken porcelain, glass bottles and timbers. Some of the recreational divers provided photographs and information about the original state of the site and the distribution of the artefacts. For example, in 1991 there were about ten unbroken and unopened boxes each containing around a thousand small cups packed in a leafy substance, which was probably tea. By 1994 on our first visit to the site, these finds had been reduced to several thin wood slats and thousands of porcelain shards. The recreational divers removed some twelve thousand complete pieces from the site. They remain intact, but are currently inaccessible. We hope that the collection will be returned to the national stewardship of Egypt’s Supreme Council of Antiquities. Description of cargo Archaeological work on the site has resulted in the recovery of some thousand complete, nearly complete, or unique examples of Chinese export porcelain (Fig. 71). The assemblage includes blue-and-white, glazed, and enamelled porcelain. About one-quarter of excavated porcelain pieces are in the blue-and-white porcelain category. The most commonly encountered blue-andwhite porcelain in our assemblage is a broad, shallow dish with a peony scroll design on the interior. Arabicspeakers have historically referred to this, and still do today, as a baqdūnis or parsley dish because of the way the peony leaves and stems are portrayed. The Sadana Island assemblage includes two sizes of the same style, one approximately 34.8 cm in diameter and the other around 37.8 cm in diameter. Several sizes and shapes of small glazed cups also were common. Light green (celadon) glaze on a cup exterior typically indicated a cup broader than it was tall. Brown glazed (café au lait) cups, some with medallions featuring blue designs over the white base, tended to be narrower. A handful of cups bore a rich blue glaze over the exterior, with gold enamel designs in delicate patterns.

Figure 70: Excavation of the 50-meter-long Sadana Island ship began in 1995. Two archaeologists work in the area of the ship’s main deck. (Photograph A. Flanigan) Our international team of archaeologists has recovered thousands of these fragile cups and dishes at Sadana Island in nearly five thousand dives between 28 and 40 m depth. This was the first time such a cargo was scientifically excavated (Haldane 1996b; Ward 2000a, 2001). The porcelain, along with coconuts, black-lipped pearl oyster shells and spices from islands in the Indian Ocean, earthenware vessels, incense, and coffee from the Hadramawt, are proof that the ship sank on a northbound journey. The ship, built in a manner not previously

More common, and more casually decorated, were several hundred cups and broken remains of cups, which belong to the group of enamelled porcelains with a blueand-white background. The enamelled colours have mostly disappeared, but include Chinese Imari and Famille Rose objects. Chinese Imari is the term given to blue-and-white wares subsequently enhanced by 166

CHERYL WARD: LUXURY WARES IN THE RED SEA

Figure 71: Chinese export porcelain on the ship was typical for the mid-18th century preferences of people living in the Islamic Middle East. (Photograph N. Piercy)

Figure 72: Although the cobalt blue underglaze is easy to identify on many of the porcelain pieces, recovering the original enameled design is much more difficult. Pencil lines on this porcelain plate demonstrate part of the recovery process. (Photograph N. Piercy)

167

TRADE AND TRAVEL IN THE RED SEA REGION In addition to porcelain decorative styles that suggest a non-European clientele, archaeologists excavated a number of personal items that contribute to answering the question of the ethnicity of the ships‘ crew. Mended and unique pieces of Chinese export porcelain found in the stern, earthenware food preparation and table wares, incense burners with char marks, glass liquor bottles, a wide variety of spices and other food remains and items such as pipes and jewellery provide physical evidence for life aboard an indigenous Red Sea trading vessel (Ward 2000a, 2001).

primarily scarlet and gold painted decoration. Only a few traces of reddish brown enamel remain on a handful of pieces to indicate precisely what the overglaze pattern might have been, but careful detective work by Netia Piercy, our expedition artist, revealed that many of the enamel applications had protected the surface beneath them enough to leave a more reflective surface where leaf veins, flower stamens, or other decorative motifs had been applied after the original firing (Fig.72). Whoever purchased the porcelain we excavated from the Sadana Island ship undoubtedly intended that it would appeal to Muslims. We know from other collections in, for example, Topkapı Saray, that Chinese Imari dominated mid-eighteenth century styles for the Islamic market (Krahl & Ayers 1986). Instructions to Europeans whose job it was to purchase porcelain in China or India also made reference to the fact that these particular styles, though more expensive to purchase, could be sold for up to four times more than comparable blue-and-white or café au lait styles (Ward 2000a). It is also significant that of the around one thousand decorated pieces, only two small cups have a design that is not floral or geometric (Fig.73). A cultural prohibition against figurative decoration was common in Muslim communities, and differentiates porcelain intended for the Muslim market from that exported to other regions. Narrative porcelains, with humans in a landscape or featuring westerners at sea, for example, were sold to Europeans and Americans, but are extremely rare in Middle Eastern collections (Staniforth & Nash 1998; Wästfelt et al. 1991; Zacharchuk & Waddell 1984). Though many of the styles and forms found in the Sadana Island assemblage have strong ties to porcelain from the later seventeenth century, other forms indicate a solid mid- eighteenth century date. It is likely that a strong demand for traditional patterns of decoration for more than a century accounts for this situation.

Archaeological excavation of part of the wreck site has provided a number of inscribed copper cooking and serving containers (Willis 2002). Both names and fourdigit numbers we assume to be hijri dates are present and serve as our primary data for assigning the shipwreck to the period no earlier than 1764 (Fig.74). Three pieces bear Arabic names and dates, but only two of the dates are legible. An Arabic inscription of 1169/1755/6 on a copper basin and 1178/1764 on a copper pan places the ship‘s final voyage in the third quarter of the eighteenth century. The ship‘s last voyage took place after 1764, a time of increased foreign activity in the northern Red Sea during a period of economic transition within Egypt. Several thousand earthenware jars, qulal, packed tightly into the stern and into cavities between ship timbers comprised another major cargo. We raised about 850 qulal that were made of a thin, yellow-brown to gray fabric with significant inclusions fired at very high temperatures. Shipped empty, the qulal functioned to cool water by evaporation through the relatively porous walls, a process still appreciated by people who live in the Red Sea area today; they also functioned as waterpipe bowls or stemmed goblets (Fig. 75). Although there are a number of different designs and shapes (up to thirty styles in three basic body shapes), most of them have a

Figure 73: Only this cup and a fragment of one other similar cup depict living animals; there are no examples of human portraiture in the Sadana Island shipwreck assemblage. (Drawing N. Piercy)

168

CHERYL WARD: LUXURY WARES IN THE RED SEA

Figure 74: Arabic inscriptions from copper serving and cooking wares in the stern provide the date of 1765 CE or later for the ship’s sinking. a) Damaged monogram including the numerals 1178, presumably a Hijri date equivalent to 1764 CE b) Inscription including the Hijri date 1169, equivalent to 1754/5 CE c) Inscription reading ‘Sāhib el raΜīs Mūsā MaΉmūd’ followed by what seems to be an illegible date. (Photographs H. Wellman; tracing author)

Figure 75: Qulal, spouted jugs, clay goblets or incense burners, and pipe bowls from the Sadana Island shipwreck. (Photograph M. Kato)

169

TRADE AND TRAVEL IN THE RED SEA REGION archaeologists. A casual visitor to the site had earlier removed a powder flask, but no small or large guns, shot, or other indications of weaponry existed. A lack of guns on the ship suggests that its voyages were confined to the Red Sea, within the boundaries of the Ottoman Empire, and that it had no need to defend itself from pirates or European merchant ships that had few compunctions about appropriating goods from other vessels in the western Indian Ocean and south of the Red Sea (Brouwer 1991, 1992).

consistent body diameter that allowed them to be efficiently packed throughout the vessel. Although they may have been relatively low-cost in comparison to the porcelain, the care with which they were packed suggests that this cargo was also highly valued. Organic components of the cargo include coffee beans and coffee berries, at least fifty black-lipped pearl oyster shells from a restricted area low in the hull, about one hundred coconuts packed between ship timbers in the stern, and a yellow aromatic resin spilled by the kilo across several areas of the site. Because only limited numbers of artefacts were present and much of the ship was ‘empty’ when we arrived, it is likely that it carried a large organic cargo, perhaps coffee, for which we have evidence, and perhaps also cotton muslin and calico cloth imported from India to Jiddah (Raymond 2002: 50; Pearson 1994: 160-161). No evidence of such fabrics was found, but historical documents contain frequent references to textiles imported from India.

Following the many threads in this pattern leads us to a better understanding of the economics of the luxury trade and the lives of the people who conducted it. Analysis of artefacts classed as personal possessions rather than cargo suggests that, like the ship, the crew was nonEuropean. The limited number of finds, their strong Islamic cultural parallels and even Arabic inscriptions point to a Muslim crew.

The Sadana Island ship

References

The largest and most complex artefact is the ship itself. The starboard side of the vessel is well preserved, and archaeologists exposed about 20 percent of it during three seasons of excavation. Study of the hull continues, but it seems to be of a hitherto unrecorded method of construction that is probably indigenous to the northern Red Sea, while at the same time reflecting the influences of Mediterranean, European, and Indian shipbuilding. For its nearly 50 m length, the ship is relatively lightly fastened with iron bolts, and its frames are spaced further apart than comparable Mediterranean or European ships. Built of imported pine and oak, the ship reflects what seems to be a common practice of importing construction materials from the Mediterranean through Alexandria and then transporting them up the Nile to Cairo and across the desert to Suez.

Brouwer C.G. 1991. Non-Western shipping movements in the Red Sea and Gulf of Aden during the 2nd and 3rd decades of the 17th century, according to the records of the Dutch East India company (Part 1). Die Welt des Islams 31: 105-67. Brouwer C.G. 1992. Non-Western shipping movements in the Red Sea and Gulf of Aden during the 2nd and 3rd decades of the 17th century, according to the records of the Dutch East India company (Part 2). Die Welt des Islams 32: 6-40. Haldane C. [Ward] 1996a. Archaeology in the Red Sea. Topoi 6: 853-68. --------1996b. Sadana Island shipwreck, Egypt: preliminary report. International Journal of Nautical Archaeology 25: 83-94. Hanna N. 1998. Making Big Money in 1600. Syracuse, NY: Syracuse University Press. Krahl R. & Ayers J. (eds) 1986. Chinese Ceramics in the Topkapi Saray Museum, Istanbul III, Qing Dynasty Porcelains. London: Sotheby’s. Niebuhr C.1792. Travels through Arabia, and Other Countries in the East, performed by M. Niebuhr (2 volumes). (transl. R. Heron). Edinburgh: R. Morrison and Sons. Panzac D. 1992. International and domestic maritime trade in the Ottoman Empire during the 18th century. International Journal of Middle East Studies 24:189206. Pearson M.N. 1994. Pious Passengers: The Haj in Earlier Times. New Delhi: Sterling. Periplus/Casson L. (ed. and transl.) 1989. The Periplus Maris Erythraei. Text with Introduction, Translation, and Commentary. Princeton: Princeton University Press. Raban A. 1971. The shipwreck off Sharm-el-Sheikh. Archaeology 24/2: 146-55.

Tuchscherer (1997) describes the complex process required for the Ottoman government to build three similar ships for the Red Sea fleet at about the same time. Each vessel took five to seven months to build, and they were built sequentially because only enough timber for a single ship was available at any given time. The supply problem was so severe that imported teak was bought locally for planking. The vessels built between 1761 and 1766 were among the last of the very large Red Sea ships. By the end of the eighteenth century, the most common type of ship at Suez was built in India, and teak was the dominant wood used for the diminished Red Sea merchant and imperial fleet. Conclusions One of the more surprising things about excavating a ship of this size from the eighteenth century is the near total lack of armament. Only a few musket balls, all retaining casting sprues, were recovered from the site by 170

CHERYL WARD: LUXURY WARES IN THE RED SEA Willis K.E. 2002. The copper artifact assemblage from an Ottoman period shipwreck in the Red Sea. M.A. thesis, Department of Anthropology, Texas A&M University, College Station, Texas. Winkler H. 1938. The Rock Drawings of Southern Upper Egypt (2 volumes)/vol. 1. London: Egypt Exploration Society. Zacharchuk W. & Waddell P. 1984. The excavation of the Machault: an 18th-century French frigate. Studies in Archaeology, Architecture and History. Ottawa: Parks Canada. Zarins Y. 1989. Ancient Egypt and the Red Sea trade: The case for obsidian in the predynastic and archaic periods. Pages 339-368 in A. Leonard Jr. & B.B. Williams (eds) Essays in Ancient Civilizations Presented to Helene J. Kantor. (Studies in Ancient Oriental Civilization 47). Chicago: University of Chicago.

----------1990. Medieval anchors from the Red Sea. International Journal of Nautical Archaeology 19: 299-306. Raby J. 1986. The porcelain trade routes. Pages 55-63 in volume 1 of R. Krahl Chinese Ceramics in the Topkapi Saray Museum Istanbul (ed.J. Ayers). (3 volumes). London: published in association with the directorate of the Topkapi Saray Museum by Sotheby’s. Raymond A. 1974. Artisans et commerçants au Caire au XVIIIe siècle (2 volumes). Damascus : Institut français de Damas. Staniforth M. & Nash M. 1998. Chinese export porcelain from the wreck of the Sydney Cove (1797) (The Australian Institute for Maritime Archaeology, Special Publication No. 12). Adelaide: Brolga Press for the Australian Institute for Maritime Archaeology. Tuchscherer M. 1994. Approvisionnement de villes saintes d’Arabie en blé d’Égypte, d’après des documents ottomans des années 1670. Anatolia Moderna/Yeni Anadolu 5: 79-100. ----------1997. A propos de l’assemblage de trois navires ottomans dans l’arsenal de Suez (1762-1767). Pages 323-334 in volume 1 of C. Villain-Gandossi, L. Durteste & S. Busuttil (eds), Méditerranée, mer ouverte. (Actes du Colloque de Marseille, 21-23 septembre 1995). (2 volumes). Malta: University of Malta, Foundation of International Studies. ----------2001. Production et commerce du café en Mer Rouge au XVIe siècle. Pages 69-90 in M. Tuchscherer (ed.) Le Café avant l’ère des plantations coloniales: espaces, réseaux, sociétés (XVe-XVIIIe siécle). Cairo: Institut français d’archeologie orientale. --------2002. Trade and port cities in the Red Sea-Gulf of Aden Region in the Sixteenth and Seventeenth Century. Pages 28-45 in L.T. Fawaz & C.A. Bayly (eds), Modernity and Culture from the Mediterranean to the Indian Ocean. New York: Columbia University Press. Ward C. 2000a. The Sadana Island shipwreck. Pages 185-202 in U. Baram & L. Carroll (eds), A Historical Archaeology of the Ottoman Empire. New York: Plenum. ----------2000b. Coconuts, coffee and commerce. Scientific American Discovering Archaeology 2/4: 3237. Ward C. 2001. The Sadana Island shipwreck: an eighteenth-century AD merchantman off the Red Sea coast of Egypt. World Archaeology 32: 371-385. Wästfelt B., Gyllensvärd B. & Weibull J. 1991. Porcelain from the East Indiaman. Götheborg. Sweden: Wikens Förlag. Whitcomb D. & Johnson, J. 1979. Quseir al-Qadim 1978: Preliminary Report. (American Research Center in Egypt Reports, 1). Cairo: American Research Center in Egypt. ---------- 1982. Quseir al-Qadim 1980: Preliminary Report. (American Research Center in Egypt Reports 7). Cairo: American Research Center in Egypt. 171

Index Abba Pantaleon, 80 Abyssinia, 16, 20, 21, 22, 23, 37, 134 Acheh, 141 Addi Galamo, 60, 62 Addi Gerametem, 72 Aden, 7, 9, 12, 13, 14, 15, 16, 19, 20, 21, 22, 23, 35, 53, 63, 72, 93, 132, 138, 139, 140, 141, 142, 157, 158, 159, 160 Adi Keyh, 89 Adulis, 7, 19, 21, 29, 59, 65, 71, 72, 73, 80, 87, 89 Aegean, 108 Africa, 11, 12, 13, 16, 17, 19, 20, 21, 22, 25, 26, 27, 28, 30, 33, 34, 37, 38, 39, 52, 57, 59, 60, 63, 65, 71, 72, 73, 74, 79, 82, 87, 93, 94, 95, 97, 100, 101, 102, 125, 128, 131, 133, 134, 135, 138, 145, 154, 155, 157, 158, 159, 160, 162, 165 Afro-Arabian, 52, 65, 67, 71, 73, 74 Agatharchides, 108 Agordat, 72 Akhenaten, 27, 30 Akkele Guzay, 72, 73, 74 Aksum, 58, 60, 71, 72, 73, 74, 75, 79, 80, 81, 82, 83, 84, 87, 89, 99, 112 Aksumite, 19, 53, 57, 58, 60, 87, 89, 93, 101, 105, 112 alabaster, 33 Albuquerque, Afonso de, 14, 15, 21, 138, 139 Aleppo, 133 Alexander the Great, 93, 132, 133, 134, 135 Alexandria, 108, 132, 137, 170 Algiers, 140 Almeida, Manoel de, 22 altars, 64, 72, 73, 111 Alvares, Francisco, 20, 21, 22 Amarna, 38 Amau, 26, 27, 28, 30 amber, 87, 159 Amenemhat II, 25 Amenophis III, 27 amphora, 82 Andalucia, 132, 134 antimony, 22 Antwerp, 137, 138 antyu [Κntyw], 25, 33, 34, 35, 37, 38, 39, 40

Aphilas, 112 Aphrodite, 111 Apis, 110 ΚAqabah, 13 Aqiq, 27, 71, 72 Arab(s), 7, 13, 16, 19, 20, 22, 23, 38, 93, 96, 101, 139, 148, 155, 157, 158, 159, 160 Arabia, 7, 9, 11, 12, 13, 19, 20, 21, 22, 23, 25, 26, 29, 30, 33, 45, 52, 53, 59, 60, 63, 65, 71, 72, 73, 74, 82, 87, 94, 95, 102, 110, 111, 123, 131, 132, 134, 138, 139, 141, 153, 155, 157, 158, 159, 160, 162, 165 Arabic, 12, 13, 62, 95, 117, 120, 131, 132, 134, 135, 157, 158, 166, 168, 170 Aramaic, 62, 105 architecture, 53, 57, 59, 60, 62, 63, 64, 72, 74, 80 Arogi, 37 aromatics, 25, 26, 27, 28, 29, 33, 38 Arsinoë (see also Suez), 7 art, 51, 53, 65, 66, 72, 111, 137 Asab, 158 ashlar, 62, 63, 79, 81 Asia, 19, 20, 27, 29, 30, 33, 35, 38, 65, 93, 95, 97, 100, 101, 102, 111, 112, 127, 128, 131, 158, 159, 161 Asiatics, 25 Asmara, 29, 57, 59, 60, 72, 73 Asoka, 108 Aswan, 12, 105 Atbai, 71 Atbara, 27 Atlantic, 95, 101, 131, 133, 134 Augustus, Caesar, 93 Aurelius, Marcus, 110, 111 Avalites, 19, 20 Awsa, 161, 162 Awwām temple, 62 ΚAydhāb, 7, 8, 9, 11, 12, 13, 15, 16, 17, 123, 148 ΚAynūnag, 7 Ayyubid, 117, 118, 124 Azania, 28, 29 al-Azhar, 157 Baahti Nebait, 82 Bab al-Mandab, 16, 28, 139, 141, 165 baboons, 26, 33 Baghdad, 132 172

Bahr el-Ghazal, 38 Balboa, 133 Balkans, 165 balsam, 133 bamboo, 111, 112 banana, 100, 102 Banāt ΚĀd, 53 Bandar Harshau, 19 Bandar Kassim, 19 Baraqish, 73 Barbaria, 20 Barbarossa, Khayr al-Dīn, 139, 140 Barbosa, Duarte, 21, 22 Barka Delta, 71 barley, 21, 60, 82, 124, 128 Barq al-Yamānī fī l-fatΉ alΚUΕmānī, 134 Barugaza, 19 basalt, 46, 47, 50, 53, 108 Bawerded, 38 Baylūl, 158, 160, 161 bdellium (see also Theban palm), 12 beads, 22, 37, 40, 45, 46, 47, 49, 51, 60, 97, 99, 100, 108, 111, 112 beer, 35 benzoin, 141 Berbera (see also Malao), 19, 20, 21, 22, 29, 157, 158, 159 Berechei, 111 Berenike, 8, 11, 12, 16, 97, 105, 108, 110, 111, 112, 113, 123, 128, 165 Bieta Giyorgis, 59, 74 al-Birūnī, 131 boats, 7, 9, 11, 12, 16, 19, 25, 27, 29, 33, 95, 131, 137, 138, 139, 165 Bombay, 9, 35 bone, 45, 49, 50, 51, 112, 124, 126 Boswellia (see also frankincense), 34, 35 Britian, 152, 153 bronze, 35, 37, 38, 53, 72, 74, 81, 111, 112 Bronze Age, 52, 53, 72 Bruce, James, 22, 152 al-BuqΚah, 139 Burckhardt, J.L., 9 Buri peninsula, 73 Burseraceae (see also antyu), 33 Cabral, 133

INDEX Cadiz, 95, 132 Cairo, 9, 12, 15, 20, 22, 117, 118, 132, 133, 134, 138, 139, 140, 142, 150, 154, 157, 159, 160, 162, 165, 170, 171 Cairo Geniza papers, 118 Cairo Museum, 33 Calcutta, 9 Calicut, 131, 132, 134, 135, 137 Cambay, 21, 22 camels, 21, 22, 111, 112, 123, 126, 165 camphor, 141 Cape Guardafui, 22, 29 caravans, 8, 13, 20, 21, 22, 39, 89, 110, 111, 157, 159, 160, 162, 166 cardamom, 123 carnelian, 97, 159 cassia, 19, 95, 100, 101 Castilian, 131 Castro, Don João de, 13, 15, 16, 145, 148 Catalans, 132 catfish, 26 caves, 95, 97, 98, 100, 101 Ceibão, 15 Celer, Aemilius, 111 cemeteries, 50, 52, 53, 72, 80, 112 cereals, 21, 23, 80, 82, 97, 124, 158 Ceuta, 132, 134 chalcedony, 81 Chaul, 138 chickpeas, 125 chilli, 153 China, 87, 123, 127, 131, 132, 134, 150, 159, 165, 166, 168 Christian Topography, 19 Christian(s), 16, 19, 20, 22, 80, 112, 133, 135, 157, 158, 160, 161, 162 churches, 22, 80, 112, 157 cinnamon, 19, 87, 95, 100, 101, 123, 141 clothing, 22, 118 cloves, 123, 160 Clysma (see also Suez), 7, 11 coconuts, 12, 95, 97, 100, 102, 112, 123, 127, 128, 166, 170 coffee, 9, 22, 158, 159, 160, 161, 162, 165, 166, 170 coins, 93, 97, 108, 111, 112 Columbus, 133 Conti, Niccolò de’, 20 copper, 22, 29, 37, 46, 51, 53, 57, 60, 81, 84, 94, 108, 165, 168 Coptic, 105, 157

Coptos (see also Qift), 12, 25, 27, 111, 128, 145 Copts, 133 coral, 87, 159, 166 Corsali, Andrea, 20, 21, 22 cosmetics, 33 cotton, 19, 21, 22, 95, 100, 112, 150, 159, 160, 162, 170 crops, 21, 82, 84, 117, 121, 124, 125, 127, 128 Crusades, 13 D site, 58, 60, 80, 81, 82, 83, 84 DΚMT (Daamat), 57, 60, 62, 63, 64, 65, 71, 72, 73, 79, 80, 81, 83, 84 Dahlak, 15, 16, 53, 71, 159 Dahlak Kabir, 53, 71 Dahshur, 33 Damascus, 118, 137 Danakil, 73, 81, 161 Darfur, 26, 159 Debarwa, 21 Deir el-Bahri, 26, 27, 29, 34, 35, 39 Delaware, 105, 111 Demotic, 105, 108 Dhofar, 21, 22, 26, 34, 35, 159 dhow, 11, 13, 17 al-Dhubayr, 155 Dilmun, 29 al-Dīn, Imām Sharaf, 139, 141 dinars, 117, 118, 121, 159 Dinqash, 12 Diodorus Siculus, 93, 94, 95, 100 Diogenes, 97, 101 dirhams, 117, 118 Diu, 16, 138 Djedkarê‘-Isesi, 38 Djibouti, 28, 29, 71 dogs, 39 donkeys, 27, 123, 126 doum palms, 26, 50 drugs, 21 durum wheat, 124 dwarfs, 25, 26, 38, 165 Early Iron Working period, 99 Early Stone Age, 79 East Africa, 25, 26, 27, 29, 30, 37, 38, 58, 80, 87, 89, 93, 94, 95, 97, 99, 100, 101, 102, 118, 131, 159 Eastern Desert Ware, 112 ebony, 26, 27, 33, 87 Edfu, 108, 148 Edom, 7 Egypt, 7, 9, 12, 13, 14, 15, 17, 20, 25, 26, 27, 28, 29, 30, 33, 34, 35, 37, 38, 39, 40, 50, 53, 64, 71, 80, 81, 87, 93, 95, 97, 98, 99, 100, 101, 105, 108, 173

110, 111, 112, 113, 117, 118, 121, 123, 124, 127, 132, 133, 134, 137, 140, 141, 142, 145, 148, 152, 153, 154, 155, 157, 158, 160, 161, 162, 165, 166, 168 Egypto-Graeco-Roman, 93, 97, 102 Egypto-Roman, 93 Eighteenth Dynasty, 35 Elath, 7 electrum, 25, 26, 29 Elephantine, 34 elephants, 19, 84, 87, 94, 108, 165 Eleventh Dynasty, 25 El-Kab, 30 emmer wheat, 82, 124 Enda Cherqos, 62 Endasellassie, 83 Eratosthenes, 95 Eritrea, 7, 26, 27, 28, 50, 53, 57, 59, 60, 65, 71, 72, 73, 74, 75, 79, 82, 87, 89, 108 escargot, 111, 112 Esna, 99 Ethiopia, 7, 16, 19, 20, 21, 22, 23, 26, 27, 29, 30, 53, 57, 58, 59, 60, 71, 73, 74, 79, 80, 82, 83, 84, 87, 89, 95, 135, 138, 157, 158, 159, 160, 161, 162 Ethio-Sabaean, 57, 59, 73, 80, 81 Eudoxus of Cyzicus, 95, 101 Euhemerus, 93, 94, 95, 100, 102 Europe, 20, 37, 38, 39, 87, 89, 132, 133, 135, 138, 159, 161, 165, 166, 168, 170 Eziongeber, 7 faience, 37, 81, 108 Farasān, 159 fava beans, 125 al-Fazza, 43 Ferrand, Gabriel, 134 Fifth Dynasty, 34, 38 Fikiya, 72 firearms, 37, 38, 138, 139, 140, 142, 159, 170 firmān, 148, 154, 155 Foro, 89 France, 131, 152, 153 frankincense (see also incense), 20, 23, 26, 34, 35, 53, 87 Fustat, 118 Gades, 95, 101 galena, 33 Gama, Estevão da, 16 Gama, Vasco da, 14, 16, 131, 132, 133, 134, 141 Gash Delta, 52, 53 Gash Group, 52, 71, 72

TRADE AND TRAVEL IN THE RED SEA REGION gazelles, 94 Gebbanitae, 28, 29 Genebtyu, 28, 29 Genoese, 131, 132 al-Ghawrī, Qānsawh, 133, 137, 138, 139 Ghubbat-Kharab, 29 Gibraltar, 132, 134 ginger, 141 giraffes, 26, 34, 165 glass, 19, 22, 23, 81, 97, 100, 165, 166, 168 Goa, 16, 138 gold, 20, 21, 22, 23, 25, 26, 27, 29, 30, 34, 45, 46, 47, 49, 51, 81, 94, 97, 117, 118, 121, 134, 140, 141, 142, 159, 160, 161, 165, 166, 168 Gondar, 22, 160, 162 Graeco-Egyptian, 19 Graeco-Roman, 93, 95, 97, 98, 99, 102, 145 graffiti, 108, 110, 111, 112 Granada, 132, 133 granite, 46, 47 grapes, 125 Grat BeΚal Gebri, 62, 73 graves, 33, 46, 47, 50, 52, 71, 74, 80, 81 Great Harris Papyrus, 27 Greece, 34, 87, 93, 95, 97, 99, 100, 105, 132 Greek, 105, 110, 111 Gujerat, 16, 138, 141, 160, 162 Gulf of Aden, 19, 20, 21, 22, 23, 71, 157, 158, 159, 160, 161, 162 Gulf of ΚAqabah, 7 Gulf of Zula, 50, 71, 73 gums (see also resins), 12, 19, 33, 34, 100, 159 gypsum, 111 Haddas river, 73 Hadramawt, 63, 157, 159, 160, 162, 166 hajj, 8, 9, 150, 153, 154, 155 Hamasien, 72 al-Hāmid, 73 Haremhab, 27 Harkhuf, 25, 26, 29, 39 Harsiyotef, 74 Hatshepsut, 26, 27, 28, 29, 35, 37, 39 hawālah, 117, 160 Hawlti, 58, 60, 62, 64, 73, 74, 80, 81, 82, 83, 84 Hays, 19, 152, 153 al-Дazm, 53 hbny (see also ebony), 39 Hebrew, 12, 105, 112

Heliopolis, 133 Hellenistic, 93, 99, 102 Henenu, 25 Hergigo, 20 Hierobol/Yarhibol, 111 Hieroglyphic, 105, 110, 111, 145 Hijaz, 7, 9, 13, 16, 118, 145, 152, 154, 155, 157, 158, 159, 160, 161, 162, 165 Himyar, 20, 62, 99 Hipparchus, 95 hippopotamus, 83, 158 Hishmale, 89 honey, 21, 89, 158 Hormuz, 14, 141 horn, 19, 29, 30, 97 Horn of Africa, 19, 20, 21, 22, 23, 40, 43, 52, 53, 57, 58, 59, 60, 62, 63, 64, 65, 71, 73, 95, 99, 132 horses, 21, 111, 126, 133, 137, 138, 158, 159 htyw (terraces), 35 al-Hudaydah, 22, 139, 160 Iam, 25, 29 Iambulus, 93, 94, 95, 100 ibexes, 62, 63 Ibn BaΓΓūta, 20 Ibn Hawqal, 20 Ibn Iyās, 132, 133, 134, 137 Ibn Jubayr, 8, 12, 13 Ibn Mājid, AΉmad, 9, 11, 13, 14, 15, 131, 133, 134, 135 Ibn SaΚīd, 20 Ichthyophagoi, 87, 112 al-Idrīsī, 20, 131 ihmut (see also incense), 38 incense (see also frankincense), 19, 21, 22, 23, 25, 26, 28, 34, 35, 38, 60, 64, 72, 159, 165, 166 incense burners, 51, 52, 60, 168 India, 9, 12, 13, 14, 15, 16, 19, 20, 21, 22, 23, 87, 93, 95, 96, 97, 98, 99, 101, 108, 111, 112, 123, 126, 127, 131, 132, 133, 134, 135, 137, 138, 139, 140, 141, 145, 158, 159, 160, 161, 162, 165, 166, 168, 170 Indian Ocean, 7, 14, 15, 65, 71, 87, 93, 95, 97, 101, 102, 111, 112, 117, 131, 132, 133, 134, 135, 137, 139, 140, 141, 157, 159, 160, 161, 165, 166, 170 Indies, 21, 140 inscriptions, 25, 29, 51, 52, 59, 60, 62, 63, 64, 65, 72, 73, 79, 80, 81, 110, 111, 112, 168, 170 Irem, 25, 27, 28, 29 174

iron, 37, 81, 94, 97, 99, 100, 102, 108, 159, 170 Irwin, E., 150, 152, 154 Isesi, 25 Isis, 111, 112 Islam(ic), 7, 8, 9, 11, 13, 22, 43, 52, 53, 87, 105, 118, 123, 124, 125, 126, 128, 131, 132, 133, 134, 145, 154, 157, 161, 168, 170 IsmaΜīlī, 142 IsmaΜiliya, 124 Istanbul, 139, 140, 141, 142, 148, 152, 154, 158, 159, 160, 161 Italy, 59 Itwer, 33 ivory, 19, 21, 23, 29, 80, 83, 87, 97, 108, 159, 165 Jabal ΚArafāt, 9 Jabal Mokram Group, 72 Jabal Zuqar, 15 Jawf, 53 Jazīrat al-КāΜir, 15 Jerusalem, 157 Jew(ish), 12, 15, 112 jewellery, 49, 51, 81, 84, 168 Jiddah, 7, 8, 9, 11, 12, 13, 14, 15, 16, 20, 21, 22, 132, 138, 139, 140, 141, 142, 154, 157, 158, 159, 160, 161, 162, 165, 166, 170 jilāb (see also boats), 12, 13 Jīzān, 71, 73, 139, 141 Jordan, 13 Juani, 98 Jupiter, 110 Kamarān, 15, 139, 140 Karak, 13 Kārimī, 123, 127 Kaskasè, 62, 72, 73, 89 Kassala, 26, 28, 52, 71 Kawkabān, 142 Kenya, 100 Kerma, 30, 64, 71, 72 Khartoum, 30, 72 Khenty-khety-wer, 25 Kidane Mehret, 58, 60 Kilimanjaro, 97 Kilwa, 118 Kitāb al-fawā’id, 9, 13 Kitāb al-Sufālīyyah, 133, 134 Koloe, 87, 89 Koptos, 128 Kosmas Indicopleustes, 19, 101 al-Kurdī, Husayn, 132, 134, 138, 139, 140 Kush, 30, 73, 93 Kyeneion, 80 Lake Nyanza, 97 Lake Tana, 84

INDEX Late Stone Age, 81, 82, 95, 98, 100, 102 Latin, 105 lead, 108 Lebanon, 118 Legatio babilonica, 132 Leiden, 105, 111 Leningrad, 133 lentils, 60, 125 leopards, 33, 34, 94 Lesbos, 137, 139 Leucos Limen, 123 Leuke Kome, 7 Levant, 53, 87 linen, 37 linseed, 60, 82 lions, 34 Lisbon, 14, 33, 135, 137, 141, 143 locusts, 95 Louvre, 137 Ludolph, Job, 22, 23 al-LuΉayyah, 160 LuΉayyah, 22, 139 Macedonia, 105 Madagascar, 97, 100, 134 madder, 140 Mafia, 97, 98, 99, 101, 102 Magdala, 37 Magellan, 133 Maghrib, 131, 165 magic, 33 Mahal Teglinos, 52, 71, 72 Mai Adrasha, 83 Mai Hutsa, 60 MaΚīn, 63, 73 makeir, 19 Malacca, 14 malachite, 33 Malao (see also Berbera), 19, 20 MaΜlayba, 52, 72 Malindi, 134 Mamluk, 118, 124, 126, 127, 128, 132, 133, 134, 137, 138, 139, 140, 141, 142, 145, 154, 155, 158, 160, 161, 165 Mārib, 45, 62, 63 Marj Dābiq, 133 Marsa Halaib, 12 Massawa, 15, 16, 17, 19, 20, 21, 22, 29, 73, 158, 159, 160, 161 al-MasΚūdī, 131 Matara, 59, 62, 65, 72, 74, 80, 82, 84, 87, 89 Matariyah, 133 MawzaΜ (see also Muza), 19 maydān, 137 Mecca, 7, 8, 9, 12, 15, 20, 21, 22, 117, 118, 121, 132, 134, 138,

139, 140, 141, 142, 155, 157, 158, 160, 161, 165 Medaguè, 74 medicine, 33, 37 medieval, 13, 20, 53, 117, 123, 124, 125, 126, 127, 128, 148, 154, 165 Medina, 7, 13, 132, 138, 155, 157, 158 Mediterranean, 7, 12, 14, 34, 71, 93, 95, 97, 98, 99, 100, 101, 102, 108, 110, 111, 112, 118, 123, 131, 132, 133, 138, 139, 142, 157, 158, 159, 160, 161, 162, 165, 170 megaliths, 43, 45, 46, 50, 52 Mekelle, 81 Melazo, 58, 60, 62, 64, 73 Memphis, 25, 26, 27, 39, 99 Mentuhotep III Sankhkare, 25 Menuthias, 96, 101 Meroe, 73, 74, 93, 97, 101 Mersa Gawasis (see also Sawaw), 25, 29 Meshwesh, 30 Mesopotamia, 53, 100 metal, 27, 31, 34, 35, 37, 38, 39, 46, 50, 52, 81, 84, 108, 159, 160 al-Midamman, 43, 45, 50, 51, 52, 53 Middle East, 93, 166, 168 Middle Kingdom, 25, 26, 29, 33, 35 Middle Stone Age, 87 millet, 21 Minā, 9 mines, 34, 37, 94, 134, 165 MKRB (mukarrib), 62, 64, 73 MLK (malik), 62, 64 Mocha, 19, 20, 22, 139, 141, 157, 158, 159, 160, 161, 166 Mokimos, Marcus Aurelius, 111 Moluccas, 123 Mombasa, 34 Monçaide, 131, 135 Mondou, 19 monsoon, 7, 9, 14, 15, 95, 165 Moors, 22, 139 Morocco, 20, 132, 134, 138 mosques, 43 Mosullon, 19 Mozambique, 96 mudbrick, 49, 50, 150, 152 MuΉammad ΚAli, 7, 153 Muhimme Defteri, 148 Muslim(s), 7, 9, 13, 14, 15, 20, 22, 23, 53, 101, 118, 131, 133, 135, 138, 157, 158, 160, 161, 162, 165, 166, 168, 170 175

muslin, 132, 170 al-Mutaynah, 139 Muwaylih, 157, 158 Muza (see also MawzaΜ), 19 Myos Hormos (see also Quseir al-Qadim), 7, 8, 11, 123 myrrh, 19, 25, 26, 27, 28, 87, 159 Mytilene, 139 Nabataea, 62, 99 al-Nahrawālī, Qutb al-Dīn, 134, 135, 139, 142 Najrān, 162 Napata, 39, 57, 64, 98 Napier, General, 89 Napoleon, 150, 152, 153 Nastasen, 74 nautilus shells, 97, 100 Navigazioni e Viaggi, 135 Near East, 51, 60, 82, 84, 97, 100, 103, 112, 117, 118, 128, 165 Necho, 93 Neolithic, 43, 45, 87, 95, 97, 98, 102 Nero, 101, 110 New Kingdom, 34, 35, 38, 39, 53 Niebuhr, Carsten, 9, 165 Nile, 7, 8, 11, 12, 13, 17, 25, 26, 27, 28, 30, 33, 34, 39, 40, 57, 64, 71, 73, 80, 83, 84, 93, 95, 97, 98, 100, 101, 105, 108, 110, 111, 117, 123, 124, 126, 134, 135, 145, 150, 152, 154, 158, 161, 170 nomads, 34, 124, 138, 140, 148, 155, 160 Noronha, Garcia de, 16 Ntanei, 72 Nubia, 20, 25, 27, 30, 33, 34, 39, 64, 71, 73, 74, 80, 81, 98, 108, 112, 134, 157, 160 Nubian C-Group, 52, 71 Nunes, João, 131, 135 nutmeg, 123 Obock, 29 obsidian, 45, 46, 50, 51, 52, 53, 71, 81, 84, 165 Okalis, 19 Old Kingdom, 25, 26, 27, 33, 34 Olympiodorus, 112 Ona, 57, 59, 60 Ona Culture, 72, 73 Ona Gudo, 60 Ophir, 29 opium, 159 Opone (see also Ras Hafun), 19 Osiris, 110 ostraca, 108, 110, 112 ostriches, 33, 45, 49, 51, 108

TRADE AND TRAVEL IN THE RED SEA REGION Ottoman, 7, 11, 15, 16, 53, 133, 134, 137, 139, 140, 141, 142, 145, 150, 154, 155, 157, 158, 159, 160, 161, 162, 165, 170 Palermo Stone, 25, 34 Palestine, 7, 132 palm, 50, 51 Palmyra, 62, 111 Palmyrene, 105, 110, 111 Panara, 94, 100, 102 Panchaea/Azania, 93, 94, 95, 96, 97, 99, 100, 101, 102 Pan-Grave Culture, 72 panthers, 20 papyri, 108, 110 Parehu, 26, 29, 37, 38, 39 Partho-Sasanian, 97, 99 Pasha, Ibrahīm, 7 Pasha, Sinān, 141, 142 Pasha, Suleyman, 16, 141 pearls, 12, 87, 159, 166, 170 pepper, 21, 22, 33, 87, 111, 112, 123, 126, 127, 128, 141, 160 Pepy II, 25, 26, 39 Pepynakht, 25 Pereira, Bernado, 21 Perim Island, 19 Periplus of the Erythraean Sea, 19, 20, 28, 29, 80, 83, 84, 87, 89, 95, 96, 97, 100, 101, 102, 165 Persia, 20, 21, 22, 95, 97, 101, 159 Philotera, 110 Phoenicia, 108 Phoenicia(n), 29, 99 pilgrimage, 7, 8, 9, 12, 13, 15, 132, 148, 154, 155, 157, 160 pilgrims, 8, 9, 12, 13, 117, 134, 154, 155, 158, 160, 162, 165 pilots, 13, 15, 134 pipes, 150, 152, 153, 168 Pires, Tomé, 21, 22 Pistacia, 33, 38 Pliny, 12, 16, 28, 87, 95, 100, 101, 105 porcelain, 150, 152, 153, 159, 166, 168, 170 ports, 19, 20, 21, 22, 23, 33, 34, 87, 105, 108, 110, 112, 113, 117, 118, 123, 124, 128, 131, 137, 138, 139, 141, 142, 145, 148, 154, 155, 157, 158, 159, 160, 161, 165 Portugal, 7, 13, 14, 15, 16, 21, 22, 33, 101, 131, 132, 133, 134, 135, 137, 138, 139, 140, 141, 142, 145, 148, 155, 157, 158, 161 potters, 80

pottery, 28, 33, 35, 43, 45, 46, 47, 50, 51, 52, 53, 60, 63, 65, 71, 72, 73, 74, 80, 81, 82, 84, 87, 97, 98, 99, 100, 108, 110, 111, 112, 150, 152, 153, 165, 166, 168 Prakrit-Sanskrit, 105 pre-Aksumite, 57, 58, 59, 60, 62, 63, 64, 65, 71, 72, 73, 74, 79, 80, 81, 82, 83, 84, 87 pre-Islamic, 8, 43, 82, 99 Prester John, 15, 16, 21, 135, 138 Procopius, 105, 112 proto-Aksumite, 57, 58, 74 Ptolemaic period, 11, 93, 97, 105, 108, 112, 113, 124, 145 Ptolemy II, 30, 105 Ptolemy the Geographer, 16, 19, 34, 95, 96, 97, 99, 100, 101, 102, 131 Ptolemy X, 30 Punt, 25, 33, 34, 35, 37, 38, 39, 40, 50, 53, 93, 95, 99, 100, 165 pygmies, 25, 26, 38 Qalat Sai, 150 Qasr Ibrim, 145, 150 Qatabān, 28, 63 Qena, 154 Qift (see also Coptos), 12, 145, 154 Qina, 118 Qohaito, 72, 73, 87, 89 Qunfudhah, 22 Qur’ān, 132 Qus, 8, 12, 118, 128, 148 Quseir, 7, 8, 15, 16, 17, 27, 28, 29, 117, 118, 123, 124, 126, 127, 128, 145, 147, 148, 150, 152, 153, 154, 155, 158, 165 Quseir al-Qadim (see also Myos Hormos), 8, 105, 123, 125, 128, 145, 148, 155 Quseir Fort, 145, 148, 154, 155 al-rahn, 117 RaΜīs, Salmān, 132, 134, 139, 140, 141 rafts, 19, 50 Raheyta, 158 Ramesses II, 27, 28, 29, 30 Ramesses III, 27, 30, 35, 38 Ramesses IX, 30 Ramusio, 135 Ras Abu Fatma, 12 Ras Banas (see also Berenike), 11, 16, 35 Ras Hafun, 19, 28, 34, 97, 99, 100 Ras Shagarah, 26 Rasulid, 161 176

religion, 33, 57, 59, 62, 63, 64, 65, 87, 105, 108, 112, 157, 158 resin (see also gums), 33, 34, 38, 165, 170 Retjenu, 33 Reynauld de Châtillon, 13 Rhapta, 29, 96, 99, 100, 101, 102 rhinoceroses, 19, 26, 30, 34, 87, 97 Rhodes, 97, 108 rhyolite, 46, 47, 49, 50 rice, 95, 100, 111, 112, 123, 127, 128, 158 Rift Valley, 95, 100 Rio Cuama, 134 rock art, 59, 72 Rohlfs, Gerhard, 89 Roman(s), 11, 87, 93, 97, 99, 100, 101, 102, 105, 108, 110, 111, 112, 113, 123, 124, 125, 126, 127, 128, 145, 148, 150, 154, 155, 165 Rome, 87, 126 roses, 21 Roteiro, 13, 16, 131, 145 Rudrasena III, 112 Rufiji, 96, 97, 101, 102 Russia, 133 Ruwenzori mountains, 97 Saba’, 43, 51, 53, 57, 59, 60, 62, 63, 64, 65, 71, 72, 73, 74, 79, 80, 81, 82, 84, 99 Sabaic, 62, 72, 73 Sabr, 52, 53, 65, 72 Sabr Culture, 72 Sadana Island, 165, 166, 168, 170 Safagah, 8, 25 Sahel, 71, 74 Sahure, 25, 34 sailors, 7, 13, 19, 25, 34, 35, 50, 87, 93, 95, 101, 111, 165 Salalah, 35 salt, 81 СanΚāΜ, 21, 139, 157, 161 sandalwood, 141 Sardes, 35 Sasanian, 93, 99, 101 Saudi Arabia, 12, 71, 72 Sawaw (see also Mersa Gawasis), 25, 26, 27, 28, 29 Saybān, 15 sculpture, 37, 52, 59, 60, 73, 74, 79, 80, 81, 108, 111, 112 seals, 72 Seglamien, 80 Selim I, 140 Selim II, 148 Selim the Grim, 133 Sembel, 60

INDEX Sembel Kushet, 60, 72 Semitic, 110, 111 Senafe, 59 Sennar, 157, 159, 160, 161, 162 Sequeira, Diego Lopes de, 21 Serapis temple, 108, 110, 111, 112 Sesostris I, 25 Set, 34 Sethos I, 27, 30 ShāfīΚīs, 118 Shawbak, 13 shell, 22, 29, 98, 100, 112, 126, 166, 170 Shenshef, 110 Shihr, 157, 158, 159, 160 ships, 7, 9, 11, 12, 13, 15, 16, 17, 19, 20, 21, 22, 25, 26, 27, 28, 29, 35, 39, 53, 108, 131, 132, 134, 137, 138, 139, 140, 141, 145, 152, 158, 159, 160, 161, 165, 166, 168, 170 shipwreck, 35, 95, 134, 165, 166, 168 Shire, 81, 83 Siberia, 87 Sihi, 65, 71, 72 silk, 19, 22, 87, 123, 141, 165 silver, 25, 26, 94, 97, 117, 118, 159, 160, 161, 162 Sinai, 33, 34, 154, 155, 157, 165 Sixth Dynasty, 34, 39 slaves, 19, 20, 21, 22, 23, 39, 158, 159, 160, 162 sntr (see also Pistacia, incense and resin), 33, 38 Socotra, 16, 138 Sofala, 131, 133, 134 Somalia, 25, 26, 27, 28, 29, 30, 96, 101, 159, 160 Somaliland, 34, 35 sorghum, 111, 112, 125, 158 South Arabia, 20, 25, 26, 28, 29, 30, 35, 50, 51, 52, 53, 57, 59, 60, 62, 63, 64, 65, 71, 72, 73, 74, 82, 84, 100, 105, 108, 111, 123, 138 Spain, 95, 101, 131, 132, 159 spices, 19, 20, 21, 87, 95, 100, 101, 123, 124, 126, 137, 141, 157, 158, 159, 160, 161, 162, 165, 166, 168 Sri Lanka, 97, 111, 112 St. John’s Island, 35 statues, 37, 111 stone, 34, 35, 38, 45, 46, 47, 49, 50, 51, 52, 53, 59, 60, 62, 63, 72, 73, 79, 80, 81, 87, 100, 108, 110, 111, 145, 165 Strabo, 95, 100, 101

Suakin, 12, 13, 15, 16, 26, 28, 30, 138, 141, 157, 158, 159, 160, 161 Sudan, 12, 16, 25, 26, 27, 30, 34, 38, 39, 52, 53, 71, 72, 73, 74, 95, 108 Suez, 7, 9, 11, 15, 16, 17, 29, 105, 124, 132, 138, 139, 154, 157, 158, 160, 165, 166, 170 sugar, 123 sulphur, 141 Sumatra, 141 Surat, 9, 22, 160, 162 al-Suyūtī, 132, 134 synagogues, 112 Syria, 132, 154, 155, 158, 159, 160 TaΚizz, 161 Taconda, 89 Tahirid, 138, 139, 140, 141 Takaze River, 72 takkīyyah, 11 Tamil-Brahmi, 105, 110 Tanzania, 28, 93, 95, 96, 99, 100, 101, 102 Tarikh-i Hind-i Gharbi, 134, 135 tea, 166 teak, 111, 112, 170 Tell Asmar, 100 Tell Defenneh stela, 30 temples, 26, 35, 53, 62, 63, 64, 73, 74, 80, 81, 110, 112, 145 termis beans, 125 terraces, 35, 37, 38 textiles, 19, 20, 21, 22, 23, 108, 111, 112, 118, 123, 137, 140, 159, 170 Theban palm (see also bdellium), 12 Thebes, 25, 26, 27, 30, 35, 39 Theodoros II, 89 Theophrastus, 35 Third Intermediate Period, 93, 99 Thomas (Brother Thomas), 20, 22 Thulā, 140 Tigray, 58, 59, 62, 65, 71, 72, 73, 74 Tihāmah, 7, 12, 43, 52, 53, 65, 71, 72, 74, 139, 141 tilapia, 26 tin, 29, 94 tobacco, 150, 152 Toconda, 87, 89 tools, 23, 45, 46, 50, 51, 52, 53, 60, 63, 71, 73, 81, 84, 87, 100 Topkapı Saray Museum, 168 tortoise shell, 19, 87, 97 trade, 7, 8, 9, 12, 13, 14, 15, 19, 20, 21, 22, 23, 28, 33, 34, 35, 177

37, 39, 62, 64, 65, 71, 87, 89, 93, 95, 96, 97, 98, 99, 100, 101, 102, 105, 108, 110, 111, 112, 117, 123, 126, 131, 133, 137, 138, 139, 140, 141, 145, 148, 155, 157, 158, 159, 160, 161, 162, 165, 166, 168, 170 Trajan, 111 trees, 12, 26, 33, 34, 35, 38, 133 Trevisan, Domenico, 137 Tripoli, 118 Tunis, 131, 140 al-Tur, 15, 16, 157 Turkey, 134, 137, 139, 140, 141, 142, 159 Turks, 15, 158, 160, 165 turquoise, 165 Tuthmosis III, 26, 27, 28, 29, 30 Twelfth Dynasty, 25, 33 Twentieth Dynasty, 35 Twenty-Sixth Dynasty, 27, 30 Ubhur creek, 13 Varthema (Ludovico di Varthema), 20, 21 Venice, 14, 20, 22, 123, 131, 132, 133, 137, 141, 159 Verus, Lucius, 110, 111 Vienna, 87 Wādī Ambagi, 145 Wādī Haddas, 89 Wādī Hammamat, 25, 27, 145, 154, 165 Wādī Kalalat, 110, 111 Wādī Komalie, 89 Wādī Lahma, 110 Wādī Sihām, 73 Wādī UrqΚ, 65, 72 Wahhabis, 153 al-Wajh, 155 weapons, 19, 35, 37, 38, 39, 40, 100, 170 Weigall, 145 Wellsted, Lt. J.R., 7, 9, 11 wheat, 21, 60, 82, 123, 124, 128, 141, 145, 152, 154, 155, 158 wine, 35, 97, 108, 153 wood, 12, 28, 38, 39, 49, 50, 111, 112, 123, 138, 158, 166, 168, 170 Yamites, 39 Yanbu, 7, 132, 154, 158 Yeha, 58, 60, 62, 63, 64, 65, 72, 73, 80, 81, 82, 83, 89 Yeha alabastron, 82, 84 Yemen, 7, 9, 12, 16, 20, 22, 43, 50, 52, 53, 59, 62, 71, 72, 73, 74, 79, 82, 123, 132, 137, 138, 139, 140, 141, 142, 145, 157, 158, 159, 160, 161, 162, 165

TRADE AND TRAVEL IN THE RED SEA REGION Zabīd, 20, 21, 43, 139, 140, 141, 142, 157, 158 Zambezi, 134 Zanzibar, 95, 96, 97, 100, 101, 102

Zaydī, 139, 141, 142, 161, 162 ZaylaΚ, 19, 20, 21, 22, 158, 159, 160, 161 al-Zayyānī, 132 Zimbabwe, 118

178

Zorzi, Alessandro, 20, 21 Zula, 29, 89 Zuqar Island, 46