The Phoenicians in Spain: An Archaeological Review of the Eighth-Sixth Centuries B.C.E. -- A Collection of Articles Translated from Spanish 9781575065298

Twelve essays, written by various scholars and originally published in Spanish, explore the ways in which Phoenician col

171 32 5MB

English Pages 320 [318] Year 2002

Report DMCA / Copyright

DOWNLOAD FILE

Polecaj historie

The Phoenicians in Spain: An Archaeological Review of the Eighth-Sixth Centuries B.C.E. -- A Collection of Articles Translated from Spanish
 9781575065298

Citation preview

THE PHOENICIANS IN SPAIN

The Phoenicians in Spain An Archaeological Review of the Eighth–Sixth Centuries b.c.e. A Collection of Articles Translated from Spanish

Translated and Edited by

Marilyn R. Bierling Associate Editor

Seymour Gitin

Winona Lake, Indiana

Eisenbrauns 2002

ç Copyright 2002 by Eisenbrauns. All rights reserved. Printed in the United States of America.

Cataloging in Publication Data The Phoenicians in Spain : an archaeological review of the eighth–sixth centuries b.c.e. : a collection of articles translated from Spanish / translated and edited by Marilyn R. Bierling in consultation with Seymour Gitin. p. cm. Includes bibliographical references and index. Contents: The Phoenician settlement of the 8th century b.c. in Morro de Mezquitilla (Algarrobo, Málaga) / Hermanfrid Schubart — The Phoenician settlement at Toscanos : urbanization and function / Hans Georg Niemeyer — Phoenician and Punic Sexi / Manuel Pellicer Catalán — Notes on the economy of the Phoenician settlements in southern Spain / María Eugenia Aubet Semmler — Phoenician trade in the West / María Eugenia Aubet Semmler — La Fonteta : a Phoenician city in the far West / Alfredo González Prats, Antonio García Menárguez, and Elisa Ruiz Segura. Contents : The ancient colonization of Ibiza / Juan Ramón — The ancient Phoenicians of the 8th and 7th centuries b.c. in the Bay of Cádiz / Diego Ruiz Mata — Some questions regarding the Tartessian orientalizing period / María Eugenia Aubet Semmler — The Phoenician impact on Tartessos / María Eugenia Aubet Semmler — The Tartessian economy : mining and metallurgy / Jesús Fernández Jurado — The beginnings of the Phoenician presence in southwestern Andalusia / Diego Ruiz Mata. ISBN 1-57506-056-6 (cloth : alk. paper) 1. Spain—Antiquities, Phoenician. 2. Phoenicians—Spain. 3. Excavations (Archaeology)—Spain. I. Bierling, Marilyn, 1949– II. Gitin, Seymour. DP44.P46 2001 936.6u02—dc21 2001040869

The paper used in this publication meets the minimum requirements of the American National Standard for Information Sciences—Permanence of Paper for Printed Library Materials, ANSI Z39.48-1984.†‰

Contents

Contents Abbreviations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

vii ix xi

Part I The Phoenicians in the Mediterranean The Phoenician Settlement of the 8th Century b.c. in Morro de Mezquitilla (Algarrobo, Málaga) . . . . . . . . . . . . . . . . . . Hermanfrid Schubart

3

University of Munich, and retired director of the Deutsches Archäologisches Institut in Madrid

The Phoenician Settlement at Toscanos: Urbanization and Function . . . Hans Georg Niemeyer

31

University of Hamburg

Phoenician and Punic Sexi . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Manuel Pellicer Catalán

49

University of Seville

Notes on the Economy of the Phoenician Settlements in Southern Spain . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . María Eugenia Aubet Semmler

79

University Pompeu Fabra, Barcelona

Phoenician Trade in the West: Balance and Perspectives . . . . . . . . . . . . María Eugenia Aubet Semmler

97

University Pompeu Fabra, Barcelona

La Fonteta: A Phoenician City in the Far West . . . . . . . . . . . . . . . . . . . 113 Alfredo González Prats, Antonio García Menárguez, and Elisa Ruiz Segura University of Alicante

The Ancient Colonization of Ibiza: Mechanisms and Process . . . . . . . . 127 Juan Ramón Conselleria de Cultura d’Eivissa i Formentera. Servei d’Arqueologia. Ibiza

v

vi

Contents

Part II The Phoenicians in the Area of the Atlantic The Ancient Phoenicians of the 8th and 7th Centuries b.c. in the Bay of Cádiz: State of the Research . . . . . . . . . . . . . . . . . Diego Ruiz Mata

155

University of Cádiz

Some Questions Regarding the Tartessian Orientalizing Period . . . . . . . 199 María Eugenia Aubet Semmler University Pompeu Fabra, Barcelona

The Phoenician Impact on Tartessos: Spheres of Interaction . . . . . . . . . . 225 María Eugenia Aubet Semmler University Pompeu Fabra, Barcelona

The Tartessian Economy: Mining and Metallurgy . . . . . . . . . . . . . . . . . 241 Jesús Fernández Jurado Diputación Provincial of Huelva

The Beginnings of the Phoenician Presence in Southwestern Andalusia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 263 With findings from the excavations at Cabezo de San Pedro (Huelva), San Bartolomé (Almonte, Huelva), Castillo de Doña Blanca (Puerto de Santa María, Cádiz), and El Carambolo (Camas, Sevilla)

Diego Ruiz Mata University of Cádiz

Index of Sites and Geographical Names . . . . . . . . . . . . . . . . . . . . . . . . 299

Abbreviations AA AAA AEArq AuOr BASOR BPH BRAH DPH EAE EH IGME MB MDOG MF MJSEA MM NAHisp Or RSF TrPrHist

Archäologischer Anzeiger Anuario Arqueológico de Andalucía Archivo Español de Arqueología Aula Orientalis. Revista de Estudios del Próximo Oriente Antiguo Bulletin of the American Schools of Oriental Research Biblioteca Praehistorica Hispana Boletín de la Real Academia de la Historia Diputación Provincial de Huelva Excavaciones Arqueológicas en España Excavaciones en Huelva Instituto Geológico y Minero de España Madrider Beiträge Mitteilungen der deutschen Orient-Gesellschaft Madrider Forschungen Memorias de la Junta Superior de Excavaciones Arqueológicas Madrider Mitteilungen Noticiario Arqueológico Hispánico Orientalia Rivista di Studi Fenici Trabajos de Prehistoria

vii

viii

Contents

Preface

Preface This volume was originally conceived in January 1993 in Barcelona when María Eugenia Aubet Semmler and Seymour Gitin were discussing ways in which the results of archaeological research in Spain could be integrated into the international research project, “The Neo-Assyrian Empire in the Seventh Century b.c.: A Study of the Interactions between Center and Periphery,” then being organized by Gitin. They agreed that in order to achieve this goal, contact between researchers in Spain and those working in the ancient Near East would have to be increased, and that the first step in this direction would be to organize the publication of an English translation of articles written in Spanish and German which dealt with Phoenician colonization in Spain in the 7th century b.c.e. It was hoped that this would encourage a dialogue between researchers working on related problems of international trade in the 7th century in the areas of the Mediterranean basin, the Levant, and Mesopotamia. It was only in 1995, however, when Marilyn Bierling, a long-time staff member of the Tel Miqne-Ekron excavations in Israel (of which Gitin is co-director) agreed to organize such a volume that work on the project actually began. The presence of the Phoenicians in Spain is a good example of the effect of Assyrian economic policies on the extended periphery of the empire and, therefore, is a vital part of the Neo-Assyrian project. The project is designed to investigate the general effect of these policies by examining the dynamics of growth and development of the Neo-Assyrian Empire in the 7th century b.c.e., which stimulated the first “world market” in history. The project will evaluate the impact of the Neo-Assyrian Empire on its center—the heartland of Assyria—on its periphery—the provinces and vassal states in the eastern Mediterranean basin—and on its extended periphery in the central and western Mediterranean basin. Among the key issues to be examined are the development of local and regional economic exchange systems, international trade, industrial and agricultural development, the process of urbanization, town planning, cultic syncretisms, ethnic identification, radical demographic movements, land management, diet, and administrative systems as tools of imperial policy. This innovative program incorporates archaeological, historical, and environmental evidence from Mesopotamia, the Levant, and the Mediterranean basin which in antiquity were either part of or related commercially to the ix

x

Preface

Neo-Assyrian Empire. Participants in this integrated international research program include 50 scholars who have worked or currently work in Bahrain, Cyprus, Egypt, Greece (mainland and Crete), Iran, Iraq, Israel, Jordan, Lebanon, Spain, Syria, Tunisia, and Turkey. The project director is Seymour Gitin, Dorot Director and Professor of Archaeology of the W. F. Albright Institute of Archaeological Research in Jerusalem (affiliated with the American Schools of Oriental Research). The administrative coordinator is Mary Ellen Lane, Executive Director of the Council of American Overseas Research Centers at the Smithsonian Institution, under whose aegis the program has been organized. Within the Neo-Assyrian project, the data from Spain can provide an important basis for the analysis of Mediterranean trade. It is the hope of all those who have made this volume possible that it will achieve one of the primary goals of the project—to establish a comprehensive scholarly approach to understanding the dynamic of trade in the 7th century. The publication of this volume is primarily due to the efforts and perseverance of Marilyn Bierling, Professor of Spanish Linguistics at Calvin College, who served as translator and editor, and to the guidance of María Eugenia Aubet Semmler, who was instrumental in choosing the articles and graciously gave of her time to assist throughout the editing process. Appreciation is also extended to the authors of the translated articles for permission to include their research, for their patience in responding to repeated questioning from the editor, for the guided tours of the sites provided to Marilyn Bierling, and for permission to photograph their excavations. Additional thanks go to the organizers of the November 1997 Symposium on Phoenician Pottery at Guardamar del Segura for their kind invitation to Marilyn Bierling to attend the symposium and discuss many of the issues presented in this volume. We are also most grateful to Calvin College for providing grants to support the translation work and travel to Spain to consult with the contributors, to Phoenix Data Systems for permission to use additional photos, and to the Albright Institute’s 7th century project for its printing subvention. S. Gitin Gota a gota, se llena la bota. Drop by drop, the wine bag is filled. (Spanish proverb)

Introduction

Introduction Ancient historians such as Strabo referred to the riches of a far-away land called “Tartessos,” but only in the 1960s did archaeologists begin to uncover extensive evidence that would lead to an information explosion in our knowledge of ancient Tartessos and the Phoenician colonies that appeared on the Iberian Peninsula in the 8th century b.c.e. While earlier chance finds and previous excavations had laid the groundwork, the excavations begun in the 1960s represented a turning point that caught the attention of international scholarship with regard to the phenomenon of Phoenician colonization on the Iberian Peninsula. Since then, intensive research has demonstrated that Phoenician colonization on the Peninsula was far more widespread than formerly believed, with known Phoenician coastal settlements extending from the mouth of the Sado River (south of Lisbon in Portugal), past Gibraltar, to the Segura River in the province of Alicante, Spain. This collection of articles translated from Spanish to English represents an attempt to bring this important aspect of Phoenician commercial activity in the West to the attention of students of ancient Near Eastern studies. It is hoped that the volume will serve as an important research tool for those dealing with the history of Mediterranean trade during the 8th–6th centuries b.c.e. and provide an entrée to the study of the Phoenicians in Spain. The Iberian Peninsula has been home to many peoples and cultures over the centuries and has served as a door from the Mediterranean to the Atlantic and as a bridge between Europe and Africa. The numerous peoples who have settled the area reflect the peninsula’s ancient history and strategic position. The contributors to this volume refer to some of the many cultures that existed in Iberia before the Phoenician colonization, such as the Argaric culture in the dry Southeast and the Cogotas culture in the high, central plateau of the Peninsula. On the eve of the Phoenician colonization in the 8th century b.c.e., the lower Guadalquivir River basin was densely inhabited by a Final Bronze Age population, following the earlier Chalcolithic and Bronze Age cultures which had occupied the area. Huelva, on the southern Atlantic coast, served as a center for the area, later called Tartessos by the Greeks. Huelva was located downstream from the Riotinto mining region, rich in mineral deposits. The Phoenician settlers preferred a certain type of topography for establishing their colonies—generally, a low hill on an island or peninsula at the mouth of a river, which formed a harbor where ships could anchor and xi

xii

1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23.

Santa Olaia Conimbriga Santarém Torres Vedras Lisbon (Olisipo) Setúbal Alcácer do Sal Abul Sines Almogrebe Ourique Rocha Branca Fonte Velha Castro Marim Safara Azougad Ratinhos Aljucén Zarza de Alanje Medellín Valdegamas Siruela La Aliseda

Introduction

24. Villanueva Vera 25. Huelva (La Joya, San Pedro, Esperanza) 26. Niebla 27. Riotinto (Quebrantahuesos, Cerro Salomón) 28. San Bartolomé de Almonte 29. Tejada la Vieja 30. Aznalcóllar 31. Gadir (Cádiz) 32. Castillo de Doña Blanca 33. Mesas de Asta 34. Lebrija 35. El Carambolo 36. Sevilla (Cerro Macareno, Valencina) 37. Setefilla (Lora del Río)

38. Carmona (Arcebuchal, Alcantarilla, Cruz del Negro, Bencarrón, Cañada de Ruiz Sánchez 39. Osuna 40. Aratispi 41. Ecija 42. Quemados (Córdoba) 43. Cástulo 44. Lixus 45. Cerro del Prado 46. Montilla 47. Acinipo 48. Cerro del Villar 49. Malaka 50. Toscanos (Cerro del Peñón, Cerro de Alarcón, Jardín, Cerro del Mar) 51. Morro de Mezquitilla (Trayamar)

52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 71.

Chorreras Frigiliana Mesa de Fornes Cerro de la Mora Cerro Infantes Galera Almuñécar (Sexi, Laurita, Puente de Noy) Salobreña Abdera Villaricos Gavilanes Verdolay Saladares Peña Negra La Fonteta (Guardamar) Sa Caleta Ebusus Vinarregell Coll del Moro Aldovesta

Introduction

xiii

provided a route to the interior. At times there was a sister settlement on the mainland, and often a necropolis was situated on an opposite shore or riverbank. Gadir (Cádiz) follows this pattern, which is reminiscent of Tyre itself. It must be remembered that today’s coastline has changed dramatically from the time when the Phoenicians first arrived. Due to deforestation and the build-up of silt from the rivers, many of the ocean bays and river estuaries have become marshlands, and the islands have become peninsulas. Spanish archaeologists generally use the term “Final Bronze” (Bronce final) to designate the period from approximately 1200 to 700 b.c.e., which is followed by Iron I (700 to 500) and Iron II (from 500 to the Roman conquest). The translator has chosen to use the term “Final Bronze” for two reasons: to avoid confusion with the chronology for the “Late Bronze” as used in ancient Near Eastern Archaeology, and to avoid confusion with the Spanish term Bronce tardío ‘Late Bronze’, which some Spanish archaeologists use to describe the period just prior to or contemporaneous with the earliest phase of the Bronce final. The high point in trade by the Phoenician colonies and the orientalizing process in the indigenous Tartessian communities occurred during the 7th century, which is also the period on which the Neo-Assyrian project focuses, as described in the preface. Around the middle of the 6th century, the Phoenician colonies and the Tartessian area seemed to suffer a general decline and change in organization. Some historians attribute this decline to the surrender of Tyre to Nebuchadnezzar in 573 b.c.e. and the resulting readjustment in Mediterranean trade. Others attribute it to internal factors. After these dates, the terms used to describe the various cultures change. The culture in ancient Tartessos evolved gradually into what is known as the Turdetanic culture, a term generally used for the period after 550 b.c.e. in the lower Guadalquivir. The term “Iberian culture” is generally used to describe the cultures of southern and southeastern Spain from the 6th century on, when cities began to develop. Greek traders and colonists gained a strong foothold in northeastern Spain. And of great importance, Carthage (with its Punic culture) filled the vacuum left by Tyre, with varying degrees of influence over the area of the former Phoenician colonies. The present volume is divided into two sections: the first section deals with the Phoenicians in the Mediterranean and the second with their presence in the Atlantic coastal areas of the Iberian Peninsula. Although some factors hold constant for almost all the Phoenician colonies (for example, the location at the mouth of a river), there are also differences between these two large geographic areas, such as availability of mineral resources and density of indigenous population. The first section begins with two articles by H. Schubart and by H. G. Niemeyer about two major early excavations on the southern Mediterranean coast

xiv

Introduction

near Málaga, Spain. Schubart worked at Morro de Mezquitilla and Niemeyer at Toscanos, both beginning in the 1960s, although a good deal of the data in these articles come from later excavations. These Phoenician settlements both date to the 8th century b.c.e. and are among the earliest on the Peninsula. The red slip plates found at the excavations on the Málaga coastline have been used to date pottery at other sites, since the width of the rim increases as the plate evolves through time. Niemeyer’s essay focuses on the question of whether, based on the archaeological data, Toscanos can qualify as a “city,” using F. Kolb’s criteria. M. Pellicer Catalán’s article discusses the discoveries from the Phoenician and Punic necropolises of Almuñécar (the ancient Sexi), also on the southern coast of Spain. The surprising discovery in 1962 of Egyptian urns with hieroglyphics in these tombs sparked international interest in eastern colonization in the “Far West.” In two articles, M. E. Aubet Semmler presents a synthesis of her ideas on the economic bases of the colonies on the southern coast, which were located at a considerable distance from primary mineral deposits. She also discusses the role of the indigenous Iberians in the overall economic system. The final two articles in Section 1 deal with the eastern Mediterranean coast and the small island of Ibiza off the eastern coast. The Phoenician colony of La Fonteta, just south of Alicante, was excavated by A. González Prats and his colleagues in the fall of 1996. Buried under the dunes of the Segura River, it is a remarkably well-preserved, early Phoenician settlement. The evidence from the colonies on Ibiza, once thought to have been founded by Carthaginians, has shown western Phoenician origins. J. Ramón describes the founding of the settlement at Sa Caleta in the 7th century, probably by Phoenicians from southern Spain, and the transition to the Bay of Ibiza where the capital city stands today. Section II deals with the Atlantic coastal area of Spain beyond the Strait of Gibraltar, which includes the region of Cádiz, Huelva, and the lower Guadalquivir River basin, the fabled Tartessos of ancient times. The first article, by D. Ruiz Mata, deals with the Phoenician colony of Cádiz (the ancient Gadir), the earliest archaeological strata of which are still covered by ruins from later periods, as well as by the modern city. For this reason the excavations at Castillo de Doña Blanca are of special significance, since this Phoenician colony across the bay was contemporaneous with Cádiz and probably mirrors its development. M. E. Aubet Semmler’s additional two articles deal with the “orientalizing” phenomenon and its influence on the indigenous cultures of Tartessos. The Tartessian economy was already based on farming and mining/metallurgy before the Phoenician colonization, which precipitated a rapid enrichment of the indigenous elite. Aubet uses a core-periphery model to explain the

Introduction

xv

relationship between the Phoenician colonies on the coast and the indigenous communities in Huelva and the interior. J. Fernández Jurado highlights the mining and metal-production aspects of the Tartessian economy. He examines the evidence from excavations of indigenous communities at Huelva, San Bartolomé, and Tejada la Vieja, leaving little doubt about the silver being sought. He also describes the process of metal extraction and the different routes by which the metal was transported to the coast. In the final article, D. Ruiz Mata presents the data from the excavations at Huelva, San Bartolomé, Castillo de Doña Blanca, and El Carambolo which provide further support for the conclusions presented in this section regarding the nature of the Tartessian economy and the orientalizing process. The new archaeological data from the Iberian Peninsula has already greatly helped to clarify Spain’s Phoenician past and the role it played in Mediterranean trade. As this story continues to unfold, Spain’s links to the ancient Near East and, through the Phoenicians, to the Neo-Assyrian Empire will become even clearer. M. Bierling

xvi

Introduction

Part I

The Phoenicians in the Mediterranean

1

The Phoenician Settlement of the 8th Century b.c. in Morro de Mezquitilla (Algarrobo, Málaga) Hermanfrid Schubart

The German Archaeological Institute of Madrid (Deutsches Archäologisches Institut) first surveyed the hill of Morro de Mezquitilla in 1967, and a large-scale excavation of the site began in 1976. 1 This project formed part of a research program to study Phoenician settlement in the area of Torre del Mar. Further excavations of a broader surface area of Morro de Mezquitilla were undertaken in October of 1981 and concluded during the spring of 1982. 2 Morro de Mezquitilla is a foothill of the coastal mountains and reaches some 30 meters above sea level. It is located just to the east of the mouth of the Algarrobo River. Below the hill lies the fishing village of La Mezquitilla, where houses extend in a row along the beach (fig. 1; pl. 1). According to the owners of the land, the hill once had gentler slopes on the south and west, but it appears that part of the dirt was hauled away to the river plain in order to improve the farmland. Moreover, at the time the house on the top of the hill was constructed, the hill was 0.8 to 1 meter higher than at present. Today terraces also extend along the sides of the hill, greatly altering the original appearance of the surface.

Editor’s note: This article appeared in an updated Spanish version under the title “El Asentamiento Fenicio del s. VIII a.C. en el Morro de Mezquitilla (Algarrobo, Málaga)” in Los fenicios en la Península Ibérica (eds. G. del Olmo Lete and M. E. Aubet Semmler; Sabadell [Barcelona]: Ausa, 1986) 59–83. The original German articles on which the Spanish article was based appeared in MM (Madrider Mitteilungen) under the titles listed in notes 1 and 2. 1. H. G. Niemeyer, M. Pellicer Catalán, and H. Schubart, MM 5 (1964) 86ff.; H. G. Niemeyer and H. Schubart, Trayamar: Die phönizischen Kammergräber und die Niederlassung an der Algarrobo-Mündung (Madrider Beiträge 4; Mainz, 1975) 3ff. (see EAE 90 [1973] 11ff.); H. Schubart, “Morro de Mezquitilla: Vorbericht über die Grabungskampagne 1976 auf dem Siedlungshügel an der Algarrobo-Mündung,” MM 18 (1977) 33ff.; idem, AA (1978) 230ff.; idem, NAHisp 6 (1979) 175ff. 2. H. Schubart, “Morro de Mezquitilla: Vorbericht über die Grabungskampagne 1981 auf dem Siedlungshügel an der Algarrobo-Mündung,” MM 23 (1982) 33ff.; idem, NAHisp 19 (1984) 85ff.; idem, “Morro de Mezquitilla: Vorbericht über die Grabungskampagne 1982 auf dem Siedlungshügel an der Algarrobo-Mündung,” MM 24 (1983) 104ff.; idem, NAHisp 23 (1985) 141ff.

3

4

Hermanfrid Schubart

Fig. 1. Map with the location of Morro de Mezquitilla.

Across the Algarrobo River, on its western bank, is the necropolis of Trayamar with its chamber tombs (fig. 1), which were explored during the 1967 excavation season. 3 These rock-cut chambers with inclined passageways demonstrate a type of architecture that is eastern in origin. The finds of amphorae, jugs, and lamps, as well as items of gold, completely confirm the Phoenician character of the tombs, also seen in the typical coexistence of cremation and inhumation. The period of use appears to have been from the middle of the 7th century to around 600 b.c. Another Phoenician site related to the mouth of the Algarrobo River is located on top of a rocky coastal promontory, slightly to the east of Morro de Mezquitilla. However, this western Phoenician settlement of Chorreras had a shorter life span than Morro de Mezquitilla, as indicated by the finds, especially the plates. Chorreras was inhabited during the second half of the 8th century and came to an end already in the 7th century b.c. In addition, the structures show only a single phase of occupation. 4 3. H. G. Niemeyer and H. Schubart, Trayamar, 59ff., and EAE 90, pp. 103ff. 4. M. E. Aubet, G. Maass-Lindemann, and H. Schubart, “Chorreras: Eine phönizische Niederlassung östlich der Algarrobo-Mündung,” MM 16 (1975) 137ff.; idem, “Chorreras – Un establecimiento fenicio al Este de la Desembocadura del Algarrobo,” NAHisp 6 (1979) 89ff.

The Phoenician Settlement in Morro de Mezquitilla

5

The 1976 excavations 5 proved that an earlier settlement had existed at Morro de Mezquitilla during the Chalcolithic period. The strata of the Chalcolithic settlement (A), which in almost all areas lay directly on top of bedrock, were perfectly distinct from the superimposed Phoenician strata. At no point could we prove that the Chacolithic settlement lasted until the Bronze Age, nor did we discover any out-of-context finds from that era. Since forms from the Chalcolithic period did not continue until the beginning of the Bronze Age, we must assume that there was a gap in the occupation of the hill during the millennium that preceded the founding of the Phoenician settlement. When the Phoenicians arrived in the 8th century, they must have found the hill of Morro de Mezquitilla completely uninhabited, which was typical of their pattern of settlement. The Phoenician buildings and their corresponding floors are found directly on top of the strata from the Chalcolithic period, as was demonstrated in all the excavation areas. In Area 7 from the year 1976, where the Phoenician layers reach their greatest thickness, we observed a stratigraphy that was also valid for Area 8 in 1976 and also for Areas 11 and 12 of the 1981/82 season— in other words, for the higher areas at Morro de Mezquitilla. The six phases (BI–BVI) of the Phoenician settlement at Morro de Mezquitilla can best be followed stratigraphically in Stratum 7/8 and show up clearly with all their subdivisions in Sector VI. 6 From Phase BI, with a stratigraphic thickness of up to 0.76 m, and from Phase BII, with up to 0.55 m of thickness, no structural remains have been preserved, with the exception of a series of surfaces in Sector VI. However, a pile of stones in Sector III could be understood as a destroyed wall from Phase BI/II. In any case, a wall from Phase BIII is situated on top of the mass of fallen stones that are clearly recognizable in the northern section of Area 7. Phase BIII of the Phoenician settlement is represented in Area 7/8 by a system of wall sections some 0.50 m wide that are set in clay. According to the cross sections, these stretches of wall are foundations, and the foundation trenches are clearly recognizable even when incomplete, reaching 0.58 m to 1.30 meters below the original surface. In the northern cross section of Sector VI we can see clearly that the surface corresponding to a wall with a deep foundation is 1.50 m above the base of the wall. Unfortunately, surface levels and wall foundations cannot generally be observed. In Sector III B we find signs that a wall was built using stones and mudbrick at the same time that these Phase III structures existed.

5. See H. Schubart, MM 18 (1977) 39ff.; idem, NAHisp 6 (1979) 183ff.; idem, MM 23 (1982) 37ff.; idem, NAHisp 19 (1984) 91ff.; idem, MM 24 (1983) 107ff. 6. See H. Schubart, MM 18 (1977) 47f.; idem, NAHisp 6 (1979) 192.

6

Hermanfrid Schubart

After the structures built during Phase BIII had been abandoned or destroyed, it appears that this part of the Phoenician settlement at Morro de Mezquitilla, which is crossed by Area 7/8, underwent no further construction, as indicated by layers up to 0.90 m thick of reddish and reddish-brown alluvium and fill (B IV). The finds from these layers must belong, at least in part, to Phase BIII of the settlement. According to the type of items found, the Phoenician structures preserved in Area 3 must also have belonged to Phase BIII/IV. Evidently, the earliest phases of the settlement (I and II) did not leave any stratigraphic evidence on the top of Morro. At a later period in the history of the settlement, construction was resumed in Area 7/8, this time on top of the reddish-brown layer BIV and in a direction that was different from the orientation of the buildings in Phase BIII (Phase BV). Greenish and brownish-gray strata, the remains of decomposed mudbrick walls, correspond to this Phase BV. The upper, most recent layers of this phase reach in part to the decomposed vegetable earth of Phase BVI, whose walls again have a different orientation. The work begun in 1976 in Areas 7/8 was continued in the excavation of Areas 11 and 12 in 1981, which resulted in a considerable addition to the base of material found in1976 and served as an important confirmation of the earlier results. The 1981 and 1982 excavations had the goal of continuing the longitudinal cut westward for more than 100 m. Also, we hoped to discover a grouping of the settlement’s structures in an especially appropriate spot, with the goal of documenting structural remains and also, if possible, parts of the urban layout. In order to meet this objective, we chose an agricultural terrace lying on the lower part of the western slope, just above the cliff. The most interesting results for the Phoenician settlement of the 8th century b.c. were obtained in this excavation zone. We will look at the stratigraphic results and then at the sequence of construction phases, as they were observed during the 1981/82 seasons. Due to the intensely dark color of the earth, it is easy to establish the boundary between the strata of the Chalcolithic period and the superimposed Phoenician strata. The earliest strata have a dark grayish-black to intense black color. The dark upper layer contains artifacts that come principally from the Chalcolithic period, but there are also some Phoenician pieces; this demonstrates that the surface of the upper layer, which is mostly Chalcolithic in origin, was leveled on the surface by the first Phoenician settlers, probably while they were constructing their buildings. During this construction, they must have done some banking and terracing work since the ground slopes downward in this area, and some Phoenician buildings have a terraced structure. This upper layer of Chalcolithic strata also differs slightly in color and

spread is 1 pica short

The Phoenician Settlement in Morro de Mezquitilla

7

consistency from the lower, earlier layers. This stratum, which could not be seen in all the areas, was designated A/B1 (fig. 2). Along with the few Phoenician pieces mentioned above, we made a very important find in this stratum: a scarab. This artifact came from the northern section of Area 11 and has already been published in a separate article. 7 We can assume that this scarab was brought by one of the first Phoenician colonizers to arrive at Morro de Mezquitilla. Either he lost it during the first construction work or the scarab was already on the surface shortly before the construction began and was then included in layer A/B1. Thus, it is evident that we can perceive the first signs of the Phoenician settlement (B) already in transitional layer A/B1, upon which the buildings of the first phase (B1) of the town were later constructed. During the previous excavations carried out at Morro de Mezquitilla, it was possible for the earliest phases of the town (8th century b.c.) to be documented by strata, but not by structures. However, in 1982 we were able to excavate surfaces and walls, rooms and buildings (fig. 3; pls. 2–5). However, it was not possible to establish a direct stratigraphic connection between Areas 7/8 and 11/12 on the top of Morro and Areas 13–23 in the western part of the excavated zone, which constitute their own stratigraphic unit. This circumstance was due to strong disturbances in the strata in the western part of Areas 11 and 12 and also to the great distance between Areas 12 and 13, caused by the different levels of the terraces and by the irrigation canals dug at the upper and lower boundaries. Nevertheless, we can already affirm, even before concluding our study, that on the basis of the strongly differentiated stratigraphy and the abundant archaeological material from both areas, we can compare these stratigraphic complexes, as shown in fig. 12. Contemporary with these first structures (and evidently with an earlier date, at least in part) is a horizon B1a with remains of workshops, located in the southwest of the archaeological surface of Areas 19 to 23. Here we found ovens, some of which show evidence of having been renovated on various occasions (pl. 4). Near the ovens, which show strong evidence of fire, we found remains of slag, and sometimes in direct connection with the ovens we found fragments of ventilating pipes, especially bellows nozzles (pl. 6), occasionally with remains of metal adhered. This group of Phoenician nozzles is the largest found in southern Spain so far and provides clear testimony that smelting was carried out and that the workshops in question were metal workshops. We also found fragments of large clay pots with droplets of molten metal adhered (fig. 11a); these pots were probably used in the workshops as containers for the molten metal. Two analyses performed on the slag have shown that it was iron 7. See I. Gamer-Wallert, MM 24 (1983) 145ff.

8

Hermanfrid Schubart

Fig. 2. Morro de Mezquitilla 1982. Area 13/14, northern section, showing the distinct phases of inhabitation (compare with fig. 12).

slag. 8 Obviously, the installations were not used for primary smelting, since in that case the ovens would show signs of a much stronger combustion and the quantity of slag would be greater. We assume, rather, that the ovens for primary smelting were located somewhat farther from town, perhaps even next to 8. This information was kindly provided to us by the Geologisch-Paläontologisches Institut der Universität Kiel. We are very grateful to Dr. H. D. Schulz for his valuable collaboration in the present and in the future.

The Phoenician Settlement in Morro de Mezquitilla

9

the mines. In the case of Morro de Mezquitilla, we believe that we are dealing with a workshop that was used for recasting and working metal, or possibly a blacksmith’s shop. Thus it has been demonstrated that a metal workshop existed at Morro de Mezquitilla during the first period of Phoenician settlement, one whose importance for the economic history of this site is obvious. The artifacts that have been found so far in the workshop area and in the corresponding strata are very similar to those found in the houses of the first period, which allows us to consider them as belonging to Phase B1. Based only on stratigraphic observations, we have designated the metal workshop as B1a in order to distinguish it from the houses of B1b, even though the chronological distance between them must be very small. As we have already mentioned above, the upper stratum of the Chalcolithic period was leveled in order to construct the buildings of Phase B1b of the

10

Hermanfrid Schubart

settlement, and this stratum was also affected by walls and stairs (fig. 2). The levels of the surfaces are easily recognizable because the yellow clay floors were renovated several times in different places (fig. 2; pls. 3, 5). Also, the surfaces were slightly terraced, following the slope of the ground (fig. 2). The walls, preserved up to a meter high, are made of mudbrick; the fill between the bricks is gray, and the first rough outside plaster is reddish brown. The second layer of plaster consists of very fine yellow clay, which is covered with several layers of very fine lime and red or greenish-yellow paint. The walls contain several openings for doors, with high thresholds and often steps on both sides (pls. 2, 5). Hearths can be observed at two points. The floors of the rooms are rectangular, but not always regular (fig. 3). Since the excavated surface is limited, we have been able to determine the exact measurements of only some of the rooms; they are 4.20 to 4.80 meters long and 2.20 to 2.80 meters wide—in one case, somewhat more than 3 meters. Two of these large, rectangular rooms are subdivided. In all, we can distinguish three structural complexes (fig. 3); the largest is designated K and is 19 meters long, 11 meters wide, and has at least sixteen rooms. However, Building K was not planned all at once, as shown by the acute and obtuse angles of its walls that at times come together irregularly, or its rooms, which also have irregular shapes and are located, for example, in the center of the building. Between Complex K and Building I, located more to the south and documented only in part, there seems to have been a narrow street. It could be observed especially in the western area (fig. 3). This street is characterized by a bank of greenish color containing a large amount of organic matter. The streets of the more recent period B2 show similar banks of greenish color. In Building I we could document only three rooms; as a result, we cannot give more detailed observations about its internal division. Building H, located to the southeast of Areas 19/20, has an orientation different from that of structures K and I. We cannot make a final determination of its date in relation to that of Building I, but surely they coexisted for some time. In the area of Complex K, we found a number of collapsed mudbrick walls, under which were buried several complete assemblages of pottery (fig. 5). Various handmade pots also correspond to this first phase of settlement. These pots were invariably found directly under the earliest Phoenician surface and sunken into Stratum A/B1, which contains principally Chalcolithic period material (fig. 2; pl. 3). The perfectly preserved state of the pots, along with their vertical position below the floor and the fact that they were almost always near a wall, clearly indicates that they had been placed there intentionally. We are still not sure whether these pots were offerings related to the building (which is unlikely because of their large number) or whether these pots were used to

The Phoenician Settlement in Morro de Mezquitilla

11

Fig. 3. Morro de Mezquitilla 1982. Partially reconstructed schematic map of first phase of Phoenician construction.

keep certain foods fresh by taking advantage of the coolness of the ground, and the handmade containers were chosen deliberately for this purpose. 9 On top of the collapsed mudbrick homes of Phase B1 and partially penetrating them, the walls of a new constructive phase (B2) were built; their orientation differs visibly from that of the earlier phase (fig. 4). Unlike the earlier mudbrick walls, simply built on top of the leveled surface, the mudbrick walls of this new constructive phase were built on a stone foundation set in a foundation trench (fig. 2, pls. 3, 4). In comparison with the earlier buildings, especially Complex K, the construction method used in this second phase of the settlement is improved and much more complex. This and other observations 9. One of the pots contained sheep or goat bones, which demonstrates that a piece of meat was kept inside (or perhaps an offering?). We are grateful to Prof. Dr. J. Boessneck (Munich) for kindly providing this information.

12

Hermanfrid Schubart

Fig. 4. Morro de Mezquitilla 1982. Partially reconstructed map of second phase of Phoenician construction.

show the progressive consolidation of the Phoenician settlement after its first phase, which was provisional in comparison. 10 Determining the different rooms and structures of Phase B2 turned out to be extremely difficult, since after the buildings were abandoned (probably at the end of the Phoenician period and the beginning of the Punic phase, and certainly before the construction of the Punic buildings), the foundation stones were torn out from above for reuse. The result was that in numerous cases the excavation could only document the robber trenches in the corresponding floors and sections (fig. 2; pl. 3). The figure clearly shows these trenches. Cartographic verification of the remains of the still-existent walls and of the robber trenches made it easier to obtain a general floor plan of the buildings constructed during Phase B2. In some cases we were able to distinguish the foundation trenches from the robber trenches—always in cases 10. H. P. Uerpmann and M. Uerpmann, Studien über frühe Tierknochenfunde von der Iberischen Halbinsel 4 (Munich, 1973) 87; H. G. Niemeyer, Madrider Beiträge 8 (Mainz, 1982) 202f.; and Huelva Arqueológica 6 (1982) 101–27.

The Phoenician Settlement in Morro de Mezquitilla

13

where the surfaces and the gravel from the street covered the wide foundation trench and came up against the narrower projecting wall. The robber trenches, however, which were narrower than the foundation trenches, affected only the width of the wall from above, though always to the bottom course (fig. 2, on the right). On some occasions we were able to differentiate between the consecutive trenches based on the different colors of the corresponding earth. The orientation of the walls in the construction phase B2 differs completely from that of the earlier phase B1 (figs. 3, 4; pl. 4). The rooms, which are very diverse in distribution and form, are grouped into three buildings, with a different floor plan in each case (fig. 4). The structure with the plan most similar to Complex K of Phase B1 is House E, which contains a long room measuring about 5.40 x 2.20 m, plus other adjoining rooms. The floor plan of Building F is characterized by the clear importance of its central part, which forms a room of 3.80 x 2.80 m, with a somewhat narrow gallery (or hallway?) along three of its four sides. Other rooms extend outward from this central room, but we were not able to determine their complete extension (fig. 4; pl. 4). Finally, in the southeast and across from Houses E and F, we find Complex G, which may have originated from two distinct buildings whose corners touched. In this complex, too, we were able to distinguish a long rectangular room; the rest of the plan could not be determined beyond the boundaries of the excavated area (fig. 4). Between E/F on one side and G on the other, runs a street that in the northeast measures 3.20 m and eventually widens to a little more than 5 m. This street, like the one mentioned earlier, is characterized by layers of greenish gravel which form a series of successive banks. The green color comes from numerous organic substances existing there (fig. 4). Also, a narrow street passes between Houses E and F, measuring 1.50 m wide along its entire course. It is also covered by layers of greenish gravel, which can still be recognized between the robber trenches along the outer walls of both houses in the northern section of Area 14 (fig. 2). Numerous strata are found in both building phases of the Phoenician settlement, and they are usually characterized by layers of floor surfaces that are yellow and green in color; these strata can be recognized in the northern section of Areas 13 and 14 (Fig. 2). Only in the northern section of Area 13 (fig. 2; pl. 3) are there four clay surfaces clearly visible, all of which belong to Phase B1. Each one of them is accompanied by rather thick banks, as well as thresholds covered by yellow plaster, which evidently correspond to openings in the walls for doors. As was already mentioned above, it is not possible to link these strata with those in the stratigraphy of Areas 7/8 and 11, nor can we carry out a definitive comparison of the strata until the general research has been finished. However, with respect to the comparison of the stratigraphy of 1976 /81 to that of

14

Hermanfrid Schubart

1982, we predict the following: according to the results of the 1982 excavation, the lower part of the 1976 Stratum BI 11 is considered to be A/B1. The rest of Stratum BI plus Stratum BII must correspond to construction phase B1(b) of the 8th century b.c. Strata BIII and BIV of the 1976 excavation, especially BIV (as well as other finds from the 7th century b.c. which were present in the stratigraphy of Area 7/8 only in some of their forms 12) correspond to construction phase B2 of 1982 (fig. 12). Buildings BV and VI of 1976 correspond to the sparse structural remains of B3 in 1982. Of the very abundant Phoenician finds, we will present only some of the most characteristic forms. The red slip plates (figs. 5a, 6–8) are especially important because they are essential for dating. Several plates with graffiti have been treated in a separate article. 13 Hundreds of red slip plates, so important for dating Phoenician settlement in the West, have appeared in the strata of Morro de Mezquitilla. 14 Their documentation is not completely finished, but after they are classified, they will constitute the best explanatory element for dating, as they have at other sites. The 142 plates that were sketched were not chosen systematically and therefore are not representative of the whole. For this article we have chosen some plates that are characteristic of the Phoenician construction phase B1 (fig. 6), others that correspond to Phase B2 (figs. 7, 8), and finally some plates from the most recent phase of the Phoenician town, B3 (fig. 8), where we documented only a few remains of walls and therefore had few findings. The forms of the plates shown in figs. 6 and 8 have an increasing rim width and are characteristic of the evolution of this Phoenician form; this evolution was confirmed again by the 1982 excavations at Morro de Mezquitilla. The plates from the first phase of the Phoenician town stand out because of their narrow rims, which usually measure less than 2.2 cm. Already in 1976, we were able to find these narrow widths in Strata BI and BII of Areas 7 and 8. 15 These plates with rims of less than 2.2 cm belong to the earliest forms discovered until now on the Iberian Peninsula, 16 and evidently they are even earlier than the group of plates found in Strata II and III at Tyre, where there is a total of ten plates with rims measuring between 2.2 and 2.8 cm. 17 The plates from 11. For this and following material, see MM 18 (1977) 61, fig. 17, and NAHisp 6 (1979) 201, fig. 17. 12. See H. Schubart, in Phönizier im Westen (Madrider Beiträge 8; ed. H. G. Niemeyer; Mainz, 1982) 220, fig. 11 “1967 Kuppe”; and Huelva Arqueológica 6 (1982) 71–97. 13. See W. Röllig, MM 24 (1983) 133ff. 14. H. Schubart, “Westphönizische Teller,” RSF 4 (1976) 179ff.; idem, in Madrider Beiträge 8, pp. 207ff., fig. 18; and Huelva Arqueológica 6 (1982). 15. See H. Schubart, MM 18 (1977) 51ff., figs. 12–14, and NAHisp 6 (1979) 197ff., figs. 12– 14. 16. See H. Schubart, Madrider Beiträge 8, p. 229, fig. 18, and Huelva Arqueológica 6 (1982). 17. P. M. Bikai, The Pottery of Tyre (Warminster, 1978) pl. 9.1–18.

The Phoenician Settlement in Morro de Mezquitilla

15

Tyre date to the final third of the 8th century b.c. Thus we must date the plates from Morro de Mezquitilla, with their visibly narrower rims, at least to the second third of the 8th century. In any case, they confirm that the beginning of the Phoenician colonization of the southern coast of Spain took place around or before the middle of the 8th century b.c. A red slip plate with a rim 1.6 cm wide (fig. 5a) was found in a grouping with other vessels (fig. 5) on the earliest floor of an eastern room in Building K (fig. 3), under one of the collapsed mudbrick walls. This floor lies directly on top of layer A/B1, which in this room alone contains five handmade pots (pl. 3). These pots were placed below the surface at the beginning of or during the time this room was used, and therefore they are earlier than the pottery found on top of the floor, which comes from the final period of use. It is not certain whether this room was reused, unlike the adjacent room to the north, which has four surfaces. Either it was used with its original surface for a long period of time, or it was abandoned after the walls collapsed. Therefore, the assemblage of pottery can surely be assigned to the first phase of Phoenician construction (B1), and most likely to its beginning (B1b1). In addition to the plate with the very narrow rim (fig. 5a), this important pottery grouping includes the following elements: two carinated bowls of the type that has rims thickened toward the outside and slightly lengthened (fig. 5b, d); 18 a red ceramic bowl with a slightly thickened rim turning inward and a distinct base (fig. 5e); 19 a fragment of a Type 1 amphora rim, which has a vertical rim and a high shoulder (fig. 5f); a circular handle of an amphora (Mo 82/1920/57); a handmade vessel with an S-shaped profile (fig. 5h); and finally two jugs, one of which has a whitish slip and belongs to the rare type of jug whose handle protrudes directly from the rim (fig. 5c). The second jug, belonging to the red pottery type, has a mushroom-shaped mouth (fig. 5g). Its rim is very narrow and the neck widens toward the opening, its double circular handle grips the central band around the neck, and its globular body rests on a concave base that is only slightly distinct from the body. All these data indicate that this is one of the oldest mushroom- mouth jugs found in a Phoenician settlement on the Iberian Peninsula. Without doubt we can date it to the middle of the 8th century. There are no known tombs from this century, which until now have been the greatest source of complete jugs. 20 In addition to this jug there are two parallel cases in the Iberian Peninsula, although we do not 18. G. Maass-Lindemann, Toscanos 1971 (Madrider Forschungen 6.3; Berlin, 1982) 39ff., pl. 5.151–159, and also 160–162; and NAHisp 18 (1984). 19. See G. Maass-Lindemann, Toscanos 1971, pp. 46f., for example, pl. 7.193, and NAHisp 18 (1984). 20. H. G. Niemeyer, H. Schubart, Toscanos 1964 (Madrider Forschungen 6.1) (Berlin, 1969) 91, 94f., and EAE 66, pp. 113, 116ff.; M. Almagro Gorbea, MM 13 (1972) 172ff., especially 180; I. Negueruela , MM 22 (1981) 211ff.

16

Hermanfrid Schubart

know their dates. One, a loose find, is conserved in the Málaga Museum, 21 and the other comes from El Carambolo (province of Sevilla). 22 Another piece of red pottery that has appeared is a concave base with prominent exterior ribbing, belonging to a vessel with relatively fine sides. Its top part has not been found, but it must have been an incense burner (fig. 9c). These incense burners are quite common in the Phoenician towns on the Málaga coast, 23 although they normally do not have this unusual striated base. Pottery with an untreated surface, which comprises the majority of the finds at Morro, is represented here by only a few drawings (fig. 9a. d–l). One of the pieces has a characteristic pithos form, with a triple circular handle and decoration consisting of broad areas painted red, interspersed with narrow grayish-black bands (fig. 9a). 24 Several bottles (fig. 9f–h) and juglets (fig. 9d, e, k) have also appeared, one of which has a special form (fig. 9i). 25 The astonishing size of a fragment that belongs to a ring-shaped stand (fig. 9l, shown here at a scale of 1:6!) catches our attention and provides evidence of the size that these stands can sometimes acquire; 26 usually we see only the upper part of these stands. Some fragments that were once thought to belong to flat serving dishes in reality formed part of one of these immense ring-shaped stands. 27 Due to this new find, it will be possible to recognize and classify such finds correctly in the future. Among the handmade vessels, we note the dozen or so pots that were placed in the leveled layer of Chalcolithic material, which was found under the earliest level of the first phase of the Phoenician town (fig. 10, 11b–e). It appears that these pots were used for storing foods, as mentioned above. 28 However, we must keep in mind that the phenomenon is limited to the earliest period of the town, which demonstrates the importance of this handmade pottery at the time. 21. See H. G. Niemeyer, M. Pellicer Catalán, and H. Schubart, MM 5 (1964) 81f., fig. 7. 22. J. de M. Carriazo, Tartessos y El Carambolo (Madrid, 1973) 609, fig. 457ff. 23. See H. G. Niemeyer, M. Pellicer Catalán, and H. Schubart, MM 5 (1964) 78ff., fig. 4.8; H. G. Niemeyer and H. Schubart, Trayamar, 131ff., pls. 12.553, 554; 20.1057, 1058; and EAE 90, p. 210, pls. 12. 553, 554; 20.1057. 24. For information on pithoi forms, see M. E. Aubet, G. Maass-Lindemann, and H. Schubart, MM 16 (1975) 152f., 173, fig. 8.113a, and NAHisp 6 (1979) 110f., fig. 8.113a; G. Maass-Lindemann, Toscanos 1971, pp. 31ff., fig. 3, pl. 1.15–18; and NAHisp 18 (1984). 25. H. G. Niemeyer and H. Schubart, Toscanos 1964, pp. 95f., 104f., pl. 17; and EAE 66, pp. 188f., 127f., pl. 17; G. Maass-Lindemann, Toscanos 1971, pp. 60ff., pl. 13.445ff.; and NAHisp 18 (1984). 26. For example, H. G. Niemeyer and H. Schubart, Trayamar, 134ff., pls. 12.548, 560; 16.607, and EAE 90, pp. 213f.; pls. 12.548, 560; 16.207; G. Maass-Lindemann , Toscanos 1971, p. 69, pl. 19.775, 779; and NAHisp 18 (1984). 27. For example, G. Maass-Lindemann, Toscanos 1971, pl. 8.239, 241; and perhaps also pl. 8.225ff., and NAHisp 18 (1984). 28. See no. 9.

The Phoenician Settlement in Morro de Mezquitilla

17

Among the pots we can distinguish those that have regular handles (fig. 10) from those which have ledge handles that are more or less pronounced (fig. 11b, d, e). There is one very large pot that lacks both (fig. 11c). The handles are attached either directly to the rim (fig. 10a, b) or just below (fig. 10f), or to the shoulder (fig. 10 c–e); the pots have an S-shaped profile (for example, fig. 10d; fig. 11d) or a more distinct and protruding rim (for example, fig. 10a; 11e). All of the pots have a flat base, which is at times slightly pronounced. The method of manufacturing, firing, and coloring these pots, especially in the earliest strata, 29 would lead us to assign them to the prehistoric category of pottery found at Phoenician sites, except that these forms are completely absent in the corresponding strata of Final Bronze peoples living in the hinterland. 30 Because of this, we must consider them to be Phoenician pottery, made at these sites by hand. Fragments of this pottery have certainly been observed in the earlier excavations at Toscanos and Morro de Mezquitilla, but because we were unacquainted with their complete form and special importance, we assigned them to the category of prehistoric pottery. 31 Today we realize that this pottery, including the assemblage of pots from Morro, belongs to a simple type of Phoenician pottery. It is remarkable that during the Phoenician period these handmade pots existed next to highly evolved, wheelmade pottery, and that they were submitted (probably in Phoenician ovens) to a stronger firing. This phenomenon probably has its explanation in the special function of these pots, and perhaps also in the greater influence that handmade pottery retained during the initial phase of Phoenician settlement. Without a doubt, the handmade pots described here constitute their own category of Phoenician cooking pots. In summary, the top part of the hill of Morro de Mezquitilla has been intensely populated since the third millennium b.c.—with a gap in the second millennium b.c. It was occupied again around the year 750 b.c. or somewhat earlier, and it continued to be occupied without interruption until the century before Christ. The duration and density of this occupation, and especially the return of the Phoenician colonists to a place that had already been inhabited during the Chalcolithic period, can be understood only if we keep in mind the particular location of Morro de Mezquitilla at the mouth of the Algarrobo 29. H. Schubart in Toscanos 1971 (ed. G. Maass-Lindemann) 75ff. with bibliographic notes, and NAHisp 18 (1984). Susana Puch (Madrid) is preparing a detailed study of this material. 30. A. Mendoza, F. Molina, O. Arteaga, and P. Aguayo, “Cerro de los Infantes (Pinos Puente, provincia de Granada),” MM 22 (1981) 171ff., especially 199ff., fig. 11ff. Spanish version in Cuadernos de Prehistoria de la Universidad de Granada. 31. See M. E. Aubet, G. Maass-Lindemann, and H. Schubart, MM 16 (1975) 176, fig. 11.154, 158 (?); and NAHisp 6 (1979) 118, fig. 11.154–158; H. Schubart, MM 18 (1977) 55, fig. 15c (the vessel in fig. 15b is wheel made); and NAHisp 6 (1979) 199, fig. 15c; H. Schubart in Toscanos 1971, pl. 22.917ff (?), and NAHisp 18 (1984).

18

Hermanfrid Schubart

River and near the coast. The choice of this location for a Chalcolithic period settlement might seem surprising at first, especially when we consider the examples known until now, though from a theoretical point of view it is not unexpected. However, this location is completely typical for a Phoenician settlement, as shown by the examples of Guadarranque, Guadalhorce, Toscanos, Almuñécar, and Adra. A direct parallel for a settlement such as this one during the Punic and Roman Republic periods can be found in the immediate vicinity: Cerro del Mar is also located near the coast at the mouth of a river, the Vélez. In addition to the summit of Morro de Mezquitilla, there are numerous possibilities for further research on the western slope, which extends for almost 300 m down toward the valley of the Algarrobo River. The docks for the Phoenician ships were probably located here at the base of the hill, where there was a bay that must have provided a very adequate harbor, similar to the harbor on the Vélez River near Toscanos. 32 This bay supported the Phoenician settlement of the 8th century b.c., which was certainly one of the earliest on the Iberian Peninsula—and perhaps the earliest, according to our present knowledge. 32. See H. D. Schulz, MM 24 (1983) 59ff., fig. 1.

The Phoenician Settlement in Morro de Mezquitilla

19

a b

d

e

c f

g

h

Fig. 5. Morro de Mezquitilla 1982. Closed find of pottery from the floor of a room in Complex K: (a) red ceramic plate (Mo 82 / 1925); (b, d) carinated bowls (Mo 82 / 1920 / 64; Mo 82 / 1920 / 61+62); (c, g) jugs, (c) with a white slip, (g) with a red slip (Mo 82 / 1923 / 1; Mo 82 / 1927 / 1); (e) red ceramic serving bowl with rim turning inward (Mo 82 / 1924 / 1); (f) fragment of a rim from a Type 1 amphora (Mo 82 / 1929); (h) handmade vessel (Mo 82 / 1926 / 1).

20

Hermanfrid Schubart

a

b

c

d

e

f

g

h

i

k

l

Fig. 6. Morro de Mezquitilla 1982. Red ceramic plates from the strata of the first phase of the Phoenician town (B1); (a) Mo 82 / 2179 / 1; (b) Mo 82 / 2260 / 20; (c) Mo 82 / 1901 / 17; (d) Mo 82 / 1906 / 2–7; (e) Mo 82 / 2256 / 11–26; (f) Mo 82 / 2256 / 36; (g) Mo 82 / 2256 / 47; (h) Mo 82 / 2031 / 39; (i) Mo 82 / 2031 / 56; (k) Mo 82 / 2031 / 27–35; (l) Mo 82 / 2031 / 55.

The Phoenician Settlement in Morro de Mezquitilla

21

a

b

c

d

e

f

g

h

i

k

l

m

Fig. 7. Morro de Mezquitilla 1982. Red ceramic plates from Complex Mo 82 / 204 from the second phase of the Phoenician town (B2): (a) Mo 82 / 2014 / 152; (b) Mo 82 / 2014 / 155; (c) Mo 82 / 2014 / 151; (d) Mo 82 / 2014 / 162; (e) Mo 82 / 2014 / 177; (f) Mo 82 / 2014 / 166; (g) Mo 82 / 2014 / 101; (h) Mo 82 / 2014 / 180; (i) Mo 82 / 2014 / 88; (k) Mo 82 / 2014 / 70; (l) Mo 82 / 2014 / 89; (m) Mo 82 / 2014 / 183.

22

Hermanfrid Schubart

a

b

c

d

e

f

g

h

i

k

l

Fig. 8. Morro de Mezquitilla 1982. Red ceramic plates from the strata of the late second phase (B2) and the third phase (B3) of the Phoenician town: (a) Mo 82 / 1248 / 40; (b) Mo 82 / 1247 / 11–13; (c) Mo 82 / 1983 / 5; (d) Mo 82 / 1663 / 22; (e) Mo 82 / 1828 / 54–55; (f) Mo 82 / 1247 / 38; (g) Mo 82 / 1247 / 30; (h) Mo 82 / 2004 / 103; (i) Mo 82 / 2005 / 14–15; (k) Mo 82 / 1985 / 1; (1) Mo 82 / 1823 / 1.

The Phoenician Settlement in Morro de Mezquitilla

23

b

c a

f e

d

i

g

h

l k

Fig. 9. Morro de Mezquitilla 1982: (a), (d-l) pottery with untreated surface: (a) pithos with areas of red paint and blackish-gray bands (Mo 82 / 1888); (d), (e), (i), (k) juglets (Mo 82 / 2241 / 1; Mo 82 / 2297 / 1; Mo 82 / 1362 / 1; Mo 82 / 2493 / 1); (f), (h), (g) bottles (Mo 82 / 1079 / 93; Mo 82 / 2172 / 11; Mo 82 / 894 / 31); (l) large ring-shaped stand (Mo 82 / 1335 / 1); (b) rim fragment from a proto-Corinthian kotyle (Mo 82 / 1915 / 1); (c) probably a fragment from the bottom of a red ceramic incense burner (Mo 82 / 1793 / 2).

24

Hermanfrid Schubart

b

a

d c

e

f

Fig. 10. Morro de Mezquitilla 1982: Handmade pots with handles that are attached to the rim or shoulder (Mo 82 / 1933 / 1; Mo 82 / 1937; Mo 82 / 1250 / 4; Mo 82 / 1921 / 1; Mo 82 / 2307 / 1; Mo 82 / 1922).

The Phoenician Settlement in Morro de Mezquitilla

25

a

c b

d

e

Fig. 11. Morro de Mezquitilla 1982: (a) rim fragment from a handmade vessel with thick sides, with stub of a nozzle and coating of metal inside (Mo 82 / 1750 / 29); (b-e) handmade pots, (b), (d), (e) with ledge handles (Mo 82 / 2084 / 1; Mo 82 / 2108 / 1; Mo 82 / 1250).

26

Hermanfrid Schubart

Morro de Mezquitilla

1976 / 81 (Areas 1–12)

1981 / 82 (Areas 13–23)

C Late Punic-Roman structures (4th–1st centuries b.c.)

wall system in Area 2

C2: added structures

(VII): few finds

C1: new structures

VI:

latest walls

B3: few structural remains

V:

system of later walls

B Phoenician-Punic settlement (8th–5th centuries b.c.)

A Chalcolithic settlement (third millennium)

IV:

decomposed strata

B2: 2d building phase

III:

system of early walls

B1: 1st building phase b

II:

early strata

I:

earliest stratum

IV:

stratum

III:

stratum

workshop a A / B1: period of leveling 6 5 4 A: strata of the settlement

II:

stratum

3 2 1

I:

stratum

Fig. 12. Morro de Mezquitilla. Chronological table for comparing the central and western archaeological areas. 1982.

The Phoenician Settlement in Morro de Mezquitilla

27

Plate 1. Morro de Mezquitilla 1982. View from the southwest toward the summit of the Phoenician settlement, with mountains in the background and excavation area in the left center. Inst. Neg. Madrid R 3-82-11 (photo P. Witte).

Plate 2. Morro de Mezquitilla 1982. Area 17 / 18 viewed from the west, with mudbrick walls and surfaces of yellow clay in Phase B1. Inst. Neg. Madrid 9-82-69.

28

Hermanfrid Schubart

Plate 3. Morro de Mezquitilla 1982. Area 13, walls and clay surface from Phase B1, viewed from the south; in the foreground three handmade pots, placed in Chalcolithic strata; in the section a piece of wall from Phase B2 containing light-colored surfaces; above, remains of walls from the phase of Punic-Roman construction. Inst. Neg. Madrid 10-82-19A.

Plate 4. Morro de Mezquitilla 1982. Area 17 / 18, walls from Phase B2, with earlier mudbrick walls below (B1), viewed from the east. Inst. Neg. Madrid R-10-82-11 (photo P. Witte).

The Phoenician Settlement in Morro de Mezquitilla

29

Plate 5. Morro de Mezquitilla 1982. Area 13, northwest corner with three surfaces from Phase B1 visible in the section, viewed from the east. Inst. Neg. Madrid 7-82-13A.

Plate 6. Morro de Mezquitilla 1982. Bellows nozzles (Mo 82 / 2146, among others). Inst. Neg. Madrid R5-82-14 (photo P. Witte).

30

Hermanfrid Schubart

The Phoenician Settlement at Toscanos: Urbanization and Function Hans Georg Niemeyer

“The introduction of urbanization is the chief virtue of the Greek colonial polis.” This phrase comes from the new synthesis on the city in antiquity by F. Kolb 1 and represents the consensus of researchers in the field of ancient history. The term “urbanization” refers principally to the planning spirit of the Greeks, which they demonstrated when they established their colonies according to prior design, especially in the region where Magna Graecia later emerged. The definition of “city” on which the historian Kolb bases his concept of urbanization is the classic definition of Max Weber: a city is a place where “the resident population, principally from an economic standpoint, satisfies its daily needs in the local market, mainly with products that the resident population and that of the surrounding area have produced or acquired for this market.” 2 In accordance with this definition, Kolb sets forth the following criteria for a “city”: 1. Topographic and administrative unity of the population, 2. population of a few thousand inhabitants (towns with around a thousand inhabitants are considered borderline cases), as a necessary condition for 3. clear division of labor and social differentiation, 4. variation in building types, 5. urban lifestyle, and 6. function of the population center as a focal point for the surrounding area. Editor’s note: This latest version of this paper (translated into Spanish by Katharina A. Niemeyer) appeared under the title “El yacimiento fenicio de Toscanos: urbanística y función” in Los fenicios en la Península Ibérica, vol. 1 (eds. G. del Olmo Lete and M. E. Aubet Semmler; Sabadell [Barcelona]: AUSA, 1986) 109–26 (= Aula Orientalis 3 [1985] 109–26). 1. F. Kolb, Die Stadt im Altertum (Munich, 1984) 99ff. 2. M. Weber, “Die Stadt,” Archiv für Sozialwissenschaft und Sozialpolitik 47 (1921) 621ff. See the reprint in Die Stadt des Mittelalters 1 (3d ed.; ed. C. Haase; Darmstadt, 1978) 41ff.

31

32

Hans Georg Niemeyer

Throughout his work, Kolb uses these criteria to analyze the Greek and Roman city of the classical world. He attributes equal importance to the cities of the ancient Near East for their part in the historical process of urbanization. The concept of urbanization is used in the study of antiquity to describe the historical development of a city as such, based on two distinct sources of information: written records and archaeological discoveries of urban remains. Perhaps the only written references to Toscanos near Torre del Mar (Vélez Málaga, Province of Málaga) are those that mention the Greek polis of Mainake and allude, in my opinion, to the site of Toscanos. 3 Thus, when we apply the concept of urbanization to the history of Toscanos, our only basis is the horizontal and vertical analysis of finds in the various occupation and construction strata that have been excavated at the site. In addition, we assume that Toscanos was a city, a condition that our site satisfies only in part, as we shall see. But we can also accept the fact that Toscanos had at least some attributes of a city, further complicating our analysis of the possible urbanization of Toscanos. Therefore, what follows is merely an attempt to classify the site of Toscanos within the methodological framework developed by the anthropological sciences. I hope to clarify the social and economic structures that dominated life in this border community that existed between the highly developed culture of the ancient Near East and the primitive or less advanced cultures of the hinterland on the Iberian Peninsula. The site was discovered in 1961 on the western bank of the Vélez River near its mouth, at the base of the eastern slope of the hill Cerro del Peñón (fig. 1). The excavations at this site have focused on a small hill located between the bank of the river and Cerro del Peñón, where today two farmhouses of the Toscanos family are located. Here, below this hill, the structural remains of a Phoenician settlement dating from the end of the 8th century to the beginning of the 6th century b.c. have been preserved. Eight seasons of excavation took place between 1964 and 1984, 4 during which houses (A, B, D, and H), as well as more humble buildings or huts (E, F, and G), have appeared. We have also found a more important structure, Building C, which very likely was a storehouse (see map, fig. 2).

Extension of the Settlement Northwest of the nuclear Phoenician settlement of Toscanos, another settlement from the same period has been found on the nearby hill called Cerro de Alarcón (79 meters high). It has been interpreted as a military outpost or outer 3. Niemeyer, Habis 10–11 (1979–80) 79ff. 4. In the following discussion of excavation results at the site of Toscanos, I will not cite bibliographical data at each step of the discussion. Refer to the extensive bibliography at the end of this work.

The Phoenician Settlement at Toscanos

33

Fig. 1. The Phoenician settlement of Toscanos at the mouth of the Vélez River. Map of the archaeological zone indicating the ancient coastline (light gray) and inhabited areas (dark gray). Among the necropolises, the one belonging to Cerro del Peñón has not yet been located. Triangle: location of metallurgic installations.

fortification, built in front of the trading post’s center for protection. We still lack specific information about the density of the population that lived in the area between Toscanos and Cerro de Alarcón, since it has not been possible to excavate more extensively between the two points. But there can be no doubt

34

Hans Georg Niemeyer

Fig. 2. Toscanos: the nuclear settlement.

about the cultural and historical connections between the structures at Cerro de Alarcón and those at the settlement of Toscanos. Remains from the Phoenician settlement have also been preserved on the eastern slope of Cerro del Peñón. These remains were examined already in 1978 and then more thoroughly during the 1984 season. Various examples of imported Greek pottery stand out among the finds: amphorae from Attica and from the island of Khios, an early Corinthian alabastron, and several fragments of Etruscan bucchero sottile, for example. It appears that most of the material at Cerro del Peñón dates to the 7th century b.c., although some dates to the 6th century. Until now we have not found material from the 8th century that would correspond to the material in Strata I and II at Toscanos. Important remains of metallurgic installations, obviously used for working iron, have also been found on the slope of Cerro del Peñón. Using this data we can reconstruct rather easily the urban development of the Phoenician settlement founded around 740/30 b.c. At the beginning, the area that we call the “nuclear zone” was urbanized, and both the “storehouse zone” and the “ashlar masonry zone” are part of this area (fig. 3). We will be-

The Phoenician Settlement at Toscanos

35

gin by examining the storehouse zone, where House A, with only one excavated room, definitely belongs to Stratum I. We still cannot be certain whether House A had a residential or commercial function, but there are various signs that support the first alternative. Immediately to the north and to the west of House A are public areas. During the first phase of settlement, the area extending 15 meters to the west of the public area remained free of any lasting construction (fig. 3). In Stratum II of the settlement we see a denser population of the land (fig. 4). Building A has been enlarged with an annex (B) and has been joined by Houses D, H, and K. These houses have preserved only the foundations of the walls, which were usually constructed from unbaked mudbrick. These buildings, all with several rooms, can be classified unequivocally as residences. It is interesting that this group of buildings does not have the same orientation; instead, the orientation varies slightly with respect to the neighboring buildings during each phase. Therefore, it is possible that a homogeneous, fixed network of streets did not exist at Toscanos, something that Chorreras also appears to have lacked. On the other hand, the clear evidence of earlier structures gives us reason to assume that a more or less obligatory division of land was in place and that an approximately uniform orientation of the different buildings needed to be observed, at least within the same stage of the settlement’s chronological development. The Phoenician settlement of Chorreras 5 was founded before Toscanos, though it was probably of short duration. Chorreras gives the impression of being larger than Toscanos, but we should remember that the initial city layout at Toscanos must have been quite extensive as well. The area designated for urbanization at the time of its founding probably began to fill up more rapidly during its second stage. In Stratum III at Toscanos, the construction of the large Building C began around 700 b.c. (fig. 5), showing characteristics of concentrated settlement. This building was constructed in the immediate vicinity of the western walls of Houses A and H, which had been enlarged with additions to their western side. The space between buildings is hardly large enough for a narrow stairway to pass through. The stairway connected the higher, archaeologicallyverified streets on the north with the lower streets that must have existed on the south, possibly along the river. In order to construct this stairway, the builders had to demolish a part of the western annex (I) of House H. Thus it is evident that they wanted to make Building C as large as possible, but they also had serious difficulties procuring the necessary space within the limits of the predetermined system.

5. M. E. Aubet, G. Maass-Lindemann, and H. Schubart, NAHisp 6 (1978) 91ff., fig. 13.

36

Hans Georg Niemeyer

Fig. 3. Toscanos: Building A of Stratum I and sketch of the defensive trench, storehouse area, and ashlar masonry area.

Fig. 4. Toscanos: Stratum II buildings (B, H, K). See caption on the previous figure.

The Phoenician Settlement at Toscanos

37

Fig. 5. Toscanos: Stratum IV buildings, with storehouse (C). In the ashlar masonry area, the defensive trench has probably been abandoned. See the caption under the previous figures.

Important clues about the original function of Building C come from comparing this building with another building with three sections, excavated by an English archaeological team at the port of Motya and thought to be a storehouse (fig. 6). We can also compare it with other warehouses of the Early Iron Age, such as those at Hazor. 6 The interpretation of Building C as a warehouse is also based on the fact that the majority of sherds found inside are of transport and storage containers. Building C, which is relatively impressive in comparison to the houses and huts studied until now, thus appears to be one of those storehouses that must have existed in every Phoenician settlement on the Mediterranean. In terms of urban evolution, it is interesting to note a visible architectural change in the use of this part of the settlement with the construction of Building C, a change which may have begun earlier, of course. Houses E, F, and G, which were constructed either at the same time or later than Building C, are dwellings of inferior quality. Perhaps these “huts” provided lodging for the personnel working in the storehouse. 6. See Niemeyer, Jahrbuch des Römisch-Germanischen Zentralmuseums 31 (1984) 46ff.

38

Hans Georg Niemeyer

Fig. 6. Storehouse C at Toscanos and the storehouse at Motya (according to Madrider Beiträge 8 [1982] 115, figs. 1 and 2).

In Stratum V we now observe a radical change in the use of the storehouse area. This change also affects the western area, where there had previously been a V-shaped defensive trench or moat. The populated areas that could be examined outside of the nuclear settlement, on Cerro de Alarcón as well as on the eastern slope of Cerro del Peñón, date from a period prior to Stratum V. We might ask whether the settlement of these areas was a necessary condition for the reorganization of the nuclear settlement, even if chronologically there were no causal relation evident between these two events. In any case, the defensive trench must have been abandoned and filled in during this period. In the old nuclear settlement, we now find ashlar structures that must have been important because of their size and quality of workmanship (fig. 7). These structures suffered a great deal of damage during the first imperial period because of the reuse and modification of the ashlars. This damage resembled that discovered by F. Rakob at Carthage, where the inhabitants of the imperial Augustan city exploited the ancient and Hellenistic structures for building

The Phoenician Settlement at Toscanos

39

Fig. 7. Toscanos: ashlar structures on top of the defensive trench, which has been filled in. The buildings in the “storehouse area” have not been retained.

stone until only the foundations remained. 7 But even in the parts of Toscanos which suffered most from this later exploitation (such as the ashlar wall, which could be traced in a straight line for 40 meters), the remains of the earliest construction phases were preserved. This is not the place to enter into detail, but these few data should suffice to demonstrate the fundamental change in the use of the nuclear settlement. It is evident that the population of Toscanos increased quite rapidly during the first two generations. Later, the following events stand out: the construction of the storehouse around 700 b.c. (Stratum III), the expansion of the populated area outside the bounds of the nuclear settlement around the beginning of the second half of the 7th century, and finally the reorganization of the settlement in the decades around the year 600 b.c. or a little after. As shown above, a fairly radical change took place twice in the use of space in the nuclear settlement. Keeping in mind this summary of the development and exterior organization of the Phoenician settlement at Toscanos, we will raise some questions 7. See the preliminary report of F. Rakob, Mitteilungen des Deutschen Archäologischen Instituts, Römische Abteilung 91 (1984) 1ff.; M. Vegas, ibid., pp. 215ff.

40

Hans Georg Niemeyer

about its urban character according to the criteria formulated by Kolb. As we discuss these questions, however, we must not forget the factors that necessarily diminish the general value of the answers. The incomplete state of preservation of the settlement and the unfinished excavations are only the most important points that must be mentioned in this context. Nevertheless, we can deduce the following points from the excavation results. (1) There can be no doubt about the “topographic and administrative unity of the population.” Two consecutive fortification systems protected the population (see fig. 1). The first consisted of a V-shaped trench, which we have observed ever since the 1971 season in the areas opened on the southwest border of the Toscanos farmhouse area. This was a trench or moat cut artificially into bedrock at about a 45o angle. The trench followed the natural lay of the land and probably marked the boundaries and protected the area of the nuclear settlement from the hill and coastal regions. Parallel examples exist in the eastern homeland, including Palestine, testifying to a tradition that goes back to the Bronze Age. The existence of this V-shaped moat or fortification system, to which we must mentally add a wall (unfortunately not preserved), sufficiently explains the density of the population within a fixed area during the phase of most intense settlement (beginning with Stratum II or perhaps I/II). The second fortification system enclosed Cerro de Alarcón as well, where its trajectory has been traced for more than 500 meters. (2) In reference to the second criterion (the number of inhabitants), it must be admitted that the population of Toscanos is very difficult to calculate. The area of Toscanos plus the eastern slope of Cerro del Peñón and the southern slope of Cerro de Alarcón covers a total of 12 to 15 hectares. If we assume a sparse population on the southern slope of Cerro de Alarcón and a primarily industrial use of the areas on Cerro del Peñón, we can begin by supposing a fairly high population density in the nuclear settlement, at least during the period of the storehouse and before the restructuring at the beginning of the 6th century. Based on an average density of 200 people per hectare and allowing that a good share of the area was industrial and not residential, we calculate that in the 7th century between 1000 and 1500 people may have lived here. This is the lower limit set forth by Kolb that would allow us to speak of Toscanos as a “city.” 8 (3) As for the clear division of labor within this population, we have already mentioned some important evidence: the impressive remains of metallurgic installations on the eastern slope of Cerro del Peñón. In order to explain these installations, we must assume a very specialized, professional group of people. There is also evidence of the production of purple dye. We know about other specialized groups through the analyses of bones by R. Soergel 8. Kolb, Die Stadt im Altertum, 15.

The Phoenician Settlement at Toscanos

41

and H. P. and M. Uerpmann. Cattle increasingly predominated in the meat supply, and sheep and goats were next in importance. Considerable accumulations of fish bones have also been found, while pigs and wild game had little importance. This is opposite of what we know about the indigenous sites on the Peninsula. We can deduce that professional fishermen coexisted with cattle raisers at Toscanos, especially during the later periods. For our purposes this is very important, since “the predominance of cattle as providers of meat, a predominance which increases throughout the successive stratigraphy, presupposes an organized and relatively wide distribution of meat.” We can conclude “that in the beginning, small groups of people largely provided for their own needs” (with smaller animals, which were more commonly used in this period), “while later the dependency on the special production of beef increased.” 9 This conclusion also answers the question about social differentiation, as well as Kolb’s first criterion on the administration of the settlement. Without social differentiation and centralized administration, what we have just explained above would have been impossible. But other questions, no less important, are raised about the internal composition of this socially differentiated and centrally governed population. Did it include the complete spectrum of population typical of an “ancient city”? Did a noble governing class of reduced size exist, from which the leaders of Toscanos were recruited? Or were the reins of government in the hands of representatives or commercial agents sent from the Phoenician homeland? We cannot give a definitive answer to the final two questions, but we will see at the end of this paper that there is strong evidence for the second alternative. (4 and 5) We have already discussed architectural diversity extensively in the previous paragraphs; there can be no doubt that there was a variety of architectural types. However, it is almost impossible to go into any detail about whether the trading post at Toscanos had an “urban” lifestyle. The same uncertainty exists about many ancient towns and especially about others of the 7th century b.c. Kolb himself has emphasized that his fifth criterion is very difficult to demonstrate empirically. (6) The question still remains of the city’s function as a focal point for the surrounding area. This criterion is the result of the further evolution of Weber’s ideas on the “market” as a constitutive factor of his concept of “city.” The answer, on one hand, could rely on particularly relevant phenomena, but, on the other hand, large uncertainties exist. In the first place, it has not yet been proven that the Phoenician settlement at Toscanos created a back country that was typically Phoenician, that is to say, its own hinterland or chora. Rather, the Phoenician settlements were strung across the narrow strip of 9. See H. P. and M. Uerpmann, Studien über frühe Tierknochenfunde von der Iberischen Halbinsel 4 (Munich, 1973), especially 83ff.

42

Hans Georg Niemeyer

coastal plain like a necklace of pearls. This phenomenon is very evident on the Iberian Peninsula, for example in the Vélez River valley and in the Málaga basin. During the same time period, just a few kilometers inland and still on the coastal side of the narrow mountain passes, we find sites whose population is entirely indigenous (fig. 8). Moreover, we lack evidence of an intensive Phoenician acculturation. For this early period of the 8th and 7th centuries b.c., we have only the fragments of wheel-made Phoenician pottery, very superior in quality, which appear in the indigenous settlements of the hinterland as unequivocal evidence of importation. These sherds testify to the supposed “cultural emanation” coming from the Phoenician trading posts. In this context, the necropolis at Frigiliana, an indigenous town a few kilometers inland from Almuñécar, is typical. The wheel-made funerary urns cannot be understood without the Phoenician models from the trading posts on the coast. However, their closest precedents must be sought in the cremation urns and painted amphorae of the Cruz del Negro type, especially as we know them from the necropolises at Carmona in the southernmost part of the Guadalquivir River valley. Of course, neither can these vessels be explained without the Phoenician models. However, trade relationships, and along with them the “cultural emanation” of the trading posts, did extend beyond the mountainous region of the coast. Among the many pertinent sites that have been at least partially researched by Spanish archaeologists, I will mention only Cerro de la Mora and Cerro de los Infantes, both of which are located on the edge of the plain west of Granada. The answer to the question concerning the “urban” function of Toscanos, namely whether it was the focal point for the surrounding area, remains ambiguous to a certain point. Although we realize that researchers of ancient urban history have attached great importance to this criterion, it is not possible to resolve satisfactorily the question of the urban nature of the settlement. In this case, then, it does not help to work with a term as generalized as Kolb’s “commercial city,” which he uses to define the Phoenician-Canaanite cities. He sees in them, in contrast to the cities of the ancient Mesopotamian kingdoms, a “new type” of city, whose “development was essentially determined by trade and craftsmanship.” 10 Certainly, trade, crafts, and industry played the dominant role in Toscanos and in the other coastal Phoenician settlements on the Peninsula, as well as in the rest of the Mediterranean where there were settlements similar to Toscanos. However, it is also obvious that we cannot compare Toscanos with examples of the “new type” that Kolb examines in this context: with Ugarit/Ras Shamra during the Late Bronze Age, or with Byblos, Sidon, and Tyre itself in the period after the invasion of the Sea Peoples. 10. Kolb, Die Stadt im Altertum, 40ff.; see also the analysis by Isserlin, RSF 1 (1973) 135ff.

spread is 9 points long

The Phoenician Settlement at Toscanos

43

Fig. 8. Phoenician settlements on Spain’s Costa del Sur, up to the point of Motril, and indigenous sites with Phoenician imports.

It is well known that sociologist Karl Polanyi has coined the term “port of trade,” referring precisely to the Phoenician settlements on the Mediterranean shore. 11 This term designates an institution that controls trade on the geographic and structural border between a society not oriented to a market economy and a mercantile economy or professional trading system. The most important characteristics of a port of trade are, first, its independence from the hinterland as well as from the merchants (in other words, its role as a “buffer”), and second, its function as a meeting point for professional merchants of diverse origins. This model, which surely includes the term “trading port” recently used by S. Frankenstein, 12 refers only to the structural and 11. K. Polanyi, “Ports of Trade in Early Societies,” Journal of Economic History 23 (1963) 30ff. See S. C. Humphreys, “History, Economics, and Anthropology: The Work of Karl Polanyi,” History and Theory: Studies in the Philosophy of History 8 (1969) 165ff.; M. Austin and P. VidalNaquet, Économies et sociétés en Grèce ancienne (Paris, 1972; German edition, 1984), 55ff. 12. S. Frankenstein, “The Phoenicians in the Far West: A Function of Neo-assyrian Imperialism,” in Power and Propaganda: A Symposium on the Ancient Empires (Mesopotamia 7; ed. M. T. Larsen; Copenhagen, 1979) 263ff., especially 278ff.

44

Hans Georg Niemeyer

economic function 13 of the settlements in question and leaves largely unconsidered the “urban” and monumental aspect. Of course, it would be very tempting to use Carthage as an example of the Phoenician “commercial city” in the western Mediterranean, something that Kolb does in the book already cited. 14 Most likely, Carthage became a true “city” relatively quickly. Especially the late accounts that deal with its general appearance during the Hellenistic period describe it in this way. As we continue, however, we will see that in the case of Carthage there are some unique reasons for this development. Carthage is surely the most important Phoenician colony in all of the western Mediterranean because of its historical role. However, it appears to behave as an ordinary Phoenician settlement only at first glance. Its geography offers not only the “typical” Phoenician location (on the coast, oriented toward maritime traffic, and set apart from the mainland), but also a hinterland of enormous size, especially when compared to the narrow coasts of Andalusia and Sardinia. Its fertile plains were still esteemed during the period of the Roman Empire as regions that exported agricultural produce of the highest quality. Even more important in this context is a second factor. It cannot be mere coincidence that within the framework of Phoenician settlement in the West we possess a foundation myth only for Carthage—a legend mingled with historical elements. E. Meyer has rejected it as an accumulation of fantastic stories. I believe, however, that behind this legend hides an extremely interesting historical nucleus that can help explain the peculiarities of the history of Carthage in contrast to that of the other Phoenician settlements. The novelistic traits of the legend are especially evident at the beginning and the end. The story begins with the dispute between Elissa (the Dido of Virgil) and the young Pumai/Pygmalion, both children of King Mettenos/Mattan of Tyre, who died without leaving an adult heir to the throne. Elissa, the older sister who was married to Akerbas, the high priest of Melkart, first assumed a type of regency. With time the dispute between siblings became an irreconcilable conflict. It appears that Pumai claimed the right of masculine succession to the throne. Up to this point the story could be a later “romantic disguise” of a historical truth, although it does not sound so fictional. However, what follows can definitely be interpreted as a description of a historical fact: the nobility split, and part of it sided with Elissa. At an opportune moment they occupied a part of the Tyrian merchant fleet, which had just returned from a 13. See, for example, the profound study by J. Renger, “Patterns of Non-institutional Trade and Non-commercial Exchange in Ancient Mesopotamia,” Incunabula Graeca 82 (1984) 31–124. I am thankful to the author for lending me his manuscript. In this work, Renger, using results from Assyriology, broadly treats Polanyi’s theories and the problem of his reception in his own discipline. 14. Kolb, Die Stadt im Altertum, 43f.

The Phoenician Settlement at Toscanos

45

trading mission undertaken by order of the king of Tyre. They sailed with all their supporters first to Cyprus, where the “secession” gained reinforcements from the Cypriot Phoenicians. Next they sailed toward Carthage, where the Phoenicians of Utica (which had probably been founded only a little earlier) and the indigenous Berber population welcomed the emigrants, or rather, refugees. Here the foundation myth again adopts some romantic-novelistic traits. I will mention only the ox skin, which Elissa/Dido astutely cut into thin strips and used to mark the new territory of Carthage, and the romantic episode with Prince Aeneas, also a refugee. 15 But if we understand the central part of the account in the literal sense of the words, we will perhaps find the most coherent explanation for the fundamental difference between Carthage and the Phoenician settlements that for preliminary and provisional purposes we will continue to call “trading posts.” In the case of Carthage we can surely see the phenomenon of a “complete” population with a broad governing noble class, unlike the representatives of the Tyrian merchants or of the king of Tyre himself that we must assume existed at Toscanos. As later history demonstrates, the Carthaginian governing class knew how to develop its own political dynamic. In addition, it is interesting to note that it is precisely to the written record that we owe this information. I will end here. But I still must mention that the reflections expressed in the previous paragraphs in no way resolve the issue. They represent an attempt to weigh results of the archaeological research (in this particular case, research on the ancient Phoenician site of Toscanos) against theoretical discussion in the fields of history, economics, and anthropology by using archaeology itself, an approach which has been long overdue. Finally, I must mention explicitly the limitations that have been implicitly present in this text. First, we have had to accept the incomplete state of preservation of the remains at Toscanos, from which we have selected only the most important features in order to stay within the intended framework of this paper. Second, the reflections have focused on the first period of Phoenician expansion to avoid the risk of having later developments (which in the western Mediterranean were essentially dominated by Carthage) eclipse the panorama. For the same reason, I have refrained from comparing our information on Toscanos with the process of urbanization in other western Phoenician settlements, such as Motya, Tharros, and Lixus. For the moment, it would be impossible to delve more deeply into these issues by making such a comparison. Thus, at the end of this paper, we are left with more questions than answers.

15. All of the literary tradition in G. Bunnens, L’expansion phénicienne en Méditerranée (Brussels/Rome, 1979).

46

Hans Georg Niemeyer

Toscanos Bibliography 1. Excavation reports 1964 Schubart, H., H. G. Niemeyer, and M. Pellicer Catalán. NAHisp 7 (1963) 150 ff. Niemeyer, H. G., M. Pellicer Catalán, and H. Schubart. AA (1964) 476ff. Niemeyer, H.G., and H. Schubart. Toscanos. Die altpunische Faktorei an der Mündung des Río de Vélez 1: Grabungskampagne 1964 (Madrider Forschungen 6.1). Berlin, 1969. Schubart, H., H. G. Niemeyer, and M. Pellicer Catalán. Toscanos: La factoría paleopúnica en la desembocadura del río de Vélez. Excavaciones de 1964 (EAE 66). Madrid, 1969. 1967 Niemeyer, H. G., and H. Schubart. AA (1968), 344ff. ____ . MM 9 (1968) 76ff. Schubart, H., and H. G. Niemeyer. NAHisp 13/14 (1971) 353ff. 1971 Niemeyer, H. G., and H. Schubart. AA (1972) 226ff. Lindemann, G., H. G. Niemeyer, and H. Schubart. MM 13 (1972) 125ff. Schubart, H., H. G. Niemeyer, and G. Lindemann. NAHisp 16 (1972) 11ff. Schubart, H., G. Maass-Lindemann. NAHisp 18 (1984) 41–205. Maass-Lindemann, G. Toscanos. Die westphönikische Niederlassung an der Mündung des Río de Vélez 3: Grabunskampagne 1971 (Madrider Forschungen 6.3). Berlin, 1982. 1973 Bakker, L., and H. G. Niemeyer. AA (1975) 96ff. ____ . NAHisp 4 (1976) 93ff. See 1976 (MM, NAHisp). 1974 (Jardín Necropolis) Maass-Lindemann, G., and H. Schubart. MM 16 (1975) 179ff. Schubart, H., and G. Maass-Lindemann. AA (1976) 181ff. ____ . NAHisp 6 (1978) 139ff. 1976 Schubart, H. MM 18 (1977) 93ff. (Jardín Necropolis) Niemeyer, H. G. MM 18 (1977) 74ff. (1973 and 1976). ____ . AA (1978) 240ff. ____ . NAHisp 6 (1978) 219ff. Schubart, H. NAHisp 6 (1978) 151ff. 1978 Niemeyer, H. G. NAHisp 6 (1978) 244–49. 1984 Niemeyer, H.G., C. Briese, and R. Bahnemann. “Die Untersuchungen auf dem Cerro del Peñón.” Pp. 155ff. in Forschungen zur Archäologie und Geologie im Raum von Torre del Mar 1983/84 (Madrider Beiträge 14). Mainz, 1988.

The Phoenician Settlement at Toscanos

47

2. Studies on Toscanos Niemeyer, H. G. “Feldbegehung bei Torre del Mar (Prov. Málaga).” MM 3 (1962) 38ff.; 6 (1965) 74ff. Niemeyer, H. G., and H. Schubart. “Ein ostphönikisches Thymiaterion vom Cerro del Peñón.” MM 6 (1965) 74ff. Solá-Solé, J. M. “Textos epigráficos de Toscanos.” MM 9 (1968) 106ff. Soergel, E “Die Tierknochen aus der altpunischen Faktorei von Toscanos.” MM 9 (1968) 111ff. Schubart, H. “Colonias fenicias en la región de Málaga.” Arbor 280 (1969) 37ff. Schubart, H., and H. G. Niemeyer. “La factoría paleopúnica de Toscanos.” Pp. 203ff. in V Symposium de Prehistoria Peninsular. Barcelona, 1969. Pellicer, M. “El yacimiento de Los Toscanos y su contribución al estudio de las cerámicas pintadas Hispanas protohistóricas.” AEArq 42 (1969) 3ff. Niemeyer, H. G. “Zum Thymiaterion vom Cerro del Peñón.” MM 11 (1970) 96ff. Hoz Bravo, J. de. “Un grafito griego de Toscanos y la exportación de aceite Ateniense en el siglo VII.” MM 11 (1970) 102ff. Niemeyer, H. G. “Zwei Fragmente ostgriechischer Schalen von Toscanos.” AEArq 44 (1971) 152ff. Almagro Gorbea, M. “Los dos jarros paleopúnicos del Museo Arqueológico Nacional hallados en la Casa de la Viña (Torre del Mar).” MM 13 (1972) 172ff. Díaz Esteban, F. “Dos notas a las inscripciones de Toscanos.” MM 13 (1972) 158ff. Uerpmann, M. “Archäologische Auswertung der Meeresmolluskenreste aus der Westphönizischen Faktorei von Toscanos.” MM 13 (1972) 164ff. Niemeyer, H. G. “Orient im Okzident. Die Phöniker in Spanien.” MDOG 104 (1972) 5ff. Schubart, H. “Phönizische Niederlassungen an der spanischen Küste und ihre Beziehungen zum Hinterland.” Pp. 1ff. in Studien über frühe Tierknochenfunde von der Iberischen Halbinsel 4. Munich, 1973. Uerpmann, H. P., and M. Uerpmann, “Tierknochenfunde aus der phönizischen Faktorei von Toscanos.” Pp. 35ff. in Studien über frühe Tierknochenfunde von der Iberischen Halbinsel 4. Munich, 1973. Boessneck, J. “Vogelknochen aus der phönizischen und römischen Niederlassung von Toscanos.” Pp. 101ff. in Studien über frühe Tierknochenfunde von der Iberischen Halbinsel 4. Munich, 1973. Lepiksaar, J. “Fischknochenfunde aus der phönizischen Faktorei von Toscanos.” Pp. 109ff. in Studien über frühe Tierknochenfunde von der Iberischen Halbinsel 4. Munich, 1973. Schubart, H. “Las excavaciones de Torre del Mar y el panorama arqueológico de las fundaciones de colonias fenicias en la costa mediterránea de la Península Ibérica.” Papeles del Laboratorio de Arqueología de Valencia 11 (1975) 199ff. ____ . “Westphönizische Teller.” RSF 4 (1976) 179ff. Pérez Die, M. C. “Notas sobre cuatro vasos egipcios de Alabastro procedentes de Torre del Mar (Málaga), conservados en el Museo Arqueológico Nacional de Madrid.” Revista de Archivos, Bibliotecas y Museos 79 (1976) 903ff. Treumann, B. W. “West-Phoenician presence on the Iberian Peninsula.” The Ancient World 1 (1978) 15ff. Baena del Alcázar, L. “Fragmentos de vasos de alabastro en yacimientos fenicios de la provincia de Málaga.” Anejos de Baetica 1 (1978) 159ff. Niemeyer, H. G. “Auf der Suche nach Mainake: der Konflikt zwischen literarischer und archäologischer Überlieferung.” Historia 29 (1980) 165ff.

48

Hans Georg Niemeyer

____ . “A la búsqueda de Mainake: el conflicto entre los testimonios arqueológicos y escritos.” Habis 10–11 (1979–80) 279ff. ____ . “Die phönizische Niederlassing Toscanos: eine Zwischenbilanz.” Pp. 185–206 in Phönizier im Westen. Die Beiträge des Internationalen Symposiums über “Die phönizische Expansion im westlichen Mittelmeerraum” in Köln vom 24. bis 27. April 1979 (Madrider Beiträge 8). Mainz 1982. ____ . “El yacimiento fenicio de Toscanos: balance de la investigación 1964–1979.” Huelva Arqueológica 6 (1982) 101–27. ____ . “La cronología de Toscanos y los yacimientos fenicios en las costas del Sur de la Península Ibérica.” Pp. 633ff. in Atti del I Congresso Internazionale di Studi Fenici e Punici, 3. Rome, 1983. ____ . “Phönizische Blasebalgdüsen?” Der Anschnitt 35 (1983) 50ff. Niemeyer, H. G., I. Keesmann, and F. Golschani. “Schlackenfunde von Toscanos.” MM 24 (1983) 65ff. Schulz, H. D. “Zur Lage holozäner Küsten in den Mündungsgebieten des Río de Vélez und des Río Algarrobo (Málaga).” MM 24 (1983) 55 ff. Niemeyer, H. G. “Una ánfora Chiota procedente de Toscanos.” Pp. 253–58 in Homenaje al Prof. M. Almagro Basch, 2. Madrid, 1983. ____ . “Die Phönizier und die Mittelmeerwelt im Zeitalter Homers.” Jahrbuch des RömischGermanischen Zentralmuseums 31 (1984) 3–94. ____ . “Griechische Keramik in Phönizischen Faktoreien. Der Befund der Kampagne 1967 in Toscanos (Málaga).” Pp. 212–17 in Ancient Greek and Related Pottery. Allard Pierson Series 5. Amsterdam, 1984. ____ . “Trade before the Flag? On the Principles of Phoenician Expansion in the Mediterranean.” Pp. 335–44 in Biblical Archaeology Today, 1990: Proceedings of the Second International Congress on Biblical Archaeology. Edited by A. Biran and J. Aviram. Jerusalem: Israel Exploration Society, 1993. ____ . “Die Phönizier im Mittelmeerraum: Expansion oder Kolonisation?” Universität Hamburg 1994. Schlaglichter der Forschung zum 75. Jahrestag. Hamburger Beiträge zur Wissenschaftsgeschichte 15 (1994) 321–47. ____ . “Expansion et colonisation.” Pp. 247–67 in La civilisation phénicienne et punique. Manuel de recherche. =Handbuch der Orientalistik 1 (Der nahe und mittlere Osten), volume 20. Edited by V. Krings, 1995. ____ . “Phoenician Toscanos as a Settlement Model? Its Urbanistic Character in the Context of Phoenician Expansion and Iberian Acculturation” in Social Complexity and the Development of Towns in Iberia. Edited by V. B. Cunliffe and S. Keay. Proceedings. British Academy 86 (1995) 67–88. ____ . “Dreissig Jahre phönizische Archäologie auf der Iberischen Halbinsel. MM 36 (1995) 103–10.

Phoenician and Punic Sexi Manuel Pellicer Catalán

Review of the Research Relating to Colonization and Orientalization at Sexi The month of March 1962 turned out to be a lucky one for research on eastern colonization in the West. The first Phoenician necropolis on the Andalusian coast was discovered entirely by accident, in spite of my earlier attempts to find the fabled colony of Sexi. Workmen were in the process of removing the side of a hill called Cerro de San Cristóbal, a promontory located one kilometer to the northwest of the Castle of Almuñécar, with the purpose of constructing a building complex for the city’s fishermen, when pits began to appear that contained large alabaster vessels with inscriptions and various other grave goods at the bottom, 3 to 4 meters down. These finds caused the expected surprise among the workers and would have disappeared, as so often happens, if it had not been for the determined intervention of Dr. Laura de Prieto Moreno, who was able to recover some of the materials that had been found. The news traveled quickly to the Dirección General de Bellas Artes (General Administration of Fine Arts, in Madrid) through the Delegación Provincial de Arqueología (Provincial Office of Archaeology) of Granada. The provincial office released a notice in a city newspaper and considered the find Roman, confusing the Phoenician red slip pottery with terra sigillata pottery, it appears. The director of the Dirección General de Bellas Artes, Gratiniano Nieto, with whom I was professionally associated, showed me the photographs of one of the alabaster urns with an Egyptian inscription, surprising me so much that I immediately left Madrid for Almuñécar in order to prepare a report on this amazing discovery. The excavation began on April 3, 1962. The small funding that we had received allowed for an excavation of only 15 days, until the 18th of that month. Editor’s note: This article appeared in Spanish under the title “Sexi fenicia y púnica” in Los fenicios en la Península Ibérica, vol. 1 (eds. G. del Olmo Lete and M. E. Aubet Semmler; Sabadell [Barcelona]: AUSA, 1986) 85–107 = Aula Orientalis 3 [1985] 85–107).

49

50

Manuel Pellicer Catalán

However, after a new grant was received, I was able to carry out a second phase during the entire month of July, supposedly exhausting the site. I spent the month of August studying the necropolis and writing my report with enormous difficulty, since it was summer vacation for the research centers and libraries in Madrid. I must thank Antonio García y Bellido, director of the Rodrigo Caro Institute of the Consejo Superior de Investigaciones Científicas (Higher Council for Scientific Research) for kindly permitting me to use the library. I delivered the report to the press at the end of the month, and it was published in September with the title Excavaciones en la necrópolis púnica “Laurita” del Cerro de San Cristóbal (Almuñécar, Granada). 1 At that time, the name “Punic” seemed to me to be the most appropriate. I did not dare to consider the site Phoenician, since I found so many similarities with the necropolises at Carthage, especially those of Junon, Dermech, and Douïmes, and also with the Sicilian necropolis at Motya. In the second publication, 2 I had already changed the adjective “Punic” to “Paleo-Punic,” which was kindly suggested by H. Schubart, a member of the Deutsches Archäologisches Institut (German Archaeological Institute) in Madrid. Today, with the huge amount of research during the past 20 years on Phoenician colonization in the West, we are able to consider the necropolis “Phoenician.” Thus “Laurita” represents the first Phoenician necropolis on the Iberian Peninsula to be systematically excavated and published. Actually the discovery was not surprising, since my teacher García y Bellido and I were already planning a vast project for surveying and excavating at Adra, Almuñécar, and Cádiz (a project which never took place). We hoped to locate and demonstrate what he had predicted in his valuable work Fenicios y Cartagineses en Occidente (Phoenicians and Carthaginians in the West) (Madrid, 1942). This prediction was later repeated by P. Cintas, who “did not doubt that very soon Spanish archaeologists would procure for us one of the most pleasant surprises of the century in the field of Carthaginian archaeology.” 3 The decade of the 50s was a preface for decisive later discoveries, which would lead to obtaining the data and conclusions that we presently possess about the Phoenician presence in Spain. In his works entitled “Orientalia” on Phoenician and orientalizing materials on the Peninsula, 4 A. Blanco presented the products and results of an intense and ancient Phoenician colonization. 1. See M. Pellicer, Excavaciones en la necrópolis púnica “Laurita” del Cerro de San Cristóbal (Almuñécar, Granada) (EAE 17; Madrid, 1962). 2. See M. Pellicer, “Ein altpunisches Gräberfeld bei Almuñécar (prov. Granada),” MM 4 (1963) 9–38. 3. See P. Cintas, La céramique punique (Tunis, 1950). 4. See A. Blanco, “Orientalia I,” AEArq 19 (1956) 3–51; idem, “Orientalia II,” AEArq 33 (1960) 3–43.

Phoenician and Punic Sexi

51

But even though Blanco skillfully analyzed the gold and silver work, many pieces of great aesthetic and scientific value regretfully did not have a context or place of origin. However, the red slip mushroom-mouth jug from Casa de la Viña at Torre del Mar (province of Málaga), included in the Visigoth collections in the Museo Arqueológico Nacional (National Archaeological Museum) and published by A. Fernández Avilés, 5 alerted scholars to Phoenician colonization on the Peninsula and signaled the importance of the mouth of the Vélez River on the coast of Málaga. It was precisely during this period at the end of the 50s when the treasure of El Carambolo was accidentally found. The magnificence of its indigenous, orientalizing gold and silver work obscured the importance of the site where it had been hidden and which J. de Mata Carriazo excavated. 6 Meanwhile, J. Maluquer deduced the dates and stratigraphy of the ancient town. 7 A decade later the results of the excavations were published. 8 The groundwork had now been laid for the discovery of Phoenician sites on the Spanish coastline. For that reason, in April of 1960 I did a stratigraphic section in the southern side of the hill where the Castle of Almuñécar stands. The section revealed five levels, all of which produced modern, medieval, and Roman materials. However, in the lowest level 1.15 meters down, there appeared several fragments of a type B2 Ionian cup, probably Samian, from the first half of the 6th century b.c., with parallel black lines on the outside, a black slip on the inside, and an unpainted rim. 9 The hurried publication on the necropolis “Laurita” certainly caused a sensation among Spanish scholars and generally throughout international scholarship dealing with eastern colonization in the West. Since that moment, many pages have been written about this famous necropolis: its dates, its eastern or Carthaginian nature, the alabaster vases and their origin and age, the reasons why the vases traveled to Almuñécar, the Semitic and Egyptian hieroglyphic inscriptions with their reading, interpretation, and age, the painted ostrich eggs, use of the proto-Corinthian kotylai as a basis for dating, the scarabs, the gold and silver work, the red slip pottery, such as plates and mushroom-mouth and pear-shaped trefoil-mouth jugs, their typology and relationship to the 5. See A. Fernández, “Vaso oriental de Torre del Mar (Málaga),” Arqueología e Historia 8 (Lisbon, 1958) 39–42. 6. See J. de M. Carriazo, Las joyas y las excavaciones del Carambolo (Archivo Hispalense 30; Seville, 1959). 7. See J. Maluquer, “El tesoro tartésico del Carambolo,” in I Congreso Nacional de Arqueología (Lisbon, 1959). 8. See J. de M. Carriazo, El tesoro y las primeras excavaciones en el Carambolo (EAE 68; Madrid, 1970); idem, Tartesos y el Carambolo (Madrid, 1973). 9. See M. Pellicer, “Actividades de la Delegación de Zona de la Provincia de Granada durante los años 1957–1962,” NAHisp 6 (1964) 303–50, fig. 33.

52

Manuel Pellicer Catalán

tombs, etc. All this vast amount of literature has obviously served to qualify, add detail to, and deepen our understanding, especially in reference to the alabaster vases, but the fact remains that the chronological-cultural approach that I proposed in 1962 has not varied substantially after more than 20 years. Part of the material was exhibited a year later at the 8th National Archaeological Congress held in Seville and Málaga, and I had the opportunity to present a report related to this topic. 10 Currently the material from the necropolis is exhibited in the Archaeological Museum of Granada, with the exception of some alabaster urns which I did not publish. Also some of the gold and silver work stayed in private hands in Almuñécar but is apparently being placed in the Museo Monográfico at Almuñécar, due to the work carried out there by F. Molina Fajardo. Scarcely a year after the publication on the necropolis “Laurita,” J. Maluquer wrote on the topic, 11 commenting on the discovery and stressing its exceptional importance. J. Leclant, an Egyptologist whom I had met during my excavations in Nubia and to whom I sent photographs of the hieroglyphics on the alabaster vases at Almuñécar for study, began to publish notes in 1964 in the journal Orientalia 12 on the unquestionable importance of these vases. F. Díaz studied the painted inscription on an alabaster urn from Tomb 3A and the writing on the plate from Tomb 16 at Laurita, 13 and F. Ferron also studied the urn. 14 Ferron believed that the inscription was Carthaginian, but his opinion has not been shared by other researchers such as J. Leclant and G. C. Picard, who consider it purely Phoenician. J. M. Blázquez has wisely compiled the results of the research carried out on the Iberian Peninsula on the Phoenician orientalizing world, especially since the end of the 50s. He has published a work which is extremely useful because of its encyclopedic and bibliographic nature. In it he compares materials and theories of different researchers on this splendid new world in which the necropolis “Laurita” has had an especially large impact. 15 W. Culican, who is very familiar with the eastern Phoenician world, studied the alabaster urns of “Laurita” in 1970, looking for their origin and for 10. See M. Pellicer, “Relaciones de la necrópolis púnica del Cerro de San Cristóbal de Almuñécar en el Mediterráneo occidental,” in VIII Congreso Nacional de Arqueología (Zaragoza, 1964) 193f. 11. See J. Maluquer, “Descubrimiento de la necrópolis de la antigua ciudad de Sexi en Almuñécar (Granada),” Zephyrus 14 (1963) 57–63. 12. See J. Leclant, Or 33 (1964) 337–404; 34 (1965) 229; 35 (1966) 174; 36 (1967) 223–24. 13. See F. Díaz, “Dos nuevas inscripciones púnicas hispánicas,” Sefarad 25 (1965) 283–87. 14. See F. Ferron, “La inscripción cartaginesa pintada en la urna cineraria de Almuñécar,” TrPrHist 27 (1970) 177–90. 15. See J. M. Blázquez, Tartessos y los orígenes de la colonización fenicia en Occidente (Salamanca, 1979).

spread is 12 points short

Phoenician and Punic Sexi

53

relationships with other parallel eastern vases, including those found in Egypt and those transported to Phoenicia and to the Assyrian world. 16 Work on the transcription, translation, and interpretation of the Egyptian hieroglyphics on the alabaster urns of Tombs 1, 15, 16, 17, and 20 was finally carried out by I. Gamer and J. Padró beginning in 1973. 17 Their work confirmed that these vessels belonged to the 22d dynasty and to the Pharaohs Osorkon II (874– 850), Takelot II (850–825), and Sheshonq III (825–773). Padró corrected the name of Sheshonq II, which I had proposed in my publication, to Sheshonq III, which had the interesting result of moving the date for some of the vessels closer to the end of the 8th century, the date which is calculated for the beginning of the necropolis. Some time ago we heard about the existence of another marble Egyptian vessel which had been discovered in Almuñécar during the past century, but locating it and then studying it were difficult because it was privately owned. Due to the excellent work of Molina Fajardo, the vessel was found, photographed, sketched, and published, with F. Daumas of the University of Montpellier 18 supervising its study (figs. 6 and 7). The painted ostrich eggs from “Laurita,” so common in the Phoenician and Punic necropolises in the West, were included by M. P. Nicolás in his general work on these materials. 19 Cintas, who dedicated part of his life to the excavation of the necropolises at Carthage and with whom I maintained a friendly and beneficial correspondence until his death, finally published his manual on Punic archaeology. 20 He included the results of my excavations, arguing for an early date for part of the necropolis “Laurita” and pushing back the date of its origin to the 9th 16. See W. Culican, “Almuñécar, Assur and the Phoenician penetration of the Western Mediterranean,” Levant 2 (1970) 28–36. 17. See I. Gamer, “La inscripción del vaso de alabastro de la tumba no. I de Almuñécar (Granada),” in XII Congreso Nacional de Arqueología (Zaragoza, 1973) 401–8; idem, Aegyptische und Aegyptisierende Funde von der Iberischen Halbinsel (Beihefte zum Tübinger Atlas des Vorderen Orients, Reihe B, no. 21; Wiesbaden, 1978); J. Padró, “Precisiones sobre la identificación del cartucho de un rey Sheshonq en Almuñécar,” in XIII Congreso Nacional de Arqueología (Zaragoza, 1975) 751–58; idem, Los materiales de tipo egipcio del Litoral Mediterráneo de la Península Ibérica (Doctoral thesis, Universidad de Barcelona, 1976; revised English version, Leiden, 1980); idem, “Las inscripciones egipcias de la Dinastía XXII procedentes de Almuñécar (provincia de Granada),” AuOr 1 (1983) 215–26; idem, “Los fenicios y la distribución de objetos egipcios en el extremo occidente del Mediterráneo,” in Atti del I Congreso Internazionale di Studi Fenici e Punici, 1 (Rome, 1983) 67–76. 18. See F. Molina Fajardo et al., Almuñécar: Arqueología e Historia (Granada, 1983) 35–44. 19. See M. P. San Nicolás, “Las cáscaras de huevo de avestruz fenicio-púnicas en la Península Ibérica y Baleares,” in Cuadernos de Prehistoria y Arqueología 2 (Madrid: Universidad Autónoma, 1975) 75–100. 20. See P. Cintas, Manuel d’Archéologie I/II (Paris, 1970/1976).

54

Manuel Pellicer Catalán

century or beginning of the 8th century b.c. His argument is based on a high chronology for the alabaster urns, an opinion that I have always considered unacceptable. I had another opportunity to discuss the geographic characteristics of Sexi’s location when I wrote about the discovery of another Phoenician colony on the bay of Algeciras, next to Carteia. 21 Two proto-Corinthian kotylai from Tomb 19B, dated to 675–650 b.c., have provided a precise date for the context of the materials at Laurita. The chronology of the tombs at Almuñécar, which were used around these dates with the possibility of slightly earlier and later burials within the first three quarters of the 7th century and during the final quarter of the 8th century b.c., helped to date the stratigraphy at Toscanos, where I participated with H. Schubart and H. G. Niemeyer during their first season in 1964. 22 Among the materials used for comparison, the red slip plates (fig. 9) have had an extremely important role in dating, based on the width of the rim and the diameter. This research was carried out by Schubart with optimal results except in Huelva, where the normal evolution of the width of the rim through time did not continue. 23 With respect to amphorae, only one rim and one Phoenician handle have appeared at the necropolis “Laurita,” coming from Tomb 19 and dated to the 7th century. However, at the Punic necropolis of Puente de Noy the amphorae were abundant and varied, as well as in the shipwrecks surrounding Almuñécar. Molina Fajardo has collected these amphorae, established a typology (fig. 10), and published the results. 24 Most of the amphorae were dated according to the chronology established at Cerro Macareno (province of Sevilla), a site with 8 meters of stratigraphic depth and an uninterrupted chronology from the middle of the 8th century to the beginning of the 1st century b.c. 25 In the publication “Laurita” I presented a typology of the tombs, which I related to others that were approximately from the same time period in the central Mediterranean. Because of this work, I thought it advisable that a doctoral thesis be undertaken on the typology, relationships, and chronology of the Phoenician and Punic tombs in the western Mediterranean, a study which A. Tejera carried out under my direction. 26 We also needed a study on red slip 21. See M. Pellicer, L. Menanteneau, and P. Ruillard, “Para una localización de colonias fenicias en las costas ibéricas. El Cerro del Prado,” Habis 8 (1977) 217–52. 22. See H. Schubart, H. G. Niemeyer, and M. Pellicer, Toscanos, la factoría paleopúnica en la desembocadura del río Vélez (EAE 66; Madrid, 1969) 1. 23. See H. Schubart, “Westphönizische Teller,” RSF 4 (1976) 183, 185. 24. See F. Molina Fajardo et al., Almuñécar, 131–58. 25. See M. Pellicer, “Tipología y cronología de las ánforas prerromanas del Guadalquivir según Cerro Macareno (Sevilla),” Habis 9 (1978) 365–400. 26. See A. Tejera, Las tumbas fenicias y púnicas del Mediterráneo Occidental (Seville, 1979).

spread is 12 points short

Phoenician and Punic Sexi

55

Phoenician pottery, in view of the abundant material from “Laurita” and from the other sites, including towns as well as necropolises of a colonial or orientalizing type in the southern Peninsula. This work was carried out by I. Negueruela under my direction, and after dating this ceramic type, he was able to establish a more precise chronology for our necropolis. 27 With respect to the dating of the necropolis “Laurita,” I have already indicated that the proto-Corinthian kotylai from Tomb 19B provided the key. In my publication I presented certain parallels that had quite precise dates; thus the kotylai probably came from the end of the 8th century b.c. when included in Kreiker’s initial proto-Corinthian typology. 28 According to Payne’s study, 29 the kotylai of Almuñécar probably correspond to the first style with black figures, from the first quarter of the 7th century b.c. But twenty years after the publication of “Laurita,” the substantial study by B. B. Shefton on Greek imports in the southern Iberian Peninsula evaluates the chronology further, with the result that although both kotylai are considered products of Pitecusas, the sample with a black slip on its lower section is probably from the end of the 8th century, while the one with a rayed foot is probably later, from the first half of the 7th century b.c. 30 Molina Fajardo, who dedicated several years to the systematic study of the archaeology at Almuñécar, recently located, excavated and published information on a new necropolis, the one at Puente de Noy, which is Punic in character and thus later than “Laurita.” He has also presented information on his excavation of the Roman cistern called “Cueva de Siete Palacios” (“Cave of Seven Palaces”), built on the site of Phoenician Sexi and containing materials that may go back to the 8th century b.c. 31 (figs. 3 and 4). Among the many works that deal in some way with the necropolis “Laurita,” we must also mention that of Niemeyer, who is studying the chronology of the Spanish Phoenician sites. 32

27. See I. Negueruela, “Zur Datierung des westphönizischen Nekropole von Almuñécar,” MM 22 (1981) 211–28; idem, “Jarros de boca de seta y de boca trilobulada de cerámica de barniz rojo en la Península Ibérica,” in Homenaje al Prof. M. Almagro Basch, 2 (Madrid, 1983) 259–79. 28. See W. Kreiker, Aigina. Die Vasen (Berlin 1951) pl. 10.185. 29. See H. G. G. Payne, Protokorintische Vasemalerei (Berlin, 1933) pl. 10.4. 30. See B. B. Shefton, “Greeks and Greek imports in the South of the Iberian Peninsula: The archaeological evidence,” in Phönizier im Westen (Madrider Beiträger 8; Mainz, 1982) 337–67. 31. See F. Molina Fajardo, Almuñécar en la antigüedad. La necrópolis fenicio-púnica de Puente de Noy (Granada, 1982); F. Molina Fajardo et al., Almuñécar. 32. See H. G. Niemeyer, “La cronología de Toscanos y los yacimientos fenicios en las costas del sur de la Península Ibérica,” in Atti del I Congresso Internazionale di Studi Fenici e Punici, 3 (Rome, 1983) 633–36.

subhead drop

56

Manuel Pellicer Catalán

Information from Ancient Writers The written sources on Sexi and especially on Phoenician Sexi are very scarce. The city is first mentioned by Hecataeus of Miletus around 500 b.c. From that time on the city of Sexi is sporadically and briefly mentioned by Diphilus in the 4th century b.c., 33 by Pliny, who tells us that the city is also called Firmum Iulium, 34 by Pomponius Mela, 35 by Ptolemy, who limits himself to giving its geographic position, by Martial, and by Stephanus of Bizantium, who says that Sexi is the city of the Mastieni. The writers constantly allude to its famous salted fish. 36 Strabo’s text holds special interest for speculating on the early Phoenician presence in Sexi. He says: Malaca is closer to Kalpe [than Mainake] and has a Phoenician settlement. Afterwards comes the city of the Exitani, from which the salted fish receive their name. After this city [Sexi] comes Abdera, which also is a Phoenician colony.37 Concerning the founding of their city, the people of Gadir remember the following: that a certain oracle commanded the Tyrians to found a colony at the Columns of Hercules. The explorers arrived at the strait of Kalpe and believed that the promontories that formed it were the ends of the inhabited earth and the end of Hercules’ labors. Supposing, therefore, that these were the columns to which the oracle referred, they dropped anchor at a certain spot on this side of the Columns, where today the city of the Exitani [italics added] is found. But after they offered a sacrifice to the gods at this spot on the coast and the victims were not propitious, they returned. Later the explorers crossed the strait, arriving at an island consecrated to Hercules, located next to Onoba, a city of Iberia and some 1500 stadia on the other side of the strait. Believing that the Columns were there, they again sacrificed to the gods, but the victims were unfavorable again, and they returned to their homeland. On the third expedition they founded Gadir and built the sanctuary on the eastern part of the island and the city on the western part. 38

Strabo’s text is sufficiently explicit for us to deduce that the first point where the Phoenician explorers stopped and disembarked on their explorations in the western Mediterranean was Sexi, then came Onoba (Huelva), and finally they founded Gadir (Cádiz). If we disregard the early dates for the founding of Utica and Cádiz, at the end of the 2d millennium according to the ancient writers and not yet docu33. Athenaeus 3.21.a; A. Schulten, Fontes Hispaniae Antiquae II (Barcelona, 1925) 85. 34. Pliny 3.8. 35. Pomponius Mela 2.9. 36. Strabo 3.4.2. 37. Strabo 3.4.2,3. 38. See A. García Bellido, España y los españoles hace dos mil años, según la geografía de Estrabón (Madrid, 1954) 200–202; Strabo 3.5.5.

spread is 6 points long

Phoenician and Punic Sexi

57

mented archaeologically in spite of the intense work carried out at these sites, it is very probable that the story of the exploration or expedition narrated by Strabo took place already at the end of the 9th century b.c. and the definitive settlement at the beginning of the 8th century b.c., according to the latest archaeological data. It is also possible that the first settlement was Sexi, next came Onoba, and then Gadir, and still later came the rest of the colonies, including those at the mouths of the Algarrobo and Vélez Rivers.

Location and Site of Sexi The Phoenician settlements, not only the metropolises of the Syro-Palestinian and Cyprian coasts, but also the colonies in the central and western Mediterranean and on the Atlantic coast of the Iberian Peninsula and Morocco, correspond to certain consistent topographic features, such as mouths of rivers, estuaries, bays, swamps, promontories, and islands. 39 The site of Sexi at Almuñécar is at the mouth of two rivers, the Seco to the west and the Verde to the east, with a promontory between them (fig. 1) that very possibly was an island some 3000 years ago. The old city center known for its castle is found there today. It was on the side of this promontory, between the Castle and the old Hotel Sexi, that I did a sondage which resulted in the discovery at the lowest levels of fragments of Type B2 Ionian cups, probably Samian, from the 6th century b.c. 40 Molina also found both rough and fine indigenous pottery from the Final Bronze close to the castle in the Roman cistern called “Cueva de Siete Palacios,” in the lowest level of his excavation. The indigenous pottery is decorated with triangular incisions and is dated to the 8th century. 41 There are carinated bowls with ungular rims thickened on the inside (rims shaped like inverted, truncated cones), 42 similar to those of Phase II at Cabezo de San Pedro in Huelva and dated to the end of the 8th century 43 (fig. 3). Alongside this pottery he found Phoenician red slip pottery with similar dates, including carinated, ungular, concave bowls with prominent beveled rims, 44 plates with narrow rims of 38 mm, which date to the 8th century according to the calculations made by Schubart, 45 a red slip, pear-shaped oenochoe with the same age, and numerous shells and fish vertebrae (fig. 4). 39. See M. Pellicer et al., “Para una localización,” 217–52. 40. See M. Pellicer, “Tipología y cronología,” fig. 33. 41. See F. Molina Fajardo et al., Almuñécar, fig. 23.1. 42. Ibid. fig. 23.2. 43. See J. M. Blázquez et al., Excavaciones en el Cabezo de San Pedro (Huelva). Campaña de 1977 (EAE 102; Madrid, 1979) fig. 25 (137), 28 (196, 202). 44. See M. E. Aubet et al., “Chorreras, un establecimiento fenicio al este de la desembocadora del río Algarrobo,” NAHisp 6 (1979) fig. 7 (80); F. Molina Fajardo et al., Almuñécar, fig. 25 (15–17). 45. See H. Schubart, “Westphönizische Teller,” RSF 4 (1976) 183, 185.

58

Manuel Pellicer Catalán

Fig. 1. Location of Sexi and its necropolises.

The nature of the nucleus of settlement at Sexi is probably similar to that of other settlements on the southern coast of Spain and on the Mediterranean and Atlantic. At first these settlements were referred to as colonies in the research, but this term did not seem to apply in all cases, which is why they began to be considered as trading posts, market centers, “relay” stations, etc. Sexi must have been founded as a Phoenician colony with all the characteristics that this term implies. It was a colony because there were colonists, people from the outside who settled there. However, it is difficult, in view of the materials from the lowest stratum of the “Cueva de Siete Palacios,” to affirm whether the colonists began from scratch, as normally happens in a colony, or if the new Phoenician population superimposed itself upon the indigenous population, which seems unlikely. If the Phoenician population was not superimposed, we need to accept that the Final Bronze pottery found in a Phoenician context must correspond to cheap trade acquisitions by the colonists. The main function and economic basis of Sexi must have been trade, which would have made it into a market center. Its location on a natural route like

Phoenician and Punic Sexi

59

Fig. 2. Inventory of materials from the tombs of “Laurita.”

that of the Jete River connecting with the Genil River and Cástulo would have permitted the transport of metals, especially silver. The settlement could also be considered a trading post, since the industry of salted fish must have played a crucial role, as mentioned later by the ancient writers and depicted on coins showing tuna. We must not rule out the agricultural possibilities of the broad, fertile river plain and the splendid climate suitable for olives, grapes, wheat, pastures, etc.

The Necropolises at Sexi: Typology of the Tombs and Burial Practices The necropolis “Laurita” is located on a coastal promontory about 800 meters to the northwest of the original center of Sexi (fig. 1). The Seco River runs between the promontory of Cerro de San Cristóbal, site of the necropolis “Laurita,” and the promontory where the castle is located and where the center of the ancient city once stood. River silt has filled in the space between the promontories. The tombs stretch along the high part of the southeast side of the hill, facing the city center, and they cover an oblong area of about 750 square meters, about 50 meters long from southwest to northeast and 15 meters wide from northwest to southeast. The area could be larger, since the

60

Manuel Pellicer Catalán

2 1

4 3

5 6

7

8

9

Fig. 3. Materials from the Final Bronze. Cueva de Siete Palacios (after F. Molina).

Phoenician and Punic Sexi

61

15

16

17

18

19

Fig. 4. Red slip Phoenician pottery. Cueva de Siete Palacios (after F. Molina).

62

Manuel Pellicer Catalán

lowest and supposedly oldest tombs are probably hidden under the urban development that brought about its discovery. The necropolis of Puente de Noy, 46 of a totally different type and coming at least two centuries after “Laurita,” is also located about 800 meters from the castle, this time to the north, and a half kilometer to the northeast of Cerro de San Cristóbal (fig. 1). The surface area has not been determined because of the construction that covers it, but could easily have been a hectare, given the density of tombs in the sectors already excavated. In “Laurita” there are four types of tombs, although the differences are not great. All have a more or less cylindrical pit with an average diameter of 2 meters and an average depth of 3.50 meters, dug into a subsoil of schist. Type 1 has one lateral niche (tombs 2, 14, 17, and 20); Type 2 has two lateral niches, converging rather than opposite each other (tombs 15 and 19); Type 3 has a lateral niche and a type of rough central cist (tombs 1 and 3); and Type 4 has a central trench at the bottom (tombs 10, 11, 12, 13, and 16). Another type of tomb is the simple pit without any burial. After analyzing the possible relationships between the topography and dates of the tombs of “Laurita,” we were unable to find any positive correlation. Types 1, 2, and 3 with lateral niches probably had their remote origin in the Egyptian mastaba of the 3d millennium b.c., which found its way to Palestine and Syria and was common in Byblos already in the 12th century b.c. This type is passed on to the central and western Mediterranean, with a pit reaching more than 20 meters deep at Gebel Mlezza (Tunis), a somewhat shallower pit in Utica, and shallower pits (only 4 meters deep) in “Laurita.” The pit tombs of the necropolis at Junon in Carthage are shallower yet than those at Almuñécar, reaching barely a meter and a half, and, in addition, the cremations are deposited at the bottom of the pits. However, in Tomb 8 at Junon (corresponding to our Type 1), the burial rite and the grave goods are identical to what we find in the tombs at “Laurita,” making it the nearest parallel. This pit, which gives access to a chamber larger than the simple niche at Almuñécar, is a frequent type in Carthage, Sardinia, Malta, Algeria, and Ibiza. Tomb 4 in the eastern sector of the necropolis at Puente de Noy in Almuñécar, the largest and oldest tomb in this necropolis, belongs to this type of large chamber built with ashlars. Tomb 4 is spectacular and consists of a prismatic pit with a square base 5.20 meters on a side and a depth of 7.50 meters, with descending steps along the wall. The chamber, which is built of ashlars, has a trapezoidal floor 3.45 by 1.90 meters and a height of 1.60 meters. A cavity or niche is cut into the western wall. Even though the tomb had been robbed in the 4th century b.c. (evidenced by a Greek lamp found in the chamber), the remains of an inhuma46. See F. Molina Fajardo, Almuñécar en la antigüedad: La necrópolis fenicio-púnica del Puente de Noy (Granada, 1982).

Phoenician and Punic Sexi

63

tion and different ceramic grave goods were salvaged. The fragmented pieces included 2 amphorae, 3 red slip plates, fragments of a mushroom-mouth jug, and other materials, all of them dating to the end of the 7th or beginning of the 6th century b.c. 47 Since 1975, Molina has excavated a hundred tombs at Puente de Noy. The dates range between the second half of the 5th century and the 1st century b.c., with the exception of Tomb 4 in the eastern sector (Sector C) described above. Molina has established the following ten types: • Type I: Simple circular pit, in the shape of an inverted, truncated cone; diameter of 1 meter and depth of 0.90 m; 1% of tombs; uncertain date. • Type II: Simple rectangular pit; 2.60 by 0.60 and depth of 1 m; inhumation; 30%; date: 3d–1st centuries b.c. • Type III: Rectangular pit with ledges along the long sides (in order to fit a flat stone that covered the tomb); 3 m by 1.20 m; inhumation; 5%; 2d century b.c. • Type IV: Rectangular pit with ledges along the long sides; 3 m by 2 m and depth of 0.80/0.30 m; inhumation; 42%; 5th–2d centuries b.c. • Type V: Rectangular pit with ledges all around; 3.20 by 2 m and depth of .80 m; inhumation; 7%; 3d century b.c. • Type Va: Rectangular pit with ledges all around and low walls of rubblework; inhumation; 1%; 2d century b.c. • Type Vb: Rectangular pit of rubblework with ledges all around; inhumation; 1%; 2d century b.c. • Type VI: Trapezoidal pit with ledges along the two long sides; 3.30 m by 2 m and a depth of 1.60 m; inhumation; 2%; 4th–3d centuries b.c. • Type VII: Ashlar chamber with dromos and niche; 3.20 m by 2.20 m with a depth of 4 m; inhumation, 1%; around 600 b.c. • Type VIII: Cinerary urn in a small pit; 3%; 1st century b.c. • Type IX: Cinerary urn with grave goods on the rock’s surface; 7%; 3d–1st centuries b.c. • Type X: Simple cremation in a small hole; 1%; 1st century b.c. Thus, the necropolis at Puente de Noy is completely different from the one at “Laurita” with regard to typology of tombs, burial practices, and dates. At Puente de Noy we see the predominance of pits (in their different versions), the practice of inhumation, and dates ranging from the end of the 5th century to the 1st century b.c. We should point out that beginning in the 3d century b.c. the practice of cremation was reintroduced, and it coexisted (at a low percentage of 10%) with inhumation, which continued to predominate. When we 47. See F. Molina Fajardo et al., Almuñécar, 57ff.

64

Manuel Pellicer Catalán

examine the number of burials chronologically, we find the following percentages: 5% in the 5th century, 17% in the 4th century, 34% in the 3d century, 34% in the 2d century, and 10% in the 1st century b.c. In other words, the greatest density of burials at Puente de Noy corresponds to the 4th and 3d centuries. In Sexi, as in other Punic sites in the western Mediterranean, the two practices of cremation and inhumation existed side by side. In Carthage, where more necropolises have been studied, Cintas believes that the less frequent practice of cremation is an indigenous tradition, 48 a theory that we cannot accept because this practice was absent before the Phoenician colonization of the 8th century b.c. On the other hand, Cintas also states that at “Laurita” we find cremations because this site was only a stop for sailors, a “relay” station, 49 and that in Gadir we see inhumations because it was a stable Phoenician colony dating from early times. These observations must be questioned. The best explanation appears to be that the burial practices depended on conditions that we do not yet understand, such as traditions of the colonists, who formed a diverse and heterogeneous group, sex, social status, and various other influences and styles. According to Vassel, 50 the cremations at Carthage may belong to sailors who traded frequently with Greece. At Motya, as at Carthage, cremations have been found in low percentages. At Junon the cremations have been found in simple pit tombs about 3 m deep and 0.80 m in diameter, similar to the pozzi of Sicily where the ashes were placed along with simple grave goods. 51 Later on the ashes were deposited in a kind of basket, to be transferred to ceramic urns at a later date. However, in spite of the resemblance between the type of pits in the tombs at Junon and “Laurita,” inhumation predominates over cremation at Junon beginning already in the 8th century. In Rachgoun, an Algerian necropolis of the 7th–6th centuries b.c., cremation predominates (with 33 examples in urns and 68 examples deposited in holes without cinerary urns). These cremations exist along with inhumations that are almost exclusively of infants. In Utica and Dermech (at Carthage), according to the archaeological record, Phoenician burials seem to begin with inhumations in the 7th century b.c.. From all these examples we can deduce that burial practices do not correspond to specific dates, although it is true that beginning in the 6th century b.c. cremation loses ground and gives way almost totally to inhumation. When we analyze the orientation of the niches in the tombs at “Laurita,” we observe an evident arbitrariness, although it is true that no tomb faces west.

48. 49. 50. 51.

See P. Cintas, Manuel d’Archéologie punique II (Paris, 1976) 292. Ibid. 290. Ibid. 122. Ibid. 292.

Phoenician and Punic Sexi

65

Material Culture Alabaster Vases Perhaps the most spectacular finds at the necropolis of “Laurita” are the alabaster vases, which were utilized as cinerary urns in all the burials. In my 1962 publication, I presented the Punic inscription and the Egyptian hieroglyphic inscriptions and attempted to ascribe the cartouches to the corresponding Pharaohs of the 22d Egyptian dynasty of the 9th century: Osorkon II (870–847), Takelot II (847–823), and Sheshonq II (847). However, specialists like Padró, 52 Gamer, 53 Díaz, 54 Ferron, 55 and Leclant have been able to evaluate and correct errors in connection with the dates for Osorkon II (874– 850) and Takelot II (850–825). What is most extraordinary is the identification by Padró of the cartouche of Sheshonq III (825–773), which I mistakenly supposed belonged to Sheshonq II (847), as mentioned above. Now with the dates for Sheshonq III, which partially fall within the 8th century, we can more closely approximate the true date of the initiation of burials at “Laurita”—at the end of the 8th century b.c. Gamer and Padró have translated in a similar way the Egyptian text on the alabaster vase of Takelot II found in Tomb 1: “I have arrived from my faraway country, I have passed through lands, and I have heard tell of your being, of the primordial state of the two lands, which has created what now exists. Your two eyes shine within you. Your word is the breath of life that allows throats to breathe. Now I am on the horizon inundated by the happiness of the oases of Bahria and Karga, with a companion. Within me there is a fountain of health and life, and the serpent Mehen has sat down on its edge.” Padró considers the vases from Tombs 3, 12, 13, 14, and 15 to be Egyptian, from the 9th and 8th centuries b.c. According to Padró, the inscription on the vase from Tomb 15 is copied from an unconnected Hieratic text, in which Osorkon is probably a high priest of Amun in Thebes. This Osorkon was a son of Takelot II and died around 785. The inscription has been translated in this way: “Your heart will become intoxicated in order to unceasingly do what gives her pleasure. Intoxicate yourself for eternity. Put yourself in an agreeable frame of mind. Complete happiness is what she wants continuously. Tie and braid garlands. Put them on your head. Anoint it with oils. Do according to your desire. Protect the Ka of Osorkon in Bubastis.” The vase from Tomb 16, with an illegible inscription and a cartouche of Sheshonq III says: “Sheshonq Meryamun Si-Bastet, Governor and God of 52. 53. 54. 55.

See J. Padró, “Precisiones sobre la identificación.” See I. Gamer, “La inscripción.” See F. Díaz, Sefarad 25 (1965) 283–87. See J. Ferron, TrPrHist 27 (1970) 177–90.

66

Manuel Pellicer Catalán

Heliopolis.” The vase from Tomb 17, a hebenet with a cartouche of Osorkon II, is transcribed as “Usimar Setepenamun/Osorkon Meryamun Si-Bastet.” The vases from Tomb 19 are Egyptian and come from the 8th and 7th centuries, according to Padró. The vase in Tomb 20, with a cartouche of Osorkon II, reads: “Usimar Setepenamun/Osorkon Meryamun Si-Bastet.” Gamer 56 has contributed more information on the alabaster vases from “Laurita.” He affirms that Takelot II was buried in the tomb of his father Osorkon II in Tanis, which, in spite of being robbed, still held two alabaster vases at the time it was excavated, with cartouches of Osorkon I, founder of the 22d Libyan dynasty. Although it is evident that some funerary vessels at “Laurita” came from the sacking of Egyptian tombs, especially of the royal necropolis at Tanis, some people believe that other vases were luxury containers for the exportation of wine. I do not share this belief, especially after having extracted the urns from the deep tombs of “Laurita,” a process complicated by the weight and fragility of the empty containers. If they possibly contained wine or other soothing substances, they would have been so utilized only as funerary offerings in the royal Egyptian burials. There is no way that they could have been used as containers for these products during trade with the West. The different typologies of the vases (fig. 5) indicate an Egyptian origin, but also on certain occasions a Canaanite origin, such as the vase in fig. 13.1 of my publication 57 (fig. 5.E.3) and the torpedo-shaped vases of Tombs 3, 12, and 15B (fig. 5.A). The vase in Tomb 15B is analogous to the ceramic cinerary urns at the necropolises of Akhziv, Khaldeh, Rechidiye, and Carthage. The vase with a wide mouth and large handles (fig. 5.F.3) is a type coming from Amarna. The vase from Tomb 19A (fig. 5. D.1) is analogous to one from Assur that we will discuss below. The vase from Tomb 20 (fig. 5.B.3) is a type that is very common during the 25th Egyptian dynasty (750–656), according to W. Culican. According to Culican, 58 a torpedo-shaped vase from Assur is similar to the vases found in Tombs 3 and 12 at “Laurita” (fig. 5.A). This vase from Assur has the following Egyptian hieroglyphic inscription: “I have come and returned from the oasis of Bahria with abundant good milk. Oh Hentkau! You should bring something to the wharf of the priest of Harsaphes, King of Lower and Upper Egypt, Prince of the house of the power of Kheper-re, royal son of Ramesses, Takelotis” (translation by Von Bissing). Clearly we can observe a certain similarity with the inscription of Takelot on the vase from Tomb I of “Laurita,” cited earlier. Also Culican 59 has provided us with another impor56. See I. Gamer, “La inscripción,” 401–8. 57. See M. Pellicer, Excavaciones en la necrópolis púnica “Laurita” del Cerro de San Cristóbal (Almuñécar, Granada) (EAE 17; Madrid, 1962). 58. See W. Culican, Levant 2 (1970) 28–36. 59. Ibid. 30.

spread is 3 points long

Phoenician and Punic Sexi

67

tant piece of information about another inscription from Assur, written in cuneiform, whose text reads: “Palace of Esarhaddon, the great king and powerful king . . . of Sumer and of Akkad. Vessel of oil, large and full, that came with all kinds of rich booty from the treasure of the ruler Abdimilkuti, King of Sidon, which my strong arm captured.” This inscription indicates that at least some of the vessels which were plundered from the royal tombs at Tanis came to Phoenicia and to Sidon, from where they probably went to Assyria as booty of Esarhaddon (680–669), and also to the West, in trade or as valuable personal belongings of the first Phoenician colonists in Sexi, who perhaps came from Tyre. Recently Molina 60 has been able to locate four more vases (fig. 5: arrows) which appear to come from the necropolis “Laurita.” These vases had disappeared at the time of the discovery of the necropolis by workmen, were saved in private homes, and have now passed to the Museo Monográfico in Almuñécar. These vases correspond to forms already known in materials I have published. 61 We cannot explain why the only necropolises where these alabaster cinerary urns appear are at Carthage and in southern Spain (“Laurita”). These urns are found throughout the Mediterranean, but with other little-understood uses (Crete, Minet-el-Beida, Puerto de Santa María, Cerro del Mar, Cerro del Peñón, Toscanos, Cerro del Villar de Guadalhorce, Retiro de Churriana, etc.). The gray marble vase in fig. 5.F.4 was a sensational and surprising find, discovered already during the past century in Almuñécar under unclear circumstances. It has a cartouche and Egyptian hieroglyphic inscription (figs. 6 and 7) of the Hyksos king Auserre Apophis I of the 15th dynasty, who ruled for some 40 years from the end of the 17th to the beginning of the 16th century. 62 The inscription has been translated by Padró: “The good God, lord of the Two Lands, whose power reaches the borders of the victorious strongholds, since Tuab is not in his service, the King of Upper and Lower Egypt, Auserre, Son of Re, Apophis, endowed with life and the royal sister Charudiyet, living.” In addition, the cartouche says, “The good God Auserre/ The son of Re, Apophis/The royal sister Charudiyet, living.” 63 Tuab refers to the kingdom of Kush in Sudan, and Apophis is the fourth monarch of the 15th dynasty, the first of the Hyksos kings to use an Egyptian name, according to F. Daumas, and a probable successor of Seuserenre Khyan, according to Padró. The vase belongs to the first part of Apophis’ reign, before Upper Egypt became independent, since the inscription refers to the victorious strongholds of Nubia. In spite of the lack of documentation, it 60. 61. 62. 63.

See F. Molina Fajardo et al., Almuñécar. Ibid. 36–37, 46–52. Ibid. 38. See J. Padró, in F. Molina Fajardo et al., Almuñécar, 38–40.

68

Manuel Pellicer Catalán

Fig. 5. Typology of the alabaster urns from “Laurita.”

Phoenician and Punic Sexi

Fig. 6. Marble vase of Apophis I (after F. Molina).

69

70

Manuel Pellicer Catalán

Fig. 7. Egyptian hieroglyphic inscription from the vase of Apophis I (after Padró).

is very possible that this vase came to Sexi along with the alabaster vases found at “Laurita.” Finally, in addition to the hieroglyphic inscriptions, the alabaster torpedoshaped urn in Tomb 3 of “Laurita” (fig. 5.A.1) has a Semitic inscription painted in black, which was translated by F. Díaz: “Burned remains of Magon, son of H [ ].” 64 According to Díaz, paleographically the inscription probably dates from the end of the 8th century. For Ferron, who has analyzed the name Magon, the inscription is most likely Carthaginian because of the frequency of this name in Carthage and because of the similarity of this inscription to another one in a burial at Sainte-Monique in Carthage. 65 In contrast to Ferron’s theory, Leclant and Picard believe that the name could easily be eastern Phoenician rather than Carthaginian.

Red Slip Pottery Red slip pottery, typically Phoenician and Cypriot from the Final Bronze in the second half of the 2d millennium to the middle of the 1st millennium b.c., 64. See F. Diaz, “Dos nuevas inscripciones púnicas hispánicas,” Sefarad 25 (1965) 283–87. 65. See J. Ferron, “La inscripción cartaginesa pintada en la urna cineraria de Almuñécar,” TrPrHist 27 (1970) 177–90.

Phoenician and Punic Sexi

71

is the normal type found at “Laurita,” with types of pear-shaped and trefoilmouth oenochoes, mushroom-mouth jugs, plates, and two-cornered lamps (figs. 8 and 9). The pear-shaped oenochoe (fig. 8, bottom) with a trefoil mouth and derived from the metallic type is of Phoenician origin. At the end of the 9th century b.c., the type is totally formed in Khaldeh, and it spreads at the beginning of the 8th century throughout the western Mediterranean, predominating in the Phoenician colonies. In addition to Sexi, it appears in the Phoenician settlements at the mouths of the Algarrobo, Vélez, and Guadalhorce Rivers, as well as at the Guadalete in Torre de Doña Blanca. This form penetrates somewhat into the interior by means of the Guadalquivir River, according to the finds at El Carambolo, Alcores, and Carmona. However, it is not common at indigenous orientalizing sites. Four of these oenochoes have been found at “Laurita.” The mushroom-mouth jug (fig. 8, top), incorrectly called oenochoe, is also of Syro-Palestinian origin but has precedents in certain Egyptian alabastra and is adapted to the Cypriot Phoenician world. It follows the colonizing trajectory of the Phoenicians throughout the central and western Mediterranean, passing through the Strait of Gibraltar to Mogador on the Atlantic. Four examples have been found at “Laurita” and a fragment in Tomb 1E at Puente de Noy. Its Iberian typology has been studied by Almagro Gorbea 66 and by Negueruela, 67 and they distinguish three types. The example from Tomb 20 at “Laurita” corresponds to Type I with a symmetrical body and a grooved neck and is dated to the second half of the 8th century. The examples from Tombs 13 and 19B at “Laurita” correspond to Type II, with a pear-shaped or ovoid body, without a groove but with a broadening of the edge on the neck. They date to the first third of the 7th century. The example from Tomb 12 at “Laurita” corresponds to Type III, also pear shaped or ovoid and with the lower part of the neck broadened, from the second half of the 7th century. The red slip plates (fig. 9) have been studied by Schubart, 68 who has attempted to establish a chronology based on the width of the rim and the ratio of the width of the rim to the diameter, using the results of the stratigraphy of the Phoenician colonies in Málaga. According to Schubart’s thesis, the plates from “Laurita,” whose rim widths range from 41 to 73 mm, probably corresponded to Toscanos IV and V. In other words, the plate from Tomb 13, with a rim of 47 mm and a quotient of 5.4, would be the oldest of the finds in context, from around 700 b.c. or earlier. The plate in Tomb 12, with 56 mm and a quotient of 4.7, would correspond to the first quarter of the 7th century. The 66. See M. Almagro Gorbea, “Los dos jarros paleopúnicos del Museo Arqueológico Nacional hallados en la Casa de la Viña (Torre del Mar),” MM 13 (1972) 172–83. 67. See I. Negueruela, “Jarros de boca de seta y de boca trilobulada de cerámica de barniz rojo en la Península Ibérica,” in Homenaje al Prof. M. Almagro Basch, 2 (Madrid, 1983) 159–279. 68. See H. Schubart, “Westphönizische Teller,” RSF 4 (1976) 183, 185.

72

Manuel Pellicer Catalán

Fig. 8. Typology of red slip jugs from “Laurita” (after I. Negueruela).

one in Tomb 15B, with 67 mm and a quotient of 4.1, would be dated to the middle of the 7th century. The plates of Tomb 17, also with rims of 67 mm, but with a quotient of 3.8, the one from Tomb 16, with 69 mm and a quotient of 3.6, and the one in Tomb 2, with 71 mm and a quotient of 3.8, would be included in the second half of the 7th century. In Tomb 1E at Puente de Noy, the width of the rims of the numerous fragmented plates in the fill ranges between 50 and 77 mm, corresponding to dates within the second half of the 7th century and including the beginning of the 6th century. These dates have been corroborated by some samples with a groove on the lip, which according to the stratigraphies at Toscanos and Cerro del Villar de Guadalhorce and according to the tombs at Trayamar, occur around 600 b.c.

Phoenician and Punic Sexi

73

a

b

c

d

e

Fig. 9. Typology of the red slip plates from “Laurita” (after I. Negueruela).

The two-cornered lamps found at “Laurita” (Tombs 2 and 15B) are typical of the western colonies. They are scarce in the East and absent on Cyprus.

Proto-Corinthian kotylai The two proto-Corinthian kotylai that were part of the burial goods in Tomb 19B formed the basis for dating the “Laurita” necropolis. I dated the kotylai to the first half of the 7th century, a chronology that at first was applied to the whole necropolis. But later analyses and studies by Shefton and others confirmed that the kotyle that had a black slip on the entire lower part of the vase was initial proto-Corinthian, with a profile suggesting a date at the end of the 8th century. The second kotyle, decorated with a circle of triangles on its

74

Manuel Pellicer Catalán

lower part, like a schematization of a lotus flower coming up from the base, was dated somewhat later, within the first half of the 7th century b.c. Additionally, Shefton 69 suggests the possibility that the vessels from “Laurita” are proto-Corinthian imitations from Pitecusas or Cumas, a possibility that could also be applied to the fragments at Toscanos. These Greek materials, like the numerous ones that appear in Huelva at a later date, should not be considered in any way as proof of a Greek colonial presence, as some suggest, but rather of Phoenician trade with the Greek world and the Iberian Peninsula (with the West). A. M. Bisi attributes some kotylai from Motya and Carthage to the early proto-Corinthian period and the samples from Almuñécar, Toscanos, and others from Carthage, Motya, and Malta, to the mid proto-Corinthian period. 70 The nearest western parallels of the kotylai from “Laurita” are those found in Tomb 52 at Motya, 71 the subgeometric proto-Corinthian example from Tomb A-143 at Byrsa (Carthage), 72 the one from Tomb 28 at Cumas (from the beginning of the 7th century), those from the Regolini-Galassi tomb, and finally other samples from Rhodes and Pitecusas.

Amphorae The Phoenician and Punic amphorae from Almuñécar present an abundant and varied repertoire, known through the work of Molina 73 (fig. 10). The amphorae that Molina compiles come from the necropolis of Puente de Noy and from shipwrecks located off the coast near Almuñécar. In “Laurita” I found only one rim of a Phoenician “de saco” (bag-shaped) amphora, and it came from Tomb 19 and was dated to the 7th century. But perhaps Molina’s Type VI and its evolution into Types VII and VIII are more interesting because of their implications. 74 Molina’s Type VI comes from Tomb 22 in Zone C of the necropolis at Puente de Noy and has an ovoid body, convex sloping shoulders, semicircular handles, and a high rim beveled on the interior. This type, along with other similar examples from the shipwrecks of Punta del Vapor and Cueva del Jarro, 69. See B. B. Shefton, “Greeks and Greek imports in the South of the Iberian Peninsula. The archaeological evidence,” in Phönizier im Westen (Madrider Beiträge 8) (Mainz, 1982) 338 n. 2. 70. See A. M. Bisi, “Importazioni e imitazioni greco-geometriche nella più antica ceramica fenicia d’Occidente,” in Atti del I Congresso Internazionale di Studi Fenici e Punici, 3 (Rome, 1983) 89–95. 71. See V. Tusa, “La Sicilia fenicio-punica: stato attuale delle ricerche e degli studi e prospettive per il futuro,” in Phönizier im Westen (Madrider Beiträge 8; Mainz, 1982) table 38.3. 72. See S. Lancel, “Ivoires phénico-puniques de la nécropole archaïque de Byrsa, à Carthage,” in Atti del I Congresso Internazionale di Studi Fenici e Punici, 3 (Rome, 1983) 687–92. 73. See F. Molina Fajardo et al., Almuñécar. 74. Ibid. 139–41.

Phoenician and Punic Sexi

75

with parallels at Cádiz, Ibiza, Kouass, etc., 75 is dated between the end of the 6th century and the beginning of the 4th century b.c., according to Levels 18– 14 of Cerro Macareno. 76 Surprisingly, this same type of Punic amphora, containing fish (sea bream and tuna), has been found in abundance in the socalled “Punic Amphora Building” in the Corinthian agora, along with other amphorae from Corinth, Chios, and Mende. These amphorae have been dated with precision to the middle of the 5th century b.c. 77 The Punic amphorae found in Corinth are probably products of the trade with the Punic colonies of Andalusia. The presence of these amphorae in Level 3b at Toscanos, 78 a level which the excavators estimated to be around 700 b.c., creates a serious problem for dating, especially because Toscanos has served as a guide and reference point for establishing the chronology of Phoenician colonization and the orientalizing process on the Iberian Peninsula. It may be necessary to simply lower the date of the end of Phoenician-Punic Toscanos to the 5th century— and preferably to the second half of that century. 79

Iron Regarding the presence of iron in Spain, we are aware of the contributions of the so-called Hallstatt Culture in the northeastern Peninsula. However, these influences have not been documented before the middle of the 7th century b.c., a date which is contemporary with or later than the date attributed to iron implements found in the region of Carmona (Mazagoso, Acebuchal A, Cañada de Ruiz Sánchez, Cruz del Negro, Bencarrón) and Setefilla. The iron fragment that was found in Tomb 19 of “Laurita” confirms the presence of iron already in the first half of the 7th century and its Phoenician origin.

75. Ibid. 139–41. 76. See M. Pellicer, J. Escacena, and M. Bendala, El Cerro Macareno (EAE 124; Madrid, 1983) 86, fig. 1072. 77. See C. K. Williams, “Corinth, 1978: Forum Southwest,” Hesperia 42 (1979) 105–44. 78. See G. Maass-Lindemann, Toscanos 71 (Madrider Forschungen 6; Berlin, 1982) fig. 18.755–56. 79. The type of amphora to which we refer is very problematic in regard to its typology and chronology, since it has many variants dating from 600 b.c. up to the Roman era. This form needs a serious study that complements the one done by R. Pascual (“Un nuevo tipo de ánfora púnica,” AEArq 42 [1969] 12–19), a work that collects examples from Villaricos (fig. 2.2), Cartagena (fig. 2.3), Almuñécar (fig. 3.3), Kouass (fig. 4.4), Ceuta (fig. 4.1), Mogador (fig. 4.5), etc. Two examples dated to the middle of the 5th century b.c. come from Cerro del Villar de Guadalhorce (A. Arribas and O. Orteaga, “El yacimiento fenicio de la desembocadura del río Guadalhorce, Málaga,” Cuadernos de Prehistoria de la Universidad de Granada 2 [1975] pl. X.b,c). There are other examples at Cerro del Prado (M. Pellicer et al., “Para una localización,” 237, fig. 5.40, 42, 44, 52, 53), at Quebrantahuesos (M. Pellicer, NAHisp 15 [1983] 70, fig. 5.1; 80, fig. 11.1), at Castañuelo (M. del Amo, Huelva Arqueológica [1978] 337, pl. 7.2), at Tejada (A. Blanco et al., Exploración arqueológica de Huelva [Barcelona, 1981] fig. 276.19, 20), etc.

76

Manuel Pellicer Catalán

Fig. 10. Typology of the amphorae from Almuñécar (after F. Molina).

Chronology When we analyze the most important materials from the tombs at “Laurita,” we conclude that the burials in Tombs 20 and 13 probably began at the end of the 8th century, based on the dates of the oenochoes, mushroom-mouth jugs, and plates. The date for Tomb 19 would fall within the first half of the 7th century, according to the proto-Corinthian kotylai found inside and the

Phoenician and Punic Sexi

77

width of a plate rim in the fill. Tombs 2, 12, 15, 16, and 17 probably belong to the period immediately following, within the third quarter of the 7th century. Tomb 1E of Puente de Noy was used at the end of the 7th century. The necropolises that correspond to the 6th century and the first half of the 5th century have not yet been found in Almuñécar. However, we have found the tombs in Sectors A, B, C, and D at Puente de Noy that correspond to the middle of the 5th century and that were used until the 1st century b.c. Molina’s excavations have sufficiently proven that the settlement of Sexi was located on the promontory where the Castle stands today and generally in the most elevated area of the city of Almuñécar. We have not yet determined the precise date of the founding of the Phoenician colony, but perhaps it was founded at the beginning of the 8th century like Morro de Mezquitilla, apparently merging with the indigenous population of the Final Bronze.

78

Manuel Pellicer Catalán

Notes on the Economy of the Phoenician Settlements in Southern Spain María Eugenia Aubet Semmler

Introduction The purpose of this article is to reopen discussion on the economy of the Phoenician settlements in the southern Iberian Peninsula, beginning with a new reading of the archaeological, faunal, and geomorphologic data obtained years ago in the region. Using this information, I will attempt to sketch a paleoeconomic reconstruction that will serve as a working hypothesis and a frame of reference for discussing the significance of the oldest Phoenician installations known in Spain. Spain is well known as one of the principal destinations for Phoenician expansion in the West. The historiographic tradition of the classical world associates the expansion to Spain with the exploitation and trade of Spain’s important metals, especially silver. The Phoenicians received highly profitable returns from this trade (Diodorus 5.20.1), as verified by the archaeological record. 1 The evidence, in fact, demonstrates that the exploitation of silver mines in the Riotinto region reached almost industrial proportions during the 8th–7th centuries b.c. 2 Moreover, only an intensive exploitation of this nature could have justified the founding of Cádiz (the Phoenician Gadir) with its location at the mouth of the Guadalete River, near the entrance to the Guadalquivir River valley and the ancient Tartessos (fig. 1). Gadir served as the principal port for loading metals obtained from Tartessian mines in the interior, probably through use of indigenous labor. Thus, Gadir was able to Editor’s note: This article appeared in Spanish under the title “Notas sobre la economía de los asentamientos fenicios del sur de España,” in Dialoghi di Archeologia (Terza Serie, Anno 5, no. 2, 1987) 51–62. 1. A. Blanco, J. M. Luzón, and D. Ruiz, Excavaciones arqueológicas en el Cerro Salomón (Riotinto, Huelva) (Seville: University of Seville, 1974); D. Ruiz Mata, “El poblado metalúrgico de época tartésica de San Bartolomé (Almonte, Huelva),” MM 22 (1981) 150–70. 2. A. Blanco and R. Rothenberg, Exploración arqueometalúrgica de Huelva (Barcelona, 1981) 26–27, 114–15.

79

80

María Eugenia Aubet Semmler

channel and control a luxury trade that benefited the indigenous aristocracy as well as the Phoenician colonists, as archaeological finds suggest. 3 We must also point out that only the large-scale trade of precious metals could compensate for the costs of a naval enterprise on this scale and attract Tyre’s interest in territories so far away. However, the metal trade does not function as the only causal factor when we attempt to explain the location of other ancient Phoenician settlements on the Iberian Peninsula. For example, neither the settlements on the island of Ibiza nor those on the coast of Málaga, Granada, and Almería are located in territories rich in metals. Only at Villaricos, the ancient Baria in Almería, can we speak of silver mining. The other southeastern settlements are normally called “trading posts,” “merchant ports,” or even “colonies,” while the preCarthaginian settlement on Ibiza (Ebusus) can be explained by its excellent strategic position on the east-west navigational routes of the era (fig. 1). 4 The Phoenician settlements on the Mediterranean coast of Andalusia have provided the earliest known dates for the arrival of the Phoenicians in the West. The settlements at Toscanos, Chorreras, and Almuñécar appear to have been founded during the second half of the 8th century b.c., while Morro de Mezquitilla may go back to the first half of that century. These settlements on the coast of eastern Andalusia belong to one of the areas most intensely excavated during the past years. They constitute an unusual archaeological grouping, not only because of their antiquity, but also because of their uniform cultural characteristics. In this article we deal specifically with this region and with some of the ancient settlements situated around three cities mentioned by the classical sources: Malaka, Sexi, and Abdera (Strabo 3.4.2–3). These sites occupy a territory where a marine handbook of the 6th century b.c. located an authentic Phoenician “multitude” (Avienus Ora Maritima 440, 459–460).

Pattern of Settlement The 170-kilometer stretch of shoreline between the ancient Abdera (presentday Adra in Almería) and the ancient Malaka (the modern city of Málaga) contains a considerable density of Phoenician settlements that were established gradually during the 8th century b.c., judging from the archaeological record. The general topography of these ancient settlements appears to correspond to a well-defined pattern of settlement with the following characteristics: 3. M. E. Aubet, “La aristocracia tartésica durante el periodo Orientalizante,” in Opus 3 (Rome, 1984) 445–56. 4. M. E. Aubet, “Los fenicios en España: estado de la cuestión y perspectivas,” Aula Orientalis 3 (1985) 9–30.

The Phoenician Settlements in Southern Spain

The Phoenician Settlements in Southern Spain

81

Fig. 1. Phoenician centers in the western Mediterranean.

(1) Settlement on low-rising coastal promontories located on small peninsulas or on islands at the mouths of rivers. This pattern for locating settlements is essentially Western, since we see it also at the Phoenician colonies of Gadir and Lixus on the Atlantic, but not on Sicily or Sardinia. 5 (2) In most cases, the corresponding necropolis is located on the opposite bank of the river and a few kilometers from the settlement (fig. 1). For example, the necropolis of the Phoenician Toscanos is found on Cerro del Mar, on the other side of the Vélez River (fig. 2); the necropolis of Morro de Mezquitilla is located at Trayamar, on the other side of the Algarrobo River; and the necropolis of Almuñécar is found on Cerro de San Cristóbal, on the other side of the Seco River. (3) The surface areas of these settlements were surprisingly small, especially if we compare them with the areas of the large Phoenician colonies of North Africa and Sardinia. For example, the approximate area of Morro de Mezquitilla was 2 hectares, and the area of the island of Villar at the mouth of the Guadalhorce River (fig. 3) was 1 hectare or even less. Thus, with few exceptions, we are dealing with small settlements, which in some cases were dominated by a large central warehouse used for merchandise (as in Toscanos). The characteristics of the known necropolises support 5. H. Schubart, “Phönizische Neiderlassungen an der Iberischen Südküste,” in Phönizier im Westen (ed. H. G. Niemeyer; Mainz am Rhein, 1982) 207–31; H. G. Niemeyer, “Die Phönizier und die Mittelmerwelt im Zeitalter Homers,” Jahrbuch des Römisch-Germanischen Zentralmuseum 31 (1984) 3–90.

82

María Eugenia Aubet Semmler

Fig. 2. Location of Phoenician settlements in the Vélez River Valley.

the hypothesis of a considerable density of population, organized by means of a network of numerous small settlements. Only approximately twenty tombs from the 8th–7th centuries b.c. have been found on the necropolis of Laurita, associated with Phoenician Almuñécar (the ancient Sexi). This contrasts with the size of the Punic necropolises in the same region (6th–1st centuries b.c.), which, like the necropolis of Villaricos, contained hundreds of burials (fig. 5). 6 6. M. Pellicer, “Excavaciones en la necrópolis púnica ‘Laurita’ del Cerro de San Cristóbal (Almuñécar, Granada),” Excavaciones Arqueológicas en España 17 (Madrid, 1962); H. G. Niemeyer and H. Schubart, “Trayamar,” Madrider Beiträge 4 (Mainz, 1975); M. E. Aubet, “Excavaciones en Las Chorreras (Mezquitilla, Málaga),” Pyrenae 10 (Barcelona, 1974) 79–108; M. E. Aubet, G. Maass, and H. Schubart, “Chorreras,” Noticiario Arqueológico Hispánico 6 (1979) 91– 134; M. E. Aubet, “La necrópolis de Villaricos en el ámbito del mundo púnico peninsular,” in Homenaje a Luis Siret (Seville, 1986) 612–23.

The Phoenician Settlements in Southern Spain

Fig. 3. Phoenician settlements in the area of Málaga.

83

84

María Eugenia Aubet Semmler

Fig. 4. The valley of the Guadalhorce River and its ecology (N. Carulla).

(4) At times, the average distance between Phoenician settlements is only 5 to 10 kilometers, as the crow flies. For example, the distance between Cerro del Villar and Malaka is 4 kilometers (fig. 4); Toscanos is 7 km from Morro de Mezquitilla; Chorreras is located at a distance of only 800 meters from Morro (fig. 2). (5) Since all of these settlements are located on bays or inlets that are well protected from wind and waves, they provide optimal conditions as natural harbors and anchoring spots. Even today when strong westerly winds blow, sailboats heading toward Cádiz and the Atlantic find shelter in the harbors at Motril, Almuñécar, Vélez-Málaga, and Fuengirola. Under such weather conditions, it is dangerous to sail through the Strait of Gibraltar, and boats may need to wait weeks or even months for the wind to change direction.

Trade and Communication Routes It is generally believed that the Phoenician colonies on the Iberian Peninsula were founded for two reasons: to obtain metals and to control maritime trade. Thus it was necessary to control points of access to principal resources and to maritime routes. According to this idea, the location of the so-called “colonies” at the mouths of rivers indicated an interest in controlling the natural routes that followed these rivers toward the interior. To a certain extent, this idea has connotations of control and territorial sovereignty. However, recent studies of the geomorphology of the coastal territory of the province of Málaga have demonstrated that important mineral deposits did

The Phoenician Settlements in Southern Spain

85

Fig. 5. Alabaster urn from the necropolis of Cerro del Villar (Guadalhorce).

not exist in this region. 7 Moreover, there are no archaeological indications of a systematic process for controlling the territory or hinterland, such as have been seen on Sardinia, for example. Neither can we speak of a hinterland that was acculturated or assimilated politically (chora) as on Sicily and Sardinia or at Carthage, 8 since the indigenous population was very small and practically nonexistent during the 8th and 7th centuries. When we consider communication routes to the interior, we find that these are limited to a few natural pathways that coincide with river flood plains. 7. M. E. Aubet and N. Carulla, “El asentamiento fenicio del Cerro del Villar (Málaga): Arqueología y paleogeografía del Guadalhorce y de su hinterland,” Anuario Arqueológico de Andalucía 2 (1986) 425–30. 8. S. Moscati, “I fenici e il mondo mediterraneo al tempo di Omero,” Rivista di Studi Fenici 13 (1985) 182–83.

86

María Eugenia Aubet Semmler

Those of the Guadalhorce, Vélez, and Motril Rivers stand out because of their importance. In most places, the Mediterranean coast of Andalusia is separated from the interior by the Penibetic Mountains, which in some areas act as a true natural barrier between the sea and upper Andalusia. This fact highlights the value of the natural river routes that permitted the Phoenicians to establish contact with the indigenous hinterland. Without doubt, the Guadalhorce River forms the principal natural route of communication between the interior and the entire coastal area of Málaga and Granada. We know that in adverse weather conditions, when it was not advisable to sail toward Gadir, a round trip land route was used between Malaka and Tartessos, taking a total of 9 days (Avienus, Ora Maritima 178–182). This route must have followed the valley of the Guadalhorce River, the only land route that directly connects the Mediterranean coast of Málaga with the countryside of Córdoba and Seville, by way of Antequera. However, we do not believe that these factors alone determined the settlement of the Phoenician population throughout this area, since many of the ancient Phoenician installations appear to be practically cut off from the interior of the country. In fact, neither mining nor the strategic factor nor territorial control can justify by themselves the founding of so many trading settlements. It has been suggested, among other things, that timber was exploited on a large scale for ship building and even that iron in the region was intensively mined. 9 Neither hypothesis appears to be viable when we try to explain a colonization carried out at a distance of 4000 kilometers from Tyre in Lebanon, which also was an area with rich forest resources. The small settlements that we are examining have the basic characteristics of commercial ports, where all the institutional, architectural, and religious elements of the large urban colonies of the central Mediterranean or of Gadir itself are absent. The archaeological record shows that various mercantile and industrial installations appeared in the principal centers of the region (Toscanos, Morro de Mezquitilla) beginning in 700 b.c. Their presence speaks in favor of intense trading and port activity: storehouses for merchandise, shops for working copper and iron on a local scale, etc. 10 All of the evidence shows that these enclaves quickly reached a certain degree of self-sufficiency in commercial and industrial materials. Along with their role as ports for necessary stops on the 9. G. Schüle, “Tartessos y el hinterland,” in Tartessos y sus problemas (ed. J. Maluquer; Barcelona, 1969) 15–32. 10. H. Schubart, “El asentamiento fenicio del siglo VIII a.C. en el Morro de Mezquitilla (Algarrobo, Málaga),” Aula Orientalis 3 (1985) 59–79; H. G. Niemeyer, “El yacimiento fenicio de Toscanos: urbanística y función,” Aula Orientalis 3 (1985) 109–24. [Both of these articles are found in English translation in this volume.]

spread is 1 pica short

The Phoenician Settlements in Southern Spain

87

route toward the Atlantic, these settlements complemented their trading and shipping activity with secondary activities such as metal and pottery production and fishing-related industry. In fact, the discovery of murex shells and fish bones at Toscanos and Almuñécar suggests the possibility of salting and dyeing industries in this region from very ancient times. 11 The strictly commercial, industrial, and harbor functions of the Phoenician enclaves along the Mediterranean coast of Andalusia and their great number also correspond to the ancient model of Phoenician settlement that Thucydides described for the island of Sicily. These settlements came before the Phoenicians focused on the three large centers of Motya, Panormo, and Solunto at the end of the 8th century b.c. In other words, the process corresponds perfectly to the oldest pattern of Phoenician settlement known in the West: a large number of enclaves situated on small islands or coastal promontories, with a commercial purpose (Thucydides 6.2.6). In Sicily, this type of port settlement precedes the second phase of Phoenician expansion on the island. During this second phase we see signs of civic and religious institutions for the first time (tophet, sanctuaries, walls) and the progressive control of the territory on the western end of the island. 12 Thus we are looking at an extremely ancient pattern of Phoenician settlement in the West, almost “pre-colonial,” which, for unknown reasons, did not turn into a colony or urban center like those we observe in other Mediterranean regions. True urban life arrived late in this region, and therefore we do not see it before the appearance of the large Punic centers in the 6th century b.c.: Malaka, Baria, Ebusus. However, the industrial and port activities, the storing of merchandise, and the role as harbors of support and refuge for ships bound for the Atlantic still do not explain why so large a Phoenician population was gathered in such a dense network of scattered and decentralized settlements on the Mediterranean coast. This is a population that must be defined as a specialized, and, in a certain way, a privileged class, judging by the content of the tombs at Trayamar and Almuñécar. 13 If it were only for the activities mentioned above, a reduced personnel in charge of maintenance for the mercantile and port installations would have been enough. Thus it is necessary to search for other 11. M. Uerpmann, “Archäologische Auswertung der Meeresmolluskenreste aus der westphönizischen Faktorei Toscanos, “MM 13 (1972) 164–71; J. Lepiksaar, “Fischknochenfunde aus der phönizischen Faktorei von Toscanos” in Tierknochen von westphönizischen und phönizisch beeinflussten Ansiedlungen in Südspanischen Küstengebiet (ed. J. Boessneck), in Studien über frühe Tierknochenfunde von der Iberischen Halbinsel 4 (Munich, 1973) 109–19; F. Molina and C. Huertas, Almuñécar en la Antigüedad, II (Granada, 1985) 26. 12. S. Moscati, “I fenici,” 179–87. 13. M. Pellicer, “Excavaciones en la necrópolis”; H. G. Niemeyer and H. Schubart, “Trayamar.”

88

María Eugenia Aubet Semmler

Fig. 6a. Aerial view of the Vélez River Valley (Photo “Paisajes Españoles”).

economic motives offered by the coastal environment that were profitable enough to secure a considerable contingent of a stable and permanent population in this area for more than 200 years.

The Countryside: Elements for a Reconstruction of the Ancient Environment An approximate reconstruction of the environment is feasible only through extensive analyses of pollen, fauna, soil, carbon, etc. Such analyses have not been carried out yet in the area in question, although some isolated studies have permitted us to sketch a hypothetical picture of the “Phoenician countryside” on the Andalusian Mediterranean coast and thereby determine to a certain extent the potential resources of the environment. From Adra to the Guadalhorce, the coastal strip where the Phoenician settlements were located is relatively isolated from the interior by a chain of mountains called the Penibetic range. This range runs parallel to the coast and is only on average some 20 kilometers from the coast. This range delimits a

The Phoenician Settlements in Southern Spain

89

Fig. 6b. Aerial view of the Algarrobo River Valley (Photo “Paisajes Españoles”).

series of small, narrow river valleys which are extremely fertile and were usually controlled by a Phoenician enclave on the estuary (figs. 6a and 6b). Even today these flood plains sustain enormous human concentrations. This is possible through intensive farming that benefits from the typical temperate/hot Mediterranean climate, characterized by its short mild winters. Moreover, the mountains prevent the penetration of cold northerly winds, and the winter rains are influenced by the Atlantic. Today all these factors favor horticulture, the cultivation of cereal grains, and the pasturing of goats and sheep. These practices have considerably degraded the forest and the countryside, since they have taken place since ancient times. Additionally, river and alluvial sedimentation has caused the coast to advance several kilometers in some places. This has completely transformed the shoreline since 700 b.c., when the sea penetrated considerably further into the interior. Several studies performed on the fauna discovered at some of the Phoenician sites permit us to reconstruct in broad terms the type of vegetation that predominated in the region during Phoenician times. Thus, for example, in the indigenous town of Cerro de la Tortuga, located 3 kilometers northwest of the urban center of Málaga, the presence of deer, wild boar, and wildcat suggests that during the 4th century b.c. a deciduous forest still survived in the mountains that surround the valley of the Guadalhorce (fig. 4). Additionally, the hunting of wild boar indicates a much higher degree of humidity than at

90

María Eugenia Aubet Semmler

present, while the existence of horses and donkeys in Tortuga points to the marked development of open spaces of semi-steppelike vegetation next to the forests. 14 In the Vélez River valley the Phoenicians of Toscanos also hunted wild boar and deer, and to a lesser extent, hares and rabbits. This implies the existence of a mixed deciduous forest near the settlement, since all of these species live in the forest. Finally, the relatively high percentage of deer in Tortuga and Toscanos indicates that during the 8th–4th centuries b.c. the degradation of the forest had not reached significant proportions.

Farming: The Example of the Vélez River Valley The study of the domestic fauna at Toscanos performed some time ago highlights the presence in that region of horses, donkeys, oxen, sheep, and goats in significant proportions. 15 Dogs, pigs, and chickens occur in much lower numbers. Cattle predominate, and their percentage varies with the period. Thus, at the end of the 8th century b.c. (Stratum 1), cattle represent 49% of all domesticated animals, compared to 42% sheep and goats and 8.5% pigs. Around 700 b.c. (Stratum III), cattle undergo a considerable increase (62%), while sheep and goats decrease (27%) and pigs remain very scarce (11%). Finally, throughout the 7th century b.c. (Strata IV–V), the percentage of cattle continues to increase (63%–80%), to the detriment of sheep and goats (18– 31%) and pigs (1.5%–6%). The percentages of fauna observed in the Algarrobo River valley are very similar. From all of this we can infer that there was pasture for sheep and goats in the vicinity of Toscanos, which implies the use of wool, meat, and milk obtained from these extremely profitable animals, since they consume little and can be a good complement for intensive agriculture. The most significant information, however, is provided to us by the high percentage of cattle, whose numbers increase beginning in the 7th century b.c. This indicates that, in addition to being consumed (given the profitability of products derived from them), the cattle probably were used for hauling and fertilizer, which indirectly suggests a progressive increase in farming and plowing with oxen. In fact, only an intensification of cultivation can explain this important increase of cattle in Toscanos. An argument in favor of this hypothesis is the low percentage of pigs in the Vélez River valley during the Phoenician period. Some have tried to explain this scarcity using ethno-ideological factors: Semitic peoples do not consume 14. H. P. and M. Uerpmann, “Tierknochenfunde aus der phönizische Faktorei von Toscanos und anderen phönizischen beeinflussten Fundorten der Provinz Málaga in Südspanien,” in Tierknochen von westphönizischen (ed. J. Boesneck) 35–69. 15. H. P. and M. Uerpmann, “ Tierknochenfunde,” 83–93.

The Phoenician Settlements in Southern Spain

91

pork. 16 However, the agricultural hypothesis might be the most reasonable. The pig is actually an animal that competes directly with humans and is normally incompatible with intensive agriculture, which necessitates moving it to oak groves and other non-agricultural areas. As an omnivorous animal, the pig eats an excessive amount, especially grain. In the long run, raising pigs can exhaust cultivated lands and soil and can constitute a serious threat for the integrity of cultivated fields. When there are no pigs, the protein deficit caused by not consuming the meat is usually compensated for with meat from hunting, as we have seen at Toscanos. In any case, a low percentage of pigs speaks in favor of a deliberate and wise ecological strategy and, in fact, points to an intensive agriculture. 17 Even today the true richness of the Málaga coast is rooted in the agriculture practiced on its riverside plains, which are particularly rich and fertile and suitable for both irrigated and dry farming. The type of cultivation practiced in the valley of the Vélez River today is similar to that practiced by its inhabitants in Phoenician times. It is important, however, to stress the increase in agriculture beginning in 700 b.c., seen in the evolution of the local fauna. Also, the few indigenous towns and dwellings located in an economic radius of just a few kilometers around Toscanos (Cerca Niebla, Vélez-Málaga—fig. 2) indicate the presence of a sparse, scattered indigenous population living on the edges of the Vélez alluvial plain, suggesting the possibility of indigenous labor in raising and pasturing cattle at the middle elevations. This type of territorial organization characterizes irrigated farming, in which the population scatters into isolated groups living near the cultivated fields. 18

The Guadalhorce River Valley: A Case of Irrigated Farming It is significant that up until now the Phoenician settlement at Cerro del Villar on the estuary of the Guadalhorce River is the only one that has provided numerous stone grinders, which testify to the grinding of wheat both at 16. W. Schüle, “Die tierische Reste der altpunischen Niederlassung von Toscanos,” in Toscanos 1964 (eds. H. G. Niemeyer and H. Schubart; Madrider Forschungen 6; Berlin, 1969) p. 124; H. P. and M. Uerpmann, “Tierknochenfunde,” p. 84. 17. D. R. Harris, “Agricultural systems, ecosystems and the origins of agriculture,” in The Domestication and Exploitation of Plants and Animals (eds. P. J. Ucko and G. W. Dimbleby; London, 1969) 3–15; H. Harris, Cows, Pigs, Wars and Witches: The Riddles of Culture (Random House, 1974; Spanish edition, Madrid, 1980); M. R. Jarman, G. N. Bailey, N. H. Jarman (eds.), Early European Agriculture (Cambridge University Press, 1982); J. Estévez, “Reflexiones sobre algunos aspectos de los complejos faunísticos de Mallorca y la Península Ibérica,” in Early Settlement in the Western Mediterranean Islands and the Peripheral Areas (eds. W. H. Waldren and R. Chapman; BAR 229, Oxford, 1984) 193–204. 18. J. L. García, Antropología del territorio (Madrid, 1976) 104.

92

María Eugenia Aubet Semmler

Fig. 7. Stone grinder from Cerro del Villar (Guadalhorce, Málaga).

the domestic level and on a large scale at this site (fig. 7). In addition, geomorphologic and paleogeographic studies of the river valley show extraordinary conditions for intensive agriculture along an area of 18 square kilometers of land fit for irrigation (fig. 4). 19 Presumably the crop was wheat. The Málaga-Almería coast is favorable for the cultivation of quality wheat, as the Neolithic discoveries at Nerja or those at the Punic Villaricos demonstrate. 20 Perhaps the toponymy of the region evokes its fields of wheat. It has been suggested that the name of the principal river of the province of Málaga, 19. M. E. Aubet and N. Carulla, “El asentamiento fenicio del Cerro del Villar (Málaga).” 20. M. Pellicer, Estratigrafía prehistórica de la cueva de Nerja (Madrid, 1962) 41.

The Phoenician Settlements in Southern Spain

93

Fig. 8. Bread oven at Chorreras, Málaga (8th century b.c.).

Guadi-al-orce, must have originally meant “river of wheat.” 21 Whatever the case, the abundance of grain that still grew on its banks in the 17th century was surprising. The presence of only one hand grinder in the 8th-century Phoenician enclave of Chorreras probably indicates that flour was usually imported and baked into bread in local ovens (fig. 8). 22 In this case the isolated grinder indicates a dependence on imported grain produced in more suitable areas, such as in the valley of the Algarrobo River located nearby (fig. 2). At this level Chorreras must be defined as a secondary center in relation to Morro de Mezquitilla. In any case, the example of Chorreras indicates that, beginning in the 8th century b.c., wheat was grown near the Phoenician settlements on the Mediterranean coast. Without the corresponding analyses, it is difficult to determine the amount of wheat which was cultivated in the valley of the Guadalhorce: whether it 21. E. Terés, “Materiales para el estudio de la toponimia hispanoárabe.” Nómina fluvial I (Madrid, 1986) 367–69. 22. M. E. Aubet, “Excavaciones en las Chorreras,” 94, tables II–III; J. M. Gran, “Excavaciones arqueológicas en la región de Vélez-Málaga,” NAHisp 12 (1981) 352, fig. 17.

94

María Eugenia Aubet Semmler

was produced strictly for local consumption or whether surpluses were produced for export. In our judgment, the enormous agriculture possibilities of the valley and the small size of the Phoenician settlement at Cerro del Villar probably indicate that a surplus of grain was produced that far exceeded the needs of the sparse Phoenician population. In the hinterland of the Guadalhorce there are no signs of mineral deposits, and therefore only the strategic location, excellent port facilities, and intensive agriculture can explain the settlement of the Phoenicians in this area. Actually Cerro del Villar was located on an island in the center of the large harbor at Málaga, one of the best anchoring spots on the entire Andalusian Mediterranean coast. Even today sailboats heading toward the Strait of Gibraltar find protection here when strong westerly winds blow. In ancient times ships sailed only in spring and summer and especially during the months of June to September. A Phoenician ship coming from Tyre or from another Mediterranean port could find itself forced to anchor in the harbor at Málaga or another suitable place for a month or longer. If the westerly winds persisted, it was not advisable to pass through the Strait on the way to Gadir. These winds can blow forcefully for weeks and even months in the region. Given that the trip between Tyre and Gadir could take three months of sailing, 23 in adverse conditions one month of delay in reaching Gadir would not only endanger a commercial transaction, but even worse, could force the whole crew of the ship or merchant fleet to delay the return trip for a year and wait on land for the sailing season. During the circumnavigation of Africa in approximately 600 b.c., it is said that the Phoenicians put in to the coast when fall arrived and cultivated wheat while waiting for a favorable season for sailing (Herodotus 4.42). The valley of the Guadalhorce provided all the necessary conditions for abundant annual crops without the need for excessive preparation of the soil. Moreover, the availability of sufficient water year long permits two crops of grain, one in the fall and the other in the spring. Thus the preparation of the soil, sowing, and harvest fall within the seasons not suitable for sailing. An irrigated region such as the Guadalhorce River valley implies high yields per cultivated hectare. On the other hand, huge investments of work and manual labor are demanded during times when the wheat must be weeded, harvested, and threshed. Moreover, irrigation requires taking water from the river or from wells, spreading it through the fields by animal or human means, and constructing irrigation ditches and terraces. In short, we must assume an increase in the rural population specifically during the fall and spring. This population could be recruited from the indigenous peoples of the interior, but 23. J. Alvar, “Los medios de navegación de los colonizadores griegos,” Archivo Español de Arqueología 52 (1979) 79–83.

The Phoenician Settlements in Southern Spain

95

also from the crews of ships anchored in the harbor. Therefore, the choice of a place like Cerro del Villar implies a desire to produce an agricultural surplus. This is the only reason, along with the strategic factor and the role as a port, that could explain a settlement in such a fertile flood plain. Intensive farming of this type does not require a huge number of laborers, except during the specific seasons mentioned above. We know that there were indigenous peoples in the region, although in small numbers, as in the Vélez River valley. The pattern of settlement at the mouth of the Guadlahorce River, with small rural communities around the Phoenician center (perhaps at San Julián, Churriana, the area of the airport, etc.; figs. 3 and 4), fits the model of a scattered population that we usually associate with irrigated farming. We do not know how the assimilation of the indigenous population was produced, whether by coercion (usurping the indigenous lands) or by sharing the land with them. An in-depth study of this territorial pattern (a central nucleus with secondary specialized settlements ) would permit us to fix the political boundaries of the Phoenician territory. If we accept the agricultural hypothesis, we must assume that the need to intensify cultivation, together with the expansion of irrigation systems through the efforts of these eastern colonists who were familiar with an advanced technology, must have carried some negative consequences for the environment in the long run. The progressive disappearance of pasture land, the regression of the forest, and the gradual degradation of the environment by converting large extensions of land into meadows and plains are suggested by the fauna at the settlement of Tortuga, located only 6 kilometers away from the valley. This is the destruction of the environment that generally follows a colonial period. 24 However, it is premature to speak about agricultural communities like those of the Greek colonies in Magna Graecia. We have already seen that the role of trading center and port was of considerable importance in the development of the Phoenician enclaves of southeastern Spain. But we must also consider the destination of wheat produced in the hinterland of Villar, and perhaps also Toscanos, and we cannot help but compare it to other Phoenician settlements dedicated to the large-scale production of wheat. We refer here to colonies on the plain of the Kishon River near Haifa, 25 whose probable purpose was to make up for the great deficit of grain suffered by Tyre from the 10th century on (1 Kings 5:25). 26 24. E. Boserup, The Conditions of Agricultural Growth (London, 1965; Spanish edition, Madrid, 1967) 32, 67–69, 88–90. 25. J. Briend and J. B. Humbert, Tell Keisan (Fribourg-Paris, 1980) 361–76. 26. [[1 Kings 5:11 in most English translations]].

96

María Eugenia Aubet Semmler

Phoenician Trade in the West: Balance and Perspectives María Eugenia Aubet Semmler

Thirty years have passed since the initiation of Phoenician archaeology on the Iberian Peninsula, during which time the theoretical and methodological debate about the political and economic significance of the colonial settlements has constituted an important part of the research. Interest in achieving a greater methodological precision in the analysis of the Phoenician question has permitted considerable progress in the resolution of these problems, and at the same time it has raised new questions as successive archaeological discoveries continue to paint a panorama that is increasingly complex and heterogeneous. Currently, several different questions continue to interest researchers, such as the place that the trade in metals occupies in Phoenician expansion to the West and the causes and significance of the great concentration of small Phoenician colonies on the Mediterranean shore of Andalusia. The numerous important findings relating to Phoenician commerce that have recently occurred on the western end of the Mediterranean suggest that the time has come to reexamine the information that has accumulated and reflect on the present state of research and the questions that are being raised for the future.

1. The Trade in Metals The written tradition of the Classical Age emphasizes without exception the fundamental importance of metals in the first Phoenician expeditions to the West. Therefore, from the beginning the Phoenician question has been approached from a strictly commercial and metallurgic perspective, thus differentiating it from the fundamentally demographic and agricultural nature of Greek colonization. The literary sources and the archaeological record reflect Editor’s note: This article appeared in Spanish under the title “El comercio fenicio en Occidente: balance y perspectivas,” in I Fenici: Ieri Oggi Domani (Rome: Consiglio Nazionale delle Ricerche, 1995) 227–43.

97

98

María Eugenia Aubet Semmler

certain features that were characteristic of the Phoenician expansion and that were demonstrated above all by the location and geographical distribution of colonial settlements. The sites were favorable for anchoring boats and strategically important for navigational routes and for controlling natural passageways of communication from the mouths of the principal rivers of Andalusia into the interior. The city of Gadir (Cádiz), the principal Phoenician colony in the West, was clearly oriented toward the Atlantic and thus toward the trade in metals. The archaeological findings of the 80s in the region of the city of Huelva and in the mining centers of Tejada la Vieja, San Bartolomé de Almonte, and Peñalosa (Ruiz Mata and Fernández Jurado 1986; Fernández Jurado, García Sanz, and Rufete 1992) reveal the existence of extraordinary activity in the mining and processing of silver in this zone beginning in the middle of the 8th century b.c. The introduction by the Phoenicians of technological innovations into indigenous metallurgy would have increased the yield of local mining and metal working, indirectly favoring the growth of Gadir’s economy. The archaeological record of the region of Huelva shows that the commercial strategy of the Phoenicians probably gave less priority to the nuclei of production in the hinterland than to the centers that received the riches from the interior, such as the Tartessian city of Huelva. Without doubt, the obtaining of metals and, in particular, precious metals such as gold and silver, constituted a prime objective for Phoenician commerce. Only in this way could the Phoenicians have compensated for the high costs of shipping incurred by a colonial enterprise so far distant. But the trade in metals was not the only objective of the Phoenicians’ commercial expansion. The early presence of eastern merchants and artisans in basically agricultural and cattle-raising regions such as the lower Guadalquivir or the hinterland of the colonies in Málaga and Granada, where few relevant mineral deposits have been documented, raises serious questions about the significance of Phoenician colonization outside of the mining and metal-working regions of Huelva and its periphery. The first call for attention came from the new excavations on the coast of Málaga when it was verified that these port settlements had at their disposal a hinterland rich in agriculture and cattle, as well as excellent routes of communication, but contained little of metallurgic interest. Moreover, until a few years ago it was generally believed that this region was practically uninhabited when the Phoenicians arrived, and thus there were apparently no serious reasons to sustain the hypothesis of commercial objectives for the founding of these colonial enclaves. Several more or less sensational explanations have been proposed, some of them incomplete or contradictory, and almost all extracted from the Greek colonial model: for example, the possible existence of a chora or a territory of political control around the trading posts (an explanation that was immedi-

spread is 1 pica long

Phoenician Trade in the West: Balance and Perspectives

99

ately discarded; see Niemeyer 1984), forms of agricultural colonization and cultivation in the areas surrounding the colonies (González Wagner and Alvar 1989: 90–92), or centers dedicated exclusively to artisanry. The existence of trade in agricultural products in the hinterland of the colonies could be perceived, and our work in the valley of the Guadalhorce gave the first clues in this direction (Aubet 1987). However, the debate soon became distorted when the very idea of commerce as a motivating factor in westward Phoenician expansion was questioned (López Castro 1993: 60; González Wagner 1993: 106). Basing their opinions on an erroneous reading of the archaeological record and relying again on the Greek colonial model, some authors tried to explain Phoenician expansion not in terms of a basically commercial movement, but rather as a phenomenon of territorial occupation with hopes of exploiting the agricultural regions of the interior, which could have originated in a need for land and demographic pressure in the land of origin (Whittaker 1974; González Wagner and Alvar 1989). According to these authors, the Phoenicians stopped being merchants for the first time in their history in order to become makeshift farmers. This explanation is as simplistic as the one that earlier sought to define the Phoenicians as simple seekers of metals. It is evident that the problem is essentially a methodological one and that an explicit theory is needed to permit us to formulate new models, beginning with a more rigorous reading of the archaeological record. We must go beyond the stage of stratigraphic sequences, chronology, and formal description of pottery. We need to incorporate a much wider context, examining, among other things, the role of the indigenous elite as co-protagonists and co-beneficiaries of the Phoenician colonial enterprise and trade in the West.

2. The First Stages of Colonial Trade The scientific community of the 70s concentrated basically on achieving a rigorous and exhaustive knowledge of the Phoenician colonial settlements and, in particular, of their stratigraphic sequence. Some authors have qualified this stage as merely descriptive, with the predominance, above all in Mediterranean Andalusia, of preoccupation with stratigraphic sequences, chronology, and taxonomic classifications. The result was a detailed and exhaustive archaeological record of the coast of Málaga and Granada that still constitutes one of the best-known regions for the diachronic study of Phoenician colonization in the western Mediterranean. During those years the principal research objective was to analyze the Phoenician colonies themselves in search of a correct reading of the stratigraphic and cultural data, which for two decades favored the proliferation of debates centered only on the character and category of the colonial settlements (trading post, port of commerce, colony), independent of their regional context.

100

María Eugenia Aubet Semmler

All considered, we believe that this stage has been judged too severely by recurring at times to easy and superficial criticism and accusing many archaeologists of adopting a purely positivist methodology, of not possessing a general and global theory of culture, and of being interested only in strictly archaeological questions and ignoring “historical” research (see González Wagner 1992: 84). However, these stages, with all that the excessive analysis of artifacts and stratigrafies implies, were necessary in order to arrive at the point where we now find ourselves. The proof is that the relative chronology established for certain ceramic types (plates, amphorae, etc.) at Toscanos, Chorreras, Almuñécar, and Morro de Mezquitilla still serves today as a necessary reference when dating ancient contexts in the Tartessian area, in Torre de Doña Blanca, Carthage, Sulcis, and in Tyre itself. Without these stratigraphies and typological sequences we could not have advanced so far in the analysis of the diachronic and synchronic relationship of the Phoenician colonial settlements, nor would we have been able to capture the richness and complexity of the data relative to Phoenican colonization in southern Spain. Thus, for example, the detailed analysis of artifacts that so irritates some historians allows us to establish with some degree of certainty that the founding of the colonies in the West did not follow a uniform spatial-temporal pattern. According to the archaeological record, the first colonial settlements were Morro de Mezquitilla and Gadir/Doña Blanca in the middle of the 8th century (see Maass-Lindemann 1990a), followed, in the second half of that same century, by a larger grouping of foundings, such as Chorreras, Toscanos, Almuñécar, Cerro del Villar (Aubet 1991), Abdera (Suárez, Aguayo, Carrilero, López Castro, and San Martín 1989), and probably also Villaricos. Within this second grouping, the relative chronology of the archaeological record suggests that these foundings also occurred at different points in time between 750 and 700 b.c. Consequently, several successive and clearly differentiated stages in Phoenician colonization can be observed, whose correct ordering in time will depend in the future on a rigorous and strictly archaeological investigation, like it or not. We still need to resolve the reasons why in some cases this process was interrupted, such as at Chorreras or at Guadiaro, where Phoenician settlements were abandoned around 700 b.c. It is even appropriate to speak of a third stage in the process of Phoenician colonization in the West, characterized by the establishment of “secondary” enclaves founded by older primary colonies, beginning in the middle of the 7th century. This is inferred from the archaeological data obtained at the Bay of Algeciras, where the Phoenician site of Cerro del Prado does not seem to emerge before the second half of the 7th century (Ulreich, Negrete, Puch, and Perdigones 1990). Also, in the valley of the Guadalhorce, the site of Loma del Aeropuerto could have been settled from Cerro del Villar (Aubet 1992), and all the indications are that in the valley of the Almanzora, the trading post of

Phoenician Trade in the West: Balance and Perspectives

101

Cabecico de Parra, located in the interior and near the silver and iron mines of Herrerías, was founded by the nearby Villaricos in the second half of the 7th century (López Castro, San Martín, and Escoriza 1987–88). In general, the latest discoveries at the Bay of Mazarrón, at Cartagena, and at the mouth of the Segura (see González Prats 1993) do not permit a continued discussion at the present time about a concentration of ancient colonies on the coast of Málaga and Granada only, but rather these newest discoveries support the existence of an authentic “Phoenician coast” between Gibraltar and Alicante (fig.1). The present state of research suggests that, between 750 and 650 b.c., Phoenician commercial and colonial strategy consisted of a progressive occupation of the coastal territories of the southern Peninsula. As this commercial enterprise turned into a lucrative activity, the old ports of call consolidated, some of them turning into authentic port cities and repositories for merchandise, and, little by little, the shore between Gadir and Villaricos became an authentically Phoenician coast. Thus, to the extent that the first trips and incursions throughout the West gave way to regular trade, the colonial territories came to constitute an extension of Phoenician domains in the East. In any case, we can say that Tyrian control over western trade was exercised by means of successive expeditions throughout more than three generations of merchants and navigators. It should not be forgotten that the Phoenician presence in the West lasted for some 200 years (750–550 b.c.). Throughout that period, changes must have taken place in the geopolitical realities of the region, as well as in the location of the markets, the demand, and the networks of supply. Thus, the circumstances at the middle of the 8th century were not the same as at the end of that century or at the beginning of the 7th. For example, in the second half of the 8th century, the international trade of the Phoenician colonies in Andalusia seemed to be oriented preferentially toward Sardinia, Pitecusas, and Carthage, judging by the presence of Phoenician and Euboic vessels of Tyrrhenian origin in southern Spain (see Ruiz Mata 1993: 52) and by the export of plates and Phoenician amphorae of Andalusian origin to Pitecusas and Carthage (see Vegas 1989: 224–25). In the 7th century, the international contacts probably were much more vast and oriented preferentially toward the Greek world and the Atlantic, while at the beginning of the 6th century international relations were once again oriented toward the area of Tyre and the East Greek world.

3. The Role of Indigenous Communities: Mechanisms of Trade and Interaction The indigenous element constituted a decisive factor in the commercial strategy of the Phoenician world. As we have noted elsewhere (Aubet 1990;

102

María Eugenia Aubet Semmler

Fig. 1. Map of eastern Andalusia and of southeastern Spain: Phoenician colonies (white circles) and indigenous settlements (black circles).

Aubet and Delgado 1998), the success and duration of Phoenician colonization in the West depended on favorable economic conditions in conjunction with the character and political structure of the indigenous communities involved. In order to arrive at this conclusion, we start from the hypothesis that a commercial enterprise such as that of the Phoenicians could have established prolonged trade relationships only with indigenous societies that were capable of guaranteeing the flow of surplus goods and of procuring native labor in the ports, mines, and fields—and all of this in a state of stability, peace, and continuity of interchanges. In addition, economic history teaches us that colonial trade is always established in territories that have a consumer hinterland and where indigenous communities exist with experience in regional trade. The political authority in these communities is capable of acting as a resource distribution center within the framework of hierarchical trading networks and is capable of controlling access to the resources of peripheral dependent territories (see Meilassoux 1971; Frankenstein and Rowlands 1978; Wolf 1982; Curtin 1984; Rowlands 1987). All of these circumstances come together in a good part of Andalusia during the 8th–7th centuries b.c., and only in this way can we understand the role played by indigenous centers such as El Carambolo, Carmona, Huelva, and Setefilla, which had already controlled interregional trade since at least the 9th century b.c. (Aubet 1990: 36–37).

Phoenician Trade in the West: Balance and Perspectives

103

The appreciation that has developed during the past few years for the indigenous horizon of the precolonial Final Bronze outside of the Tartessian area has demonstrated once again the importance of analyzing systems of interaction when reconstructing the Phoenician commercial and colonial dynamic. Contrary to what was believed earlier, the presence in eastern Andalusia of an important indigenous substratum has been verified, a substratum that appears to be a key to explaining the presence of the commercial trading posts on the coasts of Málaga, Granada, and Almería (fig. 1). Although insufficiently known up to this point, the panorama has become quite revealing, in that the existence of coastal indigenous settlements in the Final Bronze has been demonstrated in Almuñécar itself next to the mouth of the Verde River (Molina Fajardo 1985), in Salobreña next to the mouth of the Guadalfeo (Arteaga 1989), and, finally, in Montilla on the Guadiaro River (Schubart 1988). At all of these sites a Phoenician colony was established at the end of the 8th century near a preexisting indigenous center situated on the coast. In the interior and on the northern boundary of the province of Málaga, recent excavations at the indigenous towns of Ronda and Acinipo confirm the presence of powerful indigenous centers placed in strategic spots and dominating the principal interregional trade routes before the arrival of the Phoenicians (Aguayo, Carrilero, and Martínez 1987; 1991). The early interest that colonial trade showed in these centers of the interior already in the 8th century b.c. seems to be based on the enormous possibilities to be derived from the political and territorial preeminence of these settlements within trade networks and the consumption of metals and agricultural products extending from the cultivated fields of the Guadalquivir to northern Andalusia. It was only necessary to create incentives in order to reorient the indigenous economy and increase production and consumption among the local elites. That this was achieved is demonstrated by the fact that the chiefs in the interior succeeded in imposing and intensifying the cultivation of vineyards and olive groves in a vast surrounding territory formed by dependent agricultural villages in order to produce a surplus oriented toward colonial trade (Carrilero 1992: 135–36). This not only favored Phoenician commerce, but also the indigenous leaders, who in this way saw their political authority strengthened. As we will see, this was the commercial strategy adopted by Gadir in the 7th century b.c.

4. The Great Commercial Expansion of Gadir in the 7th Century The recent archaeological discoveries in Portugal and in eastern and northeastern Spain show that the direction of research initiated a few years ago is most likely correct. In these areas a broad commercial expansion, probably

104

María Eugenia Aubet Semmler

directed from Gadir beginning around 650 b.c., appears to copy the same strategy inaugurated by the eastern Phoenicians a hundred years earlier. Again the principal items of Phoenician trade seem to be wine, oil, grain, crude and manufactured metals, salt, skins, and probably also slaves. The 7th century constitutes without doubt the golden age of Phoenician trade in the West. Phoenician goods now begin to inundate the entire valley of the Guadalquivir and the valley of the Guadalhorce, and they reach territories ever more distant from the coast. The same approximate dates mark the beginning of the decline of the great Atlantic trading circuit, which had operated during the entire Final Bronze Age and whose routes the Phoenicians would now exploit. During the Final Bronze Age (1100–700 b.c.), the coastal territories of central southern Portugal were occupied by powerful indigenous communities that had long experience in maritime and overland trade. From their fortified centers they dominated the river traffic of the estuaries of the principal Atlantic rivers (fig. 2). Thus, in the estuary of the Tagus, we can confirm during this period a considerable density of population formed by small agricultural units grouped around large population centers, such as Catujal, Tapada da Ajuda, Quinta do Almaraz, Lisbon itself, and Santarém. These centers controlled the interior delta of the river and land traffic heading toward the territories of the interior (Cardoso 1990; Arruda 1993; Barroso, Cardoso, and Sabrosa 1993). Other similar concentrations of population are observed on the estuary of the Guadiana controlled by the town of Castro Marim, at the mouth of the Sado controlled by the towns of Alcácer do Sal and Setúbal, and in the valley of the Mondego controlled by Conímbriga and Santa Olaia (Correia 1993). All these centers functioned as authentic commercial ports, channeling copper from Alemtejo, tin from Beira Alta, and gold from Portuguese Estremadura toward the coast, as well as channeling Phoenician products toward the interior (toward Medellín, Aliseda, and the area of Mérida in Spanish Estremadura). The estuary of the Tagus and the surrounding area probably attracted Gadir’s commercial interests quite quickly, due to the exploitation of salt marshes in this region and probably of gold in the alluvial plain. In addition, the Tagus provides an important natural route for accessing the copper and tin zones of the interior. In effect, in the second half of the 7th century a small commercial Phoenician settlement was founded at Abul, near the mouth of the Sado River (Mayet and Tavares da Silva 1993). The Phoenician pottery discovered there points to a Gadir origin for its population and merchants, probable evidence, therefore, that the Phoenicians knew how to incorporate preexistent trading networks into their own trade routes. These indigenous networks already had their own infrastructure and were controlled from the indigenous centers of Alcácer do Sal and Setúbal.

spread is 1 pica short

Phoenician Trade in the West: Balance and Perspectives

Fig. 2. Gadir and the Atlantic region of the Iberian Peninsula.

105

106

María Eugenia Aubet Semmler

On the Atlantic Coast of Morocco a similar process probably took place. In spite of the legendary date of its founding, the Phoenician colony of Lixus, like Mogador, has not shown indications up until this moment of anything before the middle of the 7th century b.c. As in Portugal, the pottery of Lixus and Mogador has much in common with the area of Gadir (see Maass-Lindemann 1990b). Probably Gadir capitalized on all the rich potential of the Atlantic, without doubt receiving enormous profits. Perhaps we can attribute to this Atlantic-Gadir trade route the cargo of a 7th-century boat found in the waters of the Mar Menor in Murcia (the wreck of Bajo de la Campana—fig. 1), which was transporting ingots of tin and lead, along with thirteen African elephant tusks engraved with small Phoenician inscriptions (see Mas 1985; Sanmartín 1986: 89–91; López Pardo 1990: 289–92; Martín Camino and Roldán Bernal 1991; Roldán Bernal, Miñano, and Martín Camino 1993). The origin of the cargo can be deduced from the raw materials that the boat contains, surely acquired in Portuguese territory (tin and lead) and in the region of Lixus/Mogador (ivory). In eastern Spain Phoenician commerce seems to have been oriented also toward those indigenous communities which, like those at Saladares, Vinarragell, and Peña Negra, controlled the regional trading networks (figs. 1 and 3). The most illustrative example is the settlement at Peña Negra in Alicante, where the presence of an important center of the Final Bronze (900/850–700 b.c.) has been documented and which in the 8th century produced and distributed Atlantic-type, manufactured metal goods (González Prats 1990: 317–57; 1993: 146–47). It was this role of redistribution center that probably attracted Phoenician trade from the south. In the 7th century this trade led to the installation of a district of artisans and merchants next to the site (González Prats 1983) and ultimately to a trading post in nearby Guardamar, on the mouth of the Segura River (González Prats 1991). The boom in Phoenician trade in the far western Mediterranean culminated in the middle of the 7th century with the founding of small trading colonies on the island of Ibiza (fig. 3). Phoenician control of the island probably began with the founding of Sa Caleta, followed by the progressive occupation of the nearby Bay of Ibiza at the beginning of the 6th century (Ramón 1991; 1992: 472–78). Due to this colonization, the Phoenicians were able to count on a bridgehead to position themselves in the markets of northeastern Spain and the Golfe du Lion where, by the middle of the 7th century, they had initiated trade relations with the local communities of Campos de Urnas. All of our information indicates that agricultural resources and metals constituted the principal products of trade in that region. Throughout the valley of the Ebro, Phoenician trade was able to access a broad spectrum of products (cereals, cattle, skins, wine, and crude and recycled metals) that could be exchanged easily in other areas of the Mediterranean basin, by means of agreements and

Phoenician Trade in the West: Balance and Perspectives

107

Fig. 3. Ibiza and Mediterranean eastern Spain, with the principal indigenous sites for Phoenician commerce.

108

María Eugenia Aubet Semmler

certain concessions to the indigenous aristocracy in the large towns of southern Aragon, such as Coll del Moro (Molas, Rafel and Puig 1987; Rafel 1991). The identification of small river warehouses of Phoenician merchandise such as at Aldovesta and other evidence suggest that a large part of Phoenician commercial activity in these territories was oriented to the systematic recovery of scrap bronze destined for the metallurgic centers or foundries of the Mediterranean (Mascort, Sanmartí, and Santacana 1991). In the Golfe du Lion and the region of the Pyrenees, the presence since the 8th century of towns in Campos de Urnas located on the principal commercial axes linking the Mediterranean region (Angles, Mailhac) with the Atlantic metal trade by means of the Aude, Herault, and Garonne Rivers explains in and of itself the arrival of Phoenician merchants in the middle of the 7th century. Once more, Phoenician commerce limited itself to intervening in and stimulating some of the already existing trade networks, along which tin and bronze objects circulated, including scrap metal (Pons and Pautreau 1993). The discovery near Agde of the shipwreck of Rochelongue, with its cargo of ingots, weapons, and bronze objects from the end of the 7th century (Maluquer 1966), is significant for our study. This underwater discovery reveals the existence of organized forms of maritime commerce in the area and suggests a commercial operation not of a purely haphazard character, but rather with a well-defined purpose: the systematic recovery of bronze weapons and objects and their recasting for later use. Probably these stocks of metals that were temporarily excluded from circulation favored, over the long run, the abandonment of bronze by the indigenous communities and the gradual, though not disinterested, imposition of iron, whose introduction is usually associated with Phoenician trade. How these interregional networks were integrated into Phoenician commerce and how systems of economic dependence in these indigenous communities functioned are subjects for further in-depth research relating to forms of trade in the Phoenician world.

Bibliography Aguayo, P.; Carrilero, M.; and Martínez, G. 1987 Excavaciones en el yacimiento pre y protohistórico de Acinipo (Ronda, Málaga). Pp. 333–37 in Anuario Arqueológico de Andalucía, 1986. 1991 La presencia fenicia y el proceso de aculturación de las comunidades del Bronce final de la Depresión de Ronda (Málaga). Pp. 559–71 in Atti del II Congresso Internazionale di Studi Fenici e Punici (1987). Rome. Arruda, M. 1993 A ocupaçao da Idade do Ferro da Alcáçova de Santarém no contexto da expansao fenícia para a fachada atlântica peninsular. Estudos Orientais 4: 193– 214.

Phoenician Trade in the West: Balance and Perspectives

109

Arteaga, O. 1989 La transformación del medio ambiente costero de Salobreña. Causas naturales e históricas. Pp. 55–83 in V Centenario de la incorporación de Salobreña a la corona de Castilla (1489–1989). Salobreña. Aubet Semmler, M. E. 1987 Notas sobre la economía de los asentamientos fenicios del sur de España. Dialoghi di Archeologia (Terza Serie, Anno 5, no. 2): 51–62. [[Reprinted in English translation here, pp. 79–95.]] 1990 El impacto fenicio en Tartessos: las esferas de interacción. Cuadernos Emeritenses 2: 29–44. [[Reprinted in English translation here, pp. 225–40.]] 1991 Die phönizische Niederlassung vom Cerro del Villar (Guadalhorce, Málaga). Madrider Mitteilungen 32: 29–51. 1992 Nuevos datos arqueológicos sobre las colonias fenicias de la bahía de Málaga. Pp. 71–78 in Lixus. Colloque Institut des Sciences de l’Archéologie et du Patrimoine de Rabat (Larache 1989). École Française de Rome. Aubet Semmler, M. E., and Delgado, A. forthcoming Fenicios e indígenas en Occidente. Cuadernos de Prehistoria de la Universidad de Granada 18. Barroso, L.; Cardoso, J. L.; and Sabrosa, A. 1993 Fenícios na margem sul do Tejo. Estudos Orientais 4: 143–81. Cardoso, J. L. 1990 A presença oriental no povoamento da Idade do Ferro na regiao ribeirinha do Estuário do Tejo. Estudos Orientais 1: 119–34. Carrilero, M. 1992 El proceso de transformación de las sociedades indígenas de la periferia tartésica. Pp. 117–37 in La colonización fenicia en el sur de la Península Ibérica: 100 años de investigación. Almería. Correia, V. H. 1993 Os materiais pré-romanos de Conímbriga. Estudos Orientais 4: 229–83. Curtin, P. D. 1984 Cross-cultural Trade in World History. Cambridge University Press. Fernández Jurado, J.; García Sanz, C.; and Rufete, P. 1992 Prospección con sondeo en Peñalosa (Escacena, Huelva). Pp. 185–90 in Anuario de Arqueología Andaluza, 1990. Frankenstein, S., and Rowlands, M. 1978 The internal structure and regional context of Early Iron Age society in south-western Germany. Bulletin of the Institute of Archaeology 15: 73–112. González Prats, A. 1983 Estudio arqueológico del poblamiento antiguo de la Sierra de Crevillente. Alicante. 1990 Nueva luz sobre la protohistoria del Sudeste. Alicante: University of Alicante. 1991 La presencia fenicia en el Levante peninsular y su influencia en las comunidades indígenas. Pp. 109–15 in 1–1V Jornadas de Arqueología Feniciopúnica. Ibiza.

110

1993

María Eugenia Aubet Semmler

Ausgrabungen in der frühgeschichtlichen Siedlung von Herna (La Peña Negra/Crevillente, prov. Alicante). Madrider Mitteilungen 34: 142–52. González Wagner, C. 1992 Tartessos en la historiografía: una revisión crítica. Pp. 81–115 in La colonización fenicia en el sur de la Península Ibérica: 100 años de investigación. Almería. 1993 Las estructuras del mundo tartésico. Pp. 103–16 in Los enigmas de Tarteso, eds. J. Alvar and J. M. Blázquez. Madrid. González Wagner, C., and Alvar, C. 1989 Fenicios en occidente: la colonización agrícola. Rivista di Studi Fenici 17: 61–102. López Castro, J. L. 1993 Difusionismo y cambio cultural en la protohistoria española: Tarteso como paradigma. Pp. 39–68 in Los enigmas de Tarteso, eds. J. Alvar and J. M. Blázquez. Madrid. López Castro, J. L.; San Martín, C., and Escoriza, T. 1987–88 La colonización fenicia en el estuario del Almanzora. El asentamiento de Cabecico de Parra de Almizaraque (Cuevas de Almanzora, Almería). Cuadernos de Prehistoria, Universidad de Granada 12–13: 157–69. López Castro, J. L.; Carrilero, M.; Suárez, A.; Aguayo, P; and San Martín, C. 1991 La colonización fenicia en Abdera: nuevas aportaciones. Pp. 981–89 in Atti del II Congresso Internazionale di Studi Fenici e Punici . Rome. López Pardo, F. 1990 Mogador, factoría extrema y la cuestión del comercio fenicio en la costa atlántica africana. Pp. 277–96 in 5e Congrès International d’Histoire et d’Archéologie de l’Afrique du Nord. Avignon. Maass-Lindemann, G. 1990a Orientalische Importe vom Morro de Mezquitilla. Madrider Mitteilungen 31: 169–77. 1990b Die phönikische Keramik von Lixus im Vergleich mit südandalusischer Keramik. Madrider Mitteilungen 31: 186–93. Maluquer, J. 1966 Nuevos datos para el estudio del comercio prerromano en el Mediterráneo occidental. Pyrenae 2: 185–90. Martín Camino, M., and Roldán Bernal, B. 1991 Aportación al conocimiento de la presencia fenicia en el Sudeste peninsular. Pp. 355–61 in XX Congreso Nacional de Arqueología (1989). Zaragoza. Mas, J. 1985 El polígono submarino de Cabo Palos. Sus aportaciones al estudio del tráfico marítimo antiguo. Pp. 153–71 in VI Congreso Internacional de Arqueología Submarina (Cartagena 1982). Madrid. Mascort, M. T.; Sanmartín, J; and Santacana, J. 1991 El jaciment protohistòric d’Aldovesta (Benifallet) i el comerç fenici arcaic a la Catalunya meridional. Tarragona. Mayet, F., and Tavares da Silva, C.. 1993 Presença fenícia no baixo Sado. Estudos Orientais 4: 127–42.

Phoenician Trade in the West: Balance and Perspectives

111

Meillassoux, C. (ed.). 1971 The Development of Indigenous Trade and Market in West Africa. Oxford University Press. Molas, D.; Rafel, N.; and Puig, F. 1987 Orientalisierende Funde von der Grabung 1984 in der Nekropole des Coll del Moro, Gandesa (Tarragona). Madrider Mitteilungen 28: 51–57. Molina Fajardo, F. 1985 Almuñécar a la luz de los nuevos hallazgos fenicios. Aula Orientalis 3: 193– 216. Niemeyer, H. G. 1984 Die Phönizier und die Mittelmeerwelt im Zeitalter Homers. Jahrbuch des Röm.-Germanischen Zentralmuseums 31: 3–94. Pons, E., and Pautreau, J. P. 1994 La nécropole d’Anglès, La Selva. Aquitania 10: 180–210. Rafel, N. 1991 La necrópolis del Coll del Moro de Gandesa. Els materials. Tarragona. Ramón, J. 1991 El yacimiento fenicio de Sa Caleta. Pp. 177–87 in 1–1V Jornadas de Arqueología fenicio-púnica. Ibiza. 1992 La colonización arcaica de Ibiza. Mecánica y proceso. Pp. 453–78 in X jornades d’Estudis Locals: La Prehistòria de les illes de la Mediterrània occidental. Palma de Mallorca. Rowlands, M. 1987 Centre and Periphery: A Review Concept. Pp. 1–11 in Centre and Periphery in the Ancient World, eds. M. Rowlands, M. Larsen, and K. Kristiansen. Cambridge University Press. Ruiz Mata, D., and Fernández Jurado, J. 1986 El yacimiento metalúrgico de época tartésica de San Bartolomé de Almonte (Huelva). Huelva Arqueológica 8. Ruiz Mata, D. 1993 Los fenicios de época arcaica—siglos VIII/VII a.C.—en la bahía de Cádiz. Estado de la cuestión. Estudos Orientais 4: 23–72. [[Reprinted in English translation here, pp. 155–198.]] Sanmartín, J. 1986 Inscripciones fenicio-púnicas del sureste hispánico. Aula Orientalis 4: 89– 103. Schubart, H. 1988 Endbronzezeitliche und phönizische Siedlungsfunde von der GuadiaroMündung, Prov. Cádiz. Madrider Mitteilungen 29: 132–65. Suárez, A.; Aguayo, P.; Carrilero, M.; López Castro, J. L.; and San Martín, C. 1989 Abdera: una colonia fenicia en el sureste de la Península Ibérica. Madrider Mitteilungen 30: 135–50. Ulreich, H.; Negrete, M. A.; Puch, E.; and Perdigones, L. 1990 Cerro del Prado: Die Ausgrabungen 1989 im Schutthang der phönizischen Ansiedlung an der Guadarranque-Mündung. Madrider Mitteilungen 31: 194–250.

112

María Eugenia Aubet Semmler

Vegas, M. 1989 Archaische und mittelpunische Keramik aus Karthago. Grabungen 1987/88. Römische Mitteilungen 96: 209–59. Whittaker, C. R. 1974 The Western Phoenicians: colonisation and assimilation. Pp. 58–79 of Proceedings of the Cambridge Philological Society 200, n.s. 20. Wolf, E. R. 1982 Europe and the People without History. Berkeley: University of California Press. (Spanish edition: Fondo Cultura Económica, Mexico City, 1987.)

La Fonteta: A Phoenician City in the Far West

Alfredo González Prats, Antonio García Menárguez, and Elisa Ruiz Segura

The excavations and research currently taking place at La Fonteta in Alicante have highlighted the importance of Phoenician colonization on this stretch of Spanish coastline and confirm that La Fonteta was one of the most important Phoenician cities in the West. During September, October, and part of November 1996, the first season of regular excavation took place at the site known as La Fonteta, a city buried under the dunes at Guardamar del Segura. The excavation resulted in the discovery of a long stretch of defensive wall and the demarcation of an inhabited area located outside the wall. Thus, we began an effort intended to firmly establish the boundaries of the site, which greatly exceed the currently enclosed Editor’s note: This article appeared in Spanish under the title “La Fonteta: Una ciudad fenicia en Occidente,” Revista de arqueología 18.190 (1997) 8–13. Authors’ note: These excavations have been subsidized by the General Administration of Artistic Heritage (Dirección de Patrimonio Artístico) of the Council of Culture, Education, and Science (Consellería de Cultura, Educación y Ciencia) of the Valencian government (Generalitat Valenciana) and by the Very Illustrious City Council (Muy Ilustre Ayuntamiento) of Guardamar, with a total investment of some five million pesetas. This research project is linked to the project directed by Prof. González Prats, of the Department of Prehistory at the University of Alicante, with his epigraph Phoenician Colonization in the Southeastern Iberian Peninsula and Cultural Interaction with Indigenous Communities (Colonización fenicia en el Sudeste de la Península ibérica e interacción cultural con las comunidades indígenas). The work combines the efforts of the Department of Prehistory with those of the Archaeological and Ethnographic Museum of Guardamar. The resolution of July 4, 1996, with file number 73/89, initiated the research project, using the same place name given for the site in 1986 in the Ministry of Culture, within the framework of the Spanish Committee for Cultural Itineraries of the Phoenician-Punic Era (Comité Español para los Itinerarios Culturales de Epoca Fenicio-Púnica), of which one of us was named a consulting member representing the Valencian Community.

113

114

A. González Prats, A. García Menárguez, and E. Ruiz Segura

Fig. 1. Red slip plate.

area. An overall plan of action will be needed to reconcile archaeological and environmental interests, given the effects that the excavation will inevitably produce on the environment of the Guardamar dunes, due not so much to the necessary sacrifice of pine trees as to the removal and transfer of sand and earth that cover the Phoenician city. In one of the depressions along the southern stretch of the defensive perimeter, where some of the tons of rock and sand coming from the 1991 excavation at the upper site of La Rábita had been dumped indiscriminately (affecting a section of Phoenician wall mistakenly identified as Islamic), we had the good fortune to find a small area not covered by dunes. A successful sondage led to opening an area of 10 x 11 meters. The archaeological record in this area (A25) has provided a stratigraphic sequence consisting of three different sedimentary layers. The most recent phase (1A) consists of a fill of

A. González Prats, A. García Menárguez, and E. Ruiz Segura

La Fonteta: A Phoenician City in the Far West

115

Fig. 2. Corner of the defensive perimeter of the 7th century b.c.

Fig. 3. Area A7 showing the interior facade of ashlars with reinforcement, buttress, and attached dwelling or casemate.

gray strata deposited in the form of pockets that dip down, filling up pits produced by an erosive process (flooding?) that affected the remains of the earlier phase. Currently there are no signs of inhabited structures, and thus the spot appears to have been used as a dump for domestic debris. Phase 1B is

116

A. González Prats, A. García Menárguez, and E. Ruiz Segura

Fig. 4. Southern flank of defensive wall.

Fig. 5. Another view of the interior facade of the wall in Area A7.

represented by the remains of a large dwelling with walls of orange-colored mudbrick, with stones at the base forming a foundation. The stratigraphic record shows the existence of two stages in the construction of this house. The earlier stage appears to have rooms of greater size; later the number of

La Fonteta: A Phoenician City in the Far West

117

Fig. 6. Stelae found in the collapsed debris of the walls.

rooms was increased by reducing their size to an average width of 2 meters. The strata from the interiors of these rooms have produced rich collections of pottery. Finally, the earliest phase (1C) is represented by a sedimentary reddish-brown formation that gets lighter the farther down we go. This formation provides little archaeological material, but at the bottom we can detect remains of flimsy structures of ash-gray color (post holes and other holes filled with sand) that we believe may correspond to the earliest stage of the large dwelling of Phase 1B. The surprising collection of material culture recovered from A25 illustrates the entire repertoire of Phoenician pottery types: amphorae with carinated shoulders (A1), gray plates, cups with monochrome and bichrome decorations, and a rich variety of red slip and burnished pottery: lamps, rimmed plates, carinated bowls, mushroom-mouth oenochoes, tripods, etc. Along with this pottery we found a large number of fragments of ostrich eggs with red ocher on the interior surface and some with bichrome motifs on the outer surface. On the floor of one of the rooms of Phase 1B (1b8) we collected slag from the smelting of bronze. The dating of the three phases was determined by various chronological indicators, such as the red slip plates and oenochoes and the Greek imports (proto-Corinthian pottery in Phase 1B and East Greek pottery in Phase 1A).

118

A. González Prats, A. García Menárguez, and E. Ruiz Segura

Fig. 7. Red slip and burnished rimmed plates, coming from the sequence in A25.

While waiting for a detailed inventory of all the archaeological material, we would like to propose the following chronology, based on what we observed directly during the course of the season: Phase 1A: 630–590 b.c. Phase 1B: 720–630 b.c. Phase 1C: no conclusions yet The work on the section of wall has exposed 60 meters of defensive fortifications and has clearly defined one of the corners of the city perimeter, where we found a bastion of apparently quadrangular shape (although we were not able to excavate it completely during the 1996 season). The sturdy wall, between 4 and 5 meters wide, was constructed from medium-sized stones of very soft sandstone and calcarenite. The analysis of the exposed wall shows that it is composed of central vertical masses, covered by reinforcing materials set on a slant, producing widths that may exceed 7 meters at the base. The wall was constructed with a sophisticated architectural technique to protect against destruction or deterioration, possibly because of frequent earthquakes in the region of the lower Segura: buttresses, higher than the mudbrick, were placed transversely at certain points, thereby separating sections of the wall and preventing tension and problems of instability from being transmitted along the entire wall. Another phenomenon that we have observed in the wall is the reuse of ashlars from previous structures, although we do not know if the earlier structures were defensive or not. In the layer of collapsed wall, we also see the interesting reuse of cultic stelae (betyls) that surely came from an older cultic or burial area.

La Fonteta: A Phoenician City in the Far West

Fig. 8. Proto-Corinthian pottery from Phase 1B of Area A25.

Fig. 9. East Greek pottery from Phase 1A of Area A7.

119

120

A. González Prats, A. García Menárguez, and E. Ruiz Segura

Fig. 10. Fragment of an alabaster vase found on the surface in El Oral.

Fig. 11. Left: Fragment of a monetary bar.

One of the excavated areas next to the interior facade of a section of wall with projecting ashlars (their crude placement may indicate a reuse) has added to our knowledge about the series of events in the port city of La Fonteta. Area A7 has provided evidence of dwellings placed against the sloped side of the interior facade of the fortification wall. The two rooms that were discovered have stone walls reaching a height of 1.50 meters, on top of which were set gray quadrangular mudbricks 10 centimeters thick cemented together with orange clay. A complex sedimentation fills up the original height of the rooms with alternating layers of fallen mudbrick and gray pockets that contain abundant remains of food, mostly snails. The lower and upper levels of this fill present surfaces with several hearths. The mudbricks that have been excavated show signs of heat-related changes, and their standard size is 30 x 25 x 10 cm. The unbaked bricks that have been recovered show a composition of clay and seaweed (poseidoniae), while other blocks of clay show impressions of reeds or cane, proving the collapse of the roof or an upper story. The whole

La Fonteta: A Phoenician City in the Far West

121

Fig. 12. Earring in the form of a basket.

scene gives the impression of a phase of fill with loose rubble and pockets of debris, after the dwelling next to this part of the wall was ruined or destroyed. Further down an earlier phase appears, with ruins of different buildings. At this point the 1996 excavation ended. What stands out most about Area A7 is that it allows us to show how the placement of the internal row of ashlars (along with the corresponding sloped covering) is higher than the floor level of the adjacent rooms, whose sides go down significantly below the row of ashlars. The phase of earlier dwellings that we are just beginning to detect is actually underneath the defensive wall, indicating the existence of phases of habitation previous to the construction of the defensive enclosure. The East Greek pottery that appears in the strata of this fill allows us, at least initially, to synchronize it with Phase 1A of the fill from Area A25, thus deriving a date from the middle to the third quarter of the 7th century for the construction of the defensive walls, at least for the section in contact with Area A7. In spite of the full and rich archaeological record of the 1996 season, it is premature to present research conclusions, given the large size of the site. Initially, however, we are working with the hypothesis that the construction of

122

A. González Prats, A. García Menárguez, and E. Ruiz Segura

Fig. 13. Belt clasp from the final phase of La Fonteta.

the defensive walls is a relatively late event in the history of this Phoenician colonial enclave. The walls literally stand on top of Phase 1B detected in Area A25, and the reuse of architectural material of a more archaic flavor (ashlars and stelae) during the construction of the walls also appears to support this interpretation. Thus we deduce that we cannot calculate demographic extension or density simply by looking at the area bounded by the defensive walls. The existence of ancient houses outside the walls confirms this belief and suggests an original perimeter either larger or different from the perimeter of the second phase of the city. This is one of the hypotheses that we may test in future seasons. For now we can only look suspiciously at some of the dunes that lie in an unusual crosswise direction in relation to the other dunes in the pine forest. The 1996 excavation at La Fonteta has begun to fulfill the expectations that we have held ever since we were shown the first Phoenician pottery in 1985. We are dealing with a Phoenician settlement at the mouth of the Segura River whose dimensions appear to indicate an urban area that could well reach 8

La Fonteta: A Phoenician City in the Far West

123

hectares, converting it into one of the principal Phoenician cities on the Mediterranean. If, in addition to its size and the quality of its archaeological record, we consider its special state of preservation due to the cover of sand, we undoubtedly find ourselves before one of the best Phoenician cities of the 8th and 7th centuries b.c., outdoing many classic centers of Phoenician civilization in the West as well as in the homeland. The excavations at La Fonteta highlight the same phenomenon that the work of the Instituto Arqueológico Alemán (German Archaeological Institute) demonstrated in the 60s and 70s on the Andalusian coast: our knowledge of the characteristics of the Phoenician world is being made possible due to discoveries in the overseas colonies. From the beginning of its existence in the 8th century, the great port city of La Fonteta had at its disposal a typical setting that included a sanctuary located at Castillo de Guardamar (where a temple with incense offerings to Ceres-Demeter later stood during the Iberian Age). Surely the sanctuary attracted worship to Astarte, protectress of navigation (the “Venus Marina”), at a point which was crucial for nautical visibility. Behind the sanctuary, on another bend of the Segura River, the fortification of Cabezo del Estaño (an incorrect Castilian rendering of estany ‘lagoon’) was built prior to the founding of La Fonteta at the mouth of the river. Thus we can speak of the reproduction of a characteristic model of occupation, in which we lack only the proof of an ancient presence on the nearby island of Tabarca. Whatever the case, the location of this port complex, primarily commercial in nature (though we should not forget the agricultural potential of the fields now buried under the dunes) and right in the middle of the southeastern coastline, must have been crucial for the very dynamics of Phoenician colonization in the West, since we are dealing with the first stable peninsular city within the area of the island route. The role that the port of Guardamar must have played in the colonization of Ibiza is becoming clearer as we analyze cultural characteristics that are beginning to be recognized. It is evident that Ibiza was already founded in the 7th century, if not in the 8th, by people who were not Carthaginians, but rather western Phoenicians. The ceramic clays of Sa Caleta are the same as those of the cups produced in the centers on the Andalusian shore (Guadalhorce, Toscanos, Mezquitilla, Chorreras) and the same as the vessels from Guardamar. In any case, Ibiza must have served to transmit commercial and cultural currents coming from the central Mediterranean. This fact is illustrated by the burial stelae from the sanctuary or tophet at La Fonteta (which have the same style as certain examples from Motya, Carthage, and Ibiza) and samples of gold and silver work with parallels in Tharros, Carthage, and North Africa. The presence of eastern peoples at the mouth of the Segura surely coincides with the period of initial Phoenician implantation in the West, refuting interpretations that saw on the southeastern Mediterranean coast only a pale

124

A. González Prats, A. García Menárguez, and E. Ruiz Segura

reflection of the process of colonization in the southern Peninsula and that, in passing, dated these colonies to a late period of Phoenician colonization. The bronze, ivory, and glass objects found in the strata of the Final Bronze at Peña Negra (in the nearby Sierra de Crevillente) alerted us ten years ago to the early date of contacts with the Phoenician world already in the 9th century and, with more regularity, in the 8th century b.c. Today it is clear that the Atlantic-style metallurgical focus of Peña Negra I led directly to the arrival (if it did not develop as a response to the presence) of eastern peoples, who benefited from metallurgical production in the Sierra de Crevillente, a stone’s throw from the mouth of the Segura. The early Phoenician presence in Peña Negra probably led to the installation of a small, exotic pottery production center during the transition from the 8th to the 7th century, principally in sectors 7 and 8 of this important orientalizing city. In 1983 we identified Peña Negra with the Herna of the Ora Maritima, a crossroads for diverse peoples and products and a flourishing market center where various trade routes converged. Trade must have been so dynamic in the Southeast that already in the 8th century we witness the establishment of the first monetary system in the western Mediterranean to employ flat bars of metal (copper, bronze and lead), which has been confirmed by the two pieces found at La Fonteta (Areas A7 and A25). In summary, the excavation of one of the largest and best preserved cities of the Phoenician world has begun, which is a favorable sign that expectations for the research project at La Fonteta (Proyecto de Investigación La Fonteta) will be broad and ambitious. We are confronted with the only Phoenician site in the Valencian Community, and public and private entities and administrative bodies must be aware of what their support can mean for the historicalarchaeological heritage represented by La Fonteta, which lies buried beneath a marvelous natural heritage.

Bibliography García Menárguez, A. “Avance sobre las excavaciones en yacimientos con fases del Hierro Antiguo en el tramo final del río Segura (Guardamar del Segura, Alicante).” Pp. 225–29 in XXII Congreso Nacional de Arqueología (Zaragoza), 1995. González Prats, A. “Sobre unos elementos materiales del comercio fenicio en tierras del Sudeste peninsular.” Lucentum 4 (1985) 97–106. ___ . “Las importaciones y la presencia fenicia en la Sierra de Crevillente (Alicante).” Aula Orientalis 4 (1986) 279–302. ___ . “La factoría fenicia de Guardamar.” Azarbe (Guardamar, 1990). ___ . Nueva luz sobre la protohistoria del Sudeste peninsular. Alicante: University of Alicante, 1990. ___ . “La presencia fenicia en el Levante peninsular y su influencia en las comunidades indígenas.” Pp. 109–18 in I–IV Jornadas de Arqueología Fenicio-Púnica. Ibiza, 1991.

La Fonteta: A Phoenician City in the Far West

125

___ . “De la Eivissa Cartaginesa a la Eivissa Fenicia. Precisiones sobre la fundación ‘púnica’ de Ibiza.” Arqrítica 2 (Madrid, 1991) 18–19. ___ . “Quince años de excavaciones en la ciudad protohistórica de Herna (La Peña Negra, Crevillente, Alicante).” Saguntum 26 (1991) 181–88. González Prats, A., and García Menárguez (in press). “El conjunto fenicio de la desembocadura del río Segura (Guardamar del Segura, Alicante).” IV Congreso Internacional de Estudios Fenicios y Púnicos. Cádiz: Ministry of Culture.

126

A. González Prats, A. García Menárguez, and E. Ruiz Segura

The Ancient Colonization of Ibiza: Mechanisms and Process Juan Ramón

1. Introduction For several years, we have known that a series of material remains on the island of Ibiza dates back to the 7th century b.c. and that the first colonization was not by the Carthaginians, but rather by the Phoenicians of the Far West. The process that led to these conclusions is well known and does not need to be described again in this discussion. Numerous authors, myself included, have written about this topic, and there is a basic uniformity in the argument, although there are different nuances which are not always free from a cautious subjectivity, especially in non-scientific matters. However, an unusual ability to “rediscover” the same thing over and over is apparent in this cycle of research. The same hypothesis is proposed again, as if ex nihilo, and the same principles and practically the same data simply lead to the same points of view. The sequence of scientific literature, from which we quote a few paragraphs, clearly illustrates this fact: . . . the material [from the 1946 excavation] still remains unstudied, but a small collection of pottery exists that corresponds to Form 3 in Bisi’s classification. . . . we doubt that this pottery can be dated later than the beginning of the 6th century b.c., but rather should be dated to the 7th. . . . our lamp [the one from plot number 40 on Vía Romana] is no longer considered a completely isolated find in regard to its date and context. . . . Was there an area of early burials at the Editor’s note: This article appeared in Spanish under the title “La colonización arcaica de Ibiza. Mecánica y proceso,” in X Jornades d’Estudis Històrics Locals: La prehistòria de les illes de la Mediterrània occidental (ed. G. Rosselló Bordoy; Palma de Mallorca, 1992) 453–78. Author’s note: Cultural Council of Eivissa (Ibiza) and Formentera. Archaeological Services. (Conselleria de Cultura d’Eivissa i Formentera. Servei d’Arqueologia.) Many of the drawings are the work of J. M. López Garí. The author also wants to thank Rosa Gurrea for the use of some of the documents from the excavation of the street Vía Romana, number 47, which took place under her direction in 1990 and under that of the author of this article in December of 1991.

127

128

Juan Ramón

flat northern base of the hill, a zone which was later invaded by the hypogea? The indications that we have today, although scarce, appear to affirm this. We have a beginning date for the necropolis at Puig des Molins that appears to be fixed before 600 b.c. 1 Archaeology [the 1977 excavation at Puig des Molins, which was the definitive discovery of the ancient necropolis with cremations, and the survey of Sa Caleta in 1978] demonstrates that Phoenicians coming from the West [from the “circle of Cádiz”] settled at Puig de Vila . . . at least by the second half of the 7th century b.c. The necropolis with cremations found at the northern base of Puig des Molins constitutes obvious proof in this respect. . . . The origin of the Phoenician colony on Ibiza perhaps has one major explanation: maritime trade carried out by Semitic peoples from the area of the Strait of Gibraltar with the area of the Golfe du Lion. . . . The colony during this initial period could not have been more than a small group of houses. . . . 2 . . . the careful study of the materials has permitted us to distinguish a first phase centered in the 6th century b.c. We already have a group of sufficiently representative materials. . . . Thus, there are juglets of the type Bisi 3. . . . Therefore, we believe that we must seriously begin to consider the existence of an area of early burials. . ..This area was of course destroyed . . . very early on, as far as we know, since innumerable hypogea were dug there. . . . Thus, in our opinion, there was an area of the necropolis where burials took place (cremations, according to the data that we possess), from almost the first moments of the existence of the Punic colony. 3 It is clear . . . that we can affirm that a zone of early burials exists at the foot of the northern slope of the hill. . . . on Puig des Molins there exist early burials until now unknown. . . . In conclusion, the present state of research permits us to date the first settlement of the island of Ibiza to the 7th century b.c., in perfect agreement with the written sources. . . . A discovery of the greatest importance has taken place [1983] at the necropolis of Puig des Molins. . . . We have found materials that fully confirm the colonization of Ibiza in the 7th century b.c., and the parallels lead us directly to the Phoenician sites on the southern Iberian peninsula and northern Africa. 4 When we finished our study of the 1946 excavation in mid 1982, we compared the accidental findings of the 70s with the results of Mañá, stating our hopes that 1. J. Ramón, “Necròpolis des Puig des Molins: Solar núm. 40 del carrer de la Via Romana de la ciutat d’Eivissa,” Fonaments 1 (Barcelona, 1978) 81–82. 2. J. Ramón, “Sobre els orígens de la colònia fenícia d’Eivissa,” Revista Eivissa 3. època 12 (Ibiza, 1981) 30–31. 3. C. Gómez, “La necrópolis del Puig des Molins (Ibiza): Campaña de 1946,” Excavaciones Arqueológicas en España 132 (Madrid, 1984) 153. 4. J. H. Fernández, C. Gómez, and R. Gurrea, “La première periode de la colonisation punique a Ibiza,” in The Deyà Conference of Prehistory (= British Archaeological Reports International Series 229; 1984) 789–91.

The Ancient Colonization of Ibiza: Mechanisms and Process

129

Fig. 1. Map of the Pitiuses (islands of Ibiza and Formentera). apparently we had finally located the ancient sector of Puig des Molins. . . .5 Judging from the archaeological evidence that we have presented, which is the result of research developed on Ibiza during the past few years, especially at the necropolis of Puig des Molins, the traditional view of the colonization of the island has been visibly questioned. In view of the new data, we are convinced that 5. C. Gómez, “La colonización fenicia de la isla de Ibiza,” Excavaciones Arqueológicas en España 157 (Madrid, 1984) 33.

130

Juan Ramón

it is no longer possible to continue attributing to Carthage the role of first founder of the Ibizan colony. 6 The excavations which took place in 1982 and 1983 under C. Gómez, R. Gurrea, and B. Costa, respectively, in the protected sector to the west of the museum, permitted us to establish definitively that at least between the end of the 7th century b.c. and the beginning of the 6th, this sector of Puig des Molins was used as a burial place for the first Phoenician settlers from the nearby city, which was then a small trading post. . . . 7

From a purely scientific viewpoint, this old information no longer catches our attention. The objective of this article is simply to summarize the principle traits of a new perspective that has been recently proposed. 8 Thus the key focus will not be a “new” demonstration of the ancient nature of the colonization of Ibiza nor its obvious connection with the Phoenician Far West, but rather the basic mechanisms and its chronological sequence.

2. Some Elements for Consideration 2.1. The Settlement at Sa Caleta This is a Phoenician settlement that occupies a peninsula known as “Sa Mola de sa Caleta,” and the name of the archaeological site comes from this toponym. 9 Sa Caleta is located on the south-southwestern coast of Ibiza, between Puig des Jondal and the northwest end of the long pebble beach of Es Codolar. As geographic indicators, we can point out the 9.5 kilometers in a straight line by land that separate this site from Puig de Vila (the upper city of Ibiza) and the 9.4 nautical miles that stretch along the southern coastline from Sa Caleta to the mouth of the Bay of Ibiza. The peninsula is separated from the mainland by a deep stream bed, which emerges on the extreme northeast of Sa Mola and forms a landing used by small boats until the present day. The surface of the peninsula is quite flat, with a maximum elevation of about 16 meters above sea level on its western 6. B. Costa, J. H. Fernández, and C. Gómez, “Ibiza fenicia: La primera fase de la colonización de la isla (siglos VII y VI b.c.), ” in II Congresso Internazionale di Studi Fenici e Punici, vol. 2 (Rome, 1991) 794. 7. C. Gómez, “Ibiza en época arcaica: estado actual de la investigación,” in I–IV Jornadas de Arqueología Fenicio-Púnica (Ibiza 1986–1989) (= Trabajos del Museo Arqueológico de Ibiza 24; Ibiza, 1991) 22. 8. J. Ramón, “El nacimiento de la ciudad fenicia de la bahía de Ibiza,” in Coloquios de Cartagena I: El Mundo Púnico. Historia, Sociedad y Cultura (Cartagena, Nov. 1990) (Murcia: Biblioteca Básica Murciana, Extra 4) 325–68. 9. More extensive information on the site of Sa Caleta can be found in J. Ramón, “El yacimiento fenicio de Sa Caleta,” in I–IV Jornadas de Arqueología Fenicio-Púnicas (Ibiza 1986– 1989) (= Trabajos del Museo Arqueológico de Ibiza 24; Ibiza, 1991) 177–87.

The Ancient Colonization of Ibiza: Mechanisms and Process

131

Fig. 2. Sa Caleta, southern sector. Floor plan of an architectural unit with 3 rooms lined up longitudinally (E-XVII-XIX): (a) door with threshold; (b) hearth.

side, where the cliff drops precipitously due to strong waves produced by gale-force winds and the geology of the terrain. Today about 3 hectares of the peninsula’s surface area remains, although we can add at least another hectare, according to the evidence. Possibly the surface area of the town measured at least 4 hectares. In any case, we are dealing with considerable surface area.

132

Juan Ramón

Since 1986, six brief excavation seasons have taken place at the site. Although they have been of short duration, they have already permitted us to uncover a considerable portion of the preserved remains. The settlement demonstrates a type of urbanism based on the juxtaposition of units, with walls oriented in different directions in relation to each other. Between the different units and groupings of rooms, we have been able to document fully the existence of very narrow, frequently diagonal alleyways and small polygonal plazas with variable surface areas, due to the varied orientation of the surrounding buildings. This is evidently an archaic type of urbanism. In contrast, the floors of the rooms are generally rectangular with distinct lengthening; trapezoid shapes are rarer. The position and dimensions of the spaces that were constructed vary according to their nature and the function for which they were designed. Thus, we have rooms up to 10 meters long with almost 30 square meters of usable interior space. Other rooms have less than 3 square meters. Of course, there are all kinds of intermediate dimensions. Another important characteristic comes from the conception and evolution of the constructed spaces. Many dwellings, usually medium or large in size, were enlarged by juxtaposing another room or two, up to a total of seven. Others were never enlarged. Also we have some infrequent cases of construction of a “building” of three rooms, all designed at the same time. The rooms are normally connected by means of relatively wide doors with small stone thresholds. The construction techniques are the same throughout the settlement. The floors usually have a layer of earth that has been stamped down and flattened intentionally or simply through usage. The walls always have bases of rubble, composed of irregular pieces of local limestone with differing measurements. The stones were joined together with red earth made from local mud. It is not clear, however, if the upper walls were normally made of mud-brick or of stone. A very important fact is that the whole settlement presents only one stratum of occupation. From north to south on the peninsula wherever they extend, the small streets and plazas and other spaces invariably have a bottom level (normally thin) of occupation/abandonment. Over this (in the case of the rooms) lies a clayey layer of variable thickness, a product of the collapse of the roofs, and on top is a thick layer produced by the collapse of the walls. Therefore, a single level of occupation exists, qualified by the double phase represented by the horizontal enlargements in which rooms were juxtaposed. Thus, the lapse of time, judging by the general stratigraphic thickness, the composition of the dumps, etc., appears to have been quite short. Moreover, the entire site was abandoned all at once, at the same time, and in an orderly fashion. Indications are that the villagers tried not to leave behind anything

The Ancient Colonization of Ibiza: Mechanisms and Process

133

Fig. 3. Sa Caleta: (1) western Phoenician amphora T-10.1.2.1. (inventory XII-1/VIII1/XIII-1); (2) central Mediterranean Punic amphora T-2.1.1.2. (inv. XX-28).

that was minimally useful, with the hearths and floors in the rooms swept clean; in other words, it was a planned and methodical abandonment. The typology and nature of the artifacts found is absolutely clear. On the one hand, we see a broad, characteristic range of vessels of strictly early western Phoenician type, and on the other hand, some central Mediterranean Punic amphorae. Also there are several non-ceramic objects. The early western Phoenician samples include wheel-made pottery, often undecorated (especially transport amphorae T-10.1.1.1. and T-10.1.2.1. 10). Decorated pottery includes pottery with bichrome and polychrome paint ( jugs with globular bodies), red slip pottery (carinated bowls, lamps, plates, etc.), and also gray pottery (bowls and plates). Also to this group, which is 10. J. Ramón, “Las ánforas fenicio-púnicas en el Mediterráneo Central y Occidental,” Instrumenta 2 (Barcelona: University of Barcelona Publications, 1995).

134

Juan Ramón

characterized by minerals that are completely non-native (slate, schist, gneiss), we should add a series of handmade pieces. The samples from the central Mediterranean (Sardinia, Sicily, and/or Tunis) include a significant number of commercial amphorae with an ovoid shape (T-2.1.1.2.). Only a few of the handmade pieces could have been produced at the site, although this is possible rather than certain. Among the miscellaneous materials, the presence of numerous nodules of argentiferous lead sulfide has a very special significance. There were several kilograms in the levels of occupation/abandonment, found in many rooms and outside spaces along the length and breadth of the settlement. There were also artifacts such as iron and bronze knives, which are interesting in other ways. In room XV we found an oven for smelting and/or forging iron, along with numerous bellows nozzles. The presence of hand grinders of sandstone and volcanic rock and the frequent discovery of bronze fishhooks are also notable. In regard to the presence of animal life, bovines, sheep, goats, etc., have been preliminarily identified.

2.2. Finds around the Bay of Ibiza 2.2.1. Puig des Molins The area at the base of the north-northwest slope of the hill of Puig des Molins is filled with a dense grouping of ancient tombs containing cremations. 11 According to our current knowledge, the size of this necropolis does not appear to be less than 8,000/10,000 square meters (plots numbers 36, 38, 40, and 47 on the street Vía Romana, numbers 1–3 on the street Juan Ramón Jiménez, and the space between the Museo Monográfico and the street Vía Romana, number 47). Its surface area is thus greater than what was thought earlier. The date of these tombs with cremations appears to stretch from the second quarter of the 6th century b.c. to surely the first quarter of the 5th century. 11. For more information on the ancient necropolis of Puig des Molins, consult L. Baqués, “Escarabeos egipcios de Ibiza,” Ampurias 36–37 (Barcelona, 1974–1975) 87–146; J. Ramón, “Sobre els orígens”; C. Gómez, “La necrópolis”; C. Gómez, “La colonización fenicia”; J. H. Fernández, C. Gómez, and R. Gurrea,“La première periode”; J. H. Fernández and J. Padró, “Escarabeos del Museo Arqueológico de Ibiza,” Trabajos del Museo Arqueológico de Ibiza 7 (Madrid, 1982); J. Ramón, “Cuatro elementos cerámicos arcaicos de Importación hallados en Ibiza,” Información Arqueológica 40 (Barcelona, 1983) 111–20; María Pilar San Nicolás, “Dos colgantes excepcionales del Museo Arqueológico de Ibiza,” in Homenaje al Prof. Gratiniano Nieto, vol. 1 (Madrid, 1984–1985); C. Gómez, “Lekythoi samios y botellas sidonias. Estudio de un ejemplar procedente de Ibiza,” Saguntum 20 (Valencia, 1986); B. Costa, J. H. Fernández, and C. Gómez, in Ibiza fenicia; B. Costa and C. Gómez, “Las importaciones cerámicas griegas y etruscas en Ibiza,” Mélanges de la Casa de Velázquez 23 (Paris, 1987) 31–56; J. Ramón, El nacimiento de la ciudad fenicia.”

The Ancient Colonization of Ibiza: Mechanisms and Process

135

Fig. 4. Sa Caleta, Phoenician red slip pottery produced in southern Spain: (1) lamp with two wicks (inv. XVIII-5); (2) carinated bowl (inv. a-44); (3) plate (inv. p-26).

The tombs correspond to the following types: (1) Charred bones placed in a small circular hole in the rock, sometimes protected by small stone tiles placed vertically and diagonally. Often this hole is entirely artificial (roughly semicircular or cylindrical in shape and tapering

136

Juan Ramón

toward the bottom), although there are also cases in which the bones were deposited in natural cavities in the rock and covered with simple stones. These natural cavities at times were enlarged artificially. (2) Cremation jar with an ovoid body and geminate handles inside a small, round hole hewn into the rock, or placed in a natural cavity in the rock, although always protected by stones. Other vessels could also function as containers for cremations, such as pithoi and amphorae (T-10.1.2.1). (3) Small rectangular pit-graves varying in depth. At times they have side ledges where irregular stone-slab covers fit, although there are others that lack this detail. (4) Rectangular pit-graves, with a long, narrow depression running lengthwise along the bottom. These graves usually contain in situ cremations of cadavers. In these graves, it is normal to see sandstone pillars (betyls). The accompanying materials are mainly characterized by their scarcity. It can be said that the vessels that contain the cremations dominate the ceramic panorama, though they do not technically form part of the grave goods. Apart from some objects for personal use by the deceased, the accompanying artifacts are reduced to a small repertoire. Small juglets that must have contained ointments or perfumed oils appear to characterize the burial ritual, although a large percentage of tombs surely never contained an offering. These small juglets show variations on a spheroid body with a handle, and they have been published on different occasions. In some cases, lamps with double wicks, narrow rims, and wide bowls (at times treated with a red slip) are the only characteristic piece of pottery accompanying the deceased. The repertoire of hand-made vessels is also interesting. In some cases they were containers for ashes. Along with other types we can mention small globular pots with a slightly concave rectilinear rim and a handle halfway up. Others have bases with a biconical tendency, a wide mouth, and two flattened handles connected to the upper center. Also there are examples of widemouthed spherical pots with rims that are rectilinear or slanting outward. Pots of this type may have ledge handles. As elements of personal ornamentation for the deceased, we can mention the very few silver earrings that have been found so far. One example is round with a central protruding spool around which a coil-like element was wound to suspend the earring from the ear. Another earring has a square apse shape, with a similar method for hanging it from the ear. These two earrings from Puig des Molins are made of silver. Both are of the simplest style, lacking any kind of decoration, embossed motif, or granulation. To these we should add the recent find from the plot on Vía Romana, number 47. This is another “niche-shaped” earring like the second one above, and it is the only piece made of gold that has been recovered so far from an early cremation.

spread is 1 pica short

The Ancient Colonization of Ibiza: Mechanisms and Process

137

Fig. 5. Sa Caleta, Phoenician pottery produced in southern Spain. Numbers 1 and 2 are gray ware; number 3 is common ware. (1) Bowl with turned-up rim (inv. k-11); (2) Bowl with convex rim (inv. h-26); (3) Juglet (inv. XII-2/XIII-8).

Several scarabs must have come from Puig des Molins. One scarab has the hieroglyphic inscription Psmtk-snb, surely alluding to Psamtik II, 595–589 b.c. We also have a scarab with the sign nfr between two uraeuses facing each other, probably dating to the 7th/6th centuries b.c., and made from white faience with the seated figure of Horus mounted on a silver bascule ring. This scarab is perhaps a product of Naukratis. We must also mention an example of

138

Juan Ramón

blue faience mounted on a silver ring which pictures a horse-drawn chariot with its driver. This scarab also appears to be a product of Naukratis from the 7th or 6th century b.c. Another scarab was found during the 1946 excavation by J. M. Mañá. It is made of blue faience and mounted on silver, with an indecipherable motif, dated to the 7th or 6th century b.c. Finally, there is the scarab found in cremation number 5 at plot 38 on the street Vía Romana. This piece is mounted on silver, made from faience, and pictures a winged sphinx. This is, in any case, the only scarab found in a context linked to the first generations of colonizers in the bay, even though its precise dating is somewhat broad. The Greek and Etruscan pieces are another group of materials with great commercial and chronological interest. An Etruscan kantharos and a Corinthian aryballos stand out. Possibly the pear-shaped Etruscan-Corinthian aryballos, presently preserved in the Museum of Barcelona, also comes from the necropolis of Puig des Molins, an origin that can also be given to a lekythos of the so-called “Samian” type acquired by the museum in Ibiza in 1909. All of these artifacts can be dated to the first, and especially the second, quarter of the 6th century b.c. Some brief comments should be made about an excavation performed intermittently in 1990 and 1991 in the western sector of the ancient necropolis, specifically in plot number 47 on the street Vía Romana. A burial area from the 6th century was located in a very small space, on top of which were placed other sequences of Punic tombs from the 5th to the 2nd centuries b.c. Remains of globular jugs have been identified, as well as pit-graves (without ledges) hewn into the rock and oriented in a north-south direction, and holes in the ground containing charred remains. However, what especially stands out is a line of pit-graves (oriented east-west) with central depressions, which had been dug into the red muddy ground very close together, occupying all of the space. Grave goods were very scarce, as usual. There were jewels, in some cases made of gold, and a few lamps. In reality, the area had suffered equally from the continuous burials of later centuries and from an excavation directed by J. M. Mañá in 1951. The inadequate methodology led to quite a bit of confusion in later research. For example, there were numerous child burials in amphorae, especially of the type T-1.3.2.3. (PE-13). When these were found very near or coinciding with the levels of charred bones in the pit-graves, they were interpreted as cremation amphorae. However, it is now evident, according to a series of samples that have been found intact, that the very much discussed “necropolis of child cremations” never existed, since the amphorae always contain inhumations of young children. In one of the early pit-graves containing a long, central depression, three or more betyls made of local sandstone were found. In some cases their length reaches 1.50 meters. They have a square or slightly rectangular section, narrowing a bit toward the top, and the top is flat. At the base they have a narrow

The Ancient Colonization of Ibiza: Mechanisms and Process

139

Fig. 6. Sa Caleta: (1) western Phoenician tripod (inv. i-13); (2) handmade pot with four lengthened ledge handles (inv. XIV-6); (3) handmade bowl (inv. c-10).

projection in order to fit into square or rectangular supports, also made of sandstone. This is the same type that appears over and over again among the Punic stelae of the central Mediterranean beginning in the 6th century b.c., which are found singly, in pairs, or forming triads or double triads. The betyls of the tomb of Vía Romana, number 47, were thrown directly on top of the

140

Juan Ramón

ashes and coals of the cremation and thus were broken and partially burned. All of this raises the possibility that rather than being memorial stones or cippi used to identify the tombs, they could have been religious images carrying with them some specific type of burial ritual, although to comment more in this article would stray too far from the present topic. Although not linked specifically to the burials, an interesting group of pottery sherds, called deposit PM NE-83, also comes from the area of Puig des Molins. This is an important group of pottery sherds from the ancient period, although they were tossed out here at an uncertain date, together with pieces of vessels from much later periods. The material is composed of the following: more than 50 gray bowls with convex rims sloping inwards; fragments of about 25 red slip plates; fragments of untreated bowls, usually with carinated, S-shaped profiles; pieces of about seven bowl tripods; fragments of a red slip jug with a carinated shoulder; fragments of some 20 jugs with globular bodies and geminate handles; fragments of another 20 wide-mouthed jugs with two or four handles (pithoi); fragments of some 15 transport amphorae with carinated shoulders; fragments of handmade pottery, including a flat-based bowl and an oblique-convex body sherd. This assemblage of pottery is different because it does not belong to any burial ritual. Possibly it is the reflection of some industrial or home activity near the spot where the pottery was found. Even more important, in our opinion, is the appraisal, totally contrary to what had been believed earlier, that more than 90% of these fragments are native proto-Ibizan pottery, all of which we have examined appropriately.

2.2.2. The Castle and Fortress (Almudaina) in Ibiza These two adjoining military structures occupy the broad oval that defines the high point of Puig de Vila, where the city of Ibiza was founded. The emergency excavations performed in 1988–1989, which took place in both the castle and fortress areas, verified the presence of several fragments of early pottery, 12 all from later strata in both areas and in the form of residual material. These fragments, however, hold an enormous interest because they demonstrate, beyond any doubt, that from the earliest period in which we can detect a Phoenician presence in the Bay of Ibiza, Semitic peoples had already built some type of installation on the summit of the hill. Strictly speaking, the nature of the occupation of this high point on Puig de Vila remains difficult to clarify since we do not yet have corresponding architectural levels or structures. However, it is not difficult to imagine that the best use of a high point such as this would be a temple and/or fortification. 12. For more information on ancient materials from the castle and Almudaina of Ibiza, consult J. Ramón, “El nacimiento de la ciudad fenicia,” 352–53.

The Ancient Colonization of Ibiza: Mechanisms and Process

141

Apart from several shapeless fragments, we should point out some pieces of wide-rimmed plates covered with a red slip on the upper surface, fragments of Phoenician gray ware bowls with the type of rims that are convex or droop slightly, and pieces of carinated amphorae with globular bodies (T-10.1.2.1.). Although the documented material of this type is not very abundant, it is significant that all the fragments that have been examined are pieces made in Ibiza, in other words, red slip, gray, and common proto-Ibizan pottery. None of these materials are imports from Phoenician southern Spain.

2.2.3. The Point of Joan Tur Esquerrer This site is located on rocky outcroppings which end in a cliff on the southern slope of Puig de Vila. This sector is outside the walls and traditionally uninhabited. If we consider the material recovered from its surface, 13 it is evident that the Phoenicians performed some type of activity on top of the rocky double hill or promontory. It is even possible that they had some fixed installation that has completely disappeared today. This could have been a place inhabited by fishermen who used the rocky and progressively deteriorating cove below for a landing. We must keep in mind that this sector, where presently the limestone and the eroded dolomites barely permit the survival of a few bushes, was located just below the area occupied by the castle/fortress that we have already described, and that no natural obstacle separates the two areas. The material gathered from the surface consists of pottery fragments. Transport amphorae stand out, with 16% of southern Iberian or “Strait of Gibraltar” production (T-10.1.2.1.), 78% of the same type but of Ibizan fabrication, and 5% Phoenician-Punic, from the central Mediterranean (T-2.1.1.2.). We must also mention fragments of wide-mouthed jugs of the pithoi type, produced by the Phoenicians of southern Iberia. The serious fragmentation of the material prohibits precise identification of the exact type to which we can assign these vessels with two or four circular geminate handles. In any case, we are dealing with a form that is well represented in Algerian, Moroccan, and Iberian sites of the 6th century b.c. Also, we should mention the existence of jugs with ovoid or globular bodies, necks with slight carinations, ring bases, and triangular rims with two circular geminate handles. These are examples of proto-Ibizan production, like those at Puig des Molins. Other more fragile pieces, such as red slip and grayware pottery, may have largely disintegrated because of the strong erosion at this site.

13. For more information on the point of Joan Tur Esquerrer, see J. Ramón, “Sobre els orígens”; J. Ramón, “El nacimiento de la ciudad fenicia,” 353–56.

Fig. 7. The Bay of Ibiza, indicating the ancient sites: (1) ancient necropolis of Puig des Molins; (2) deposit PM NE-83; (3) Bastion of Santa Lucía; (4) Castle-Fortress (Almudaina) of Ibiza; (5) Point of Joan Tur Esquerrer; (6) Illa Plana.

142

spread is 12 points short

Juan Ramón

The Ancient Colonization of Ibiza: Mechanisms and Process

143

2.2.4. Finds on the Acropolis of Ibiza In the fill of the bulwarks and stretches of Renaissance fortifications in the city of Ibiza (second half of the 16th century), we find abundant archaeological material without stratigraphy, especially pottery. The material comes from all of the historical phases of the city up to the time of the construction of the walls. 14 Specifically, we should mention two interesting sherds coming from the bastion of Santa Lucía, which dominates the entrance to the port of Ibiza. The first is a fragment of an SOS amphora. The second is a piece of a tripod with a characteristic composition of schist, slate, etc., and a whitish coating on the exterior. Its smooth triangular rim fits in perfectly with production in southern Spain (Málaga-Granada?). Also, during some work performed years ago on the neighboring bulwark of San Juan, an interesting lead figure was found out of context. It represents a male character, 14.7 cm. tall, on a rectangular pedestal. His head is covered by a cap, and he has a pointed beard and bulging eyes. The left foot is further forward than the right. The object that the figure carries in his left hand is probably a torch, unless it is the upper part of a lance or lightening bolt. It appears that we are dealing with a representation of the so-called Smiting God. It may have been a very late and distinct version, possibly produced in Ibiza.

2.2.5. Illa Plana Illa Plana is a well-known islet with flat contours, except for a small promontory that juts out on its northwest end next to the northeast side of the Bay of Ibiza. Since the archaeology of this site is well known, we will refer to only three cylindrical figures coming from the bothros (cultic pit) and belonging to the type IIA classification of J. Ferrón and M. E. Aubet. The material from which they are made is particularly interesting and can be attributed to the proto-Ibizan group, between 575 and 525 b.c. They could be the earliest figures from Illa Plana or at least the earliest made on the potter’s wheel. Also we have been able to verify the presence of some fragments of proto-Ibizan pottery on the islet’s surface, specifically of amphorae. 15 14. For more information on the different materials in the upper city of Ibiza, see J. H. Fernández, “Una estatuilla de plomo del Museo Arqueológico de Ibiza,” in Estudis de Prehistòria, d’Història de Mayûrqa y d’Història de Mallorca oferts a Guillem Rosselló Bordoy (Palma, 1982) 51–60; J. Ramón, “El nacimiento de la ciudad fenicia,” 356–57. 15. For more information on Illa Plana, see, for example, J. Ferrón and M. E. Aubet, “Estatuillas de orantes del mundo cartaginés. Tipología y cronología, “Trabajos de Prehistoria 31 (Madrid, 1974) 253–76; E. Hachuel and V. Marí, “El santuario de la Illa Plana (Ibiza). Una propuesta de análisis,” Trabajos del Museo Arqueológico de Ibiza 18 (Ibiza, 1988).

144

Juan Ramón

Fig. 8. Puig des Molins, ancient necropolis: (1) Etruscan kantharos from plot 38 on Vía Romana; (2) Corinthian aryballos from “Sector B”; (3) (eastern?) Phoenican juglet, 1946 excavation; (4) Egyptian aryballos made of faience.

The Ancient Colonization of Ibiza: Mechanisms and Process

145

Fig. 9. Puig des Molins, Phoenician pottery of proto-Ibizan fabrication: (1) red slip lamp with two wicks, ancient necropolis, plot 40 on Vía Romana; (2) to (6) deposit “PM NE-83”; (2) and (3) red slip plates; (4) gray-ware bowl; (5) and (6) carinated bowls of common ware.

Subhead drop 146

Juan Ramón

3. Chronology According to what we know from the analysis of pottery, the settlement of Sa Caleta lasted for only a short period of time, something which is corroborated by the structure and sequence of the stratigraphy. The wheel-made pottery is entirely of outside production, and therefore its dating is intrinsically linked to the place of origin. However, we must remember that the types represented in this Ibizan settlement, such as globular jugs or pithoi, bowls and plates of gray ware, and amphorae T-10.1.2.1., are difficult to date precisely, with margins of 50 years or more. In any case, we find red slip plates with rims reaching up to more than 7 cm in width. Other very significant elements are the Punic amphorae of the central Mediterranean (T-2.1.1.2.), 16 which are abundant at the site and should have reached their high point around 598 b.c. and immediately following, if we judge by the discoveries at Camarina, founded that year from Syracuse. Sa Caleta was probably inhabited between the final third of the 7th century b.c. and at least the first decade of the 6th century. As for an initial date for Phoenician settlement in the Bay of Ibiza, we believe that this cannot be studied adequately without considering two key factors. On the one hand, we must keep in mind the case of Sa Caleta, already mentioned above, and on the other hand, we must take into account the local character of most of the Phoenician pottery found at Puig des Molins and on the acropolis, which until now was believed to be of outside production. The sudden but peaceful, uniform, and, above all, organized abandonment of Sa Caleta in the first quarter or decade of the 6th century b.c. clearly indicates that this was a planned move. The possibility that its people moved to the Bay of Ibiza is more than logical. The only question, then, is whether they moved into a town that already existed or founded the town ex nihilo. It is true that a group of Phoenician pottery produced in southern Iberia has been found in the area of the bay (Puig des Molins, Bastion of Santa Lucía, point of Joan Tur Esquerrer), similar to some materials found in the level of abandonment at Sa Caleta. In reality, this pottery constitutes a low percentage, which we could interpret as material carried from Sa Caleta when it was abandoned, or as a logical continuation in the first years of occupation of the bay area of the same system of transactions and supplies seen at Sa Caleta. The imported Greek and Etruscan pottery, which is loosely dateable, gives dates not generally earlier than the second quarter of the 6th century b.c. or the late first quarter. Thus, the proto-Ibizan pottery appears to correspond to a stage that is clearly later, although not much later, than the abandonment of Sa Caleta, where not one sample of Ibizan production has been found—an extremely important piece of data. 16. J. Ramón, “Las anforas fenicio-púnicas.”

The Ancient Colonization of Ibiza: Mechanisms and Process

147

Fig. 10. Puig des Molins, proto-Ibizan Phoenician pottery: (1) and (2) deposit “PM NE-83”; (3) and (4) ancient necropolis. (1) Carinated jug with red slip; (2) pithos of common ware; (3) globular jug from the 1977 excavation; (4) globular jug with monochrome bands, found in 1976.

If the settlements at Sa Caleta and the Bay of Ibiza had been contemporaneous during the second half of the 7th century, most likely the process of change would have started with autoproduction in the bay area and would have spread directly to the neighboring town of Sa Caleta as a first point of economic expansion of the “proto-city,” contributing to the downfall of the

148

Juan Ramón

site on the southwestern coast of Ibiza and the migration of its population to the enclave on the bay. But for many years we have looked for even the smallest trace of proto-Ibizan materials at Sa Caleta, specifically for amphorae PE-10 (T-10.1.2.1.), 17 but absolutely in vain. When this type of material began to be produced in the bay area, Sa Caleta had already been completely abandoned by the Phoenicians. For these reasons, it is legitimate to propose the hypothesis of a Phoenician settlement at the Bay of Ibiza no earlier than the first quarter of the 6th century, but with the understanding that it was an intense and organized occupation motivated by reasons that we will attempt to analyze below.

4. Mechanisms and Process 4.1. The Phases and Nature of the Colonization 4.1.1. The Settlement at Sa Caleta and Its Context At the site of Sa Caleta all the wheel-made pottery and surely a significant percentage of the handmade pottery is imported. This fact is one of the most significant pieces of information because it shows the fundamentally “intermediate” nature of a great part of the economic activity verified in this settlement. Just as important was the handling of great quantities of natural lead sulfide by these Phoenicians. It is possible that the ore was brought (all or in part) from the known mines at s’Argentera, which were undoubtedly exploited during later periods. However, the location of these mines on the northeastern coast of the island, almost at the opposite geographical extreme, makes it more likely (even though the different possibilities are not mutually exclusive) that the lead sulfide came from a point on the Iberian peninsula not far from the eastern coast and/or from Catalonia. It is precisely this fact that gives us the decisive clue in favor of Ibiza’s function as an “intermediary.” When sailing conditions were favorable, it is logical to assume that traders from Sa Caleta traveled frequently to certain Phoenician enclaves on the southern Iberian peninsula. There they probably acquired the items that were typical of their trade: wheel-made pottery, cloth, and prestige goods such as wine, which probably was transported in amphorae T-10.1.1.1. and T-10.1.2.1. These amphorae are very frequent at Sa Caleta and in a multitude of preIberian sites on the Spanish coast, whose continuously-broadening relationships are well known. From the southern Iberian coast the traders were able to sail directly toward the northeastern coast of Spain in order to exchange some 17. J. Ramón, “Las ánforas púnicas de Ibiza,” Trabajos del Museo Arqueológico de Ibiza 23 (Ibiza, 1991) 100–102.

The Ancient Colonization of Ibiza: Mechanisms and Process

149

Fig. 11. Phoenician pottery of proto-Ibizan fabrication: (1) tripod, deposit “PM NE-83”; (2) amphora T-10.1.2.1. (= PE-10), coast of the Pitiuses (islands of Ibiza and Formentera).

150

Juan Ramón

of these products with the indigenous peoples. Or perhaps they returned first to Sa Caleta, and from there they undertook the voyage to the eastern Iberian coast and to Catalonia. In any case, after these traders had left the merchandise in the hands of diverse indigenous buyers, they returned to Sa Caleta with loads of metal that they had obtained in exchange, including lead sulfide, and in this way they completed the circuit. Leaving aside the different implications of contacts that Sa Caleta had with other western Phoenician enclaves, including those in Morocco and Algeria, we must emphasize the fundamental importance of contacts with the central Mediterranean region. A large quantity of ovoid amphorae from the Punic central Mediterranean exists at Sa Caleta, which highlights the very significant trade connections with areas farther east. These economic relationships between Sa Caleta and the East must have been direct; it is precisely the site at Sa Caleta that is closest to the area in question. Sailing conditions from Ibiza to Sardinia, for example, are highly favorable. On the other hand, the internal economy of the settlement at Sa Caleta can be deduced from the abundant presence of bronze fish hooks. Fish must have been an important source of food. The frequency of cereal grinders is also significant. Together with other items such as a large oven surely used to bake bread in the small plaza “j” of the southern sector, they give evidence for a production of this important food for internal consumption. In spite of the hand grinders, it is still unclear to what extent the people of Sa Caleta grew wheat on the plain just to the north or to what extent they purchased this product. This question also affects the matter of domestic animals that appear among the fauna found in several dumps at the site. We should also point out that looms have been found in some of the rooms. The looms used weights made from handles of amphorae (T-10.1.2.1.) that had been cut and filed down. In room XX, separate from a group of these weights, two spinning whorls were found. Many rooms were designated almost entirely for the smelting of argentiferous lead sulfide with the purpose of obtaining ingots of lead. Proof of such smelting, performed at low temperatures and possibly in handmade ceramic containers, appears in a considerable number of workshop environments. It was evidently a generalized activity. Thus, we believe that the Phoenicians of Sa Caleta must have resold the already smelted lead in the southern Iberian Peninsula, and possibly also in Sardinia and other Phoenician areas of the central Mediterranean. However, the smelting and perhaps the forging of iron that so far appears only in room XV indicates that the artisans possibly produced instruments and weapons for personal use by the inhabitants of the town, rather than commercially. Moreover, it appears that the iron smelted in room XV comes from iron nodules that are found abundantly in their natural state in the area, including on the peninsula of Sa Caleta itself.

spread is 1 pica long

The Ancient Colonization of Ibiza: Mechanisms and Process

151

The town of Sa Caleta had a surface area of about 4 hectares, about the same as the majority of Phoenician towns in the southern Iberian Peninsula. During the period of its greatest size, several hundred people must have lived there. This is what we know about the site so far. Sa Caleta can be considered the key that explains the beginning of the Phoenician presence on Ibiza, its nature, and its function. Its founding probably initiated the authentic first phase of the ancient colonization of the island, in contrast to certain hypotheses that have tried to see in the occupation of Ibiza the establishment of a bridge for Phoenician east–west navigation in the Mediterranean from the 8th century b.c. If this were true, the explanation would be different—that the first Semitic presence in Ibiza was the work of a large group of adventurer-traders looking for new sources of metals to make their fortune. In reality, the economy of Sa Caleta appears to have been in place throughout its entire existence, a period of several decades, and it was fully integrated with the economy of its place of origin, the central Andalusian coast and/or several sites in the Far West. Sa Caleta’s founders must have left the Andalusian area with the vision of new sources of goods on the eastern coast of Spain, but also in the central Mediterranean—aside from the island of Ibiza itself. The settlement at Sa Caleta appears to firmly demonstrate that during the third (or final?) quarter of the 7th century, the basic transactions were the purchase of foreign materials and the sale of far-western products in return. These transactions were not exempt from some intermediate handling by these trading agents. One final aspect deserves mention: the settlement at Sa Caleta covered at least 4 hectares, a considerable surface area. After lengthy surveys of an island coast which is really not very long, we find no signs of other parallel, contemporary Phoenician settlements on Ibiza, apart from the special case on the Bay of Ibiza, which we will discuss below. This is evidence for a concentrated system, in contrast to what in theory could have been a dispersion into several smaller nuclei. At Sa Caleta we find a large, unified human mass, capable of planning significant actions, such as its organized move to another site on the island.

4.1.2. The Settlement on the Bay of Ibiza and the Founding of the City When we consider the nature of the first occupation of the Bay of Ibiza, it becomes evident that we cannot continue speaking of a mere port that served as a stopover for expeditions with no interest in the bay except as a place for taking on provisions. On the contrary, key factors must have been considered when this settlement was founded, such as the obvious suitability of the topography. Not only was there an excellent port, but the surrounding territory also had high agricultural potential. Also, the splendid hill (Puig de Vila) that dominates the whole southern side of the bay provided an extraordinary site for the construction of a strong maritime fortress.

152

Juan Ramón

The discoveries around this great harbor highlight the fact that already before the middle of the 6th century b.c., and practically since its very beginning, this site was not a tiny settlement, but rather a stable colony, a city or proto-city conceived as such. The occupation of the cliffs to the southwest of the port near the northeastern slope of Puig des Molins and of places such as the point of Joan Tur Esquerrer, the construction of a possible center for defense and/or worship on the high point of the hill, the ancient remains from the sector of Santa Lucía, and the establishment of a sanctuary on the neighboring island of Illa Plana are factors that do not permit us to continue believing that Ibiza was only a small “trading post,” a simple bridge for Phoenician sailors, during the second half of the 7th century and the first half of the 6th. The almost immediate functioning of pottery workshops that produced transport amphorae in the style of the Phoenicians of southern Spain (T-10.1.2.1. or PE10), along with other types of vessels that were clearly western Phoenician, underscores this point. Why did this “urban” occupation of the bay take place? Without doubt, it is the complicated result of several factors that came together. The settlement on the bay is the true second phase of the ancient colonization of the island and should be seen as the consequence and end result of the previous phase. The significant self support possible on this island (with agricultural, marine, and possibly metal resources) permitted enormous economic savings and additional sources for work. This provided an economic base for a larger number of Phoenician immigrants and surely other Semitic and non-Semitic peoples, all of whom added to the dynamic of internal growth. Possibly only the Bay of Ibiza could meet the demands that were made on the Phoenicians only a few decades after the founding of Sa Caleta. It is also possible to think of other factors. If it is true that the Phocaean Greeks traded with “Tartessos” from the final years of the 7th century b.c., their commercial navigation by the late first quarter of the 6th century would have been considerable, contributing to the substitution of Sa Caleta by another site of greater strength and resources that were as much economic as defensive. The archaeological data also demonstrate that contacts with the central Mediterranean were on the increase, and the port on the Bay of Ibiza was much more suitable, topographically and geographically, for trade. Thus, it appears obvious that the ancient city of Ibiza was planned in advance as a center that would look toward the future. The city functioned as a port for maritime expeditions to the southern, eastern, and northern coasts of the Iberian Peninsula, Africa, and the islands of the central Mediterranean. Within the framework of the island itself, it took advantage of the splendid topographical and economic realities found around the bay. Surely we are looking at an exigency that arose during the first third of the 6th century: the need for a “historical city” at a key point in the western Mediterranean, such as on the island of Ibiza.

page is 1 pica long

The Ancient Phoenicians of the 8th and 7th Centuries b.c. in the Bay of Cádiz: State of the Research Diego Ruiz Mata

I. The Founding of Cádiz According to ancient historians, the first Phoenician colony in the far western Mediterranean was Gadir (the present-day city of Cádiz), founded around 1104/3 b.c. Strabo, writing during the time of Augustus, recounts this event with the following words: Concerning the founding of Gadeira, this is what the Gaditanians say they remember. A certain oracle commanded the Tyrians [inhabitants of the city of Tyre] to found a settlement at the columns of Hercules. Those who were sent to explore arrived at the strait next to Kalpe, and they believed that the promontories forming the strait were the ends of the inhabited earth and the end of Hercules’ labors. Supposing, therefore, that these were the columns about which the oracle had spoken, they anchored in a certain place on this side of the columns, where the city of the Exitani stands today [the present-day Almuñécar]. However, since they offered a sacrifice to the gods at this spot on the coast and the victims were not propitious, they returned. Later, the envoys crossed through the strait and arrived at an island consecrated to Hercules, located near Onoba [Huelva], an Iberian city about a thousand five hundred stadia distant [about 270 km.]. They believed that the columns of Hercules were there, so they again sacrificed to the gods. However, the victims were unfavorable again, and they returned to their homeland. On the third expedition they founded Gadeira and erected the sanctuary in the eastern part of the island and the city in the western part (3.5.5).

Editor’s note: This article appeared in Spanish under the title “Los fenicios de época arcaica—siglos VIII/VII a.C.—en la bahía de Cádiz. Estado de la cuestión,” in Estudos Orientais IV (Lisbon: Instituto Oriental, 1993) 23–72.

155

156

Diego Ruiz Mata

According to ancient belief, the founding of Gadir occurred around 1104 b.c. Velleius Paterculus, a Roman historian of the 1st century, places the date for the founding of Gadir some eighty years after the fall of Troy: “During this epoch, the Tyrian fleet, which controlled the seas, founded Gadir, at the end of Spain and at the end of the earth, on an island surrounded by the ocean, separated from the continent by a very narrow strait. A few years later, the same people founded Utica” (Historia Romanorum 1.2.3, around 30 a.d.). According to Timaeus, a Greek author of the 3d century b.c. who used Phoenician or Carthaginian sources, the founding of Utica occurred around 1101 b.c. Later, Carthage was founded in 814. Strabo, who places the founding of Gadir a little after the Trojan War, and Mela, who was born in the little town of Tingentera near Cádiz, also agree on this date. Mela also affirms that the temple of Melkart (on Sancti Petri, Chiclana) was built by the Tyrians and that its years were counted beginning with the Trojan War: “On the side of the mainland, it is almost straight [referring to Gades]. On the side that faces the sea, it rises and forms a curve in the middle of the coast, ending in two promontories. On one of these is a flourishing city with the same name as the island, and on the other a temple of Hercules Aegyptius, renowned for its founders, for its veneration, for its antiquity, and for its riches. It was constructed by the Tyrians, and its sanctity arises from the fact that it guards the ashes [of Hercules]; its age is counted from the Trojan War” (3.6.46). The archaeological record suggests that the Phoenician interest in the Iberian Peninsula was the control and trade of its mineral resources. In this regard, Diodorus of Sicily, a contemporary of Augustus, states that “the Phoenicians since ancient times founded many colonies in Libya, as well as several in the regions of Europe that extend toward the west. Because these commercial enterprises developed according to their plans, they accumulated great riches and began to sail along the part that is beyond the columns of Hercules, which they call the ocean. First they founded a city in Europe at the strait with the columns. They called the city Gadeira, because it was a peninsula. On the peninsula they put everything that was suitable to the nature of the place, as well as a splendid temple dedicated to Hercules, introducing magnificent sacrifices conducted in the Phoenician manner” (25.10.1). With this information, which the archaeological record does not contradict, we can conclude the following: (1) the primary motive of Phoenician colonization was to obtain metals, gold, tin, and especially silver, that were later traded in the Mediterranean and the Near East; (2) in this context, Gadir was the western metropolis that channeled the trade; (3) the most controversial point is the date of the founding of Gadir, around 1104 b.c. according to the ancient sources, and in the 8th century according to the archaeological data. According to Strabo’s account, the founding of Cádiz appears to be the end result of initial exploratory navigation. The first attempts at settlement took

The Ancient Phoenicians in the Bay of Cádiz

The Ancient Phoenicians in the Bay of Cádiz

157

place at Sexi (Almuñécar) and later at Onoba (Huelva), beyond the columns of Hercules. The founding of Gadir (Cádiz), a name which means “fortress” or “walled enclosure,” occurred on the third trip. If we consult the map to study the coast along which the Phoenicians sailed, it is evident that before the founding of Gadir they conducted a systematic exploration from Almería to Huelva, searching out possibilities for trade and settlement. On this stretch of coast they finally chose Cádiz, an island near the coast and near the estuaries of the Guadalete and Guadalquivir Rivers. At that time the Guadalquivir emptied into the ocean a little south of the present city of Seville, and it had a wide estuary whose banks were between El Rocío and Sanlúcar de Barrameda. These banks, especially the eastern one, were densely inhabited by indigenous peoples of the Final Bronze Age. The Guadalquivir could be navigated almost to Córdoba, and it was close to fertile agricultural zones and the mining region of Aznalcóllar, where silver was mined at that time. Gadir, therefore, controlled access to the Guadalquivir and to Huelva, which was home to an indigenous population that must have dominated the mineral resources of the Riotinto region, another point of extraordinary interest for mining. The excellent strategic location of the island of Gadir was responsible for its metropolitan nature and its development as a trading center. Ships embarked from its port carrying products directed toward the Mediterranean markets. In addition, Gadir was the political center which represented the Tyrian state, and the other western colonies revolved around it. The date for the founding of Gadir is still a problem, with a three hundred year discrepancy between the date given by the ancient historians and the archaeological record. Some writers support the accuracy of the 1104 date that was transmitted by Timaeus and accepted by Velleius and Pliny, although this date has little archaeological backing. A. M. Bisi believes that the origins of Phoenician colonization in the West go back to the end of the 2d millennium, even though there are no findings earlier than the 8th century. 1 P. Cintas and J. M. Blázquez have dated a bronze figurine, known as the Priest of Cádiz (which we now know can be identified with Ptah), to the second half of the 2d millennium, proving, according to them, the accuracy of the earlier date. However, D. Harden dates the same figurine to the 5th century b.c., demonstrating the weak arguments used to back up the ancient sources. According to D. Collon, another bronze figurine from Medina de las Torres in the province of 1. A. M. Bisi, “Fenici e Micenei in Sicilia nella seconda metá del II millenio a.C.,” I Congresso Internazionale di Micenologia (Rome, 1967) 1156ff.; idem, “Le ‘Smiting God’ dans les milieux phéniciens d’Occident,” Studia Phoenicia IV (1986) 169ff.; see also S. Moscati, “Precolonizzazione greca e precolonizzazione fenicia,” RSF 11 (1983) 1ff. For the state of the research, M. E. Aubet Semmler, Tiro y las colonias fenicias de Occidente (Barcelona: Bellaterra, 1987) (chap. 7, “Los fenicios en Occidente: Cronología e historiografía,” 175ff.) and J. M. Blázquez, Tartessos y los orígenes de la colonización fenicia en Occidente, 2d ed. (Salamanca, 1975).

158

Diego Ruiz Mata

Badajoz appears to represent the god Hadad. He finds a resemblance to prototypes from the 2d millennium, whereas Blázquez finds similarities to ancient Greek bronzes of the 7th century b.c. Another sitting figurine of bronze known as the Astarte of El Carambolo has been dated to around 800 b.c., especially due to the inscription on the footstool which supports her feet, thus proving the existence of a cult to Astarte in the lower Guadalquivir. As noted, the arguments are weak, and the dates for these materials oscillate. Excavations performed since 1962 in the Phoenician settlements of Málaga have provided consistent support for the founding dates of these Mediterranean colonies. The oldest date is for the town of Morro de Mezquitilla, which, in the opinion of its excavator, was inhabited “from about the year 750 b.c. or a little before.” 2 He bases his conclusion on a sealed find on the floor of a room in Construction Complex K, consisting of a plate with a narrow rim (16 mm.), two carinated bowls, the mouth of an amphora, a handmade pot, and a mushroom-mouth jug, all found in close association and dated to the 8th century. The nearby town of Chorreras 3 and the Phoenician colony of Sexi at Almuñécar 4 may date to the same period. In other words, the Phoenician implantation on the Mediterranean coast took place at the middle of the 8th century or a little earlier. The necropolises that have been excavated in Carthage and Utica have not provided materials earlier than the 7th century. At the sanctuary of Tanit in Carthage, the oldest level of occupation appears to belong to the 8th century, 5 the same date that the recent German excavations in the ancient city have provided. 6 The same dating occurs in Sicily, where, according to Thucydides, the Phoenician colonization followed the Greek one. The Phoenician finds at Motya (in Sicily) and those at San Antioco in Sardinia are dated to the end of the 8th century.

2. H. Schubart, “El asentamiento fenicio del siglo VIII a.C. en el Morro de Mezquitilla (Algarrobo, Málaga),” Aula Orientalis 3 (1986) 59ff., especially pp. 69 and 79 with discussion of the chronology (this article appears in English in the present volume); H. G. Niemeyer, “La cronología de Toscanos y de los yacimientos fenicios en las costas del sur de la Península Ibérica,” I Congresso Internazionale di Studi Fenici e Punici (Rome, 1983) 633ff. 3. M. E. Aubet Semmler, “Los fenicios en España: estado de la cuestión y perspectivas,” Aula Orientalis 3 (1986) 9ff.; idem, “Excavaciones en las Chorreras (Mezquitilla, Málaga). Campaña 74,” Pyrenae 10 (1974) 79ff.; M. E. Aubet Semmler et al. “Chorreras. Eine phönizische Niederlassung östlich der Algarrobo-Mündung,” MM 16 (1975) 137ff.; J. M. J. Gran Aymerich, “Excavaciones arqueológicas en la región de Vélez-Málaga. Campaña 1973,” NAHisp 12 (1981) 299ff. 4. F. Molina Fajardo, “Almuñécar a la luz de los nuevos hallazgos fenicios,” Aula Orientalis 3 (1986) 193ff., including bibliography and work done recently; M. Pellicer, “Sexi fenicia y púnica,” Aula Orientalis 3 (1985) 85ff. [[Reprinted in English translation here, pp. 49–77.]] 5. P. Cintas, Manuel d’Archéologie punique (2 vols.; Paris 1970 and 1976); D. Harden, “The Pottery of the Precinct of Tanit at Salambó, Carthage,” Iraq IV (1937) 59ff. 6. M. Vegas, “Archaische Keramik aus Karthago,” MDAI Römische Abteilung (1984) 215ff.

Spread is 6 points long

The Ancient Phoenicians in the Bay of Cádiz

159

As noted, archaeological research has not yet been able to document the date of 1104 for the founding of Gadir, and, consequently, for the beginning of Phoenician colonization in the southern Peninsula. It can be claimed that the discrepancy between the mythical date and the archaeological record is a consequence of the very process of colonization, which probably began with intermittent contacts for trading purposes only, before the founding of any urban settlement. These contacts would not have left any perceptible traces. If we accept this hypothesis, it is probable that the practice followed was similar to the one recounted by Herodotus in a passage about the Punic Phoenicians of North Africa: The Carthaginians recount another story. In Libya, beyond the columns of Hercules, there is a certain populated place where they [the Carthaginians] often appear and carry their merchandise to land. They leave it there on the seashore and reembark, giving a smoke signal from their boats that they have arrived. As soon as the people of the country see it, they come to the shore, leave gold next to the merchandise, and depart again toward the interior of the land. Then the Carthaginians return to the shore to look at the gold, and if they think that what has been left is a fair price for the merchandise, they pick up the gold and leave. However, if it does not seem enough, they board their boats again and wait. When the natives see this, they come back to add more gold until they have added enough to satisfy them. It is understood that the first party does not touch the gold until a fair price has been reached for their wares, nor does the second party touch the merchandise until the gold has been carried away. (4.196)

This information does not contradict other data obtained from several points in the Mediterranean near the eastern Phoenician cities. As J. N. Coldstream has pointed out, 7 the Phoenicians were present in the Aegean long before the Greek founding of Al-Mina. Traces of this presence are found in Athens, Euboea, Crete, and Kos, demonstrated by craftwork of a Phoenician nature exhumed from several necropolises at these places and dated to the middle of the 9th century by the association of the Phoenician objects and pottery with Greek geometric pottery. These are the first traces of contacts between Phoenicians and other Mediterranean peoples. The reason for these contacts, at least at Athens, could have been the exploitation of the silver mines at Laurion, which were worked during this period, like those near Thorikos. Around this same time, during the second half of the 9th century, Kition was founded on Cyprus, and by the end of the century the Phoenicians were firmly established there. 8 In summary, the earliest Phoenician foundings and vestiges are located at points near the eastern metropolises, beginning in 7. J. N. Coldstream, “Greeks and Phoenicians in the Aegean,” in Phönizier im Westen (MB 8) (1982) 261ff. 8. V. Karageorghis, “New Phoenician Discoveries in Cyprus,” I Congresso Internazionale di Studi Fenici e Punici (Rome, 1983) 173ff.

160

Diego Ruiz Mata

the middle of the 9th century, and later the Phoenicians sailed toward the central and western Mediterranean. In regard to the Phoenician presence in the Bay of Cádiz and the founding of Gadir, we believe that this event was related to the dynamics of the Phoenician expansion and colonization in the Mediterranean and was its most western consequence. It must be understood within an economic and political context that is complex and broad, not as an isolated and accidental event. It was part of a planned expansion, with stages that were probably not far apart in time. We must keep in mind the dating of the Phoenician evidence in the central Mediterranean, which logically explains and gives significance to what we find in the West. Given the Tyrian nature of the founding of Gadir, as indicated by the texts, it is logical to assume that similarities between ceramic types from Gadir and from Tyre should definitively clarify the dating problem. Using data from the excavations at Tyre, P. M. Bikai suggests that the western Phoenician pottery corresponds to that coming from Strata III and II in Tyre. In other words, the Phoenician colonization should be dated to the second half of the 8th century b.c. 9 However, the proposed date for the strata of the founding of Morro de Mezquitilla is around 750 b.c., or a little earlier. Thus, it is still difficult to establish an exact date for the initiation of these western foundings, due to the fact that the dating of Phoenician tableware has not been fixed with the precision that research demands.

II. Phoenician Settlements on the Spanish Coast (Fig. 1) A permanent and considerably dense Phoenician population on the Iberian coast stretched from the east of Gibraltar all the way to Almería from the 8th century on. The existence of the following 8th-century colonies has been verified: Toscanos, Chorreras, Morro de Mezquitilla, and Guadalhorce on the coasts of the province of Málaga, Adra in Almería, and Sexi on the coast of Granada. The colonies in Málaga were established on virgin soil, in a region sparsely inhabited by indigenous populations. However, at Almuñécar (Sexi) and Adra we see the existence of an indigenous population. It seems that Morro de Mezquitilla is the earliest founding, in the first half of the 8th century. Chorreras was founded toward the middle of this century, and Toscanos was founded in the second half. The remaining sites (Almuñécar, Adra, and Guadalhorce) are also dated to the 8th century.

9. P. M. Bikai, “The Late Phoenician Pottery Complex and Chronology,” BASOR 229 (1978) 47ff.

The Ancient Phoenicians in the Bay of Cádiz

161

Fig. 1. Distribution of the Phoenician colonies ∑ ; indigenous towns „ ; and metallurgic sites + „.

In marked contrast, the coast to the west of Gibraltar has not provided many colonies, except for the one at the mouth of the Guadarranque River (Cerro del Prado), the metropolis of Gadir, and Castillo de Doña Blanca. Probably there was also a Phoenician settlement at Huelva. In the interior of the region of the lower Guadalquivir, we find El Carambolo and Carmona. With the exception of Chorreras, which was abandoned at the beginning of the 7th century b.c., the majority of the ancient Phoenician settlements lasted into the 6th century. Beginning around 700 b.c. the political and economic importance of these colonies and trading posts peaked. This promoted a notable demographic growth on the coast between Villaricos and Gadir and an enrichment of certain sectors of society, as suggested by the splendid necropolises at Almuñécar (the necropolis Laurita) and at Trayamar. Also during the 7th century, Phoenician expansion took place throughout northern Africa, the Portuguese coast from Lisbon south (according to the recent data of this Colloquium), the southeastern Peninsula, and the lower Guadalquivir. In the opinion of M. E. Aubet Semmler, 10 the establishment of the Phoenician colonies can be explained in relation to Gadir. They served as strategic 10. M. E. Aubet Semmler, Tiro y las colonias fenicias.

162

Diego Ruiz Mata

Fig. 2. Ancient topography of the Gaditanian islands and the ancient bay during the Phoenician period (according to J. L. Escacena).

centers of support for navigation and for the control of access to the metals of the Tartessian region, meaning that the settlements were defined by their economic purpose. The problem is that the abundance of these trading posts on

The Ancient Phoenicians in the Bay of Cádiz

163

the coast of Málaga, located on the estuaries of rivers, is more suggestive of direct trade with the interior than of a simple stop on the way to Tartessos. In summary, it appears certain that the 8th century was the crucial period for determining the areas of economic influence in the central and western Mediterranean by the Phoenicians and Greeks. The towns of Morro de Mezquitilla, Chorreras, Toscanos, and Guadalhorce (province of Málaga), Gadir and Castillo de Doña Blanca (province of Cádiz), and Adra and Sexi prove the existence of a Phoenician population in the area of the Strait of Gibraltar at a very early date, perhaps the first half of the 8th century. From this point on, Phoenician populations controlled access to the Atlantic, which meant that the metal trade was in Phoenician hands. Gadir, the political and economic center, controlled all of this commercial traffic.

III. The Metropolis of Gadir The island of Cádiz has not yet provided early remains, but part of the problem is that we do not know the exact site of the Phoenician city. In spite of the paucity of ancient descriptions, we know that Gadir was plural for the ancient Greco-Roman authors (Strabo 3.5.3, Pliny119–20, and Mela 3.6.46). The ancient writers refer to the Gaditanian islands and consider the bay to be a partially inhabited archipelago. It was separated from the coast by a small arm of sea and had promontories on each end, occupied by the city and the temple. It was a little more than 18 kilometers long and at times narrowed to 200 meters. These data are fairly consistent with the present situation. Except for precise details, we see today the reflection of the ancient descriptions. The Phoenicians chose an island, as they had done at Tyre, and as at Tyre, they occupied two promontories. However, we need not search for a large area for the location of the ancient city, since it probably did not exceed 7/8 hectares. For this reason it is difficult to determine the exact location of the Phoenician city. Also, it lies buried under Roman structures and the more recent city, which prevents intensive surveys for this purpose (fig. 2). If we wish to locate the Phoenician city, we need to clarify some of the ancient topographic features. These do not correspond completely to the present topography. Through Pliny we know of the existence of two islands: Erytheia, the smaller island, where the Phoenician city was located, and Kotinoussa, larger in size and elongated, with the temple of Melkart at its end. The two islands were separated by a channel, which is presently filled in. The mining engineer Juan Gavala y Laborde detected the channel in 1926 11 and suggested that it was an ancient mouth of the Guadalete River. F. Ponce Cordones and 11. J. Gavala y Laborde, “Cádiz y su bahía en el transcurso de los templos geológicos,” Boletín del Instituto Geológico de España 49 (1927).

164

Diego Ruiz Mata

other authors 12 have related this channel to the ancient description. It formerly joined the bay to the ocean by way of Puerto Chico in Campo del Sur, and in recent soundings during salvage excavations, its existence has been confirmed. The channel could have had a width of 150 to 200 meters and a depth of up to 9 meters, with sloping banks. 13 Also the port may have been built here (fig. 3). Even if we assume that the texts and the proposed ancient topography are accurate, we still do not know the exact location of the Phoenician city. The smaller island (Erytheia) took up nearly the whole old city center of Cádiz, and the Phoenician town may have been located at the Torre de Tavira, which is the highest hill. 14 From the surrounding area came the so-called Priest of Cádiz (identified as Ptah), found 5 meters under the present surface in 1928, as well as a structure of ashlars another 5 meters lower down. 15 However, excavations have not yet been performed in this extremely interesting area, which may be the site of the Phoenician Gadir of ancient times. Others believe, however, that the Phoenician Gadir is submerged underwater, and they place it near the neighborhoods of Pópulo and Santa María and the narrow strip of land that joins them to the Castle of San Sebastián. 16 However, there have been no finds under the water or on the existing dry land to support this hypothesis. Thus the island of Gadir has not yet provided archaeological remains from its most ancient past. However, we can sketch its cultural development in very broad strokes as a result of the present salvage excavations. During the decade of the 60s, C. Peman lamented that “incessant building projects constantly uncover vestiges of the past wherever they dig, only to cover them up immediately with new construction.” Up until the present day, the urban rhythm of the city of Cádiz has necessitated a considerable number of salvage excavations, especially in the modern city center. These excavations have provided important data for outlining the cultural development of the island. The first occupation took place during the Chalcolithic period, perhaps at an advanced phase. Excavations during 1986 on a plot of land on the street Ciudad de Santander (at the corner of Avenida de Andalucía) uncovered a level on top of virgin soil, which has provided abundant ceramics from that 12. F. Ponce Cordones, “Consideraciones en torno a la ubicación del Cádiz fenicio,” Anuario de la Universidad de Cádiz 2 (1985). 13. L. Perdigones and A. Muñoz, “Excavaciones de urgencia en un solar de la calle Regimiento de Infantería esquina Abreu (Cádiz),” AAA III (1986) 45ff. 14. J. L. Escacena, “Gadir,” in Los fenicios en la Península Ibérica (Sabadell: AUSA, 1986) 39ff. 15. J. L. Escacena, “Gadir,” p. 43; D. Harden, The Phoenicians (London, 1962) 147, 204. 16. R. Corzo, “Paleotopografía de la bahía gaditana,” Gades 5 (1980) 5ff.; idem, Los fenicios, señores del mar (Historia del Viejo Mundo no 8, Historia 16; Madrid, 1988) 111–12.

The Ancient Phoenicians in the Bay of Cádiz

165

Fig. 3. Reconstruction of the ancient channel Bahía-Caleta and the location of the ancient city at the Torre of Tavira (according to J. L. Escacena).

period. 17 Materials from the Chalcolithic and Middle Bronze Ages have been found on another lot on the street Dr. Marañón. These materials were found 4 meters down and underneath tombs of ashlars from the Punic period, in strata filled with sand from the dunes. 18 It appears that a later phase of the Final Bronze has been located on Santander Street (mentioned above), with remains that come from the interior of a well 19 and from the area of the Torre de Tavira, 20 where we assume the ancient Phoenician city was located. We still have not found early Phoenician strata from the 8th century. However, during the past few years, Phoenician burials from the end of the 7th and beginning of the 6th centuries b.c. have been exhumed from lots on the streets of Ciudad de Santander and Tolosa Latour. These burials contained cremations in rectangular pits dug into the ground. 21 They are the only Phoenician testimonies from an already advanced period. 17. L. Perdigones, A. Muñoz, and A. Troya, “Excavaciones de urgencia en un solar de la calle Ciudad de Santander esquina Avda. Andalucía (Cádiz),” AAA III (1986) 41ff. 18. L. Perdigones and A. Muñoz, “Excavaciones de urgencia en un solar de la calle Doctor Gregorio Marañón (Cádiz),” AAA III (1985) 55ff. 19. L. Perdigones et al., “Excavaciones . . . calle Ciudad de Santander esquina Avda. Andalucía.” 20. R. Corzo, “Paleotopografía de la bahía gaditana.” 21. L. Perdigones, F. Blanco, and A. Muñoz, “Excavaciones de urgencia en un solar de la calle Ciudad de Santander esquina Brunete (Cádiz) en 1985,” AAA III (1985) 53ff.; L. Perdigones and

166

Diego Ruiz Mata

During the past century a grouping of cist tombs in Punta de Vaca was discovered. In one of these tombs was an anthropoid sarcophagus of a male, dated to around 400 b.c. 22 The later excavations of P. Quintero, beginning in 1914, and of F. Cervera 23 discovered a sector of the necropolis of the 5th–3d centuries b.c. Recently, the salvage excavations performed by a team of archaeologists from the Provincial Office of Culture have documented three necropolises from the Punic period on the streets of Asdrúbal, Tolosa Latour, and Avenida de Andalucía, with burials consisting of ashlar cists and pit graves. 24 This is practically all we have from this period on the ancient island of Cádiz. The descriptions of the island of Cádiz always mention that the temple of Melkart was located on the eastern end and that it was founded at the same time as the city. Strabo writes that “the city lies in the western part of the island, and nearby, on the end that goes toward the islet, rises the Kronion. The Herakleion [or temple of Melkart] is in the other direction, toward the east, in the place where the island approaches the coast, where it is separated only by a channel 1 stadium wide. They say, moreover, that the city is 12 miles from the sanctuary. . . .” (3.5.3). The description given by P. Mela is similar (3.6.46). A complete study of the temple was done by García y Bellido, 25 who has placed it on the island of Sancti Petri, to the southeast of Cádiz and some 18 kilometers from the city, as the sources indicate. The sanctuary was built upon a promontory, which today is nearly submerged and which Philostratus in the 3d century a.d. saw circled by water as an island, thus very similar to the present topography. The most abundant finds from this spot are from the Roman period, but a few years ago five bronze statuettes representing eastern A. Muñoz, “Excavaciones arqueológicas de urgencia en un solar de la calle Tolosa Latour, extramuros de Cádiz,” AAA III (1987) 59ff. 22. M. L. Ramos, Estudio sobre el ritual funerario en las necrópolis fenicias y púnicas de la Península Ibérica (Madrid: Ediciones de la Universidad Autónoma de Madrid, 1986) 24, 25. 23. P. Quintero, “Necrópolis Anterromana de Cádiz,” BRAH 22 (1914) 81ff.; idem, “Excavaciones en Punta de Vaca (Cádiz),” en MJSEA 5 (Madrid, 1915); idem, “Excavaciones en Punta de Vaca y en Puerta de Tierra (ciudad de Cádiz),” MJSEA 12 (1916); idem, “Excavaciones en extramuros de la ciudad de Cádiz,” MJSEA 26 (1918); MJSEA 30 (1920); MJSEA 76 (1926); MJSEA 84 (1926); MJSEA 95 (1928); idem, “Excavaciones en Cádiz,” MJSEA 99 (1929); MJSEA 117 (1932); MJSEA 122 (1933); MJSEA 129 (1934); MJSEA 134 (1935); Cervera y Jiménez, “Excavaciones en extramuros de Cádiz,” MJSEA 57 (1923). 24. L. Perdigones and A. Muñoz, “Excavaciones de urgencia en un solar de la Plaza de Asdrúbal (Cádiz 1985),” AAA III (1985) 58ff.; L. Perdigones and R. Baliña: “Excavaciones de urgencia en un solar de la calle Tolosa Latour (Cádiz) en 1985,” AAA III (1985) 63ff.; L. Perdigones, A. Muñoz, and A. Marcos, “Excavaciones de urgencia en un solar de la Avda. Andalucía esquina plaza de Asdrúbal (Cádiz),” AAA III (1986) 38ff.; L. Perdigones and A. Muñoz: “Excavaciones arqueológicas de urgencia en un solar de la calle Tolosa Latour. Extramuros de Cádiz,” AAA III (1987) 59ff. 25. A. García y Bellido, “Hercules Gaditanus,” Archivo Español de Arqueología 36 (1964) 70ff.

The Ancient Phoenicians in the Bay of Cádiz

167

deities were dredged up from the area surrounding the little island and have been dated to the 8th-7th centuries b.c. For the moment they constitute the oldest remains from the island of Cádiz. 26 Even when we consider all these data together, it is evident that they cannot explain in detail the characteristics of the ancient city, its cultural development, its political status, and its economic activity. Nevertheless, they demonstrate the importance that the city must have had, and they agree with GrecoRoman descriptions.

IV. The Ancient Geography The coastline that we see today, as well as certain regions in the interior, has changed notably from prehistoric times to the present day. Thus, if the Phoenicians or natives from the 1st or 2d millennium b.c. could sail along the coast today, they would hardly recognize many of the topographic reference points. If we want to understand the pattern of orientalizing settlement and its economic and commercial structure, the reconstruction of the landscape is a fundamental task in this area which has changed so much in so little time. In summary, the most significant aspects are the following (see fig. 4): (1) The mouth of the Guadalquivir River was located further inland in the vicinity of Coria del Río. The river emptied into an extensive estuary and later, in Roman times, into an inland lake (Lake Ligustinus). In protohistoric times, the estuary was bounded by Matalascañas and Sanlúcar de Barrameda. The map of this ancient geography is reconstructed in fig. 4, and on it we note the borders of the present day marshland, occupied earlier by the lake mentioned above. The marshland stretches toward the north in what must have been the ancient mouth of the Guadalquivir River, which flowed along the eastern slope of the Aljarafe plateau, as can be deduced from the concentration of settlement in this area. The coast along the eastern side of the lake, as well as the Atlantic coast by Cádiz at the mouth of the Guadalete River, is clearly defined by the tertiary terrain in which numerous pre-Roman and Roman towns are found. The western shore of the lake is bounded by diluvial terrain and marked by a smaller number of protohistoric settlements. The navigability of the lake and its importance as a route for penetrating into the interior are demonstrated by the pattern of settlement on the shores, with numerous important towns between Lebrija and Sanlúcar de Barrameda. (2) Later, perhaps during the first half of the 1st millennium b.c., the entrance to the lake began to silt up with a shoreline of dunes, where the nature preserve of Doñana is located today. This is where A. Schulten placed the city

26. A. Blanco, “Los nuevos bronces de Sancti Petri,” BRAH 182.2 (1985) 207ff.

168

Diego Ruiz Mata

Fig. 4. Reconstruction of the coast at the end of the 1st millennium b.c. (according to Gavala y Laborde).

The Ancient Phoenicians in the Bay of Cádiz

169

of Tartessos, in spite of the fact that the archaeological work performed at Cerro del Trigo did not produce results. (3) Navigable estuaries formed in the eastern part of this lake, described by Strabo as “scooped-out areas similar to ravines of regular size, or river valleys, by which the sea penetrates many stadia into the land” (3.2.4), through which “boats can sail as if they were going up a river” (3.2.4). (4) Around the estuaries and along the whole Guadalquivir River valley, a dense nucleus of population materialized beginning in the first centuries of the 1st millennium b.c. There were several reasons, which Strabo mentioned much later at the beginning of the Christian era, but which were also valid for this earlier period. He points out that “the shores of the Betis [Guadalquivir] are the most populated” (3.2.3), due to the ease of navigation. “One can sail an approximate distance of 1,200 stadia [more than 200 kilometers] from the sea to Córdoba, and even a little further” (3.2.3), almost to Cástulo in Linares. For this reason “the natives, who understand the nature of the region and who realize that the estuaries can serve the same purpose as rivers, have constructed their cities and towns on these, just as they do on the banks of rivers” (3.2.5). (5) As for the estuary of the Guadalete River, the water extended to El Puerto de Santa María, bathing the feet of the Sierra de San Cristóbal and bordering the tertiary terrain of Puerto Real and Chiclana. Thus it formed a broad bay, larger than today, with two important islands (Cádiz and San Fernando) at the mouth of the river (fig. 5). Some writers believe that few people lived on the coast at the time the Phoenicians arrived. This is the belief that J. L. Escacena expresses in his work on Gadir. 27 He states that “when the Phoenician colonists arrived on the southwestern Atlantic coasts of the Iberian Peninsula in general, and on the coast of Gadir in particular, supposing that they arrived around the year 1100 b.c., they would have found the interior of the territory sparsely populated and the coast completely uninhabited. From Gibraltar to the cape of San Vicente in Portugal, we do not know of any large Final Bronze precolonial population on this coastline.” He goes on to state that “the archaeological record in large areas of the lower Guadalquivir valley suggests that around the year 1100 b.c. the Tartessian territory was largely unpopulated. Thus . . . it does not seem plausible to simply accept, as is frequently the norm in the current bibliography, the founding of a trading settlement, almost at the very mouth of the Guadalquivir, in order to trade with a population that barely existed.” Because of the lack of data on the region of the lower Guadalquivir, there have been reasons to present the issue in this way. For a long time we believed in the existence of a cultural vacuum stretching from the final period of the 27. J. L. Escacena, “Gadir,” in Los fenicios en la Península Ibérica (Sabadell: AUSA, 1986) 49–50.

170

Diego Ruiz Mata

Bell Beaker horizon until the change of the millennium. The problem has been due, as is logical, to the paucity of research in this area. The situation is different now, since we have been able to organize projects to research the location of sites, based on this question, and fill the vacuum with information and explanations. Systematic surveys reveal a considerable number of Final Bronze towns along the coastline opposite Cádiz, in the countryside, and on the eastern shore of the ancient estuary of the Guadalquivir. The towns of the Cogotas horizon would have been contemporary with the mythical colonization of 1104 b.c. The more recent towns of the Final Bronze (9th–8th centuries) are numerous, and the Phoenician colonists would have found a land that was densely inhabited, since, according to the arguments presented above, the 8th century appears to be the period when the Phoenician presence and settlement began in the West.

V. The Results from Castillo de Doña Blanca In spite of Gadir’s importance in protohistoric times, the obvious gap in the research and archaeological record prevent us from learning the details of this city. Research on Cádiz is complicated, as has been pointed out, and the modern city and its Roman and medieval history constitute a serious obstacle for the development of systematic and regular work. However, Gadir is a necessary point of reference for the understanding of the orientalizing period in western Andalusia. This gap in knowledge was the reason for our undertaking excavations in 1979 in the nearby site of Castillo de Doña Blanca, which could mirror the protohistoric development of the metropolis of Gadir. In Castillo de Doña Blanca we find a landscape that has been greatly transformed (fig. 5). What originally was a port on the coast has become a desolate place today, the waves of the sea replaced by the mud of the Guadalete River. A vast marsh at the base of the site extends all the way to Valdelagrana. Castillo de Doña Blanca sits on a low natural elevation (some fifteen meters above sea level) on the spurs of the Sierra de San Cristóbal, which have a maximum elevation of 124 meters. Next to the town is a little inlet that must have served as a port from the beginning. The proximity to the estuary of the Guadalete River, whose left bank bordered the side of the hills, and to the ancient coastline of that period, justifies the selection of this site for the Phoenician city. The port was located on the eastern side of the city. Perhaps there was another port on the west, which a limestone quarry probably destroyed recently. We must point out that from this chosen point the only visual control was toward the sea, toward the ancient islands of San Fernando and Cádiz. The hills obstructed the view to the north. This situation can only be explained by the Phoenician interest in finding a site especially suitable for a port or landing, given the nature of the trading center that this nucleus must have

The Ancient Phoenicians in the Bay of Cádiz

171

Fig. 5. Ancient geography of the Bay of Cádiz.

represented. The choice of location was not capricious, but rather deliberate. In the case of Castillo de Doña Blanca, the principal reason for the choice was the facility that the small inlet offered as a safe port and shelter from the easterly winds. It is probable, although not yet verified archaeologically, that a watchtower was placed on the highest point in the hills, since from the high point it is possible to get an extraordinary view of Cádiz, the northern hills, and the countryside. The supply of fresh water from the hills was enough for the city itself and probably provided enough to transport to the metropolis of Cádiz, where there must have been a short supply. Moreover, this site may have been chosen for its proximity to the Guadalete River, which was navigable for several

172

Diego Ruiz Mata

Fig. 6. Location of Castillo de Doña Blanca: (1) position of the town of La Dehesa during the middle Chalcolithic period; (2) docking area; (3) excavations of 1987 and 1989 and the location of the Phoenician quarter; (4) excavations of 1982–83 and location of a sector from the Punic Period, 4th–3d centuries b.c.; (5) ancient coastline; (6) location of the excavations of 1979–81.

kilometers into the interior. Castillo de Doña Blanca was also near the Guadalquivir River, the important river route that connected upper and lower Andalusia and the gentle countryside. Transportation was easy, and the area was extensively settled by indigenous peoples during the Final Bronze. Beginning very early, Castillo de Doña Blanca must have been a trading center of extraordinary importance, where products from the East were collected in order to be traded among the indigenous peoples in exchange for the coveted metals. The recent research in this area suggests the development of an active trade proceeding from Gadir and Castillo de Doña Blanca, evidenced by the numerous remains of amphorae and luxury pottery found in this region.

The Ancient Phoenicians in the Bay of Cádiz

173

The site has formed an artificial mound with a roughly rectangular shape that comes from the layout of its defensive system, about 340 meters long from east to west and approximately 200 meters wide from north to south. It is 31 meters above sea level, with the archaeological strata taking up 7 to 9 of these meters. An elongated ramp of earth at the southeast corner and another smaller one at the northeast corner stand out, which must be related to the port area in some way (fig. 6). Behind the site and on the slopes of the Sierra de San Cristóbal, we find the necropolis of Las Cumbres, which occupies more than 100 hectares. At the highest point of the Sierra de San Cristóbal stood a town dating to the Punic period, some 3 hectares in size (also known as Las Cumbres ‘The Peaks’) and in the lowest strata and at its sides we find an earlier settlement from the Final Bronze, dating to the 10th and 9th centuries b.c. To the east of Castillo de Doña Blanca, an even older settlement from the 3d millennium occupied a large part of the site of La Dehesa, and a little later, during the Bell Beaker horizon, this site expanded to the area below where the Phoenician town stands. We have also found Roman ruins at the foot of Castillo de Doña Blanca and tombs from the 2d to 4th centuries a.d. at the base of the hills. In summary, during the 3d and the first half of the 2d millennia b.c., this area was settled primarily on the shore, close to the water (at La Dehesa and below Castillo de Doña Blanca). The pre-Phoenician indigenous population of the Final Bronze settled on the highest point of the hills at Las Cumbres, which was abandoned at about the same time that the Phoenician colonists arrived and founded the colony. Thus, the area known as Castillo de Doña Blanca was uninhabited when the Phoenicians arrived and built their first homes and defensive wall. Later, during the 4th and 3d centuries b.c., another town was built over the earlier Final Bronze site in the hills at Las Cumbres, demonstrating that the city of Castillo de Doña Blanca was growing during the Punic period. The abandonment of Castillo de Doña Blanca during the Punic period took place at the end of the 3d century b.c. as a consequence of the Second Punic War and the Roman presence in the bay of Cádiz. The population probably moved to the area where El Puerto de Santa María stands today. In Roman times we note the existence of a villa at the foot of Castillo de Doña Blanca, which was the beach area during the protohistoric period. Much later, during the Almohad period (12th–13th centuries a.d.), a country house was built at Castillo de Doña Blanca. A hermitage was built during modern times, perhaps at the beginning of the 15th century, and it stood there until recently. A few years ago we suggested 28 that Castillo de Doña Blanca was an indigenous town that rapidly assimilated Phoenician cultural characteristics. We 28. D. Ruiz Mata, “Castillo de Doña Blanca (Puerto de Santa María, prov. Cádiz): Stratigraphische Untersuchung einer orientalsierenden Andsiedlung,” MM 27 (1986).

174

Diego Ruiz Mata

considered the proximity of the metropolis of Gadir and the economic interests that the Phoenicians must have had in this area. Our view changed, however, after the 1987 season, when the excavation of the wall and the Phoenician quarter began. The data permit us to argue that we are dealing with a Phoenician founding. The first Phoenician urban remains are found on top of a sterile level where vestiges of occupation during the Final Bronze cannot be seen (we have already pointed out that the population of this period was located in the Sierra de San Cristóbal). The construction techniques are strictly Phoenician and have no resemblance to indigenous huts. The Phoenician materials are not the result of simple and sporadic trading contacts, since the complete set of functional tableware typical of Semitic settlements is present. In addition, we must consider the relatively abundant Phoenician graffiti on amphorae and the red slip pottery with names of owners and cities of origin, including an inscribed alphabet. 29 The archaeological record thus suggests a Phoenician founding, probably contemporary with Gadir, and the pottery suggests a rapid process of assimilation and a high degree of coexistence. When analyzing Castillo de Doña Blanca and indigenous settlements in the area next to the bay and in the lower Guadalquivir, we can point out that the Phoenician strategies of infiltration and occupation of this territory may be different from those in the settlements on the Mediterranean coast. Colonization in the Bay of Cádiz took place in an area that was already inhabited, where indigenous peoples controlled the territory. Given the commercial nature of these eastern colonists and their need for metals, it appears logical that they employed subtle and self-interested means conducive to friendly relations, in order to insure continuous trade and realize their objectives in the most efficient way. An analysis of Phoenician colonization must consider the existence of the indigenous population as a primary factor. Probes made at different points around the perimeter of Castillo de Doña Blanca have helped to determine its size during its early phase. It appears that in a short time the town reached a size of 5 hectares, almost the same size as in later periods. This surface area is considerable for the period. We estimate that the city could house between 1000 and 1200 inhabitants, if we calculate 200 people per hectare, which is not an exaggerated amount. We are not dealing with a small trading settlement or simply a trading post connected to the metropolis of Gadir, but rather with a true “polis,” which was fortified almost from the beginning. We do not yet know the surface area of Gadir during the 8th–7th centuries. However, soundings have been performed continuously in the city center (although they have not yet discovered the ancient city), and we know that the available space did not exceed 7 or 8 hectares, assuming that 29. J. L. Cunchillos, “Las inscripciones fenicias del tell de Doña Blanca (II),” Sefarad 51, fac. 1 (1991) 13ff.

The Ancient Phoenicians in the Bay of Cádiz

175

the ancient city was located in the area of the Torre de Tavira. It is probable that Phoenicians settled in Castillo de Doña Blanca shortly after the founding of Gadir, thus constituting a dual population on the island and on the mainland and following the pattern of settlement seen in their home metropolis of Tyre. If we study the Phoenician installations in the Bay of Cádiz, we can infer that a political and economic structure developed that was divided between three points. The first was the temple of Melkart with probable religious and economic significance. It was located on the present islet of Sancti Petri and served as a point of reference for sailors headed toward the Guadalquivir River. The second was the city of Gadir, a metropolis and key political center at the other end of the island. The third was Castillo de Doña Blanca, which was on the mainland and next to the mouth of the Guadalete River. Located near the indigenous population in the countryside, it was an excellent spot for an outpost and market center. Thus a political structure took shape that was divided among these sites, and all of them came to be known as “Gadir.” Until now the excavations at Castillo de Doña Blanca have focused on the following points (fig. 6): (a) the defensive system (the walls of the 8th, 5th, and 4th–3d centuries b.c.), (b) the so-called “Phoenician quarter” of the 8th century, (c) several stratigraphic probes with abundant material from the 7th century, associated with levels of habitation, (d) the levels of fill from Phoenician dwellings of the 8th century, (e) rooms from the 4th–3d centuries in Castillo de Doña Blanca and the town of Las Cumbres, with about 2000 square meters excavated, and (f) Mound 1 of the necropolis of Las Cumbres, from the 8th century. In addition, we are developing a systematic research project for the study of the region. This project analyzes the cultural development from the Neolithic to the Roman periods in the area between the Guadalquivir and Guadalete Rivers and between the Gibalbín and Gamaza mountains on the north and the coast on the south. We hope to see important results regarding settlement of the territory during the Final Bronze and orientalizing periods. The earliest strata at Doña Blanca are associated with structural remains and come from the northern wall and from Areas 4 and 5 of Square B.12 of the SE sector, or the “Phoenician quarter” (containing an area of more than a thousand square meters). The first Phoenician rooms and the defensive wall were constructed on a sterile layer. Below this layer vestiges from the 3d and 2d millennia have been found. We have not seen evidence of a settlement during the Final Bronze, which was located instead on the highest point of the range of hills, as mentioned earlier. In the Phoenician quarter we can distinguish two phases of construction during the 8th century. The second phase is better preserved, with terraced dwellings facing the north that take advantage of the uneven areas of a small hill (see fig. 6). The dwellings are composed of several rooms, with walls of

176

Diego Ruiz Mata

rubblework and occasionally of mud-brick, floors of reddish clay (frequently interspersed with thin surfaces of lime), and plastered walls. The roof was probably made of plant matter with a framework of wooden beams. On the upper terrace we have excavated a narrow street, a little more than a meter wide, which served as a simple entranceway to the dwellings. The door is usually found at the corner of the house, and one or two steps give access from the street to the interior. The door jambs are normally constructed of ashlars, and at times a pillar of ashlars is seen in the center of a section of wall, a technique that evidently comes from the East. 30 This construction technique was used in Huelva for a retaining wall between the hills of San Pedro and Cementerio Viejo. 31 The height of the rooms was 2–2.5 meters, as demonstrated by a well-preserved wall from a dwelling in this quarter of town. The majority of the dwellings had their own bread oven, consisting of an arched clay structure a little more than a meter in diameter and 0.5 meters in height, with a stone slab floor. At the base of the lower terrace (some 8 or 9 meters lower than the floor of the upper dwellings), we excavated a V-shaped ditch that is almost 3 meters wide and 2 meters deep. Its function is still unknown; perhaps it conducted water or served as a defensive device. If the latter is true, the defense would consist only of the ditch, since the dwellings are next to it, and we have not seen remains of walls or even of a wooden enclosure. During the 1991 season, just a few meters in front of this ditch, we excavated another wider ditch (10–12 meters wide), with a depth of 4–5 meters, partially dug artificially and partially taking advantage of the lay of the land. This ditch also could have had a defensive character. Most likely it was dug during the 8th century, but more evidence is needed. In the 8th century the city was fortified with a strong wall. During the 1989 and 1991 seasons, we excavated a little more than 2 meters of the 8th century wall on the northern flank and quite far away from the Phoenician quarter. The outside part of this wall has provided a stratigraphic sequence of extraordinary interest for its dating. It was constructed on a well-mortared rubblework foundation, on virgin soil where we find some remains from the 3d millennium b.c. On top of the foundation, the wall is constructed of irregular rubblework cemented with clay and is more than 3 meters wide. In front of the wall and along a length of 2 meters, we have detected and excavated a ditch or moat, which in this area is almost 20 meters wide and about 4 meters deep. This section of 8th-century wall is found on the end opposite the Phoenician quarter, about 300 meters away. Therefore, we can see that from the beginning the city was almost the same size as it was in the 4th–3d centuries b.c. 30. J. Elayi, “Remarques sur un type de mur phénicien,” RSF 8.2 (Rome, 1980). 31. D. Ruiz Mata, J. M. Blázquez, and J. C. Martín de la Cruz, “Excavaciones en el Cabezo de San Pedro (Huelva): Campaña de 1978,” Huelva Arqueológica 5 (1981).

The Ancient Phoenicians in the Bay of Cádiz

177

Other remains of walls, also associated with the defensive system, extend from the SE corner. Although these remains are still scarce and we know little about their plan, they may contain a section with casemate structures of eastern design. The most abundant and best-stratified archaeological information comes from the Phoenician quarter and the sections of wall. Red slip pottery is almost exclusive in the repertoire of Phoenician materials, and painted decorations are scarce. Thus, the most characteristic trait of the initial Phoenician phase is the almost exclusive use of forms of red slip pottery. The most common bichrome and polychrome painted forms are practically absent. Plates (figs. 7 and 8) are numerous and have rims between 20 and 30 mm, although in some cases they may be wider. They are made from very refined clays, and they are covered with a high-quality thick red slip over the whole interior and almost always the exterior rim. The bases are flat, though at times they have a concave center. We emphasize the exterior slip on these plates because it is characteristic of Doña Blanca and not frequent at other Phoenician sites such as Málaga and Huelva. 32 The plates from Huelva are peculiar because the slip is left off the inside bottom, 33 constituting a style that is seen only at that site. Open ceramic forms are found in almost the same percentage as the plates. They consist of carinated bowls or páteras (figs. 7 and 8), which come in different sizes and wares and have slips that are equal in quality to those of the plates. The entire inside surface is covered, and on the outside the slip reaches to the carination. Generally, they have an angled rim on the exterior, of different widths depending on the types, a carination on their middle section, and a flat base similar to the base of the plates. The diameters range between 20–25 cm and 10–15 cm. We also note variations in the formation of the rims, as the figure shows, although they always show the external carination. The forms depend on the function that they may have had, as deep plates or as cups. H. Schubart has analyzed and dated the plates found in the Phoenician sites on the Spanish coast according to their types. 34 According to his typological criteria, the plates from Doña Blanca correspond to those of Strata I and II at Toscanos and Chorreras. They also correspond to the lower strata at Morro de Mezquitilla, which are dated to at least the middle of the 8th century b.c. 32. The outside coating appears to be a peculiarity of the potters’ workshops at the Bay of Cádiz and specifically at Castillo de Doña Blanca. For the plates from Huelva, see P. Rufete Tomico, “Las cerámicas con engobe rojo de Huelva,” Tartessos y Huelva. (ed. J. Fernández Jurado, Huelva Arqueológica 10–11, vol. 3,1988–89) 9ff. 33. P. Rufete Tomico, “Las cerámicas con engobe rojo.” 34. H. Schubart, “Westphönizische Teller,” RSF 4 (1976) 179ff.; idem, “Asentamientos fenicios en la costa meridional de la Península Ibérica,” Primeras Jornadas Arqueológicas sobre Colonizaciones Orientales, Huelva Arqueológica VI (1982) 71ff.

178

Diego Ruiz Mata

Generally, the plates from the early period at Castillo de Doña Blanca have rims with widths between 20 and 35 mm, depending on the total diameter. Plates with a larger diameter tend to have wider rims and plates with a smaller diameter usually have narrower ones. In addition to measuring the width of the rim of the plate, we also need to keep the total diameter in mind. The carinated bowls are found with the plates and are dated to this same period at Chorreras, for example, and also at eastern sites such as Hazor, 35 Tyre, 36 Tell Keisan, 37 and the necropolis of Khaldeh. 38 A large number of red slip mushroom-mouth oenochoes come from the Phoenician homes. These oenochoes have been found in almost all of the excavated dwellings, which is unusual since they are normally found in burials and are of a funerary nature. The typical oenochoe at this site has a globular body with grooves (frequently two) at the height of the shoulder, a concave base, a narrow cylindrical neck with a raised band around the middle, small double handles that are attached to the shoulder and band, and a rim consisting of a narrow ring. This ring constitutes a distinguishing characteristic used for dating. All of the samples that have been found have a very brilliant, highquality reddish or brown slip (figs. 7 and 8). A second type differs from the former because of its wider conical neck, trefoil mouth, and longer handle from the shoulder to the edge of the rim (fig. 7.16); it has the same quality slip. In only one case was the neck decorated with red and black bands on the outside. The most frequent type found in the Phoenician rooms is the mushroommouth oenochoe, while in the stratigraphy outside the defensive wall the most frequent type is the oenochoe with the trefoil mouth. The lamp which we find in this phase has only one wick, a flat base, a greater depth than the later lamps with two wicks, and normally no red slip decoration (fig. 7). At times, its diameter is so great that it could be confused with the small plates. We also find flasks or juglets which are functionally associated with the lamps as oil containers. These juglets have globular or ovoid bodies; small, concave bases; conical necks; and small round handles. Their gritty ware is less pure than that of the plates (fig. 7). 35. Y. Yadin et al., An Account of the First Season of Excavations, 1955 (Jerusalem, 1958), coming from Strata VIII (9th century) and VII to V (end of 9th century to 732 b.c.); idem, Hazor II (Jerusalem, 1960), coming from Stratum VIII (9th century), where the most characteristic form is the carinated bowl without glaze and burnishing. However, some have a burnished slip, from Stratum VII (9th century) and low carinations, from Stratum VI (first half of 8th century) and from Stratum V (between 760 and 732). 36. P. M. Bikai, The Pottery of Tyre (Warminster: Wilts, 1978), from Strata IV (760/740) and III/II (dated between 740 and 700). 37. J. Briend and J. B. Humbert, Tell Keisan (1971–1976): Une cité phénicienne en Galilée (Orbis Biblicus et Orientalis, Serie Archaeologica 1, 1980). 38. K. Saidah, “Fouilles de Khaldé: Rapport Préliminaire sur la premier et deuxième campagnes (1961–62),” BMB 19 (1966) 69ff.

The Ancient Phoenicians in the Bay of Cádiz

179

1

2 3 4 5

6 7 9

8

14

17

10

11

12

13 15

16

18

19

Fig. 7. Phoenician pottery of the 8th century: (1–4) red slip plates; (5–7) red slip carinated bowls; (8–13) types of rims of amphorae; (14) and (15) juglets; (16) and (17) fragments of red slip oenochoes; (18) and (19) lamps with only one wick.

180

Diego Ruiz Mata

1 2

3 4 5

6 7

8

Fig. 8. Phoenician pottery of the 8th century: (1) red slip mushroom-mouth oenochoe; (2–4) plates; (5) and (6) carinated bowls; (7) and (8) amphorae. Number 7 corresponds to an imported amphora of the Sagona 2 type.

The incense burner is another form which is characteristic of this period. It is made from two pieces (such as carinated bowls) placed on top of each other. A red slip covers the outside and the upper inside surface of the lip. As in the previous cases, it comes from inside the rooms. At times, the stem that joined the two bowls has been cut apart in order to use them as cups.

The Ancient Phoenicians in the Bay of Cádiz

181

Among these Phoenician pottery types, we also find animal-shaped cups (rhytons or askoi) with distinct representations of animal heads, sometimes covered with a red slip. The species are hard to identify, but in one example we can make out the head of a pig. While excavating the wall, we found a group of undecorated jugs in association with the types described above. These jugs have trefoil mouths about 8–10 cm wide (those that have been reconstructed), ovoid bodies, flat bases, a handle inserted into the rim and shoulder, and light-colored, simply polished surfaces from which the grits frequently protrude. Similar types have been found at the Phoenician site of San Antioco 39 on Sardinia, dating to the second half of the 8th century b.c. We must note, however, that these jugs have not been seen among the materials from the homes in the Phoenician quarter, suggesting perhaps a slightly later date. Other kitchen ware consists of various sizes of rough-surfaced pots, with narrow, concave, polished rims; ovoid bodies; and solid, flat bases. We assume that these pots are common cooking ware (fig. 9.14 and 9.15), and similar vessels come from the lower strata of Morro de Mezquitilla. Amphorae are numerous, which is natural for a trading center. Although they are still being studied and require a closer analysis, we note at least three well-defined types. The most frequent type is the de saco amphora, with sloping shoulders, a very prominent carination, and a wide, pear-shaped body with round handles. This is the most common amphora of the Mediterranean and southern Iberian Peninsula. The differences are seen in the structure of the mouths, which are short and narrow during the 8th century, at times with a groove below the rim (fig. 8.8). A second type has a short neck smoothly slanting slightly inward (a type which may correspond to the eastern amphora with an ovoid body and shoulders unmarked by the characteristic carination). Its cylindrical or oval handles are also longer and are attached at the height of the shoulder (fig. 7.12 and 13). The third type has a narrow rim, molded on the outside; a nearly horizontal and strongly-carinated shoulder; and an elongated, cylindrical body, with small handles (fig. 8.7). This is a frequent type on the Levantine coast of the Near East, and it corresponds to Type 2 in A. G. Sagona’s classification. This type circulated especially between 760 and 700 b.c. 40 Thus, the early levels of the 8th-century town contain the entire Phoenician typological repertoire of functional and daily tableware. This phenomenon does not occur in indigenous settlements, which have only a limited repertoire 39. P. Bartoloni, P. Bernardini, C. Tronchetti, and L. Usai, “S. Antioco: Area del Cronicario (Campagne di scavo 1983–86),” RSF 18.1 (1990) 37ff. 40. A. G. Sagona, “Levantine Storage Jars of the 13th to 4th Century b.c.,” Opuscula. Atheniensia 14 (1982) 75ff.

182

Diego Ruiz Mata

of exotic forms. We must also point out types that were surely imported, such as some cups with fine sides covered with a thick red slip. As a distinctive characteristic, the cups are decorated with narrow grooves on the outside, and they belong to the type known as “Samaria fine ware.” Also we find short, flatbased bowls that have rims decorated on the inside with a wide red band on the upper edge of the lip and two black lines below. The design is repeated at the bottom of the bowl. These bowls are characteristic of Strata VII–V at Tyre and are dated to the first half of the 8th century b.c. 41 Our earlier doubts about the Phoenician nature of the founding of the town were justified because of the presence of indigenous handmade material from the Final Bronze. However, we now believe that the Phoenicians founded the town. As we continue learning about the protohistoric peoples surrounding the bay and the Phoenician strategy of approaching them, the indigenous material can be explained without contradiction. It appears that Phoenician settlement on the coast of Málaga took place in a sparsely populated environment, and the pottery there is exclusively wheel made. The panorama is different on the Atlantic coast, especially between Cádiz and Huelva, where there was a dense indigenous population. This fact could have translated into different Phoenician behavior. In the case of Castillo de Doña Blanca, the handmade pottery can be explained by the integration of the indigenous community that inhabited the hills into the Phoenician town. The handmade material consists of cooking vessels, cups, and large storage containers for solids and liquids, of the type known as à chardon vessels. These bell-shaped vessels with wide necks have been found frequently in Phoenician rooms (fig. 9). The cooking vessels have bulging rims on the inside and smooth carinations on the outside, a trait characteristic of the late Final Bronze. The pottery found in the settlement of Las Cumbres is different. The carinations are more pronounced, which is the norm for pre-Phoenician indigenous sites. The vessels are decorated on the interior with burnished, crossing lines, which form crosshatches of different sizes, made with precision on a background that has only been polished. The designs are divided into quadrants by broad or narrow burnished bands according to a pattern frequently found in the lower Guadalquivir. In addition, we have found large cups and vessels decorated with incised geometric designs, made with a needle on a brown or black well-burnished surface. The open forms, and sometimes the closed forms as well, are covered on the inside with a layer of red paint or ochre. The decorative designs show the prevailing Geometrism of the 9th and 8th centuries, as reflected in the painted, red monochrome pottery from the lower Guadalquivir. 41. P. M. Bikai, The Pottery of Tyre.

spread is 12 points short

The Ancient Phoenicians in the Bay of Cádiz

183

1

2

3

4

5

6

7 9

10

8

11

12

13

14

15

Fig. 9. Indigenous handmade pottery of the 8th century: (1–7) cooking vessels with an interior decoration of burnished crosshatching; (11) and (12) carinated bowl and support with an exterior decoration using geometric designs incised on a burnished surface; (13–15) rough-surfaced pots.

184

Diego Ruiz Mata

The handmade material suggests that, rather than an indigenous town, we have the result of peaceful coexistence with the large indigenous population that lived around the bay, the Río Salado, the estuaries of the Guadalquivir River, and in the countryside. The Phoenician colonization at the Bay of Cádiz cannot be explained without considering the indigenous factor and the process of assimilation and acculturation. The Phoenician quarter was abandoned at the end of the 8th century, and the spot was not occupied again until the 5th–4th centuries, when a new section of defensive wall was built over it. We do not yet know the cause for the abandonment. A possible hypothesis is an earthquake, based on the state of collapse that is seen, the condition of the walls of the rooms, the pottery found inside that is quite complete, and the skeleton of an individual beneath a wall. Various strata of fill were deposited over the collapsed walls. These strata contain abundant ceramic and organic material because the area was used as a dump. This area is where we centered the archaeological work of 1991, so the results are provisional. The materials in the fill have been dated to the end of the 8th century because of the appearance of an Euboean cup rim and a Corinthian amphora neck from the Late Geometric Period. These materials differ notably from the earlier type of pottery. The most salient characteristic is the novel appearance of vessels with a polychrome decoration. Generally, the decoration consists of wide red bands framed by black lines, which cover the greater part of the surface of urns and pithoi. Also we frequently find decorations of rows of concentric circles, in which black and red circles alternate, and of black lines on wide red bands, known as “black on red.” On large-capacity vessels we occasionally find motifs depicting fantastic animals and exotic flowers. The discovery of these fragments in a welldefined context dates these much-discussed vessels to the end of the 8th or beginning of the 7th century b.c. The plates now have a wider rim that reaches up to 4–5 cm or more, similar to the earliest at Toscanos. The carinated bowls (or páteras) are larger and continue to be covered with a red slip layer of high quality. And for the first time we find cups with narrow, straight rims and prominent carinations, covered with a red slip, and bowls with rims that are straight or that slope smoothly inward. The bowls are made from of a poorer quality clay and are decorated with red and black bands on the outside. The juglets continue to be used and generally maintain the previous forms. However, the lamps frequently have two wicks, and they have lost the flat base that characterized the earlier phase. They are covered with a red slip, although occasionally the older undecorated types with only one wick survive. We note the beginning of production of gray pottery with light or blackish tones. The forms are carinated bowls and dishes, plates, and bowls with more or less thickened rims. Their percentage is not very high, and their production

The Ancient Phoenicians in the Bay of Cádiz

185

1

2

3

4

5

6

7

8

9

10

Fig. 10. Phoenician pottery of the 7th century: (1) and (2) red slip plates; (3–5) carinated bowls; (6–10) bowls with red paint and black lines. Number 7 has grooves below the rim.

186

Diego Ruiz Mata

appears destined mainly for the demands of the indigenous market, as suggested by the huts excavated in the countryside around Gadir and Huelva. The amphorae are of the de saco type, with carinated shoulders and pearshaped bodies. They differ from the earlier ones because of their thicker rims. In this period the type Sagona 2 and the ovoid amphorae have disappeared. In regard to the handmade pottery, the open type is now a simple bowl with an indistinguishable rim, decorated on the inside with burnished designs of poorer quality than the earlier ones. The closed forms are large storage vessels and rough-surfaced pots, which at times are decorated on the shoulders with fingerprints. In summary, at the end of the 8th century the repertoire of wheel-made pottery grew with the appearance of new open forms and painted bichrome and polychrome vessels. From this point on, the typical Phoenician forms in the West became set, and these products of the “Western Cultural Province” or the “Circuit of the Strait” spread throughout northern Africa and the Iberian Peninsula. In spite of the similarities between many of the ceramic forms, the differences between this and the earlier period are significant: first, because of the appearance of painted pottery with bands and concentric circles, and second, because of the growth and evolution of the formal repertoire. The fact that there are so many differences in so little time perhaps suggests the presence of new stimuli toward the end of the 8th century, coming from Cyprus or from the Phoenician cities on the coast. The material from the 7th century and the beginning of the 6th does not show great changes and is similar to that of the previous phase. The information comes from Areas C.1.a., C.1.b, C.4, F.30, and B.12. In the first three areas we have detected three phases of superimposed construction during the 7th century without perceptible changes in tableware. This period constitutes a crucial moment of expansion toward North Africa by sea, and through the interior of the Iberian Peninsula toward Estremadura, the Guadalquivir, and the eastern coast up to Catalonia. The excavations realized during the past few years in Portugal suggest an expansion toward that region by land or by sea during the 7th century. This is also the period when many of the indigenous towns of the Final Bronze turn into true cities. We can see economic prosperity in the grave goods found in necropolises (as, for example, at La Joya in Huelva) and in the amphorae, scattered over numerous sites, that were used to transport goods. All of this points to unprecedented trade. The typology of the plates constitutes a dating element to keep in mind, although the guidelines are not extremely rigid. In Doña Blanca during the first half of the 7th century, the plates have rim widths between 50 and 58 mm, while later they reach 60–70 mm. According to the values established for Toscanos, plates that have rims of up to 55 mm are dated to the first quarter of

The Ancient Phoenicians in the Bay of Cádiz

187

1

2

3

4

5

6

7

Fig. 11. Phoenician pottery from the 7th century.

the 7th century. Those with greater widths have been dated at the necropolis of Trayamar to the middle or to the second half of that century 42 (figs. 10, 11). Other elements characteristic of this period are the carinated bowls, which have three basic forms based on the construction of their rims and decoration. One form has a sloping and carinated rim, a type that survived from the 8th century with small differences in size and quality of slip (fig. 10.3). A second form has a narrow rim that is vertical or slightly inclined toward the interior. This bowl was deeper during the earlier period and flatter during the 7th century (fig. 10.4), and it was common also after the end of the 7th century. 42. H. Schubart and H. G. Niemeyer, Trayamar: Los hipogeos fenicios y el asentamiento en la desembocadura del río Algarrobo (EAE 90) 1976.

188

Diego Ruiz Mata

Finally, a third form is a bowl with a concave and carinated rim (fig. 10.5). The surfaces of the first two forms are covered with a thick red slip, while the form with the concave rim is decorated with an application of paint that is quite diluted at times, covering a surface that has received less care and in which the grit protrudes. This final form is also broader and has a greater capacity than the others. Deep or hemispheric bowls are common and are made from less-refined clays. Their surfaces are either decorated with red and black painted bands, or the upper half has narrow black lines painted over red. This is probably a local western version of the Cypriot and eastern “black on red” decoration. Several types have been distinguished based on the structure of the rims. The first has a narrow rim and a prominent carination on the outside (fig. 10.6), the second has a rim curved in toward the interior (fig. 10.8), and the third (also with a rim curving inward) has grooves on the lip as a distinguishing characteristic (fig. 10.7). Other open forms are small bowls which are hemispheric or shallower than those above, also made from less refined clays. Usually they are undecorated, but when decorated, they have wide red bands and black lines only on the outside (fig. 10.9, 10). Large capacity pithoi are frequent among the closed forms. They have geminate handles attached to the rim and shoulder, and their bodies are decorated with red and black bands or with concentric circles (fig. 12.1 and 2). We also see urns of the Cruz del Negro type with cylindrical necks and a central band around the neck. Their bodies tend to be spherical during the earliest period (end of the 8th century), and they also have geminate handles (fig. 12.3– 5). The 7th century forms that we see came from prototypes of the 8th century and were very popular. They had a wide range of distribution throughout the Phoenician colonies and Tartessian towns, as well as in northern Africa. Their origin appears eastern, although they have not been documented at Castillo de Doña Blanca during the first period of exclusively red slip pottery. The absence of mushroom-mouth and trefoil-mouth oenochoes is paradoxical in comparison with the frequent use of these forms during the 8th century. It is probable that during the 7th century this form acquired an almost exclusive funerary use, as can be deduced from the necropolises of Trayamar in Málaga and Laurita in Almuñécar (Granada). Until now only one mushroommouth oenochoe from the 7th century has been found. It has a neck formed by two cones with a pronounced groove at its central joint, corresponding to a type that has been dated to the beginning of the 7th century 43 (fig. 12.1 bis). 43. I. Negueruela, “Jarros de boca de seta y de boca trilobulada de cerámica de engobe rojo en la Península Ibérica,” in Homenaje al Prof. Martín Almagro Basch, vol. 2 (Madrid: Ministerio de Cultura, 1983).

The Ancient Phoenicians in the Bay of Cádiz

189

1

1 bis

2

3

4

5

6 7

8

9

Fig. 12. Phoenician pottery of the 7th century: (1) and (2) pithoi with polychrome decoration; (1 bis) oenochoe; (3–5) polychrome urns; (6) tripod; (7) and (9) juglets; (8) red slip lamp with two wicks.

190

Diego Ruiz Mata

The lamps all have two wicks and are shallower than those of the 8th century. All of the known samples are covered with a thick red slip burnished on the inside and outside (fig.12.8). Other articles are juglets, small jars, and tripods, made from clays that have not been well purified. The juglets repeat the earlier forms of the 8th century, with an ovoid body and conical neck, or with a more cylindrical body and a short, turned-out neck. The most common tripods during this century have a flange below the rim (fig. 12.6), and their prototypes are from the end of the 8th century b.c. The gray pottery maintains the same percentage as earlier, and the same types are produced: the wide carinated vessels, the plates with turned-out rims, and the bowls in blackish and light gray tones with thickened rims (fig. 13.4). In summary, the great quantity of amphorae found in strata from the 7th century provides the most significant testimony to the productive development reached during this century—and of outside demand. Typologically, the classical de saco forms and thickened rims continue, offering a broad spectrum (fig. 13.1–3). We can summarize with the following points: (1) Castillo de Doña Blanca was a Phoenician settlement founded in the 8th century, probably during its first half. According to the distribution of archaeological remains from its earliest period, it had a size of approximately 5 hectares from the beginning, as well as a defensive wall. Thus we are dealing with a colony established next to the Guadalete River under the auspices of Gadir, according to a political and commercial plan. In this respect, we must consider the plurality of the concept of Gadir, which was distributed among three sites: the temple of Melkart, Gadir, and Castillo de Doña Blanca. (2) The analysis of Phoenician colonization in the Bay of Cádiz must necessarily include a parallel study of indigenous settlement: its distribution, initial contacts, trading characteristics, and degree of assimilation and acculturation. (3) The archaeological record has fully demonstrated that a phase existed in which red slip ceramic tableware predominated, almost to the exclusion of other pottery. This ceramic style is the oldest that is known up until now in the Bay of Cádiz. Any study on the mythical date of the founding of Gadir should be made from this perspective and using the evidence of material culture. (4) Again from the perspective of ceramic analysis, we note a second stage with more stylistic variation. Some forms survive, and new kinds of forms and decorations are introduced. We refer especially to the painted pottery, which constitutes an unprecedented novelty. This is the period of greatest dynamism and represents the high point of the orientalizing period, with both interior and exterior expansion.

spread is 1 pica short

The Ancient Phoenicians in the Bay of Cádiz

191

1

2 3

4

5 6

7

8

9

10

11

12

13

14

15

Fig. 13. Phoenician and indigenous pottery of the 7th century: (1–3) amphorae; (4–6) gray-ware plates; (7–9) handmade bowls with burnished decoration; (10–15) handmade pots.

192

Diego Ruiz Mata

The problem is determining the length of this second period, but it probably lasted until the second quarter of the 6th century b.c. During the 6th century, and probably toward the middle if we consider the dates given by the ancient Greek pottery, we see the evolution of certain Phoenician forms that are actually a prelude to the beginning of Turdetanian culture in the Bay of Cádiz. It is a crucial period which raises many questions. During this period Tartessos suffers an economic and political decline. We see the influence of Carthage in the West, especially on the Mediterranean coast after the fall of Tyre, and we note a Greek presence in Tartessos during the first half of the century. As a consequence of all of this, the state of Tartessos is dismembered, and the Turdetanian kingdoms emerge. In this respect, a stratigraphic sequence at the Phoenician colony of Guadalhorce (Cerro del Villar) has helped to focus the issue. 44 Two phases have been distinguished that correspond stratigraphically to a period of construction of a storehouse in the 7th century (Phase I), and another later phase of abandonment (Phase II). The second phase shows significant changes in the material context and is interpreted as a new political and economic situation. This observation has been related in turn to other sites: the abandonment of the trading post at Toscanos and the hurried and technically careless construction of the defensive wall at Alarcón, as if there were an emergency. It has also been related to the ceasing of Phoenician contacts with the southeastern Peninsula, reflected in the eastern settlements of Los Saladares, Vinarragell, and Peña Negra, and the decay or abandonment of various North African trading posts closely linked with the Peninsula, such as Mogador on the Atlantic coast of Morocco. As a probable cause we can suggest the economic disequilibrium provoked in the West by the fall of Tyre around 573 b.c., toward which a large part of the Phoenician market was directed. In addition, beginning in the 6th century, the East Greeks initiated an active trade with Tartessos, as a result of the weakening of Gadir’s maritime power in the area of the Strait. 45 Another possible reason for the changes is that, after the fall of Tyre, Carthage assumed political and economic control of the Mediterranean, as M. E. Aubet Semmler has expressed in the following terms: “The period between the 6th and 3d centuries b.c. corresponds to that of the Carthaginian pre-Barca empire and coincides with that period in which Carthage assumes, gradually and militarily, the control of the old territories of the western Phoenician population. This is a period of profound change in the heart of Phoenician society 44. A. Arribas and O. Arteaga, El yacimiento fenicio de la desembocadura del río Guadalhorce (Málaga) (Cuadernos de Prehistoria de la Universidad de Granada, Serie Monográfica 2, 1975). 45. For the bibliography and the state of the question, see P. Cabrera, “El comercio foceo en Huelva: cronología y fisionomía,” in Tartessos y Huelva (ed. J. Fernández Jurado, Huelva Arqueológica 10–11, vol. 3, 1988–89) 41ff.

The Ancient Phoenicians in the Bay of Cádiz

193

1

2

3

4 6

7

5

9

8

10

11

12

13

14

15

17 16

18

Fig. 14. Phoenician pottery of the 6th century: (1) and (2) plates; (3) and (4) carinated bowls painted in red; (5–7) bowls with thickened rims and a red slip on the inside; (8) imitation of Ionian cup; (9) bowl with decoration of red paint; (10–12) gray pottery; (13) handmade bowl with burnished decoration; (14–17) earthenware pots without decoration.

194

Diego Ruiz Mata

in the West, whose development no longer depends on the strictly economic and trading objectives that had been promoted by the Phoenician cities in the East, but rather is conditioned by a totally different political situation.” 46 Beginning in the second half of the 6th century, the archaeological record in the central and western Mediterranean and along the North African coast shows a series of changes in burial practices and ceramic tableware. We see the introduction of a group of pieces (terra cottas, masks, shaving blades, and ostrich eggs, for example) of indubitable Carthaginian character, which are the clues that allow us to perceive the Carthaginian influence. However, the signs which we see clearly in Sicily, Sardinia, and Ibiza are less transparent in the Iberian Peninsula, due perhaps to the strongly-rooted orientalizing culture already present and to a smaller degree of Carthaginian influence. If we contrast the cultural characteristics of the Phoenician colonies of the 8th–7th centuries with the later ones of the 6th century, we observe small changes that in some cases have their origin in the Carthaginian presence. The ceramic types and their restrained decoration are more similar to those of Carthage (although the decorations of the 7th century were also restrained). In burial practices, tombs are of a North African type, and inhumation substitutes for cremation (although inhumations were also present in the 7th century at Trayamar, La Joya in Huelva, and Carmona, for example). The same can be said of religion, with sanctuaries dedicated to gods of the Carthaginian pantheon. These are cultural traits that acquire different intensities in the different regions of the Iberian Peninsula, according to the degree of acceptance or imposition. In general, from the Guadiana River to Almería we find the footprints of Carthage, but it is necessary to determine the degree of influence and the chronology. In many cases, this chronology has not been established, and we must analyze the internal development of the different areas more carefully. From a stratigraphic perspective, we note some changes beginning in the 6th century, though they often do not appear to be too substantial. In Castillo de Doña Blanca, the Carthaginian influence is not extremely clear during the 6th century. During this century handmade bowls still survive, though they are scarce. The red slip plates widen their rims so that the inner edge of the rim reaches almost to the base, which forms a small central depression as a characteristic trait (fig. 14.1 and 2). This is not a new form, but it results from the evolution of older plates, perhaps as a consequence of new eating habits associated with fish. The slips lose their thickness, they are of poorer quality, and their colors tend more toward light red tones. The carinated bowls with concave rims continue, without decoration or with a poor quality slip. The shallow bowl also continues, but now loses its characteristic carination and quality of its earlier slip (fig. 14.3, 4, and 9). Both types of bowls are now scarce. The 46. M. E. Aubet Semmler, “La necrópolis de Villaricos en el ámbito del mundo púnico peninsular,” in Homenaje a Luis Siret (Consejería de Cultura de la Junta de Andalucía, 1986) 612ff.

spread is 6 points long

The Ancient Phoenicians in the Bay of Cádiz

195

1

2

3

4

5

6

7 8 10

11

9

12

13

14

15

16

17

18

19

21

20

22

23

Fig. 15. Phoenician pottery of the 6th century: (1–9) amphorae; (6) amphora with red paint on the outside; (8) and (9) Corinthian amphorae; (7) and (10) urns; (11– 15) pithoi; (16–19) rough-surfaced pots; (18) and (19) pots with incisions on the outside; (20) and (21) lamps; (22) and (23) small vessels with polychromatic decoration.

196

Diego Ruiz Mata

gray pottery reaches its highest percentage, and the bowls with thickened rims and concave rims are common (fig. 14.10–12). The urns show small typological differences, but basically the earlier prototypes continue. The classic pithos of the 7th century, with a short vertical or slightly slanting neck now has a more open and slanting neck, while the rim is less pointed (fig. 15.7–15), and the urn of the Cruz del Negro type tends to have the same slant. In other words, the classic types of Phoenician-style pottery of the 7th century are maintained, with variations only in certain formal aspects. A type of pot now appears that will survive for a long time without fundamental variation until the 4th–3d centuries. Its characteristic traits are a short and slanted rim, an ovoid body, and grooved lines on the shoulder area (fig. 15.16–19). Another new element is a high percentage of bowls with thickened rims, either without decoration or covered with a red slip on the inside (fig. 14.5–7). We also find deep containers with large diameters (between 30 and 40 cm) and vertical or slightly concave rims (fig. 14.14–18), usually without decoration, but sometimes painted with simple bands. This is a new and characteristic form of this period, which will last, with certain typological variations, until the 3d century b.c. or later. Also new are the deep cups with concave rims and decorated with red and black bands, which are surely imitations of Greek Ionian cups (fig. 14.8). The amphorae show variations in the construction of their rims and shoulders, which are generally more slanted. Amphorae with a short, slanted rim and slightly rounded or “drooping” shoulders are common (fig. 15.1–2), as are those with thickened rims on the interior (fig. 15.3) or thickened and beveled rims (fig. 15.4). A case that stands practically alone until now is fragment 5 of figure 15, which has a short, vertical rim, a raised band around the lower portion of the rim, and shoulders that are very low and rounded. Equally noteworthy is the amphora fragment in the same figure that is decorated with reddish paint on the shoulder area (fig. 15.6). In this context we should point out the Corinthian amphorae of Koehler’s Type A (fig.15.8, 9) from the first half of the 6th century, as well as amphorae of the Greek à la brosse type. 47 In this article we have presented the stratigraphic sequence of the Phoenician town of Castillo de Doña Blanca from the period of its founding, at a date still not precisely determined in the first half of the 8th century, until the middle of the 6th century, when we note obvious changes and the beginning of the gestation of the Turdetanic period in the lower Guadalquivir. During the final decades of the 5th century and during the 4th, the site of Doña Blanca acquired the urban features shown by the structures on its surface. At the same time, pottery types at Doña Blanca evolved toward classic Turdetanic forms, preserving their orientalizing and Punic characteristics to a lesser degree. 47. C. G. Koehler, Corinthian A and B Transport Amphorae (Univ. Microfilms International, 1979).

page is 6 points long

The Ancient Phoenicians in the Bay of Cádiz

197

Plate 1. General view of the 1987 excavations in the area of Espigón. In the center, the defensive wall of the 4th century; to the left, the beginning of the excavation of the “Phoenician quarter.”

Plate 2. Beginning of the excavation of the “Phoenician quarter” of the 8th century b.c.

198

Diego Ruiz Mata

Plate 3. Phoenician red slip pottery from the 8th century b.c. (plate and mushroommouth oenochoe).

Some Questions Regarding the Tartessian Orientalizing Period María Eugenia Aubet Semmler

The Orientalizing Phenomenon The excavations that have taken place during the past few years in the basin of the lower Guadalquivir and in Huelva (the territory of Tartessos according to written tradition) have contributed a series of archaeological sequences and new developments that in a very short time have permitted us to demythicize the traditional concept of the Tartessian world. We can now appreciate an indigenous cultural horizon rooted in the Bronze Age that came into contact with the Phoenician colonial enterprise during the 8th–6th centuries b.c. To what degree the establishment of Phoenician trade on the shores of the Strait of Gibraltar stimulated a series of sociocultural, economic, and artistic transformations among these indigenous populations and the problems that emerge from the study of this Iberian orientalizing period constitute important and timely questions in Peninsular protohistory. This study will attempt to approach these problems, but in doing so does not claim to give solutions, but rather to raise questions and set forth working hypotheses. In the Mediterranean, the cultural horizon known as “orientalizing” corresponds to a concrete period in protohistory that affected Greece, Italy, and southern Spain, all to different degrees. The development of this Mediterranean horizon was conditioned by diverse geopolitical and economic circumstances that logically differed from one region to another, and for this reason the response of the indigenous substratum to eastern influence took on diverse forms and characteristics. If we look at the whole picture, however, the orientalizing period represents a decisive transitional phase between protohistoric geometric cultures of the Mediterranean and historic Classical civilization Editor’s note: This article appeared in Spanish under the title “Algunas cuestiones en torno al período orientalizante tartésico,” Pyrenae 13–14 (1977–78) 81–107. A German version of this text is published in the proceedings of the symposium Die phönizische Expansion im Westlichen Mittelmeerraum (Cologne, April 1979).

199

200

María Eugenia Aubet Semmler

(Etruscan and Iberian) and the definitive access of these populations to urban ways of life. If we accept that the different orientalizing foci in the Mediterranean each present their own peculiarities within some relatively uniform characteristics, we will understand that this period is one of the most difficult to define. In Greece and Italy, for example, we deal with a phenomenon of diffusion and imitation of eastern objects, themes, and ideology born of the great currents of maritime traffic in the 8th and 7th centuries b.c. This phenomenon principally affected those places where a greater concentration of riches and a greater demand for luxury goods existed, such as the Greek sanctuaries or the mining zones of Etruria. In Greece this phenomenon was intimately related to the establishment of Greek trade centers on the Syrian coast beginning in the middle of the 8th century b.c., and, consequently, related to the intensification of economic and cultural relations with the East. 1 In Latium and Etruria the orientalizing phenomenon appears much more complex and originated from the combined stimuli of Greek, Syrian, and Phoenician influences that, directly or indirectly, penetrated Italy from the end of the 8th century. Without doubt, the Greek component was the one that predominated, with the Euboic colonies of Pitecusas and Cumas playing a major role as intermediaries between East and West. 2 In central Italy this orientalizing horizon carried with it such an acceleration in the economic and sociocultural development of local communities that it would stimulate the conditions essential for the formation of early 6thcentury urban society. At the same time that it distanced itself from its original sources, the orientalizing phenomenon grew increasingly complex, influencing more forcefully the indigenous substratum. In contrast to Italy, this horizon in Tartessos originated almost exclusively from the Phoenician component, which changed it 1. For these issues see especially T. J. Dunbabin, The Greeks and their Eastern Neighbors (London, 1957) 25–28, 44; F. Poulsen, Der Orient und die frühgriechische Kunst (Leipzig, 1912) 83f., 100, 108ff.; M. Pallottino, “Orientalizzante,” in Enciclopedia Universale dell’Arte X (1963) 223–27. 2. I. Ström, Problems concerning the Origin and Early Development of the Etruscan Orientalizing Style (Odense Univ. Press, 1971) 201–5; G. Colonna, “Prehistoria e Protostoria di Roma e del Lazio,” in Popoli e Civiltà dell’Italia antica III (Rome, 1974) 309; B. D’Agostino, “Tombe ‘principesche’ dell’Orientalizzante Antico da Pontecagnano,” in Monumenti Antichi della Accademia dei Lincei (Sèrie Miscelánea 2.1, 1977) 45; A. Bedini and F. Cordano, “L’ottavo secolo nel Lazio e l’inizio dell’orientalizzante antico, alla luce di recenti scoperte nella necropoli di Castel di Decima,” in Parola del Passato 175 (1977) 303–8; for Pitecusas-Ischia, see G. Buchner, “Nuovi aspetti e problemi posti dagli scavi di Pithecusa,” in Contribution à l’étude de la société et de la colonisation eubéennes (Cahiers Centre Jean Bérard 2, Naples, 1975) 85; G. Buchner and G. Garbini, “Testimonianze epigrafiche semitiche del’VIII secolo a.C. a Pithekoussai,” in Parola del Passato 179 (1978) 130–50; A. M. Bisi, “La presenza fenicia in Italia nei primi tempi della colonizzazione greca,” Magna Grecia, anno XIII, 5–6 (1978) 12.

The Tartessian Orientalizing Period

The Tartessian Orientalizing Period

201

into a phenomenon that was more “oriental” than “orientalizing.” In the south of the Iberian Peninsula, the permanent presence over several centuries of Phoenician cities near the Strait and their direct contact with the Tartessian communities in the immediate interior stimulated a “purer” cultural process, that, nevertheless, poses numerous problems of interpretation. For example, it is not always easy to distinguish between the Phoenician cultural component of eastern origin and the western or “provincial” root that had time to develop in the West, and at the same time differentiate both with respect to the orientalizing Tartessian component and the local traditional Tartessian component. 3 This is why interest in this period in Tartessos goes beyond the simple identification of artistic foci or the identification of imports and local imitations, since the orientalizing horizon of the 7th and 6th centuries b.c. implies a much more vast and complex process. This period should be considered the decisive transitional step of some of these proto-urban communities toward urban life. It is necessary to study not only the material culture (bronze jars, ivories, gold and silver work), but also the internal economic and social transformations that the Phoenician influence provoked, stimulated, or accelerated among the local indigenous populations. 4

Phoencian Settlements and Tartessian Communities of the Final Bronze Age One cannot doubt the importance of the establishment of Phoenician colonies and trading posts on the shores of the Strait of Gibraltar for determining the origin and development of the orientalizing Tartessian horizon. In spite of the lack of data relative to ancient Gadir (Cádiz), and, consequently, about the beginnings of Phoenician colonization in the far West, we know today that the Phoenician colonies on the shores of Granada, Málaga, and Cádiz had already been founded by the middle of the 8th century b.c. This early activity is amply documented at Toscanos, Chorreras, and Mezquitilla, and the archaeological record shows that the economic high point of the colonies did not begin until

3. About “regionalisms” in Phoenician art of the Mediterranean, see S. Moscati, in RSF 1 (1973) 127; idem, “Problematica della civiltà fenicia,” Studi Sardi 45 (Rome, 1974) 105, 111–12; idem, RSF 2 (1974) 6–9; idem, I cartaginesi in Italia (Rome, 1977) 307; M. E. Aubet Semmler, “Los marfiles fenicios del Bajo Guadalquivir I,” Studia Archaeologia 52 (1979) 66–67. 4. The first assessment of the orientalizing Tartessian facies is due to the studies of A. García Bellido, “Materiales de arqueología hispano-púnica,” AEArq. 29 (1956) 85–104; idem, “Algunas novedades sobre la arqueología púnico-tartessia,” AEArq. 43 (1970) 3–49; A. Blanco, “Orientalia,” AEArq. 29 (1960) 3–51; J. Maluquer, “Nuevas orientaciones en el problema de Tartessos,” in I Sympos. Prehistoria Peninsular (1959) 282. In general, all of the bibliography collected by J. M. Blázquez, Tartessos y los orígenes de la colonización fenicia en Occidente (Salamanca, 1975) especially 211.

202

María Eugenia Aubet Semmler

approximately 700 b.c. and that they enjoyed this economic power during the entire 7th century. The geographic location of the Phoenician colonies is of great interest for these studies (fig. 1). Even if the majority of these colonies are situated next to the mouth of a river, like Adra, Mezquitilla, Toscanos, Guadalhorce, and Guadarranque, these locations do not seem to have been chosen for exclusively strategic reasons, such as routes of access to the interior. A rugged chain of mountains (the first peaks of the Penibetic range) rises to the back of the Phoenician colonies, lining the coast and forming a natural barrier. Without doubt, the passes of Vélez-Zafarraya in Toscanos and, above all, the valley of the Guadalhorce River in Málaga were important routes of prehistoric communication, but, judging from the finds, they do not appear to have carried too much importance for the lives of the colonial Phoenicians. Even though certain contacts between the coast and the populations of the interior have been documented, the Phoenician finds in the immediate hinterland (Granada and bordering areas) do not at all reflect an intense trade in the region. 5 The sporadic Phoenician imports in this region, at times discovered along with Tartessian pottery, appear to back up the hypothesis of a greater utilization of communication routes which proceeded from the valley of the Guadalquivir River, that is, from Tartessos. One must not forget that the most important Phoenician colony on the Iberian Peninsula was situated precisely at the entrance to the valley of the Guadalquivir, which provided the ideal natural access toward the interior and, without doubt, the route most frequented by Phoenician commerce. Written tradition testifies that the primary goal of this trade was to obtain metals from the Sierra Morena and the Northwest, that is, silver, copper, and gold. This intense trade along the Guadalquivir, its tributaries, and the Guadiana, is fully documented by archaeological findings, and especially by the immediate result of these economic contacts—the orientalizing Tartessian horizon. Thus, the establishment of the Phoenician colonies to the east of Gadir, in a zone not very conducive to land trade, seems to have been to serve the great western metropolis of Gadir. These colonies surely acted as bases of support for navigation, and above all, for strategic control of access to Tartessian territory by way of the Phoenician capital. When the Phoenicians arrived, the Tartessian area of the lower Guadalquivir and Huelva was occupied by populations with a Final Bronze Age culture, whose ethnic-cultural roots must be sought in the Bronze Age communities that lived in the southwestern Peninsula during the second millennium b.c. Because of studies and excavations undertaken during the last few 5. An excellent general overview in H. G. Niemeyer, “Orient im Okzident,” MDOG 104 (1972) 31, 38ff.; see also H. Schubart, “Morro de Mezquitilla 1976,” MM 18 (1977) 56–57.

The Tartessian Orientalizing Period

203

Fig. 1. Phoenician colonies (w) and Tartessian settlements of the orientalizing period (∑) in the southern Iberian Peninsuula.

years, we now have at our disposal a large amount of information to add to the Tartessian sites known for some time (Mesas de Asta, Carmona, and Carambolo). Stratigraphic sequences have been established at the sites of Quemados, the hills of Huelva (San Pedro and Esperanza), Cerro Macareno, Setefilla, and Valencina de la Concepción, all of which have permitted us to obtain a broader and more scientific perspective for studying the development of these indigenous communities during the 8th–6th centuries b.c. In addition, the excavation of two interesting Tartessian necropolises (La Joya and Setefilla) has amplified our knowledge of the burial practices of this period, which up until a short time ago remained restricted to the materials from the old excavations at Cruz del Negro, Acebuchal, Entremalo, Bencarrón, and Osuna. At a certain point in the history of these peoples, around 700 b.c., the generalized infiltration of Phoenician imports into the interior began, initiating a process of acculturation that throughout the 7th and 6th centuries would lead to a series of cultural changes. These changes are known as the orientalizing Tartessian horizon and are considered the cultural apogee of the indigenous population under Mediterranean influence. Without doubt, the most spectacular aspect of this process is the appearance of “Tartessian” gold and silver work (treasures of Aliseda, Carambolo, Ebora, Sines), bronze and glass oenochoes, and ivories from the region of Carmona. These objects, along with imported pottery, are found widely beyond the borders of Tartessian territory and

204

María Eugenia Aubet Semmler

mark the great natural land and water routes to the mining basins of the southwestern and northwestern Peninsula. It is precisely these luxury items that have caused the greatest controversies among specialists when they cannot determine whether they are Phoenician imports or simple local imitations. Generally, the analysis of these materials focuses on local evolution and uses criteria that are exclusively stylistic. After a phase of imports, local Tartessian artisans supposedly continued with a period of imitation of Phoenician works, which constituted the orientalizing phase proper. However, the problem is much more complex because of how much we still need to know in order to delimit the “western” cast in Phoenician art, which was ever more removed from its original sources. This style could have evolved in the area of Gadir through reduction and exhaustion of eastern elements, or, as might be the case in the production of ivories or pottery in the workshops of the interior, through maintenance of archaic traits that had already disappeared in the Mediterranean. 6 Nevertheless, the orientalizing horizon implies a much wider cultural phenomenon, with artistic production only one of its aspects. Studying this period necessarily presupposes the search for the roots of the process, the sociocultural situation of the Tartessian communities beginning with the Phoenician colonial stimulus, and, subsequently, the transformations that develop as much in the structures of the habitat as in the economic, social and artistic bases, with all the problems that an expansion of this type implies.

Questions about the Tartessian Orientalizing Period 1. Population, Habitat, and Economy The geographic framework in which the development of the Tartessian communities of the Final Bronze Age took place is very significant. On the basis of the relative chronology that the first imported Phoenician pottery furnishes us, we are in a position to form an approximate picture of what happened in the Spanish Southwest during the 8th–6th centuries b.c. Without exception, the known Tartessian towns were situated along principal routes of access to the mineral resources of the interior. These routes consisted of river networks and ancient cattle trails, well worn since prehistoric times. The site of Huelva, situated on the high hills of San Pedro and Esperanza, clearly dominates the estuary of the Tinto and Odiel Rivers and, consequently, the principal route of access to the silver and copper mines of Río 6. S. Moscati, “Interazioni culturali nel mondo fenicio,” RSF 2 (1974) 5ff.; M. E. Aubet Semmler, “La cerámica púnica de Setefilla,” Studia Archaeologia 42 (1976) 23–24; idem, Studia Archaeologia 52 (1979) 64; H. Schubart, “Westphönizische Teller,” in RSF 4 (1976) 191–92.

The Tartessian Orientalizing Period

205

Tinto. The Tartessian towns located on the banks of the Guadalquivir River (Mesas de Asta, Coria, Carambolo, Macareno, Valencina, and Setefilla) controlled the principal artery of communication in all of southern Spain, with the easiest access to the mining basins of the Sierra Morena and Estremadura (silver, copper, and gold). Situated on medium-high promontories, they controlled movement on old cattle trails heading toward the central plateau, as in the cases of Setefilla and the area of Los Alcores de Carmona. It becomes evident that the location of the Tartessian towns and others like them situated on the periphery (Quemados, Cástulo) had a dual purpose involving both strategic and economic objectives. The northernmost findings, such as those in Medellín, Valdegamas, and Siruela, show that the route of the Guadiana, the tributaries of the Guadalquivir, and the cattle roads that leave from the area of Alcalá and Lora del Río, were controlled in the same way by nuclei of Tartessian populations. Another example is the famous Roman road from Hispalis (Seville) to Asturica Augusta (Astorga), which united the Guadalquivir River valley with northwestern Spain and followed the old cattle trail, on which the town of Aliseda must have represented an important link. 7 Thus it seems clear that access to the mineral resources of the interior was controlled from the beginning by Tartessian communities. Due to progress obtained through Phoenician trade, these communities appear to have been in a position to impose conditions on the penetration of the Phoenicians toward the interior of the country and on their access to its natural resources. However, this situation does not necessarily presuppose the existence of an antagonism between the two communities. On the contrary, there are clear indications that both benefited from interchanges in which the indigenous population provided the raw material and the Phoenicians traded it. 8 The principal issue lies in determining how and when this indigenous population settled its towns with primarily economic objectives. We still know little about the origins of the Tartessian world in the lower Guadalquivir, and there are hardly any archaeological data from the second half of the second millennium b.c. and from the beginning of the first. During the Bronze Age it appears that the population was reduced to small isolated groups, known by their cist burials in Huelva and in parts of the province of Sevilla. 9 At the turn 7. M. E. Aubet Semmler, “La necrópolis de Setefilla, en Lora del Río,” Consejo Superior de Investigaciones Científicas (1973) 159; M. Almagro Gorbea, “El Bronce final y el período Orientalizante en Extremadura,” BPH 14 (1977) 204, 484. 8. See J. Maluquer, “Nuevas orientaciones,” 274–81. 9. M. del Amo, “Enterramientos en cista de la provincia de Huelva,” in Huelva: Prehistoria y Antigüedad (1975) 109ff.; M. Belén, M. Fernández Miranda, and J. P. Garrido, “Los orígenes de Huelva,” Huelva Arqueológica 3 (1977) 371; F. Fernández, D. Ruiz Mata, and S. Sancha, “Los enterramientos en cistas del cortijo de Chichina (Sanlúcar la Mayor, Sevilla),” TrPrHist 33 (1976) 351ff.

206

María Eugenia Aubet Semmler

of the millennium, the record is practically null, and it seems that the lower Guadalquivir was almost uninhabited after the Megalithic and Bell Beaker period, an epoch that had shown a special splendor in this area. 10 Recent archaeological excavations and the acquisition of stratigraphic sequences during the last few years in the provinces of Huelva, Sevilla, and Córdoba, allow us to glimpse a situation of profound cultural transformation in the region, beginning in the 9th–8th centuries b.c. In Colina de los Quemados, after a prolonged hiatus that followed a small settlement in the second millennium b.c., a population with high-quality burnished pottery settled down (stratum 16/17), living in dwellings with round foundations. This culture corresponded to the characteristic Tartessian horizon that preceded the first Phoenician imports of the 7th century b.c. 11 A similar phenomenon is observed in the settlement at Cabezo de San Pedro in Huelva. Here, after a considerable hiatus that followed weak vestiges of habitation during the second millennium b.c. in the region, a Final Bronze population flourished during the 9th–8th centuries b.c. This population was characterized by its pottery with burnished decorations and geometric painting of extraordinary quality, which preceded the first Phoenician pottery imported to the region around 700 b.c. A little after this point in time, which represents the oldest nucleus of population in the first millennium b.c. known in Huelva (designated Horizon 5 of San Pedro, strata XIII–XII), the town notably grew with the expansion of the settlement to the neighboring Cabezo de la Esperanza, at the middle of the 8th century. Beginning with level XI of San Pedro, the local community entered fully into the orientalizing period, well represented by the necropolis of La Joya. 12 The Tartessian towns of Carmona, Cerro Macareno, and Setefilla reflect a similar development. In Carmona the orientalizing period of the 7th century b.c. (Stratum IV) was preceded by a level (Stratum V) with burnished pottery from the central plateau and dated to the 8th century or earlier, apparently the period in which the settlement of the first millennium began. 13 In Cerro Macareno the first human activity is documented in Stratum 26, with handmade burnished and painted pottery of great quality, dated to the 8th century. The first Phoenician imports appear in Stratum 25. 14 Finally, in the town of Sete10. A. M. Muñoz, “La civilización pretartésica andaluza durante la Edad del Bronce,” V Sympos. Prehistoria Peninsular: Tartessos y sus problemas (1969) 33ff.; R. J. Harrison, T. Bubner, and V. Hibbs, “The Bell Beaker pottery from El Acebuchal, Carmona,” MM 17: 79–141. 11. J. M. Luzón and D. Ruiz Mata, “Las raíces de Córdoba. Estratigrafía de la Colina de los Quemados,” Consejo Superior de Investigaciones Científicas (1973) 11–12. 12. J. M. Blázquez, J. M. Luzón, and F. Gómez, Huelva Arqueológica 1 (1970) 13–16; M. Belén, M. Fernández Miranda, and J. P. Garrido, “Los orígenes de Huelva” 24, 370. 13. J. de M. Carriazo and K. Raddatz, “Ergebnisse einer ersten stratigraphischen Untersuchung in Carmona,” MM 2 (1961) 97–98. 14. M. Pellicer, J. L. Escacena, and M. Bendala, El Cerro Macareno (Excavaciones Arqueológicas en España 124; 1983).

The Tartessian Orientalizing Period

207

filla, the period of the first ceramic imports from the region of the Strait (Strata IX–VIII) was preceded by a rich stage in the 9th and 8th centuries b.c. with burnished pottery and structures of great quality. 15 Thus there are clear signs of the existence of a horizon in the Final Bronze that was relatively homogeneous in the Tartessian area, preceding the orientalizing period and corresponding to the moment in which the best quality Tartessian pottery was produced in the southwestern Peninsula. This is the same cultural horizon that is documented in the so-called Carambolo Alto, which years ago provided the famous handmade, painted geometric pottery, associated with a dwelling with a circular foundation. 16 According to the available data, this protohistoric horizon began with the relatively rapid development and flourishing of a population that chose to establish its settlements in areas practically uninhabited until then, but strategically located from an economic perspective. The causes of this demographic and geographic change should not be tied to the problem of the origin of the Tartessian people (since the cultural and ethnic antecedents of this population are perfectly documented in the Southwest from the second millennium b.c. on), 17 but rather we must search for the causes of such a radical transformation in the characteristics and social and economic structure of these populations. The people who established themselves permanently in Huelva and in the valley of the Guadalquivir during the 9th–8th centuries b.c. lived in a region of considerable strategic value. Their burial practices are currently unknown, but they lived in round dwellings of adobe and stonework, and they fashioned extraordinary burnished and painted pottery. Their presence presupposes not only an important territorial transformation of habitat, but also a marked concentration and increase of population, with all of the implications for social and economic order. That the selection of habitat responded to strategic and economic objectives is demonstrated by the continuity of settlement in these same places up through Iberian and Roman times. Traditionally, demographic changes of this magnitude are associated with a series of external or internal phenomena involving transformation of a social and economic order. Thus, a demographic increase can develop rapidly if minimal favorable conditions exist, due primarily to social, economic, biological, and ecological factors. A better means of survival, a decrease of internal conflict, a decrease in the death rate due to technological advances, and a greater sedentariness in fertile zones rich in pastures are primary catalysts for 15. M. E. Aubet Semmler, “La Mesa de Setefilla: la secuencia estratigráfica del Corte 1,” in Tartessos. Arqueología protohistórica del bajo Guadalquivir” (ed. M. E. Aubet Semmler; Barcelona, 1990) 297–338. 16. J. de M. Carriazo, Tartessos y el Carambolo (1973) 188ff. 17. H. Schubart, “Die Kultur der Bronzezeit im Südwesten der Iberischen Halbinsel,” MF 9 (1975) 147–50.

208

María Eugenia Aubet Semmler

a process of this type and may be applicable to the region of the lower Guadalquivir. In its turn, an increase in population stimulates production and technical and cultural development, intensifies the economy and social organization, and consequently accelerates the process of cultural change. 18 Therefore, the concentration of population and the demographic increase that are observed in the Tartessian area during the 9th–8th centuries b.c. have important historical consequences. It is important to point out the contrast between the peaceful nature of the Tartessian communties of the lower Guadalquivir and the Bronze Age populations of southern Portugal and Estremadura, which, judging by their necropolises of cists with decorated stelae from the 10th–8th centuries b.c., 19 were characterized by their warlike spirit. Even though they came rather late in this development, the necropolises of Carmona, La Joya, and Setefilla are notable for the almost total absence of weapons, and no important fortifications have been found. 20 The establishment and growth of these populations in the Tartessian area coincide, therefore, with the initiation of a long period of peace, which surely is one of the explanations for the population increase during this period. This peaceful period permitted not only a greater sedentariness, but also the development of a more stable economy. These populations had a long mining and cattle-raising tradition dating back to the Aeneolithic and Bronze Ages, and by settling in the Tartessian area, they secured one of the most fertile and mineral-rich regions in the southern Iberian Peninsula. The possible economic and social factors that occasion a demographic phenomenon of this type are multiple, as we have briefly enumerated. The causes of a territorial restructuring like this one, which implies the transfer of the principal focus of technological and cultural innovation from the area of the lower Tagus and the mid Guadiana to the valley of the lower Guadalquivir, could have come largely from the outside. We shall now examine corroborating signs seen in the Southwest during the 9th and 8th centuries b.c. To this period we can date the arrival in the Tagus and Guadiana area of cultural objects such as the shield with a “V”-shaped armhole and the elbow fibula. Although these objects belong to a still ill-defined horizon, they are 18. About prehistoric demographic changes, causes, and repercussions, see, in general, C. Renfrew, ed., The Explanation of Culture Change: Models in Prehistory (London, 1973) 296– 300, 329ff., 344, 529; P. J. Ucko, R. Tringham, and G. W. Dimbleby, eds., Man, Settlement and Urbanism (London, 1972) 386ff., 579, 827ff.; G. de G. Sieveking, I. H. Longworth, and K. E. Wilson, Problems in Economic and Social Archaeology (London, 1976) 502ff.; see also Civiltà del Lazio primitivo (catalog of the Rome exposition,1976) 27–28. 19. M. Almagro Basch, “Las estelas decoradas del Suroeste peninsular,” in BPH 8 (1966) 199–200. 20. See J. P. Garrido and E. Orta, “Excavaciones en la necrópolis de La Joya, Huelva II,” EAE 96 (1978) 202; one exception might be the fortification discovered in 1976 in the lower levels of the settlement at Setefilla, whose date, however, has not been determined yet.

spread is 6 points long

The Tartessian Orientalizing Period

209

signs of the presence of Mediterranean influences beginning in the 10th century b.c., influences that additionally coincide with the arrival in the Southeast and the lower Tagus of the first cremations known in the southern Peninsula. 21 Moreover, around 850–800 b.c. a maritime movement began in the Southwest directed to the large-scale recovery of manufactured bronze objects undoubtedly destined for overseas. Surely this was not yet true trade, but rather a traffic directed toward obtaining discarded metal, judging by the finds at the estuary of Huelva. 22 This traffic coincided with the period in which the Tartessian population settled on the estuary of Huelva and in the valley of the Guadalquivir. The first vestiges of activity in the Phoenician colonies at the Strait of Gibraltar date to the first half of the 8th century b.c., and by the middle of this century the density of Phoenician population on the shore was already considerable. 23 It is more than probable, therefore, that the metropolis of Gadir had already existed for a long time. The presence of Phoenician, Cypriot, Greek, and Italic elements in the area of the Strait of Gibraltar beginning in the 9th–8th centuries 24 denotes a very complex panorama, but one that fully coincides with the findings of this period in all of the central and western Mediterranean. The areas that the Greeks and Phoenicians controlled economically still had not been clearly defined, and maritime routes were frequented, indistinctly, by peoples from Chalcis, Euboea, Rhodes, Corinth, and Tyre, in a coexistence of economic interests very different from what would be seen in the 7th century b.c. 25 Thus, territorial restructuring and selection of new places to settle in the lower Guadalquivir by Tartessian populations in the Final Bronze Age coincided with the arrival of the first metal prospectors in the area of the Strait. These events clearly preceded the period of greatest economic activity on the part of the Phoenician colonies on the Andalusian shore and therefore occurred prior to the beginnings of eastern influence in the interior. 21. M. Almagro Basch, “El hallazgo de la Ría de Huelva y el final de la Edad del Bronce en el Occidente de Europa,” Ampurias 2 (1940) 108–22, 138–41; H. Hencken, “Carp’s tongue swords in Spain, France and Italy,” in Zephyrus 7 (1956) 137; M. Almagro Basch “Las estelas decoradas,” 156ff. About the elbow fibulas, see G. Delibes de Castro “Una inhumación triple de facies Cogotas I en San Román de la Hornija, Valladolid,” in TrPrHist (1978) 244–46; about cremations of the Final Bronze in the Southeast and Tagus region, H. Schubart, “Die Kultur der Bronzezeit,” 147, fig. 22; F. Molina, La cultura del Bronce final en el sudeste de la Península Ibérica (summary of doctoral thesis, Granada, 1977) 15–16. 22. M. Almagro Basch, “Depósito de bronces de la Ría de Huelva,” in Huelva: Prehistoria y Antigüedad (1975) 213–20; M. Almagro Gorbea, “El Bronce final,” 123–24, 491. 23. M. E. Aubet Semmler, “Excavaciones en las Chorreras (Málaga), Pyrenae 10 (1974) 107– 8; M. E. Aubet Semmler, G. Maass-Lindemann, and H. Schubart, “Chorreras,” MM 16 (1975) 158; H. Schubart, MM 18 (1977) 52–53. 24. See P. Rouillard, “Fragmentos de estilo geométrico y corintio medio en Huelva,” Huelva Arqueológica 3 (1977) 399. 25. See B. D’Agostino, “Tombe ‘principesche,’ ” 46–51; A. M. Bisi, “La presenza fenicia,” 16.

210

María Eugenia Aubet Semmler

In very little time, beginning in the 7th century b.c. when Phoenician imports began, we see signs of transformation in the interior of the Tartessian habitat. These changes are tied to a notable cultural and economic development that is reflected in a higher standard of living, in an economic surplus, and in the evolution of social organization. At this time the construction of rectangular dwellings with luxuriously plastered adobe walls and with clear Mediterranean influence began to predominate in Setefilla, reaching its highest architectural development around 600 b.c. 26 The same improvement in construction technique is observed at Cerro Macareno, especially from the beginning of the 6th century, 27 in Quemados, and in Carambolo Bajo. 28 Beginning in the 7th century b.c., we see a rapid enrichment of these populations due to the trade in metals. As mentioned earlier, it is evident that control of mineral and cattle resources was in the hands of the Tartessian population, and there is no indication of hostility or political tension with the Phoenician colonies on the coast. On the contrary, the establishment during this period of Phoenician artisans in Tartessos who fashioned ceramics, ivories, and bronzes on site, the geographic proximity of Tartessian centers such as Mesas de Asta, almost at the gates of Gadir, and the absence of signs of violence in these indigenous towns are obvious signs of a period of peaceful coexistence between the two communities. The absence of permanent Phoenician urban nuclei in Tartessian territory, together with the archaeological data that we possess on the indigenous settlements, indicate that from the beginning the Tartessian towns were the principal foci of metallurgic activity, while the trade in metals was in the hands of the Phoenician colonies. Undoubtedly, the acquisition and trade of metal stimulated, beginning in the 7th century, the rapid growth of a Phoenician mercantile middle class that we know by means of its tombs at Almuñécar and Trayamar. Strictly speaking, we cannot apply to this period a fundamentally modern critique of competitive markets and mechanisms based on the laws of supply and demand. Phoenician trade seems to have been based simply on the acquisition of basic merchandise, such as gold, silver, tin, copper, grain, wood, and skins, in exchange for oil, wine, and luxury goods destined for the indigenous clientele, serving its economic needs and immediate provision. Therefore, it was probably not a true example of mercantile activity in the modern sense of the word, but rather a mechanism of exchange that, according to the demands of the moment, must have evolved with time. 29 26. M. E. Aubet Semmler, “La Mesa de Setefilla.” 27. M. Pellicer, J. L. Escacena, and M. Bendala, El Cerro Macareno. 28. J. M. Luzón and D. Ruiz Mata, “Las raíces de Córdoba . . . Quemados,” 15–16, 20–21; J. de M. Carriazo, Tartessos, 236. 29. About these issues, see especially E. Lepore, “Osservazioni sul rapporto tra fatti economici e fatti di colonizzazione in Occidente,” Dialoghi di Archeologia 3.1–2 (1969) 180–85.

The Tartessian Orientalizing Period

211

Beginning in the second half of the 7th century b.c. we observe a great unfolding of metallurgic activity in the Tartessian towns of the interior, demonstrated by the presence of mineral slag in indigenous dwellings and by the beginning of silver and copper mining in Riotinto. 30 Along with these activities and surely due to a large economic surplus, the Tartessian population maintained an economy based primarily on livestock, which must have consituted another considerable source of riches and was favored by the location of the towns. The excavations at Huelva, El Carambolo, and Setefilla prove that sheep-raising and, more importantly, cattle-raising provided a strong economic base. 31 The economic benefits obtained by the Tartessian communities (and especially the dominant social classes) through these two principal resources undoubtedly stimulated attempts to broaden their sphere of economic influence. It is to the beginning of the 6th century that we date the greatest share of Tartessian finds of a Phoenican type in the periphery of the lower Guadalquivir. The Tartessian population was surely the main intermediary in the “orientalizing” expansion toward the north and east. Tartessian pottery with burnished decorations, bronzes, jewels, and pottery of Phoenician tradition or fabrication have been found in Medellín, Valdegamas, Aliseda, Cástulo, Castellones de Ceal, and also in Frigiliana. 32 The economic apex reached by the Tartessian communities at the beginning of the second half of the 7th century b.c. and, above all, at the beginning of the 6th century would imply a greater complexity in social organization, which probably led to a continuous and increasingly marked hierarchization.

2. Burial Practices and Social Organization Apart from the burials with decorated stelae, we unfortunately know nothing about the burial customs of the pre-orientalizing Tartessian horizon during the 9th–8th centuries b.c. Consequently we are unfamiliar with the socioreligious structure of this important transitional period stretching from the necropolises of cists in the Southwest to the first true Tartessian burials. The principal characteristics of the burial practices of the orientalizing period are deduced from three large groups of findings: the necropolises in the 30. M. Belén, M. Fernández Miranda, and J. P. Garrido, “Los orígenes de Huelva,” 373; A. Blanco, J. M. Luzón, and D. Ruiz, Excavaciones arqueológicas en el Cerro Salomón (Riotinto, Huelva) (1970) 13; J. M. Luzón and D. Ruiz Mata, “Las raíces de Córdoba . . . Quemados,” 17. 31. M. E. Aubet Semmler, “La necrópolis de Setefilla, en Lora del Río, (Sevilla)” Consejo Superior de Investigaciones Científicas (1975) 158–59; J. P. Garrido and E. Orta, “Excavaciones . . . La Joya II,” (1978) 199–201; J. de M. Carriazo, Tartessos, 453, 459. 32. A. Blanco Freijeiro, “Orientalia II,” AEArq. 33 (1960) 26–31; A. Arribas and J. Wilkins, “La necrópolis fenicia del Cortijo de las Sombras (Frigiliana, Málaga),” Pyrenae 5 (1969) 5–7; M. Almagro Gorbea, “El Bronce final,” 239–42, 317ff.

212

María Eugenia Aubet Semmler

region of Carmona (Cruz del Negro, Acebuchal, Entremalo, etc.), with serious problems of interpretation due to the period in which they were excavated, 33 the necropolis of La Joya in Huelva, 34 and that of Setefilla in the province of Sevilla. 35 The burial practices that we see in these necropolises are notably complex, with such a confluence of elements that they seem to correspond to a culture that was very open to external suggestion. This openness, however, is common to every period of transition in general and characteristic of the Final Bronze Age in particular. 36 During the orientalizing period of the 7th and 6th centuries b.c., the most characteristic burial in the lower Guadalquivir was inhumation or cremation under a mound. The tombs form veritable fields of mounds, some of them reaching considerable proportions, as in Alcores de Carmona, Setefilla, and perhaps also in Mesas de Asta and La Joya. 37 Contrary to the traditional interpretation that attributes this type of burial, along with cremation, to northern influences of a Central European origin, we know today that the Tartessian mounds are characteristic of southern Iberia and seem to be an extension of earlier Bronze Age burial practices. The Portuguese necropolises of the second half of the second millennium b.c., such as at Atalaia, form areas of mounds around a central tomb that evidently served as a final resting place for a distinguished person of the community. This type of grouping apparently served as a precedent for the fields of cists with warlike decorated stelae of the Final Bronze Age. 38 This system of clustering mounds and cists clearly demonstrates an ancient communal tradition in which clan relationships constitute the principal base of social organization. This is the same sort of grouping that we see, for example, in the Tartessian mounds of Setefilla during the 7th–6th centuries: relatively large mounds that cover cremations that are placed in circles. These circles are perfectly differentiated from each other and at times are defined by slabs of stone or sunken stelae, some of them decorated. In mounds A and H, large central burial chambers have been superimposed on the cremation area, chambers 33. G. Bonsor, “Les colonies agricoles pré-romaines de la Vallée du Bétis,” Revue Archéologique 35 (1899). 34. J. P. Garrido and E. Orta, “La tumba orientalizante de La Joya, Huelva,” TrPrHist 11 (1963) (tomb 1); idem, “Excavaciones en la necrópolis de La Joya,” EAE 71 (1970) (tombs 2–10); idem, “Excavaciones . . . La Joya II,” (1978) (tombs 11–19). 35. M. E. Aubet Semmler, “La necrópolis de Setefilla, en Lora del Río (Sevilla),” Consejo Superior de Investigaciones Científicas (1975) (mound A); idem, Consejo Superior de Investigaciones Científicas (1978) (mound B). 36. See F. Molina, La cultura del Bronce final,” 8. 37. M. Esteve Guerrero, “Asta Regia: una ciudad tartésica,” V Sympos. Prehistoria Peninsular, Jerez 1968 (1969) 112. 38. H. Schubart, “Die Kultur der Bronzezeit,” 165–66; M. Almagro Gorbea, “El Bronce final,” 177, 185.

The Tartessian Orientalizing Period

213

that were probably intended for a privileged person from the group, who, in contrast to the rest of the population, practiced inhumation. 39 Without doubt, this system of burial “circles” indicates a social structure based on family groups with blood lines, from which a local aristocracy emerged in some cases. The aristrocracy affirmed its social prestige and perhaps its ancient family ancestry by maintaining a burial tradition going back to earlier Bronze Age practices in the Southwest and that had already fallen into general disuse by this time. On the other hand, the burial practices that are observed in the Tartessian necropolises of Alcores de Carmona indicate a panorama so heterogeneous that during the orientalizing period it seems as though the relatively uniform beliefs of the preceding periods had been lost. In Alcores, graves with both inhumations and cremations beneath mounds coexist without clear distinction (Bencarrón, Carmona, Entremalo), even though cremations in mounds (Acebuchal, Alcantarilla, Alcaudete, Cañada de Ruiz Sánchez) or in pit-graves (Cruz del Negro) predominate. 40 All of this seems to indicate that the diversity in burial practices is due more to social differentiation (or to different family traditions) than to chronological evolution. The most important funerary monument in Alcores (tomb G of Acebuchal, 6th century b.c.) preserved in its center a rectangular chamber built of masonry (analogous to those in Setefilla) containing a double inhumation. 41 In contrast, the rich tombs at Cañada de Ruiz Sánchez from the 7th century b.c. and at Alcantarilla contained cremations. 42 These findings could indicate that beginning in the 6th century b.c. there was a resurgence of the practice of inhumation on the part of the socially dominant elite. A similar confluence of rites and traditions is seen in the excavated sector of the necropolis of La Joya, a sector that undoubtedly corresponded to the socially privileged stratum of the Tartessian community of Huelva. The richness of the funerary offerings indicates a society with great buying power and marked social stratification, characteristic of a Tartessian settlement that had been very favored by Phoenician trade due to its geographic location. In La Joya we see such diversity in burial practices that we can affirm that no two 39. G. E. Bonsor and R. Thouvenot, “Nécropole ibérique de Setefilla,” in Bibliothèque de l’École des Hautes Études Hispaniques XIV (Bordeaux-Paris, 1928) 21–25; M. E. Aubet Semmler, “Materiales púnico-tartesios de Setefilla en la Col. Bonsor,” Boletín Seminario de Estudios de Arte y Arqueología 39 (1973) 6ff.; idem, “La necrópolis de Setefilla, en Lora del Río (Sevilla),” Consejo Superior de Investigaciones Científicas (1975) 66–71, 157; idem, “La necrópolis de Setefilla, en Lora del Río (Sevilla),” Consejo Superior de Investigaciones Científicas (1978) 169–71. 40. G. Bonsor, “Les colonies agricoles,” 128f. 41. G. Bonsor, “Les colonies agricoles,” 24–27; W. Schüle, “Die Meseta-Kulturen der Iberischen Halbinsel,” MF 3 (1969) 147. 42. G. Bonsor, “Les colonies agricoles,” 50–59.

214

María Eugenia Aubet Semmler

Fig. 2. Alabaster vases from the necropolis of La Joya (Huelva). Photo: Phoenix Data Systems; courtesy Fondo del Museo de Huelva.

identical graves exist: there are simple cremations, large tombs containing cremations and chariots, graves with both inhumations and cremations, inhumations with bodies in violent positions, appearing as if they were stoned, etc. 43 The significance of such a marked ideological heterogeneity is not easy to interpret, even if it appears evident that this phenomenon coincides with the culmination of eastern influence on the Tartessian population, and apparently, with the beginning of the decay and progressive disintegration of earlier burial traditions. The interpretation depends on knowing up to what point Phoenician influence had a bearing on Tartessian burial practices, and, consequently, on the social and religious ideology of these “orientalized” communities. If cremation in the Southwest does not necessarily originate from northern influences, the possibility exists of an influence proceeding from the Phoenician colonial world, where this rite is documented for the first time in Almuñécar around 700 b.c. A change in burial practice always imples profound ideological transformations, especially in a population that adopts cremation for the first time: belief in reincarnation, in the purifying role of fire, and a very specific attitude toward death and the afterlife. This presupposes a slow process, and additionally one would expect that the adoption of a new burial practice would be accompanied at the beginning by the same rituals that 43. J. P. Garrido and E. Orta, “Excavaciones . . . la Joya II,” (1978) 17–18.

The Tartessian Orientalizing Period

215

Fig. 3. Bronze axle ends (“hub caps”) with lion heads, from a light chariot found in a tomb at the necropolis of La Joya (Huelva). Photo: Phoenix Data Systems; courtesy Fondo del Museo de Huelva.

characterize the practice in the community of origin. Nevertheless, the Tartessian cremations and burial customs at La Joya, Cruz del Negro, and Setefilla, besides being verified from the beginning of the 7th century on, do not bear a close resemblance to the Phoenician cremations at Almuñécar and Trayamar. 44 The alternative explanation may be that the change in burial practices that occurred among the Tartessian population was gestating for some time in southern Spain. Peoples who practiced cremation had arrived at the estuary of the Tagus and in the Southeast (Parazuelos, Campos, Caldero de Mojácar) already in the 9th and 8th centuries b.c., 45 and, consequently, at a time clearly prior to the orientalizing period. A second issue relates to the chronology and the social structure that are reflected in the orientalizing tombs of the lower Guadalquivir and of Huelva. It appears that already during the second millennium b.c. a process of social transformation had begun among the populations of the southwestern Peninsula. The subsequent appearance of structural differences in the Portuguese mounds indicates the birth of a ruling class that evolved from the groups of 44. M. E. Aubet Semmler, “La necrópolis de Setefilla, en Lora del Río (Sevilla),” Consejo Superior de Investigaciones Científicas (1975) 155–65; M. Almagro Gorbea, “El Bronce Final,” 385, 503. 45. H. Schubart, “Die Kultur der Bronzezeit,” 147; F. Molina, “La cultura del Bronce final,” 15–16.

216

María Eugenia Aubet Semmler

warriors and herdsmen. These were the tribal chiefs who were the owners of the decorated stelae and who were occasionally entombed in great burial monuments. 46 Thus, it is logical to believe that the great Tartessian mounds are no more than the end of a process of social differentiation that was accelerated and accentuated under eastern influence. The economic transformations driven by Phoenician trade, together with the demographic changes mentioned above toward specialization in work and social hierarchization, would have given rise to a surplus of production that converted these leaders of herdsmen into true foci of economic and social power. All of the Phoenician luxury imports, like all eastern items considered to be prestige items, were absorbed by this restricted sector of Tartessian society, now in control of the natural resources of the territory. These Tartessian rulers differentiated themselves from the rest of the population by their material culture and by a series of ideological factors that make it possible to speak of a “cultural unity among the ruling groups.” A social elite depends for survival on being able to maintain its political dominance, its superior social status, and its control over economic mechanisms. All of this can be demonstrated by constructing monumental tombs that stand out from the others and by adopting a lifestyle (imports, luxury goods, symbols of prestige) that sets them apart from the rest of the population. 47 In Alcores and in Setefilla the large central chamber tombs beneath the mounds contained the richest offerings ( jewels, ivories, bronze jugs, etc.), while the rest of the population, who usually cremated, absorbed imported items only sporadically, mainly pottery. In addition, the construction of gigantic mounds like those at Acebuchal, Alcantarilla, and Setefilla, and the central burial chambers themselves exalted the triumph of the aristocracy that held power, claiming in its own way a role equal to the great Phoenician mercantile bourgeoisie buried in the chamber tombs of Trayamar. With their burial practices, the Tartessian rulers made themselves stand out from the community, either by adopting an eminently traditional ritual, such as inhumation, or by equating themselves to the Phoenician colonial class by imitating the rite of cremation in amphorae and using the chariot as a symbol of prestige, as in Tombs 12, 17, and 19 in La Joya. 48 It is evident that the circulation of Phoenician luxury imports and prestige goods was restricted to this privileged social stratum. The gold and silver 46. K. Spindler and O. Da Veiga Ferreira, in MM 14 (1973) 60ff., fig. 9 (reconstruction). 47. About this theme, see C. Renfrew, The Explanation of Culture Change, 474, 582. 48. J. P. Garrido and E. Orta, “Excavaciones . . . La Joya II,” (1978), 167. A similar phenomenon is observed among the Italic aristocracy during the orientalizing period; in this case, the local leaders imitated the burial practices of the Euboean colonial class of Cumas, thus claiming an equal role; B. D’Agostino, “Tombe ‘principesche,’ ” 58–61; A. Bedini and F. Cordano, “L’ottavo secolo . . . Castel di Decima,” 295.

spread is 6 points short

The Tartessian Orientalizing Period

217

work of the orientalizing period were the exclusive domain of these local leaders. It is no coincidence that the most important group of findings of this type, that of Aliseda, came from a cremation probably analogous to those of Acebuchal and Setefilla—and was found in an area of strong Tartessian social influence. 49 All of this presupposes that the orientalizing horizon, and with it, the process of acculturation originating from contact between the Tartessian world and the colonial Phoenician world, constituted a phenomenon with special impact on the indigenous aristocracy. This aristocracy was more open to external cultural influences and was the principal beneficiary of these interchanges, while the rest of the population remained relatively passive in the face of cultural change. Consequently, this was a cultural transformation of a selective nature. Without exception, all of the large chambered burial mounds in Setefilla and Acebuchal and the most spectacular tombs at La Joya are dated to the beginning of the 6th century b.c. and, at the earliest, to the end of the 7th century. 50 Thus they indicate that the high point of political and social power of the Tartessian aristocracy developed at a relatively late date during the orientalizing period (around 600 b.c.), and practically at the onset of a decline, or, if you will, an interruption in Phoenician trade in the area of the Strait. Without doubt, the significance of this whole process is extremely interesting. Important economic transformations accompanied this orientalizing period, with the consequent accumulation of riches in a few hands and a social differentiation that was ever more accentuated, based on local rulers who coordinated economic, social, and religious activities. These transaformations were significant signs of a proto-urban structure and a prelude to the process leading toward an urban state, a process that was in a very advanced stage around 600 b.c. 51 Throughout all this development, the Phoenician factor was probably limited to accelerating these socioeconomic changes. However, the culmination of the process appeared very late in the lower Guadalquivir, without progressing beyond the structures typical of a transitional period. This can be inferred from the burial practices described above and from the sociocultural structure that were not yet fixed, but rather remained in a stage of transformation and still open to multiple influences and traditions. This process toward urban life probably suffered a regression during the 6th century b.c., one of the least-known periods of Spanish protohistory, and, nevertheless, the period of true transition to the urban society of the Iberian Age. 49. M. Almagro Gorbea, “El Bronce final,” 219–20. 50. One exception could be the mound of Cañada de Ruiz Sánchez near Carmona, whose treasure might date to the middle of the 7th century b.c. (G. Bonsor, “Les colonies agricoles,” 55–56.) 51. For the significance of a stage with these characteristics (the stage called “chiefdoms”), see C. Renfrew, The Explanation of Culture Change, 542.

218

María Eugenia Aubet Semmler

3. Material Culture and Artisanry Studies carried out during the last few years in the provinces of Sevilla and Huelva allow us to glimpse a very heterogeneous Tartessian cultural panorama. However, instead of resolving problems of interpretion, the samples of material culture such as pottery, bronzes, and gold and silver work become ever more difficult to classify within a set terminology. In reality, it is difficult to determine the boundaries between Phoenician art and Tartessian art of the orientalizing period, and on occasion it is impossible to distinguish a Phoenician import from a local imitation, since their assessment depends on contexts and criteria that at times are less than objective. For several years it was fashionable to interpret all artistic material of an eastern style as “orientalizing Tartessian,” a practice that emerged as a reaction to a stage in which all the material of this period was attributed to Phoenician art. However, an intermediate route was set some years ago, recognizing two distinct and successive chronological phases within the orientalizing phenomenon: an initial phase of Phoenician imports, and a second consecutive phase of indigenous imitiations, or a true orientalizing horizon, all of this as part of a single process. 52 In the last few years, however, we have witnessed a process of reassessment of provincial Phoenician art, in the form of workshops with specialized artisans rooted in an indigenous environment and operating on the basis of local demand. 53 This interpretation, which actually could underrate the artistic capacity of the Tartessian population as a whole, still does not resolve many of the problems raised by the artisanry of the orientalizing period. What follows is the briefest possible discussion of the present state of the question. The jewels of Aliseda, the ivories from the region of Carmona, and the bronze oenochoes constitute, without doubt, the most significant and spectacular groups of findings from this period. The treasure of Aliseda, discovered, it appears, in an indigenous tomb of Estremadura and dated to around 600 b.c., was studied in detail by Blanco in 1956. 54 This author, basing his arguments on technical irregularities that he saw in the fabrication of the golden jewels and pointing out elements of Etruscan influence in their decoration, judged the jewels to be the product of an orientalizing Tartessian workshop. Without forgetting that among the finds at Aliseda there are pieces of authentic eastern fabrication, such as a glass jug, we have raised several questions about Blanco’s interpretation. One such question is the significance of the 52. See S. Moscati, “Problematica della civilità fenicia,” Studi Sardi 46 (1974) 105–7; idem, I Cartaginesi in Italia, 307. 53. H. G. Niemeyer, in Bonner Jahrbücher 176 (1976) 454; M. E. Aubet Semmler, “Los marfiles fenicios,” 66. 54. 55. A. Blanco, “Orientalia,” AEArq. 29 (1956) 27–32.

The Tartessian Orientalizing Period

219

Etruscan factor in the decoration of the jewels, especially if we take into account that Etruscan gold and silver work from the orientalizing period presents some of the same eastern influences as the work from Aliseda, with the additional possibility that the Aliseda jewels were fabricated in Pitecusas. 55 Recently a western Phoenician origin has been proposed (perhaps Gadir) for the whole complex of findings at Aliseda. 56 As an example, a great local gold and silver production center in Phoenician Tharros fashioned pieces inspired by eastern models beginning in the 7th century b.c., as well as fully autonomous pieces with technical irregularities in their fabrication. 57 The center in Tharros demonstrates the existence of western Phoenician gold and silver workshops within a vast and complex panorama of provincial and native Phoenician art, a fact that obliges us to work with caution when evaluating these western pieces. The ivories of the Carmona region, found at the end of the 19th century and the beginning of the 20th, have also been the object of diverse interpretations based once more, in the majority of cases, on stylistic or aesthetic criteria. 58 That there was a gradual evolution of styles in these pieces (beginning with a Phoenician workshop that probably produced its samples of highest quality during the 7th century b.c. and followed by Tartessian production of more mediocre objects during the 6th–4th centuries b.c.) is no longer a viable hypothesis. Everything appears to indicate that the ivories form a fairly homogeneous group coming from the same school of artisans in the 7th and 6th centuries b.c., centered in the lower Guadalquivir and clearly inspired by the western Phoenicians. 59 With respect to the Peninsular bronze oenochoes, the problem of their origin is generally presented by employing the same criteria used for the silver and gold work and the ivories. The examples of Phoenician fabrication, which are older and “purer” ( jugs from Cañada de Ruiz Sánchez, Alcalá del Río, Aliseda, Coca, and Torres Vedras), are clearly differentiated from those of Tartessian production, which are more recent and went through a process of evolution ( jugs from Mérida, Niebla, Siruela, Valdegamas, Villanueva de la Vera, and La Joya). 60 Of great interest for this question is the recent study of 55. G. Buchner, “Nuovi aspetti,” 81–85. 56. M. Almagro Gorbea, “El Bronce final,” 221. 57. S. Moscati, “Centri artigianali fenici in Italia,” RSF 1 (1973) 49–51. 58. G. E. Bonsor, Early engraved ivories (New York, 1928); A. Blanco Freijeiro, “Orientalia II,” AEArq. 33 (1960) 3–25; M. E. Aubet Semmler, “Los marfiles fenicios,” 14. 59. B. Freyer-Schauenburg, Elfenbeine aus dem samischen Heraion (Hamburg, 1966) 109; M. E. Aubet Semmler, “Los marfiles fenicios,” 60–67. 60. A. Blanco, “Orientalia,” AEArq. 29 (1956) 3–11; A. García Bellido, “Los bronces tartéssicos,” V Sypos. Prehistoria Peninsular 1966 (1968) 170–71; B. Grau-Zimmermann, “Phönikische Metalkannen in den Orientalisierenden Horizonten des Mittelmeerraums,” MM 19 (1978) 161–218.

220

María Eugenia Aubet Semmler

the group of bronzes from the necropolis of La Joya, precisely the site that has produced the greatest number of oenochoes of a late or Tartessian “orientalizing” type. In fact, the rich funerary offerings of La Joya suggest a permanent presence of Phoenician artisans on the estuary of Huelva, satisfying on site the demand for luxury goods by the local indigenous population. 61 The spectographic and metallurgic analyses performed on the bronzes from the necropolis have revealed that the fabrication of these orientalizing objects implies an excellent knowledge of the techniques of fusion and welding, as well as a superior command of metallurgy by these supposedly indigenous artisans. Their knowledge of the properties of the alloys and the process of producing these pieces presupposes the existence, therefore, of authentic specialists in this field. 62 This expertise is surprising since Huelva is not an area where a strong tradition of bronze metallurgy existed earlier of the magnitude that we see, for example, in pre-orientalizing Etruria. In this respect, we should point out that a western Phoenician origin has also been recently proposed for the late jugs of the 6th century b.c. from La Joya, Mérida, Siruela, and Valdegamas. 63 Finally we will examine the questions raised by studying the pottery of the orientalizing period. We have indicated that the arrival of the first Phoenician imported pottery in the interior should be dated to around 700 b.c. As with the gold and siver work and the bronzes, it is logical to believe that this phase of imports was followed by a period of local ceramics produced on the potter’s wheel, which gradually replaced the burnished handmade pottery. However, this evolution has not been clearly documented in any Tartessian settlement, since in its initial period it is not easy to differentiate authentic imported pottery from the first local imitations. In addition, the chronology of the earliest Phoenician ceramic imports in the Tartessian area has not been sufficiently clarified, and, like other cultural aspects of the orientalizing period, the evidence appears to indicate that the general adoption of the potter’s wheel was not immediate. Currently the red slip plates are among the best-dated Phoenician ceramic forms, and they constitute an important element for dating the sequence of Tartessian settlements in the interior. 64 However, it has been verified in Setefilla, for example, that plates of an archaic type, which can be dated in Toscanos and Chorreras to the second half of the 8th century b.c., appear here in a context belonging to the beginning of the 6th century or perhaps the end of the 7th century. 65 On the necropolis of La Joya, especially in the ceramic grave goods of Tombs 1 and 9, something 61. 62. 63. 64. 65.

See J. P. Garrido and E. Orta, “Excavaciones . . . La Joya II,” (1978) 199–200. A. Escalera Ureña, in J. P. Garrido and E. Orta, “Excavaciones . . . La Joya II,” (1978) 225. M. Almagro Gorbea, “El Bronce final,” 242. H. Schubart, “Westphönizische Teller,” 179ff. M. E. Aubet Semmler, La cerámica púnica de Setefilla, 22–24.

The Tartessian Orientalizing Period

221

similar takes place. 66 We also see these archaic 8th-century plates appearing in the lower strata of Cabezo de San Pedro, or mixed with later types from the 7th century in additional finds in Huelva. 67 This particular development of Phoenician ceramic typology in the Tartessian area, and especially the continuation of archaic forms into the 6th century, appears to be a phenomenon specific to the area west of Gibraltar, in other words, the Tartessian territory itself. In addition, we note that the Phoenician ceramic forms that are strongly rooted in the Mediterranean during the 8th century and even earlier, such as the à chardon vases, the painted cups, and the globular urns of the Cruz del Negro type, are extremely rare in the Phoenician colonies in the area of the Strait, while they appear in relative abundance in the Tartessian area. 68 The fact that this ceramic typology does not fit the evolution observed at the Phoenician sites on the coast does not necessarily presuppose the existence of Tartessian imitations of eastern wheel-made pottery. It is difficult to imagine that, when the Tartessian artisans produced their first wheel-made pottery during the 7th and 6th centuries b.c., they imitated styles that had disappeared long ago. It is more logical to suppose the presence of Phoenician workshops of an archaic nature, operating in the midst of indigenous territory from the 8th century b.c. on. These workshops probably did not undergo the same evolution as those near the colonies of the Strait of Gibraltar, which perhaps had relationships with Gadir. The problems that surround Phoenician pottery found in the Tartessian towns and necropolises certainly render more difficult the dating of the first local wheel-made pottery produced in the interior. In this respect we should point out that even if these exceptional ceramic types constituted the first local Tartessian imitations, this would not explain, among other things, the fact that during the 7th century the Tartessian artisans used the potter’s wheel only to fashion a few eastern-style cups, while continuing to produce the greater part of their pottery by hand. Normally, the introduction of the potter’s wheel constitutes a relatively quick process because of its simple dynamics, because it is learned rapidly, and because of its high levels of productivity. Nevertheless, throughout the 7th century, handmade pottery with or without the traditional burnished decoration clearly predominated in the Tartessian towns, and in some cases may have even continued until the 5th century b.c. 69 It is evident, 66. J. P. Garrido, “Excavaciones en la necrópolis de La Joya,” EAE 71 (1970) 52–55; H. Schubart, “Westphönizische Teller,” 188. 67. M. Belén, Fernández Miranda, and J. P. Garrido, “Los orígenes de Huelva,” 149f., fig. 83; 89, figs. 94, 112; H. Schubart, “Westphönizische Teller,” 189. 68. M. E. Aubet Semmler, “La cerámica a torno de la Cruz del Negro,” Ampurias 38–40 (1976–78) 267–87; A. M. Bisi, “La presenza fenicia,” 16. 69. M. Belén, M. Fernández Miranda, and J. P. Garrido, “Los orígenes de Huelva,” 24; M. E. Aubet Semmler, “La Mesa de Setefilla 1976,” 297–338.

222

María Eugenia Aubet Semmler

therefore, that local Tartessian pottery was deeply rooted among the peoples of the interior, even in the face of the innovation that the potter’s wheel represented. The potter’s wheel did not seem to become generalized in the interior until the 6th century b.c., on the eve of the transition to the Iberian world. 70 The supposed profound process of Phoenician influence on the material culture of the Tartessian interior remains as problematic and ill defined as the origin of social differentiation and cremation among the Tartessian peoples of the Southwest.

Conclusions In spite of the great number of unsettled issues that the orientalizing Tartessian period raises, its cultural characteristics demand that we ask a fundamental question. Did Phoenician influence on the indigenous populations of the Spanish Southwest promote an authentic process of acculturation during the 7th and 6th centuries b.c., a process which concluded with a profound cultural change and involved the definitive step of these peoples toward urban life? Without doubt, the establishment of Phoenician colonies in the area of the Strait accelerated a process of economic and social transformations that had been gestating in the Southwest since the beginning of the first millennium b.c. The rapid enrichment of these populations beginning in the 7th century b.c. due to their privileged location with respect to the Phoenician trade in metals helped these sociocultural structures to reach a very advanced stage within the evolution toward urban life. Nevertheless, the high point of these populations seems to have developed at a relatively late period, around 600 b.c., when the Phoenician markets on the coast were at the point of losing their economic and commercial function and when the intense traffic in the area of the Strait, uninterrupted since at least the 8th century b.c., was about to cease. If we accept that urban society in the southern Peninsula was not reached until well into the 6th century b.c., and, above all, beginning in the 5th century with the appearance of the Iberian cities, it does not seem plausible that either the Phoenician colonial stimulus or its most immediate effect, the orientalizing Tartessian period, constitute basic cultural factors in the birth of Iberian culture. If the principal stimulus in the process toward urban life had been the Tartessian orientalizing horizon, we could not explain why the first and most powerful foci of Iberianism appeared in upper Andalusia and southeastern Spain, that is, on the periphery of the Tartessian world, and not, as one would expect, at its epicenter on the shores of the Strait of Gibraltar or in the provinces of Sevilla and Huelva. Everything appears to indicate that a change 70. See M. Almagro Gorbea, “El Bronce final,” 394ff.

The Tartessian Orientalizing Period

223

took place in the currents of trade in the western Mediterranean beginning in 600 b.c., similar to what occurred in the central Mediterranean. The appearance of the Greek element seems to coincide with a change in the foci of economic attraction, now toward the upper Guadalquivir and the Southeastern shore, to the detriment of the Tartessian area. In the Tartessian towns during the 6th and 5th centuries b.c., we see a “parenthesis” or, if you will, a cultural regression and the beginning of a process of generalized decay, according to the archaeological data that we possess. These centuries constitute a very obscure period, with little in the archaeological record. The population was notably reduced, for example in Huelva, where signs of recovery are not seen until the 4th century. 71 Also in Cerro Macareno, Carmona, and Setefilla, the 6th and 5th centuries b.c. reflect a certain cultural hiatus that preceded an Iberian culture already fully formed. 72 Let us see, then, what significance the orientalizing period had for the development of proto-urban communities in southern Spain. In the first place, we cannot deny that the impact of a clearly urban and more advanced culture like that of the Phoenicians upon the essentially cattle-based Tartessian communities, beginning in the 8th and 7th centuries, influenced the economic structure and material culture in a special way. However, the consequent social and ideological transformations probably conserved a profoundly traditional root. The changes in the economy of these settlements most likely accentuated the social differentiation already existing in the population, and it can be said that the immense majority of the cultural loans introduced to the Tartessian world from the Phoenician coast affected mostly the aristocratic stratum of society. The technical advances characteristic of this process of “acculturation” came about gradually and very late, as we have seen in the adoption of the potter’s wheel and the improvement of construction techniques. Moreover, the diffusion of eastern elements did not extend to the entire Tartessian population, where traditional forms of the Final Bronze continued until the 6th century b.c. Therefore, the dynamics of cultural change reached a very limited sector of the population, and for this reason we do not believe that we can speak either of a true “process of acculturation” nor of profound cultural change. Burial practices and ideological traditions, which entered into an authentic phase of transition or disintegration during the orientalizing period, were not basically affected by Phoenician influence. Everything appears to indicate that the greatest transformation in the habitat and population of these communities, as well as the beginning of the process of 71. J. P. Garrido and E. Orta, “Excavaciones en . . . La Joya II,” (1978) 14; M. Belén, M. Fernández Miranda, and J. P. Garrido, “Los orígenes de Huelva,” 370–73. 72. M. Pellicer, J. L. Escacena, and M. Bendala, El Cerro Macareno; J. de M. Carriazo and K. Raddatz, “Ergebnisse . . . in Carmona,” 101–4.

224

María Eugenia Aubet Semmler

social differentiation, had started long before the Phoenician influence from the East. The desire for prestige on the part of the Tartessian aristocracy was the strongest factor in stimulating the process of diffusion and assimilation of eastern forms and customs. Therefore it is more appropriate to speak of a change in the lifestyle of the economically most powerful sectors of society rather than a generalized cultural change in the Tartessian area under Phoenician influence. During the orientalizing period, the historic proto-urbanizing evolution of the peoples of the interior accelerated. However, the Tartessian communities represent a typical transitional phase within the evolution toward forms of urban life, with cultural structures that are characteristic of a proto-urban period: tribal society, a social system based on tribal chiefs or rulers, a clear social hierarchy, cultural and technological advances, territorial expansion, demographic growth, greater productivity, greater specialization in production, some organization of collective labor, and, perhaps, more defined territorial boundaries. 73 In our judgment, the orientalizing period implies a decisive advance in the evolution of these southern proto-Iberian populations. The internal process toward forms of urban life developed at a relatively late moment in relation to the cultural influence of the Phoenicians, and thus it was interrupted by political changes that occurred in the western Mediterranean beginning in the 6th century b.c. The beginnings of Iberian culture and urban life were probably made possible, in large part, by the socioeconomic bases established by the Tartessian intermediary during the 7th and 6th centuries b.c., but we cannot affirm that Iberian culture and urban life were its immediate and direct consequences. As another author has pointed out, 74 it was most likely the Greek matrix from southeastern Spain that stimulated urban Iberian culture in upper Andalusia. 73. For these themes, see J. Evans, “Village, town, city,” in Problems in Economic and Social Archaeology (ed. G. de G. Sieveking; 1976) 509–10; C. Renfrew, The Explanation of Culture Change, 542–43; P. J. Ucko, R. Tringham, and G. W. Dimbleby, Man, Settlement and Urbanism, 601, 735, 827ff.; see also “La Formazione della città nel Lazio,” Dialoghi di Archeologia 1–2 (2 vols., 1980). 74. A. Blanco, “Orientalia II,” AEArq. 33 (1980) 36.

The Phoenician Impact on Tartessos: Spheres of Interaction María Eugenia Aubet Semmler

The study of trade and cultural interactions between Phoenicians and Tartessians has usually been approached in terms of acculturation or cultural diffusion, since the transformations produced in the Tartessian communities during the “orientalizing” period are generally considered to be a direct result of Mediterranean trade. It is especially interesting to examine the degree of internal hierarchy within Tartessian society before the Phoenicians arrived in the West and to define the mechanisms of interaction between colonists and natives (based on either equal or unequal trade) which led to the historical reality of Tartessos. This historical process is understood as the emergence of indigenous societies which, favored by trade with the coastal Phoenician colonies, developed various economic and social structures that were some of the most advanced in all of the western Mediterranean at the time (Aubet Semmler 1977–78). During the 8th century b.c.e. on the Iberian Peninsula, only the territories of western Andalusia and their immediate periphery offered a significant concentration of raw materials. The economic structure of the various societies of the southeastern Peninsula (the preeminent mining and metallurgic region of Iberia) had been eclipsed as a result of the crisis in Argaric social institutions. The archaeological record shows that Phoenician trade with Tartessos generated an increase in the exploitation of local resources and a boom in the production of raw materials. This benefited the local aristocracy, which controlled a territory with abundant resources and which took full advantage of the Mediterranean trade that had quickly generated a growing demand by Mediterranean markets for raw materials. At the same time, the local aristocracy saw the necessity of stimulating demand among peripheral communities for the purpose of increasing surplus

Editor’s note: This article appeared in Spanish under the title “El impacto fenicio en Tartessos: Las esferas de interacción,” Cuadernos Emeritenses 2 (1990) 29–44.

225

226

María Eugenia Aubet Semmler

production. In short, this aristocracy knew how to take advantage of a favorable economic opportunity by profiting from the concentration in its territory of those raw materials (silver, gold, iron, copper, tin, grain, and wood) that Mediterranean commerce coveted during the 8th and 7th centuries b.c.e. and by favoring the incorporation of additional peripheral regions (Portugal, Estremadura, the central plateau, Upper Andalusia) into a network of integrated regional economic systems. The dynamics of this network were similar to those known as “core-periphery,” a system which has been analyzed during the past few years with excellent results in other European areas equally affected by Mediterranean trade during Iron I (Rowlands, Larsen, and Kristiansen 1987; Brun 1987; Champion 1988; Cunliffe 1988; Wells 1988; and Ruiz Zapatero 1989). While channeling and absorbing all these riches from the Tartessian hinterland, the Phoenicians (in a process that was repeated later by the indigenous communities themselves in the Lower Guadalquivir) acted as economic and social intermediaries. They integrated Tartessian society, especially its ruling elites, into the system of structures characteristic of civilized market societies, that is, societies ruled by a centralized political power as in the eastern Mediterranean. As a consequence, Phoenician mediation favored the gradual abandonment by Tartessian society of its traditional “barbarous” forms, which belonged to more egalitarian or transitional structures, in favor of more “modern” forms, meaning the institutionalization of a social hierarchy (with “princedoms”) within defined political territories. We begin with the hypothesis that the development of the Tartessian communities during the Iron Age can be understood only in terms of interaction with the Phoenician colonial world. The dominant elite played the role of intermediary between the market centers (the Phoenician colonies) and the periphery (Estremadura and other regions), based on sociopolitical and territorial control (fig. 1).

Mechanisms for Mutual Relationships For a long time it was believed that the Phoenicians came to a territory inhabited by indigenous peoples who did not recognize the enormous potential of resources that could be found in the region of the Lower Guadalquivir. Thanks to the “eastern miracle,” these people learned to value and profit from their silver ore resources in the mountainous areas and from their agricultural resources in the valleys. According to this long-held belief, the Phoenician colonization acted as a catalyst within these passive and highly receptive communities. Due to their contact with the colonists, these indigenous peoples gradually learned new

The Phoenician Impact on Tartessos: Spheres of Interaction

227

w Phoenician colonies ∑ Indigenous settlements

Fig. 1. The Tartessian area and its zones of influence.

irrigation techniques for intensive agriculture, the market value of precious metals, and the metallurgic and mining techniques taught by eastern specialists, such as smelting to extract silver from ore. They also learned to use the potter’s wheel and to decorate pottery with painted, eastern-inspired geometric motifs. However, there are alternatives to this strictly diffusionist view of cultural change in Tartessos during the 8th and 7th centuries. The field of economic anthropology and the theory of colonial trade offer two proposals more in agreement with the data obtained during the past few years from the archaeological record in western Andalusia (Suret-Canale 1962, Meillassoux 1971, and Curtin 1984): (1) Colonial trade enterprises like that of the Phoenicians rarely deal with territories where indigenous societies have an egalitarian social organization or are governed by decentralized political institutions. On the contrary, trade networks are established where hierarchical societies exist with economic activity minimally coordinated by centers of political power. This is because the very nature of colonial trade demands social structures that guarantee surplus production, regulate the stability and continuity of the interchanges, and have the capacity to provide native manpower in the mines, agricultural fields, and commercial ports. In short, the society must possess slaves or forced labor, and, therefore, political authority. This was true for the European colonies in

228

María Eugenia Aubet Semmler

Southeast Asia, the Spanish colonies in America, the Greeks in Southern Italy and Sicily, and the Phoenicians in Cyprus, Carthage, and Sardinia. (2) Every trading enterprise tends to establish mechanisms that guarantee the efficient functioning of trade. In general, such methods take the form of pacts with local rulers. In exchange for funneling riches toward the indigenous rulers in the form of exotic products and prestige goods (which result in greater power for these leaders), the colonists assure themselves of a regular supply of raw materials. Three sociopolitical mechanisms for mutual relationships were commonly used in Antiquity: (a) Pacts using religious formulas. This mechanism is generally seen in common worship at indigenous sanctuaries, where both natives and colonists worship together indiscriminately. This is probably the case with the temple of Astarte in Kition or Tas Silg on the island of Malta, where beginning in the 7th century b.c.e. we see the Phoenician presence at an ancient native sanctuary (Ciasca 1982). Stelae and commemorative inscriptions, like those at Nora or Bithia on the island of Sardinia, and sacrificial rites performed when new colonies are founded, such as at Carthage, testify to the existence of alliances between the first Phoenician colonists who arrived in the West and the indigenous leaders of the interior. The most ambitious and long-lasting religious formula seems to have been the construction of a temple. The temple served several purposes: a guarantee of the good intentions of the colonists, a symbol of the Phoenician state, and common ground for carrying out and centralizing transactions and for sanctioning trade agreements. Only in this way can we explain the recurring presence of temples dedicated to Melkart in strategic market centers and in territories of special economic interest for the aspirations of Tyre, such as Gadir, Lixus, Utica, and Malta (Aubet Semmler 1987: 130–38, and Bonnet 1988). (b) Ideological integration of privileged segments of indigenous society into the world of eastern social and political structures. This type of assimilation is seen especially in indigenous princely tombs, one of the most characteristic aspects of the “orientalizing” process in Tartessos, Etruria, and Latium. The status of the local aristocracy was exhibited by adopting the burial practices of the Greek and Phoenician colonists, such as at the Tartessian necropolises of La Joya, Setefilla, and Carmona (Aubet Semmler 1984). (c) Dissuasion or coercion brought to bear upon the local population. These methods are usually seen in contexts where the colonists intend to occupy territory or perhaps foment competition among the elite, which is often surmised from the appearance of fortifications and local confrontational dynasties. This probably was the case with the Phoenician colonies in southern Sardinia, and perhaps also in eastern Andalusia. However, this was not the case in Tartessos.

The Phoenician Impact on Tartessos: Spheres of Interaction

229

Up to this point we have presented the possible formulas used by the Phoenicians to make contact and begin trade in the Tartessian area. Next we will review the archaeological data in order to examine some of the above proposals: a hierarchical Tartessian society during the Final Bronze as a necessary condition for the establishment of regular Phoenician trade in the region, and formulas for alliances with the natives which used socioreligious institutions and the ideological integration of the Tartessian elite into the framework of Phoenician social and economic structures in order to consolidate their own social status. All of this will be analyzed in chronological order beginning with the first colonial contacts and continuing to the period in which Tartessian leaders succeeded in controlling access to Mediterranean products and then redistributing these prestigious goods toward the periphery. Their purpose was to obtain new resources and raw materials for themselves within a framework of interactions dominated by a hierarchical trading system (see Frankenstein and Rowlands 1978, Brun 1987, and Champion 1988).

Trading Centers There are several indications that a hierarchical social organization existed among the Tartessian communities during the Final Bronze. These communities, organized around villages and towns and inhabited by chiefs, metalworking specialists, shepherds, and farmers, appeared during the 9th and 8th centuries b.c.e. First of all, it is important to point out that the distribution of indigenous peoples prior to the arrival of the Phoenicians showed a clear preference for strategic spots that dominated the communication routes and agricultural lands near the Guadalquivir River valley (Carmona, Carambolo, Lebrija, Montemolín, Ecija). Settlements were also found at points of access to sources of silver and copper ore (Cabezo de San Pedro and San Bartolomé de Almonte, in Huelva), cattle passes (Setefilla), and the Atlantic shipping trade (Huelva). In comparison to preceding periods, the Final Bronze was characterized by a spectacular increase in the number and size of towns. Moreover, the pattern of settlement indicated a tendency toward a structured hierarchical organization around the centers of power (Huelva, Carambolo, Carmona), toward which control of the territory gravitated. Because the archaeological record provides little information on burials (absence of tombs, lack of significant differences in the content of grave goods, etc.), it has been difficult to recognize the existence of chiefs with territorial and sociopolitical control during this period. However, the production and circulation of luxury pottery with painted or burnished decorations of the type

230

María Eugenia Aubet Semmler

found in Huelva and Carambolo indicate the presence of specialized artisans and professionals (characteristic of every society with a certain sophistication) and the existence of long-distance trade. Far from reflecting a discontinuous or unplanned nature, this trade indicates some centralized social structures capable of maintaining regular trading contacts with Estremadura, the central plateau, and the plain of Granada (Mendoza, Molina, Arteaga, and Aguayo 1981; Almagro Gorbea 1977; Pereira 1989). As for the mining of precious metals, it must be noted that recent works on this topic demonstrate the existence of silver production in Río Tinto ever since the 2d millennium (Pérez Macías and Frías 1989). Moreover, the techniques of cupellation and extraction of silver, far from being introduced along with Phoenician commerce, were known in Andalusia from the middle of the 2d millennium b.c.e. In connection with the production and trade of metals, we can mention the find of bronzes in the Huelva estuary, dated to the middle of the 9th century b.c.e. (Almagro Gorbea 1977). These bronzes are significant as possible remains of a sunken ship or a fluvial offering or as an example of hoarding metal in response to the growing difficulties of obtaining copper, tin, and bronze, which would lead to the adoption of iron. However, this find also implies an early maritime traffic at the Tartessian port of Huelva and a practice of accumulating surplus metal that is typical of hierarchical societies (see Ruiz Gálvez 1987; 1986). All these data must be compared empirically. We lack archaeological studies of space, as well as carpological, pollen, and carbon studies which show the volume and function of agricultural activities; necropolises still need to be discovered. However, the evidence points toward the existence of a hierarchical society during the Tartessian Final Bronze which was sufficiently organized to respond to the challenge of Mediterranean trade in the 8th century. To confirm this fact, we have the stelae of Carmona, Setefilla, and Ecija, no doubt symbols of status and rank, and also some vestiges of fortifications from the Final Bronze in Setefilla and Carmona, indicating the existence of elites that controlled small territories (Aubet Semmler 1989b). The dates attributed to the latest archaeological finds in Cádiz and its bay confirm the arrival of the Phoenicians in the Tartessian area during the first half of the 8th century b.c.e. (Escacena 1986). The principal ideological element used by Phoenician traders to assure good relations with indigenous communities of the interior appears to have been the construction of the temple at Gadir (Cádiz), built to honor the god Melkart and to symbolize the Tyrian monarchy and state. In this regard, the temple probably offered full political and religious guarantees to the Tartessian communities, given that no signs of tensions or conflicts during the 8th and 7th centuries exist in the region. At the same time, the existence of the temple of Melkart assured the efficient functioning of transactions and backed up, as a symbol of Tyrian

The Phoenician Impact on Tartessos: Spheres of Interaction

231

political institutions, the agreements and alliances arranged between the Tartessian leaders and the Phoenician daughter colony of Gadir. In the example of Gadir, the association between trade and temple demonstrates a clear predisposition on the part of the Phoenicians toward the establishment of long-term trading networks. Therefore, in Andalusia the temple implies the existence of peaceful and strictly commercial relations, explaining one of the constants in the Tartessian area during the “orientalizing” period: the absence of fortifications or of other coercive or dissuasive devices that characterize situations in which colonizers intend to occupy, subdue, or control a territory. Such coercion appears to be the case in the Phoenician colonies in southern Sardinia and probably also in eastern Andalusia, where vestiges of both Phoenician and indigenous fortifications exist from at least the 7th century b.c.e. The Phoenicians could not hope to control the Tartessian territory, where significant nuclei of indigenous peoples dominated the coast and the mouth of the Guadalquivir River, probably since the Final Bronze. The geographic location of Mesas de Asta, Ebora, and Lebrija, which controlled from the south all navigation through the ancient lake Ligustinus, prevented any attempt to occupy Tartessian territory (Bendala 1985: 625). Consequently, the colony at Gadir, which was located on an island near the mouth of the Guadalquivir River, realized the necessity of developing its commercial activity by means of strictly coastal ports at border points such as Mesas de Asta or Torre de Doña Blanca, located at the mouth of the Guadalete River. A mixed population of Phoenician merchants and Tartessian traders must have lived together at these ports, which functioned as true bulwarks of trade. Although the archaeological evidence tells us nothing yet in this respect, the presence of a necropolis at Torre de Doña Blanca (Ruiz Mata 1989), whose mounds show the same characteristic structure as those in the Tartessian necropolises in the area of Seville, as well as the architecture and internal organization of the settlement (Ruiz Mata 1986), hint at the possibility that a market center existed at Doña Blanca, where trade developed and where agreements were made between the residents of Gadir and the leaders from the interior. The discovery years ago of an alabaster urn that appears to have come from the vicinity of Puerto de Santa María (García Bellido 1970: 11–21; Gamer-Wallert 1978: 80–85) indicates that there could be an ancient Phoenician necropolis, similar to those at Toscanos, Lagos, and Almuñécar, near the indigenous necropolis. This would demonstrate the close relationships and intermingling that occurred between the Phoenicians and indigenous communities in border areas such as Torre de Doña Blanca.* * Editor’s note : For more recent information on Torre de Doña Blanca, see the article by Ruiz Mata in this volume: “The Ancient Phoenicians of the 8th and 7th Centuries b.c. in the Bay of Cádiz: State of the Research,” especially the section on the results from Castillo de Doña Blanca, pp. 170–98.

slight subhead drop

232

María Eugenia Aubet Semmler

Spheres of Interaction The high point for the Tartessian communities came during Iron I (7th–6th centuries), a period that is also known as “orientalizing” because of a similar process that took place in Etruria and Latium around the same dates. In terms of socioeconomic transformation, the “orientalizing” phenomenon was expressed by means of the definitive ideological integration of the Tartessian elite into the framework of eastern Mediterranean social structures and by the incorporation of its economic systems into Mediterranean trade circuits. During the 7th and 6th centuries the Tartessian area constituted the most densely populated territory of the Peninsula and had the greatest volume of surplus production and circulation of luxury goods. The effects of trade with Gadir are specifically seen in the normal contents of the so-called Tartessian “princely tombs” (Ruiz Delgado 1989), which express the degree of power and ostentation of riches reached by some princes or chiefs who took advantage of their favorable position to accumulate imported items. Examples of these tombs are found at La Joya in Huelva (fig. 2), where the grave goods have a notably eastern character, and the burial chambers of Mounds A and H at Setefilla (fig. 3), whose architecture evokes that of the Phoenician tombs on the shore of Málaga. As in Etruria, we must speak of a true “princely culture” (Torelli 1981) within the framework of the so-called “prestige” economies. The presence of cremation urns and other imported Phoenician elements in Tartessian necropolises like the one at Cruz del Negro, and the transformation into bronze of exclusively Phoenician tableware in the Tartessian tombs of the interior (fig. 4b), show to what extent other segments of the Tartessian population gradually accepted eastern imported goods in a process that crystallized into forms of more complex social organization during the 7th century. The towns were organized around large centers such as Setefilla, Carambolo, Carmona, and Huelva, which specialized in various economic activities (silver production, intensive agriculture, livestock, shipping activities, etc.) and grew at the expense of other secondary centers. Thus, princely residences emerged, probably associated with a social structure of “princedoms” centralized in territories of political control, which benefited from their position as intermediaries in the trading transactions between the Phoenician colonies and the interior (see Frankenstein and Rowlands 1978; Rowlands 1987; Mohen, Duval, and Eluère 1988). If Gadir acted as the driving force for trade with Tartessos, the communities of the Lower Guadalquivir acted in a double role: principal consumers and recipients of Mediterranean products, and producers of raw materials obtained within their own territory and from neighboring regions through varying forms of exploitation. To accomplish this, the Tartessian aristocracy must have promoted demand and consumption among the communities located on its

The Phoenician Impact on Tartessos: Spheres of Interaction

a

b

Fig. 2. Tomb 17 at La Joya: (a) floor of the tomb, with metallic remains of a light chariot on the north; (b) bronze horse bits; the sketches of the horse’s head show how the bits may have been used.

233

234

María Eugenia Aubet Semmler

Fig. 3. Floor of Mound A at Setefilla.

periphery (Estremadura, the Upper Guadalquivir, the fertile plain of Granada, the western central plateau) with the goal of increasing the volume of raw materials (tin, gold, copper, lead) that were the best guarantee for perpetuating

The Phoenician Impact on Tartessos: Spheres of Interaction

235

a

b

Fig. 4. Bronze jugs from (a) Torres Vedras and from (b) Alcalá del Río.

and reproducing their power in Tartessos. The geographic distribution of orientalizing objects in Portugal (fig. 4a) and, most of all, the distribution of bronzes and metal and glass jugs in Estremadura, such as those at Aliseda, Valdegamas (fig. 5b), Siruela, and Zarza de Alange (fig. 5a) (Esteban 1985;

236

María Eugenia Aubet Semmler

a

b

Fig. 5. Bronze jugs from (a) Zarza de Alange and from (b) Valdegamas.

The Phoenician Impact on Tartessos: Spheres of Interaction

237

Alvarez and Gil 1988), demonstrate the intense interactions between Tartessos and its immediate periphery. We note, however, that sociocultural changes occurred more slowly in the periphery than in the Lower Guadalquivir. Thus, due to the enormous potential of the resources in their territory and their privileged geographic location, the Tartessian communities acted as true intermediaries between the center (the Mediterranean market economies) and the periphery (or secondary zone where resources and raw materials were obtained). In this system of concentric, hierarchical economic circles, Estremadura constituted the great reserve for the Tartessian economy, which needed to widen its territory of exploitation during the crisis that probably occurred during the second half of the 6th century b.c.e. Tartessos accepted Mediterranean civilization and market economies at a cost. It renounced its ancient egalitarian institutions and suffered ecological degradation of its territory, due to the overuse of its soil and to the harvesting of its forests as a consequence of the intense mining and artisan activities that required enormous quantities of wood for combustion. Usually the Tartessian crisis is attributed to external factors (the fall of Tyre at the hands of Nebuchadnezzar, exhaustion of the silver in the Huelva mines, hostility of the natives toward Gadir). However, none of these factors has been rigorously proven. The fall of Tyre occurred long before the events at the end of the 6th century, the mines of Riotinto would be exploited again during the Roman period (so we cannot speak about the exhaustion of silverbearing lodes), and there is no verified sign of tension between Tartessians and Phoenicians during the 6th century. In any case, the decline of Huelva and Setefilla itself, citing two settlements that are relatively well known, became evident between 550 and 500 b.c.e. On the other hand, the abandonment of some of the Phoenician colonies on the shore of Málaga, such as Cerro del Villar on the Guadalhorce and Toscanos, occurred at the full height of their production, which contradicts a crisis or decline brought on by external sociopolitical factors. Perhaps we must search for the causes of the crisis in internal factors, most likely those of a distinctively social type. In any case, the disappearance of princely tombs in the Lower Guadalquivir and in Huelva, the decline of metallurgic activity at the port of Huelva, and the collapse of the “orientalizing” phenomenon coincide with a displacement of the principal centers of production toward the “periphery,” where we see the birth of the first true urban centers in the southern Peninsula, such as in the region of the Upper Guadalquivir. The analysis of this process, like that of so many of the ideas suggested in this present paper, is one of the topics calling for further archaeological research in the area of Iberian protohistory.

238

María Eugenia Aubet Semmler

Bibliography Almagro Gorbea, M. 1977 El Bronce Final y el período Orientalizante en Extremadura. Bibliotheca Praehistórica Hispana XIV. Madrid. Alvarez, A., and Gil, J. 1988 Aproximación al estudio de las vías de comunicación en el primer milenio a.C. en Extremadura. Trabajos de Prehistoria 45: 305–16. Aubet Semmler, M. E. 1977–78 Algunas cuestiones en torno al período orientalizante tartésico. Pyrenae 13– 14: 81–107. 1984 La aristocracia tartésica durante el período orientalizante. Opus 3: 445–68. 1987 Tiro y las colonias fenicias de Occidente. Barcelona: Bellaterra. 1989a ed., Tartessos: Arqueología protohistórica del Bajo Guadalquivir. Sabadell (Barcelona): AUSA. 1989b La Mesa de Setefilla: la secuencia estratigráfica del Corte 1. Pp. 297–338 in Tartessos: Arqueología protohistórica del Bajo Guadalquivir, ed. M. E. Aubet Semmler. Sabadell (Barcelona): AUSA. Bendala, M. 1985 Tartesos. Pp. 595–640 in Historia General de España y América 1.1. Madrid: Rialp. Bonnet, C. 1988 Melqart: Cultes et mythes de l’Héracles tyrien en Méditerranée. Studia Phoenicia 8. Leuven-Namur. Brun, P. 1987 Princes et princesses de la Celtique: Le Premier Age du Fer (850–450 av. J.C.). Paris: Errance. Champion, T. C. 1988 Introduction. Pp. 1–21 in Centre and Periphery, ed. Champion. London. Ciasca, A. 1982 Insediamenti e cultura dei fenici a Malta. Pp. 133–41 in Die Phönizier im Westen, ed. H. G. Niemeyer. Mainz am Rhein. Cunliffe, B. 1988 Greeks, Romans and Barbarians. Spheres of Interaction. London: Batsford. Curtin, P. D. 1984 Cross-Cultural Trade in World History. Cambridge: Cambridge University Press. Escacena, J. L. 1986 Gadir. Pp. 39–58 in Los fenicios en la Península Ibérica, eds. G. del Olmo and M. E. Aubet. Sabadell (Barcelona): AUSA. Esteban, J. 1985 Algunas consideraciones sobre los poblados orientalizantes extremeños. Norba 6: 19–28. Frankenstein, S., and Rowlands, M. 1978 The internal structure and regional context of Early Iron Age society in south-western Germany. Bulletin of the Institute of Archaeology (Univ. London) 15: 73–112.

The Phoenician Impact on Tartessos: Spheres of Interaction

239

Gamer-Wallert, I. 1978 Agyptische und ägyptisierrende Funde von der Iberischen Halbinsel. Wiesbaden. García Bellido, A. 1970 Algunas novedades sobre arqueología púnico-tartessio. Archivo Español de Arqueología 43: 3–49. Meillassoux, C., ed. 1971 The Development of Indigenous Trade and Market in West Africa. Oxford: Oxford University Press. Mendoza, A.; Molina, F.; Arteaga, O.; and Aguayo, P. 1981 Cerro de los Infantes (Pinos Puente, Granada). Madrider Mitteilungen 22: 171–97. Mohen, J. P.; Duval, A.; and Eluère, C., eds. 1988 Les princes celtes et la Méditerranée. Paris: La Documentation Française. Olmo, G., and Aubet Semmler, M. E., eds. 1986 Los fenicios en la Península Ibérica. Sabadell (Barcelona): AUSA. Pereira, J. 1989 Nuevos datos para la valoración del hinterland tartésico. El enterramiento de la Casa del Carpio (Belvis de la Jara, Toledo). Pp. 395–409 in Tartessos. Arqueología protohistórica del Bajo Guadalquivir. Ed. M. E. Aubet Semmler. Sabadell (Barcelona): AUSA. Pérez Macías, J. A., and Frias, C. 1989 La necrópolis de cistas de La Parrita (Nerva, Huelva) y los inicios de la metalurgia de la plata en las minas de Río Tinto. Cuadernos del Suroeste 1: 11–21. Rowlands, M. 1987 Centre and periphery: a review of a concept. Pp. 1–11 in Centre and Periphery in the Ancient World, eds. M. Rowlands, M. Larsen, and K. Kristiansen. London: Cambridge University Press. Rowlands, M.; Larsen, M.; and Kristiansen, K., eds. 1987 Centre and Periphery in the Ancient World. London: Cambridge University Press. Ruiz Delgado, M. M. 1989 Las necrópolis tartésicas: prestigio, poder y jerarquías. Pp. 247–86 in Tartessos: Arqueología protohistórica del Bajo Guadalquivir, ed. M. E. Aubet Semmler. Sabadell (Barcelona): AUSA. Ruiz Gálvez, M. 1986 Navegación y comercio entre el Atlántico y el Mediterráneo a fines de la Edad del Bronce. Trabajos de Prehistoria 43: 9–42. 1987 Bronce atlántico y “cultura” del Bronce atlántico en la Península Ibérica. Trabajos de Prehistoria 44: 251–64. Ruiz Mata, D. 1986 Castillo de Doña Blanca (Puerto de Santa María, Cádiz). Madrider Mitteilungen 27: 87–115. 1989 El Túmulo 1 de la necrópolis de Las Cumbres (Puerto de Santa María, Cádiz). Pp. 287–95 in Tartessos: Arqueología protohistórica del Bajo Guadalquivir, ed. M. E. Aubet Semmler, Sabadell (Barcelona): AUSA.

240

María Eugenia Aubet Semmler

Ruiz Zapatero, G. 1989 Centro y periferia: la Europa bárbara y el Mediterráneo en la Edad del Hierro. Trabajos de Prehistoria 46: 331–40. Suret-Canale, J. 1962 Afrique noire. L’ére coloniale. Paris: Editions Sociales. Torelli, M. 1981 Storia degli Etruschi. Rome-Bari: Laterza. Wells, P. S. 1988 Granjas, aldeas y ciudades. Comercio y orígenes del urbanismo en la protohistoria europea. Barcelona: Labor.

The Tartessian Economy: Mining and Metallurgy Jesús Fernández Jurado

A discussion on the economy of the ancient world in the light of archaeological data is a difficult and complicated undertaking, one that is not exempt from a certain subjectivity, as is every historical interpretation. In this case, however, the subjectivity is not bound exclusively to ideology, but rather it is firmly linked to the interpretation that we make of the tangible data that excavations provide. In this sense, attempts to approach the Tartessian economy are possible up to a certain point, even though we do not possess sufficient data, either quantitative or qualitative, on agricultural production and fishing. Often we must rely on what the ancient written sources (which are relatively late) tell us about Tartessos and the rest of the Iberian Peninsula and then attempt to draw parallels with what we think the Tartessian economy must have been like. In regard to mining, and especially in regard to metallurgy, we possess more concrete and measurable data. Mining, however, is still difficult to study, because even though we know that this activity was fundamental to the Tartessian economy, we cannot quantify the levels of mining activity—and we may never be able to do so. It is well known that every mining activity, as a consequence of technological advance, destroys to a large extent the evidence of previous mining.

Editor’s note: This article appeared in Spanish under the title “Economía tartésica: Minería y metalurgia” in Huelva en su historia: Miscelánea histórica (Seville: Colegio Universitario de La Rábida, Universidad de Sevilla, 1986) 149–70. Author’s note: This manuscript and the research that we are conducting on economic issues of mining and metallurgy are made possible by the work and collaboration of technicians M. Lamela, J. Martínez, and A. Alonso, chief chemist and members respectively of the Research Center of Riotinto Minera, S.A., in Huelva. They have performed analyses, interpreted results, and advised and instructed us continuously on these subjects. For this reason we give them our most sincere thanks.

241

242

Jesús Fernández Jurado

In order to address Tartessian mining and metallurgy, we must study activities directed toward obtaining silver. And, based on the data we currently possess, we must focus this study on the region of Huelva. These are points of departure for research on this fundamental aspect of the Tartessian economy. In addition to these general issues, we must detach ourselves from the political and administrative regions in which we are immersed, thereby allowing us to discard our provincial mindset and forcing us to consider the mining zone as a whole. This zone begins in the western region of the present province of Sevilla, crosses the province of Huelva, and extends well into Portugal (fig. 1). There are many areas not yet studied from this perspective, since the mining zone linked to Tartessos covers such an immense stretch of territory. In addition to the size of the territory, we must also take into account the conditions surrounding its origin. Although multiple studies and interpretations exist, 1 no definitive thesis has been accepted as a global and final explanation of how mining began in this territory. This extensive mining zone contains complex ores formed by conglomerations of minerals. Pyrite predominates, but there are also large quantities of copper, lead, zinc, gold, and silver. The order given here does not presume a greater or lesser presence of these minerals, with the exception of pyrite, which is the most prominent mineral in these polymetallic ores. 2 It is an obvious and well-known fact that this region was mined heavily in antiquity, given the great purity of these ores. Presently, the average purity of the mineral ore is approximately 40 grams of silver per metric ton and 2.5 grams of gold per metric ton. 3 We need to keep in mind, however, that in antiquity both figures must have been higher. This supposition is based on both the analytic data that we will present later and the secondary enrichment that is produced in areas surrounding the water table, where there are high levels of purity between the oxidized caps and the unaltered primary deposits of pyrites. These highly enriched areas can reach up to 100 meters of thickness. 4 Thus we can affirm that these ores contained gold and silver. 5 1. Mentioning all the studies that have been done since ancient times exceeds the purpose of this work. However, we refer to two sources, one already classic, which have collected nearly all the works referring to the origin of the mineral deposits of the zone we are studying: Pinedo, I. Piritas de Huelva. Madrid: Summa, 1963. Flores, M. Las antiguas explotaciones de Riotinto. Huelva: Diputación Provincial de Huelva (DPH), 1981. 2. G. K. Straus, “Sobre la geología de la provincia piritífera del SW de la P. Ibérica y de sus yacimientos, en especial sobre la mina de pirita de Lousal (Portugal),” IGME 77 (Madrid, 1970). 3. Figures provided by the technicians at Riotinto Minera, S.A. 4. I. Pinedo, Piritas, 19; M. P. Jones, “Los depósitos minerales de la provincia de Huelva,” Exploración Arqueometalúrgica de Huelva (EAH) (Barcelona: Labor, 1981) 31. 5. I. Pinedo, Piritas, 21.

The Tartessian Economy: Mining and Metallurgy

243

Fig. 1. Mining area of the southwestern Peninsula. See M. Flores Caballero: Las antiguas explotaciones de Riotinto. Diputación de Huelva, 1981.

Calculating the volume of ore that was mined is difficult, even though there have been various attempts to do so. 6 The region has been mined continuously, and it is well known that every mining activity presupposes, to a great extent, the destruction of the evidence of earlier work. We must also keep in mind that much of the ancient slag has been used for road surfaces and railroad ballast, and that during the Roman era and especially since the 19th century, it has also been used as flux. We also need to emphasize that researchers have focused on the large and historic mining centers, such as Riotinto, and to a lesser degree on other nearby areas, 7 leaving at the margin those areas that are farther away or where 6. I. Pinedo, Piritas, 25. 7. A. Blanco and B. Rothenberg, in Exploración Arqueometalúrgica de Huelva (EAH) (Barcelona: Labor, 1981).

244

Jesús Fernández Jurado

Fig. 2. Excavations in Huelva (1982–1985).

modern mining activities have held less interest. This situation has led to varying theories about the metal-related economy of the Tartessian era, which traditionally has been linked to Riotinto and in a lesser degree to Tarsis (based on an inaccurate interpretation of the place name). Other mining zones have been forgotten, and trade routes have been “created” that are difficult to accept if we take into account the lay of the land. Although the metallurgical technology of the Tartessians was at a high level, the technological level of the mining process was probably not so high. For this reason there was intense mining of the oxidized caps and of every

The Tartessian Economy: Mining and Metallurgy

245

possible outcropping that could be found. When we consider all of these points, we can understand the difficult nature of estimating the volume of the ore that was mined, if not the futility of trying. We have so far presented a series of points about the nature and importance of the mining area. Next we will deal with metallurgic activity. It is difficult to judge the volume of ore that was mined, but it is just as difficult to judge the volume of metallurgic production. However, when we consider quality and type of production, it is possible to come closer to reality, especially after the excavations that we have performed, and in some cases are continuing, in Huelva, San Bartolomé of Almonte, and Tejada la Vieja (Escacena del Campo). These excavations have provided information about the mineral ores that were mined (principally gossan) and the metals that were sought (principally silver). Archaeological researchers have tried, at least in the southwestern Peninsula, to link metallurgic production to the place where the ore originated. Even if it is true that at times the metallurgic activity took place next to the mines, it is also true that in numerous cases the metallurgic process took place at locations different from those that were strictly for mining. This can be explained by several factors: the need for abundant combustible materials, the search for nearby trading centers and coastal distribution points—in short, a greater economic profitability. 8 The use of distinct locations for mining and for metallurgic production, confirmed archaeologically in Huelva and San Bartolomé (Almonte), will lead us to a comparison of these sites with a third site, Tejada la Vieja, where both activities were performed.

Huelva Since 1982 we have excavated at several spots in the center of the city of Huelva 9 (fig. 2). These excavations have revealed the existence of an important Tartessian city, with evidence of Phoenician trade since at least the beginning of the 8th century b.c. and with an abundant presence of Greek pottery from 8. It must be remembered that trees are often not abundant in mining zones. Logically it is more profitable to transport the ore closer to the trading centers, once it has been washed, than to bring the combustible materials to the mine. The latter would necessitate two moves: first the combustible material to the mine and then the resulting metal to the point of sale. To this reasoning, we can add the fact that the transport of ore does not require excessive security measures, whereas the transport of metals does. 9. First sponsored by the Ministry of Culture and later by the Cultural Council (Consejería de Cultura) of the Junta de Andalucía, the Archaeological Service of the Provincial Office (Diputación Provincial) of Huelva has performed excavations and archaeological research under my direction in Huelva’s city center from 1982 until the present.

246

Jesús Fernández Jurado

Plate 1. Ovens in Puerto, 6 (Huelva).

the end of the 7th century. 10 These excavations have provided ample evidence of the metallurgic activity that took place in Huelva, given the discovery of large amounts of slag, pottery containing slag, and remnants of ore, as well as metal-working zones in Botica and Puerto, 6, where it was possible to excavate a smelting oven and part of a second one. The oven had a circular floor, with an approximate diameter of 1.50 meters, and a slightly-inclined base oriented to the SW, the direction from which the prevailing winds of the region blow. At a slightly lower level we found a ledge, where the container that collected the molten metal was probably placed. The oven was constructed of limestone blocks with some smaller stones, and slabs of slate served as a base (pl. 1). This oven was probably loaded with alternating layers of fuel and crushed ore, and the top was covered by a clay dome with a hole for a vent. The area 10. Fernández Jurado, J. “La presencia griega arcaica en Huelva.” Col. Excavaciones en Huelva (EH) 1 (DPH, 1/1984). ___ . “Phönizische Spuren in Huelva.” MM 26 (1985) 49–60. ___ . “La influencia fenicia en Huelva.” Aula Orientalis 3 (1985). ___ . “Fenicios y griegos en Huelva.” Pp. 562–74 in Homenaje a Luis Siret (1934–1984). (Seville: Consejería de Cultura, 1986). ___ . “El poblamiento ibérico en Huelva. Pp. 315–26 in I Jornadas sobre el Mundo Ibérico (Jaén, 1985). Seville, 1987.

The Tartessian Economy: Mining and Metallurgy

247

for emptying the oven was on the west side and the tuyeres (bellows nozzles) were probably inserted on the east side. 11 The clay dome had to be removed each time the oven was used. The stone base described above was the only fixed and stable part. At the present moment we can offer analytic data on the excavations at Puerto, 6, Botica, and Puerto, 9. These data are interpreted later in the article.

San Bartolomé (Almonte, Huelva) This site 12 presented serious difficulties at the time of excavation because of its large area (fig. 3) and because of the absence of solid dwelling structures. The simple huts at San Bartolomé were made from plant matter and were slightly sunken into the tertiary loam that forms the soil at this site. These huts had a brief existence, as shown by their single stratum of archaeological fill, leading us to consider the possibility that they formed part of a seasonal town. By studying the horizontal stratigraphy of the site, we have been able to determine that the dates for this settlement stretch from the 9th to the 6th centuries b.c. The existence of a town based on metal production 40 kms. away from the nearest mines (those in the area of Aznalcóllar, province of Sevilla), certainly calls for attention. However, we previously discussed some of the reasons that may explain the location, and in the case of San Bartolomé, the type of soil where the town is located provides an additional explanation. As we have already mentioned, the soil is composed of tertiary loam. This soil has a high content of salts, gypsum and especially lime—an indispensable element for producing metal, as we shall see below.

11. J. F. Healy, Mining and Metallurgy in the Greek and Roman World. (London, 1978) 184, fig. 26. The oven found in Puerto, 6, cannot be much different in its function from the one described by this author. 12. Dr. Ruiz Mata and I excavated this site from 1979 to 1983. The first three seasons were dedicated to the study of the Final Bronze town; the last two were oriented toward learning about the Chalcolithic habitat. Ruiz Mata, D. “El poblado metalúrgico de época tartésica de San Bartolomé (Almonte, Huelva).” MM 22 (1981) 150–70 Fernández Jurado, J. “El yacimiento metalúrgico de San Bartolomé de Almonte.” Revista Huelva 4 (DPH, 1982) 40–46. ___ . “San Bartolomé de Almonte: yacimiento metalúrgico de época tartésica.” Revista de Arqueología 26 (1983) 24–33. Fernández Jurado, J., and D. Ruiz Mata. “La metalurgia de la plata en época Tartésica en Huelva.” Pyrenae 21 (1985) 23–44. Ruiz Mata, D., and J. Fernández Jurado. El yacimiento metalúrgico de época tartésica de San Bartolomé de Almonte (Huelva). Huelva Arqueológica 8 (1986).

248

Jesús Fernández Jurado

Fig. 3. Location of San Bartolomé.

The Tartessian Economy: Mining and Metallurgy

249

Plate 2. Oven at San Bartolomé (Almonte).

In San Bartolomé it was possible to excavate huts and refuse pits, as well as ovens of various sizes and purposes within the complex metallurgic process. These ovens (pl. 2) differ from the one found in Huelva in that they are simple holes in the ground, sometimes with a lining of clay at the bottom, but with a use and system of loading similar to the oven in Puerto, 6. Given the generally larger dimensions of the ovens of San Bartolomé, we cannot rule out the possibility that the process took place directly in the open air and that the remains of clay found inside the ovens are pieces of protective covering from the manipulating tools. Given the geographic location of San Bartolomé and the results that the analysis of its samples have provided, its link with the area of Aznalcóllar, via Tejada la Vieja (Escacena del Campo, Huelva) is obvious. 13

Tejada La Vieja (Escacena del Campo, Huelva) This protohistoric city is one of the urban complexes from which we can expect important results in regard to Tartessian culture and its later evolution within what is called the Turdetanic world. This entirely walled site, 14 with an 13. The site was acquired by the Diputación Provincial of Huelva in 1984. 14. I have been directing the excavations in Tejada la Vieja since 1982.

250

Jesús Fernández Jurado

area of 6.5 hectares, was excavated already in 1973. 15 Later, in 1982, we began new research on the city, 16 with an ambitious program of which one of the objectives was to open the site to the public. Tejada was clearly an orientalizing city, and its location was intimately related to the mineral richness of the area. We have not yet found large metalworking zones within the city, although we cannot rule out the possibility that they existed. The amount of slag that has been found is small. Therefore, if we compare Tejada with Almonte and/or Huelva, we must consider Tejada more a principal mining center and less a metal-working center, with metallurgic activity taking place in towns like San Bartolomé in the marshy environment near the mouth of the Guadalquivir.

Metal Technology We cannot be absolutely certain of the system employed by the Tartessians for extracting silver, but we do have sufficient evidence to establish, at least as a working hypothesis, that the technique employed was cupellation, which had spread throughout the mining centers of the Mediterranean world during the time period on which we focus our work. It is a double process that can be summarized in two stages: fusion and cupellation proper. Fusion consists of mixing the crushed ore with the flux and subjecting the mixture to high heat. Once this operation is finished, we obtain two products: the regulus, which is a mixture of lead, silver, and gold, with some traces of other elements, and the slag. In this first step the lead attracts the precious metals. The second stage or “cupellation” consists of placing the regulus (lead, silver, and gold) in a crucible (or cupel) and again subjecting it to high temperatures. The purpose of the crucible is to attract the lead and release a second regulus, in this case formed by gold and silver. The lead separates during the process of cupellation; part of it adheres to the crucible and the rest passes to the air in the form of oxide. Today the crucible is prepared with magnesite, although it is not rare to use ashes of bones, especially in laboratory work, a traditional technique that has been widely used. The fashioning of crucibles using a mixture of crushed bone, lime, and water is a technique known since antiquity, referred to in works on metallurgic themes.

15. A. Blanco and B. Rothenberg, in Exploración. 16. J. Fernández Jurado, “Tejada la Vieja: una ciudad protohistórica,” Huelva Arqueológica 9 (1987).

The Tartessian Economy: Mining and Metallurgy

251

In some of his works, 17 Juan de Arfe, a Spanish metallurgist of the 16th century, discusses the suitability of different materials used to fashion crucibles, as well as the container in which the crucibles were prepared: . . . the cupels . . . are made with different types of ashes. Some are made from vine shoots, or from the broom plant, mixed with the burned and crushed marrow of rams’ horns; others use cabbage stalks, but the best are made using only the shinbones of cows or rams (one is as good as the other), burned until they are white. After these bones are crushed and sifted, they are tempered with water into which has been mixed a small amount of quicklime; just enough water should be added to make the ash stick together. Once the ash is ready, it should be put into a bronze mold . . . ; and this mold should not have a top; and when the mold is full, a second brass mold is placed on top . . . which sits on top of the ash. Two or three taps of the hammer are given on top until it is well packed. After this, it is taken out of the mold. . . .

The type of container described above is similar to some ceramic pieces that we find frequently at these sites, especially at San Bartolomé de Almonte (fig. 4). These gritty, handmade ceramic pieces have an exterior surface that has been only superficially worked and an interior without any treatment whatsoever. All have a large number of perforations. We cannot be certain of their use, but in view of the description given by Juan de Arfe referring to the mold for making crucibles, as well as the countless examples of well-developed metallurgic activity in Huelva and Almonte, we might consider the possibility that these ceramic pieces were the containers in which crucibles were prepared. In this way we could explain the perforations, which would have facilitated the “sweating” of the excessive water in the mixture of bone and lime (as Arfe says, “just enough water should be added to make the ash stick together”). If, in addition, we consider the absence of bone fragments resulting from cooking at San Bartolomé, as well as the traces of gold and silver in samples of lime coming from the site, it seems apparent enough that cupellation must have been the method used for extracting silver. In addition to the technique described above, it is possible that other methods were used. Perhaps the most simple is to place the regulus resulting from fusion (gold, silver, and lead) in an open container and to subject it to high heat. This process causes the lead to oxidize and pass to the air, although a small quantity (along with impurities from the regulus) is deposited on the surface of the container. In this way, a new regulus is released, consisting of silver and gold. This technique could explain the abundance of pottery to which slag has adhered. The pottery also may have been used for the phase of fusion during the process of cupellation. 17. Juan de Arfe, “Quilatador de la plata, oro y piedras,” in Col. Primeras Ediciones 5 (Ministerio de Educación y Ciencia, 1976) book 1, sheet 9.

252

Jesús Fernández Jurado

Fig. 4. Ceramic pieces from San Bartolomé.

Laboratory Results The following lab techniques were used to determine the different elements and compounds: 18 a. Atomic absorption: iron, copper, aluminum oxide, zinc, and magnesium b. Gravimetry: silica and sulphur c. Inductively Coupled Plasma (I.C.P.): lead, antimony, arsenic, bismuth, nickel, and selenium d. Assays by fusion: gold and silver

18. The technicians of Riotinto Minera, S.A., provided us the information about these techniques.

The Tartessian Economy: Mining and Metallurgy

253

The samples were grouped into five types: a. b. c. d. e.

Mineral ore Slag Pottery with slag Limestone Metallic lead

We were able to analyze all five types in San Bartolomé, while in Huelva and Tejada it has been possible to study only the first three. In spite of this fact, the comparison of data among the three sites permits us to determine the distinct place of origin of the ore found at each site. The data from the analysis of limestone and metallic lead from Almonte also confirm that the system of extraction was cupellation.

1. Mineral ore Mineral ore is the raw material from which metals are extracted, and below we present the ranges for the samples from the three sites. Huelva Au

0.6

Ag

32.00

-

San Bartolomé

73.8

- 2,400.00

SiO2

4.40

-

69.57

Fe2O3

0.37

-

88.24

Al2O3

0.50

-

6.89

Ca

0.1770 -

Bi

0.062

As

0.3

-

6.7

83.00

-

2,183.00

7.08

-

72.90

Tejada 0.5

-

15.00

grs./metric ton

18.00

-

482.00

grs./metric ton

--0.41

5.2030

-

2.55

---

---

%

---

%

---

%

0.2610 -

0.6584

%

>0.01

-

0.03

0.0486 -

0.1076

%

0.0139 -

3.2284 >0.01

-

1.56

0.1775 -

1.1556

%

Se

>0.0001 -

0.0989 >0.01

-

0.13

0.0874 -

0.1995

%

Sb

>0.0001 -

0.3150 >0.01

-

0.12

0.0408 -

0.1209

11.45

>0.0001

Te Pb

-

0.22

-

>0.01

73.39

0.01

-

>0.0001 2.47

% %

0.0847 -

1.4297

%

Cd

>0.0001 -

0.0048

---

0.0045 -

0.0072

%

Ni

>0.0001 -

0.0115

>0.01

0.0052 -

0.0141

%

0.039

0.286

%

Cu

0.01

-

9.66

>0.01

-

0.40

-

Fe

---

1.66

-

45.87

---

%

S

---

0.55

-

1.43

---

%

254

Jesús Fernández Jurado

2. Slag This is a blackish-gray or black vitreous material. Many varieties exist, but they are easily identifiable during field work. The ranges of the samples are shown in the following table: Huelva Au

3.4

Ag

232.00

-

San Bartolomé

Tejada

27.2

0.1

-

1.0

0.7

- 805.0

184.0

-

639.0

151.0

-

1.0

grs./metric ton

- 244.0

grs./metric ton

SiO2

35.31

-

63.23

Fe2O3

12.01

-

60.80

Al2O3

1.28

-

14.07

Ca

0.7591 -

2.6397

---

0.6492 -

2.038

%

Bi

>0.0001 -

0.9620

>0.01

0.0317 -

0.0656

%

As

0.0270 -

0.0971

0.07

-

0.13

0.0674 -

0.1544

%

Se

>0.0001 -

0.0369

>0.01

-

0.03

0.0297 -

0.0418

%

Sb

0.0082 -

0.1094

0.04

-

0.07

0.0571 -

0.0951

%

31.60

64.82

-----

%

4.02

---

%

--1.89

>0.0001

Te

-

-

%

>0.0001

---

%

Pb

0.19

-

4.03

2.1151 -

3.2302

%

Cd

>0.0001

-

0.0027

---

0.0023 -

0.0031

%

Ni

0.0006

-

0.0053

>0.01

0.0065 -

0.0071

%

Cu

0.05

-

2.91

0.076

0.128

%

Fe

---

S

---

Plate 3. Slag.

1.72

-

2.93

0.01

-

0.09

26.63

-

33.04

>0.5

---

%

---

%

The Tartessian Economy: Mining and Metallurgy

255

3. Pottery with Slag This pottery holds great interest as we attempt to understand the metallurgic process, including the possibility of silver extraction by direct fire. Some of the pieces, in many cases small navettes of adobe, may have been used to collect slag. These samples present the following ranges: Huelva Au

14.9

Ag

3,512.0

SiO2

-

San Bartolomé

210.7

0.1

-

6.0

grs./metric ton

- 3,713.0

194.0

-

692.0

grs./metric ton

9.0

-

49.56

-

50.95

Fe2O3

6.26

-

14.13

Al2O3

3.73

-

8.29

Ca

1.1520 -

1.83

Bi

0.0354 -

2.28

>0.01

-

0.03

%

As

0.0109 -

0.03

0.09

-

0.21

%

--1.15

Sb

0.0051 -

0.12

0.02

Pb

1.39

-

9.69

3.44

Ni

0.0005 -

0.0006

Cu

0.06

2.91

Fe

---

S

---

-

-

% %

% 0.20

---

>0.0001

Cd

-

% %

9.76

>0.0001

>0.0001

Te

---

>0.0001

Se

54.42

% %

5.25

%

---

%

>0.0001

%

0.01

-

0.25

%

0.43

-

15.48

%

-

0.14

%

>0.5

4. Limestone These samples are not abundant and do not contain significant quantities of metallic elements such as Cu, Pb, Au, and Ag, demonstrating that they were used as flux and not as a material from which metal was extracted. However, they may have been used in the production of crucibles. These samples have been found only at San Bartolomé de Almonte and show the following ranges:

256

Jesús Fernández Jurado

San Bartolomé Au

0.1

-

0.3

grs./metric ton

Ag

10

-

29

SiO2

12.80

-

49.80

%

0.82

-

8.65

%

grs./metric ton

Fe2O3 Al2O3 Ca >0.01

Bi

%

As

0.04

-

0.05

%

Se

0.02

-

0.04

%

Sb

0.02

%

Te >0.01

Pb

-

0.10

%

Cd Ni

>0.01

Cu

0.03

-

0.11

%

Fe

0.01

-

10.88

%

S

0.10

-

0.5

%

%

5. Metallic lead In San Bartolomé de Almonte several burrs and droplets of metallic lead have been found (pl. 4), as well as a small sheet of lead. In this analysis we present the weight of each piece of lead and give its content in gold and silver (in grams per metric ton). As already mentioned, lead serves to attract the precious metals. This is probably the function of the pieces found, rather than as raw material to be extracted. The analysis of these samples gives the following results: Samples

1

Weight in grs.

26.08

Au

>0.1

Ag

49

2

3

4

5

9.853

9.405

4.706

6.300

>0.1 124

>0.1

>0.1

>0.1

53

53

73

The Tartessian Economy: Mining and Metallurgy

257

Plate 4. Leads.

Discussion of the Results After examining the results of the analyses, we are able to reach several different conclusions about the nature of the mineral ore, as well as about the metallurgic techniques that were employed—and progress in these techniques. The samples of ore can be clearly identified as “gossan,” a mineral of reddish or yellowish tones and easily identified by sight. We must emphasize the high content of gold and silver in these samples; however, we observe that even though most of the components in the samples are similar, there is a notable difference in the gold and an appreciable difference in the bismuth content. The samples from Huelva are richer in these substances than those from Almonte and Tejada, with more similarity between the samples from these last two sites. The data appear to show that the ore that arrived in Huelva had a different origin than that of San Bartolomé and Tejada. 19 It also seems evident that these peoples knew the importance of mining an ore that was rich in both lead and silver, in order to be able to extract the silver. However, we should not forget the samples of metallic lead found at San Bartolomé. The discovery of these pieces, in addition to some marginally archaeological considerations which will be discussed later, indicates a high level of knowledge about the metallurgy of silver and an evident progression 19. In addition to the difference in gold content, we must also emphasize the difference in bismuth (Bi). We are now initiating research in this area by analyzing pieces of silver. Our goal is to evaluate the traces of Bi that may remain in the silver, thus helping us to better define the hypothesis relating to the different points of origin of the ore.

258

Jesús Fernández Jurado

in this knowledge. These samples indicate that Tartessian metallurgists knew the necessity of adding lead to those ores that had a low lead content and that nevertheless contained silver. However, this high level of knowledge contrasts with the lack of skill in other phases of production. This fact may be deduced from the high contents of precious metals in the slag, along with quartz that had not melted—indicating methods of smelting that were not totally successful. This fact must have been known by the Tartessian metallurgists, as shown by the discovery in Almonte of a “storehouse” next to an oven (installations 1.2 and 1.1 respectively). 20 Here they stored slag from which the metals had not been totally extracted, along with blocks of limestone, presumably for resmelting. The high content of gold, silver, and lead in the slag that adhered to the pottery seems to be related to the procedure explained earlier, in which the regulus of gold, silver, and lead resulting from fusion was subjected to high temperatures. The content of SiO2 and Al2O3 in the slag corresponds to its clayey composition. The samples of limestone, from San Bartolomé only, clearly show its use as a flux, but without ruling out its use in the production of crucibles. However, since we have not found any crucibles in the strict sense, we need to keep this idea on the level of a hypothesis while we wait for more objective proof.

Health Problems Related to Metal Production Certain professions give rise to specific diseases, which are known medically as “industrial poisoning.” Within this field are included the toxic effects produced by metals, and more specifically and of interest to us here, are those produced by copper, silver, and lead. In this case we will focus our attention on poisoning produced by lead. Lead poisoning, or “Saturnism,” may not cause a high number of deaths today; nevertheless, it is a common disease among workers whose occupations force them into contact with lead (painters, manufacturers and handlers of batteries, metallurgists, plumbers, etc.). Colic and other gastrointestinal and hepatic conditions are among the different symptoms which may indicate Saturnism in an individual and call for a medical exam. 21 Already in antiquity Hippocrates and Pliny alluded briefly to these ailments in relation to those who worked in metallurgic activities. In addition to gastrointestinal problems, a person with Saturnism may suffer anemia, kidney disorders, encephalopathy, neuropathy (diseases of the ner20. D. Ruiz Mata and J. Fernández Jurado, “El yacimiento metalúrgico.” 21. M. R. Cullen et al., “Adult Inorganic Lead Intoxication: Presentation of 31 New Cases and Review of Recent Advances in the Literature,” Medicine 62 (1983) 240ff.

The Tartessian Economy: Mining and Metallurgy

259

vous system), muscular lesions, and stomatitis, which is an inflammation of the mouth with a precipitation of lead sulphide in the gingival mucous membrane, producing the so-called bluish-gray “Burton’s line.” 22 Lead poisoning can also produce impotence, and it can affect pregnant women with negative consequences for the fetus. Of the symptoms mentioned above, the only ones useful for our study are stomatitis and neuropathy, because they are the only ones whose consequences can be analyzed in skeletal remains. 23 Neuropathy from lead poisoning causes paralysis of the muscles of the hands and fingers, and later it can lead to a more wide-spread and intense paralysis, causing the hands to remain bent and hanging. Flexion of the fingers is weak, and the thumbs appear bent and pressed against the palm of the hand. 24 Thus, an individual shows a series of symptoms in the joints of the hands and fingers as a result of the paralysis, which can be studied on the skeleton. In like manner, by analyzing the concentration of lead in the teeth, we can discover if the individual suffered from stomatitis resulting from lead poisoning. This type of research is at present only a working hypothesis. We can test it only with the discovery and excavation of a necropolis from the Tartessian period, belonging to a settlement with metallurgic activity. In addition to information concerning the occupational hazards for Tartessian metal workers, a study of the above-mentioned symptoms in a necropolis would provide us with a broader knowledge of Tartessian society and economy. Thus, it would be possible to better study the social organization and division of work, as well as the life expectancy for different individuals according to their professional activity. We might learn about social stratification based on type of work performed—for example, whether the metal worker was a preeminent member of Tartessian society or not, and if metallurgic activity was performed by men or by women. In short, a broad and suggestive field of research is possible.

Conclusions The data from the excavations that we have performed in San Bartolomé de Almonte and from those in progress in Huelva and Tejada permit us to claim that Tartessian metallurgists worked an ore with high contents of precious metals, both gold and silver. However, all the ore did not have the same origin, as we see from the large discrepancy in gold content between the

22. G. Marañón and A. Balcells, Manual de diagnóstico etiológico, 13th ed. (Madrid: EspasaCalpe, 1984) 445–46. 23. W. J. Williams et al., Hematología (Barcelona: Salvat, 1975) 102. 24. G. Marañón and A. Balcells, Manual, 743, 767–68.

260

Jesús Fernández Jurado

samples from Huelva, which are very rich in this metal, and those from San Bartolomé and Tejada, with a content that is considerably lower. In view of these data, we can conclude that the mineral ore that arrived in Huelva came from the area of Riotinto, by means of the river with the same name. 25 However, the ore that was worked in Almonte probably came from the area of Aznalcóllar and arrived in Almonte by way of Tejada. 26 The final destination of this metal was the trading zone at the mouth of the Guadalquivir 27 (fig. 5). But this full, rich, and diverse production of silver in the Tartessian area, which is securely documented from the 8th century b.c. on, declined in the final decades of the 6th century. Determining the causes of this decline is not easy, and for the moment we lack the data that might help us to understand completely. However, it is clear that beginning at the end of the 7th century b.c., a profound transformation in trade occurred within the Tartessian world. This transformation reflected events in the Mediterranean, where the commercial prominence of the eastern Phoenicians was yielding before the onslaught of the Greeks and the western Phoenicians. Various economic events and wars contributed to the changes that occurred in the Mediterranean beginning around these dates. These changes were clearly marked by modifications in commercial style. Aristocratic trade lost its power and prominence before the new concept that was winning ground in the eastern Mediterranean: trade that functioned on supply and demand with a monetary basis. These events would affect the southwestern Peninsula intensely, and in the first years of the 6th century b.c. we see open settlements, such as Almonte, begin to disappear. Greater urban concentrations, such as in Huelva and Tejada la Vieja, begin to appear. However, these changes, which occurred principally in the transition from the 7th to the 6th century b.c., did not transform the Tartessian economy yet, which continued to maintain a great vitality. Nevertheless, its decline was decreed nearly a century in advance. Certainly the events in the Mediterranean affected the Tartessian world, 28 but it is probable that a series of factors combined to bring about the disap-

25. We must point out that the western part of the province of Huelva has not yet been researched in depth and from the perspective that concerns us. It is possible that these ores arrived on the coast (specifically southern Portugal and Ayamonte) by means of the Guadiana River. 26. See notes 13, 14, 15, and 16. 27. Along with this proposed land route, the river route of the Guadiamar should not be overlooked. This river is a tributary of the Guadalquivir on the west and joins the Guadalquivir in the marshlands near its mouth. Ore from the area of Aznalcóllar could have also gone to the Atlantic via this route. 28. The traditional link that historiography has made between the decline of Tartessos and the battle of Alalia should not be forgotten.

spread is 6 points long

Me d tra i terra din n ea g-po n-A sts nd a

The Tartessian Economy: Mining and Metallurgy

261

ian lus

nean

Eastern Mediterra

Fig. 5. Possible routes and nuclei for trade. Area covered by sea in antiquity and trading nucleus Hinterland of Gadir Trading nuclei

§

Trade routes

pearance of Tartessos. In our opinion, one of the keys to this decline was mining and metallurgic production. In this article we have discussed the type of mineral ore exploited (gossan), the metallurgic technology employed (cupellation), and the different mining centers (Riotinto and Aznalcóllar), which we know provided the basis for the richness in silver of the Tartessian world. We believe that the problem basically became one of lack of economic profitability in the mining and metallurgic enterprise, which began to be felt at the beginning of the 6th century.

262

Jesús Fernández Jurado

The mining technology that the Tartessians possessed did not show the same progress and improvement that occurred in the area of metallurgy. We know that the Tartessians extracted ore from the oxidized caps and, consequently, once their technology became incapable of mining ore at the necessary depth when the surface concentrations ran out, it became necessary to search for ore in other places. We should not be surprised that it is precisely during the 6th century b.c. that we see the formation of the socioeconomic and population bases which give birth to the Iberian world in the southeastern and eastern Peninsula. We must remember the important mines and the richness in silver that this region possesses. The Tartessians must have known about these mines, given the contacts that they presumably had in order to obtain metallic lead, which is not found in the mining region around Huelva. All of these factors probably caused, among other things, a large transfer of population from the Tartessian region to the peninsular Southeast. 29 In summary, we believe that the nonprofitability of the mining-metallurgic enterprise, together with the new prevailing trade concept, put an end to the riches of Tartessos and provoked its collapse. 29. J. M. Blázquez et al., “Evolución del patrón de asentamiento en Cástulo. Fases iniciales,” Arqueología Espacial 4 (1984) 241ff.; J. Fernández Jurado, “El yacimiento metalúrgico de San Bartolomé de Almonte,” Revista Huelva 4 (DPH, 1982) 40–46.

The Beginnings of the Phoenician Presence in Southwestern Andalusia With Findings from the Excavations at Cabezo de San Pedro (Huelva), San Bartolomé (Almonte, Huelva), Castillo de Doña Blanca (Puerto de Santa María, Cádiz), and El Carambolo (Camas, Sevilla)

Diego Ruiz Mata

Posidonius’s account, passed on to us by Strabo (3.5.5), is often used to explain the initiation of the Phoenician presence in the southern Iberian Peninsula. During his stay in Cádiz, Posidonius heard an ancient legend that had been preserved in the temple of Melkart. According to this legend, two other trips, to Sexi (Almuñécar) and to Onoba (Huelva), had preceded the founding of Gadir, the oldest Phoenician city in the West. According to Velleius Paterculus (Historiae Romanae1.2), the founding of Cádiz took place around 1104 b.c. However, this date has not been corroborated archaeologically, and the most ancient remains date to the beginning of the 8th century b.c., at the earliest. Due to archaeological excavations in various necropolises and towns, we currently have a more accurate knowledge of the pre-Phoenician cultural environment, the economic factors that motivated the Phoenician seafarers, and a more accurate date for their arrival. Around the year 900 b.c., or possibly earlier, we see a culture in western Andalusia that we refer to as the Tartessian Final Bronze. This culture possessed an economic base that provided the principal attraction for enticing the Phoenicians to their western overseas experience and their settlement on Iberian shores. Using only the data from pottery and its range throughout the area, we can delimit the geographic boundaries of the Tartessian Final Bronze. The Editor’s note: This article appeared in Spanish under the title “Aportación al análisis de los inicios de la presencia fenicia en Andalucía sudoccidental, según las excavaciones del Cabezo de San Pedro (Huelva), S. Bartolomé (Almonte, Huelva), Castillo de Doña Blanca (Puerto de Santa María, Cádiz) y El Carambolo (Camas, Sevilla).” The article was published in Homenaje a Luis Siret (1934–1984) (Seville: Consejería de Cultura, 1986) 537–56.

263

264

Diego Ruiz Mata

results are shown by the map in fig. 1. The Tartessian settlements were located in the Spanish provinces of Huelva, Sevilla, Córdoba, and Cádiz, especially on the most level land. In addition to lowland fields and a broad front along the Atlantic coast, this area contains the basin of the Guadalquivir River, which is bounded by the foothills of the Sierra Morena on the north and the chain of the Béticos on the south. Although these boundaries define the area of the Tartessian Final Bronze, we also note influences that reach as far as the southeastern Peninsula and Estremadura. These influences are a result of trading relationships that most likely intensified under Phoenician stimulation beginning in the middle of the 8th century b.c. In the region of Huelva, the settlements were concentrated where the city of Huelva stands today (on its various hills) and along the Río Tinto (or Red River), up to its source and in the surrounding area of lowlands. Some settlements were established as a result of metallurgic activity and trade. In this way we can explain the settlements at Cerro de Salomón and Quebrantahuesos (in the region of Riotinto), and Niebla, as well as the market center that must have developed in the hills of the city of Huelva. Farther to the east we find the fortified towns of Tejada la Vieja (Escacena) and Mesa del Castillo (Villalba del Alcor), the metal-working town of San Bartolomé (Almonte), and probably El Rocío, on the shores of the marshland. Several geographic features bound the lower Guadalquivir: the plateaus of El Aljarafe and Los Alcores, la Campiña, and the foothills of the Sierra Morena. Settlements were located on the eastern side of El Aljarafe (Escacena 1982) near the Guadalquivir, and it is also the eastern side of Los Alcores that is most populated (Amores 1982). The Tartessian communities settled on low hills between the two plateaus, perhaps because of the possibilities for farming and cattle-raising in the area. We must also mention the excellent route that passed through here to the interior and to the metal-yielding region of the Sierra Morena, suggested by the location of the town of Setefilla (Aubet Semmler, Escacena, and Ruiz 1985, 135–140). Until now, not much field work has been done in the area of Gadir (Cádiz). The data come from Mesas de Asta, on the border of the marshes of the Guadalquivir, and from the recent excavations at Castillo de Doña Blanca. But in surveys within the city limits of Puerto de Santa María and the western zone of Jerez de la Frontera, several settlements have been found that fill the gap that existed previously. Likewise, little research has been done in Córdoba, but the known sites show influences from the lower Guadalquivir going up along the course of the river, as Colina de los Quemados inside the city of Córdoba itself suggests. Other sites are linked to the countryside around Sevilla, as in the case of Ategua. In the area where the Tartessian Final Bronze Age developed, not all the settlements can be dated to the same period. We see, of course, some nuclei that

The Phoenician Presence in Southwestern Andalusia

The Phoenician Presence in Southwestern Andalusia

265

Fig. 1. Map of the distribution of Tartessian and Phoenician sites. „ Tartessian sites ∑ Trading posts—Phoenician colonies W Sites with Tartessian materials

were inhabited since Phase I of the pre-Phoencician era, but the population grew under Phoenican economic stimuli. For example, the region that we are surveying between the Guadalete and Guadalquivir Rivers became more intensely populated, as was to be expected, beginning in the middle or at the end of the 8th century. According to its excavator, Jesús Fernández Jurado, the site of Tejada la Vieja dates from the same period as Cerro Macareno (Pellicer, Escacena, and Bendala 1983). Nevertheless, during Final Bronze I, the basic economic structures were forged that would shape the Tartessian economy and the Phoenician interest in the area, although it is possible that these structures were not yet well developed. In addition to agriculture and cattle raising, we refer to silver metallurgy. Silver slag has been found in the strata of the pre-Phoenician phase in Huelva, and possibly also in Almonte.

266

Diego Ruiz Mata

The information from Posidonius and passed on by Strabo (3.5.5) about the founding of Cádiz is now being confirmed objectively and can be explained in terms of economic interests and colonial possibilities. Likewise, the date given by Velleius Paterculus for the founding of Cádiz, and consequently for the Phoenician presence in the southern Peninsula, continues to be clarified by the ongoing field work. The purpose of this article is to present data on the beginnings of the Phoenician presence in the southwestern Peninsula based on the results of recent excavations. Huelva can be identified as the Onoba mentioned by Strabo and probably was the heart of the ancient Tartessos. San Bartolomé was a metalworking town where silver was extracted from the 8th century b.c. to the beginning of the 6th, and it was probably linked to Cádiz. Castillo de Doña Blanca provides data about Cádiz, since it is close to that metropolis and acts as a mirror image. El Carambolo, located on the plateau of El Aljarafe in the province of Sevilla, provides evidence of the Phoenician presence in the lower Guadalquivir, as seen in the considerable amount of red slip Phoenician pottery coming from that site.

I. Huelva: Cabezo de San Pedro The city of Huelva is located at the confluence of the Tinto and Odiel Rivers, on the southern part of a peninsula formed by these two estuaries. It is at the foot of a chain of hills that rise to the north and northwest, with elevations of 50 meters on the average. On these hills (cabezos), or more specifically, on the slopes of these hills, we find the area of greatest archaeological interest. The early settlement, and even the later one, was spread over these hills, all of which presented similar conditions for habitation. The data acquired through several archaeological excavations hint at a nuclear settlement on the slopes of Cabezo de San Pedro, as well as on the other hills. This pattern has produced nuclei of stratigraphic sequences in several zones, but we have not yet found an uninterrupted sequence from the first phase of occupation until abandonment. It was not necessary, however, to concentrate the population in only one location, since each hill provided identical conditions for human habitation. In this context we find a stretch of wall that represents the earliest structure reflecting the Phoenician presence in Huelva (Ruiz Mata, Blázquez, Martín de la Cruz 1981: 179ff.). This wall, whose function and date have at times been debated, is located on the western slope of the hill of San Pedro, at the beginning of a small stream bed that separates this hill from the hill of Cementerio Viejo (which has since disappeared).

The Phoenician Presence in Southwestern Andalusia

267

Plate 1. 8th-century Phoenician wall at Cabezo de San Pedro (Huelva). The central supporting column in this wall was constructed with ashlars (1 to 1.2 m long and 0.5 m wide), using the header-stretcher method.

The wall runs from east to west and has been excavated along a stretch of approximately 10 meters, with a gap of almost a meter on its western end. Its width is not uniform, reaching 2 meters in the center (where a column, or pillar, of ashlars is located) and narrowing at its ends to 1.30 meters. As for its construction, the wall contains a column of large ashlars, 1 to 1.20 meters long and 0.50 meters wide. Its courses alternate, using a header and stretcher arrangement. The column supports two extending wings made of rough slabs of slate, lightly squared, along with rubble to fill the cracks. The order of construction appears to be the following: first, the column was erected, taking advantage of the lowest part of the depression and serving as a foundation; afterwards, the rough slabs of slate were placed against the column, without digging a foundation trench, but rather adapting them to a previously leveled surface. The southern back face of the wall simply leans against the hill’s wall of tertiary loam. The function has not been clarified, but most probably it served as a retaining wall to contain the earth and erosion resulting from rainfall. It is less likely that it served as a defensive wall, and certainly it did not form part of an area divided into rooms. We have already mentioned how the settlements were spread over the hills of Huelva, reaching to the zone where the sea touched these promontories. It

268

Diego Ruiz Mata

Plate 2. The Río Tinto, which flows from the mining region of Riotinto to Huelva. The river has a striking reddish hue.

seems that there was no layout of streets and blocks, but rather the settlements were adapted nuclearly to the slopes of the hills of San Pedro, Cementerio Viejo (which has since disappeared), Molino del Viento (lowered by modern construction at some points), and La Esperanza. According to the existing archaeological data, the oldest nucleus of settlement corresponds principally to the hill of San Pedro (Blázquez, Ruiz Mata, et al. 1979; Ruiz Mata, Blázquez, and Martín de la Cruz 1981), which contains strata from the pre-Phoenician Final Bronze (9th century b.c.). From the 8th to the 6th centuries the settlements spread across the hills of Molino del Viento and La Esperanza, as a result of economic and, therefore, population growth in Huelva, based on bronze and silver metallurgy. The excavations directed by Jesús Fernández Jurado at different points on Molino del Viento confirm this organization of the habitat. They also confirm that its chronology is more recent (beginning in the 8th century b.c.) than that of the strata of San Pedro. These excavations have also shown that there are frequent layers of erosion and mud brought down from higher areas in the hills by torrential rains, a situation similar to the present day after periods of intense rainfall. The use of the slopes, the continuous danger of landslides of crumbly clay, and the torrential rainfalls objectively confirm the purpose of the wall of San

The Phoenician Presence in Southwestern Andalusia

269

Pedro as a way to contain erosion and to protect the inhabited areas below. This is also the function of a wall using the same construction technique (columns or pillars with extending walls of rubblework) discovered a few years ago in Tyre (Elayi 1980; Bikai 1978). Let us now look at the technique used to construct the wall in Tyre. J. Elayi (1980) has published material on a type of Phoenician wall that he calls “pillared,” which uses a technique similar to that of the wall of San Pedro. He distinguishes various types, but all have in common the use of columns of ashlars and extending walls of rubble work. This technique originated in the North, perhaps in Hittite construction during the Empire Age (1460–1190 b.c.), and developed later in areas under Hittite influence. This technique reached its high point on the central coast of Phoenicia between Joppa and Sarepta, although it is seen along the whole coast. This technique was used not only for public buildings, but also for dwellings, casemates, and certain temple zones. In Tyre, it was used for a retaining wall. Chronologically, the earliest walls are found at Ras Shamra in the 14th and 13th centuries b.c.; next are those of level IV of Megiddo (1000/800 b.c.), level IX of Tyre (around 850 b.c.), those of Hazor (level VB, 760–740, reused in level VA, 740–732, and level IV, from the end of the 8th century to the middle of the 7th century b.c.), and Aroer (8th–7th centuries b.c.). At Motya they are found in the 6th–5th centuries b.c., and at Nora and Carthage they are found in the 3d–2d centuries b.c. Area A.2.3 provides the best-defined stratigraphy for the dating of the wall of San Pedro (Ruiz Mata, Blázquez, and Martín de la Cruz 1981). The wall lies on top of levels of the pre-Phoenician Final Bronze I (Phase I.A). It was constructed during Phase I.B, and the strata that rest against its worked side (the northern face) also belong to Final Bronze I. In other words, its construction took place toward the end of Final Bronze I and represents one of the first proofs of the Phoenician presence on the hills of Huelva. It was probably an excellent way to establish friendly relations, and it provided a technical solution at the source of the small stream that separated the hills of San Pedro and Cementerio Viejo to the constant problem of torrential rainfalls during the spring and fall. It was built at the beginning of the 8th century b.c. (around 800), and the connection between this wall and those that were built on the Phoenician coast is obvious. Stratigraphically this date appears certain, since at the level of Area A.2.1 we found only one fragment made on the potter’s wheel, a body sherd which appears to belong to a small oenochoe (Ruiz Mata, Blázquez, and Martín de la Cruz 1981: figs. 41, 153). The rest of the material is characteristic of the pre-Phoenician Final Bronze I. The retaining wall in Tyre is dated to around 850 b.c. and thus probably predates that of San Pedro. The dates obtained by Jesús Fernández Jurado in several excavations at Molino del Viento in connection with the Phoenician presence also approach

270

Diego Ruiz Mata

Plate 3. Walls and square buttress tower of Tejada la Vieja.

the beginning of the 8th century b.c. And, although it was not found in context, a fragment of Attic crater from the mid-Geometric II, dated between 800 and 760 b.c. and brought to this spot by Phoenician seafarers (Shefton 1982, 342), indicates an early date for the initiation of Phoenician-Tartessian relations.

II. San Bartolomé (Almonte, Huelva) The settlement of San Bartolomé (Ruiz Mata 1981; Fernández Jurado 1983) is located within the city limits of Almonte and corresponds to an indigenous community of metal workers who extracted silver during the Tartessian period. The population was spread over four hillocks overlooking the stream of San Bartolomé. This type of settlement pattern with dwellings in several nuclei was normal for settlements of the Final Bronze. So far, more than thirty huts from the 8th and 7th centuries have been excavated. The location and the metal-related activities of San Bartolomé can be explained only in relation to mining centers and trade routes, since San Bartolomé was not a mining town itself (there are no mines in its immediate vicinity, nor for geologic reasons could there have been). Instead San Bartolomé must have been a metallurgic town which received ore for refining, as also occurred in the city of Huelva (Garrido and Orta 1982).

The Phoenician Presence in Southwestern Andalusia

271

Plate 4. Tejada la Vieja: possible area for washing ore.

San Bartolomé is located on a route that began, perhaps, in the region of Riotinto, although more probably in Aznalcóllar. The route reached Tejada la Vieja, and went from there, via Manzanilla, to San Bartolomé and El Rocío, on the banks of the Guadalquivir estuary. Remnants of slag from silver refining and other metallurgic activities are documented in Tejada la Vieja and El Rocío. This route probably served the Cádiz metal trade. San Bartolomé’s existence was justified by the necessity of maintaining a route different from the one that led to Huelva (the location of Tartessos) along the Río Tinto. The analyses of the slag coming from the region of Riotinto and from San Bartolomé provide data that do not contradict the results obtained by comparing the ceramic typology and analyzing the ware. It has been calculated that twenty million tons of slag resulted from mining the Riotinto area in ancient times, of which four million tons probably came from the Tartessian period (Pinedo Vara 1963, 25). When the slag from Riotinto was analyzed, the following results were obtained: Ag: 97%; Au: 0.10%; Bi: 0.42 %; Fe: 0.09%; Pb: 0.14%; Cu: 2.25%. The results from the analyses performed on the ores and slag of San Bartolomé (Ruiz Mata and Fernández Jurado) suggest the possibility that the ore at San Bartolomé did not come from Riotinto, but perhaps from the mines of Aznalcóllar (province of Sevilla), based on the lead and copper contents that differ notably from one site to the other. Riotinto provides

272

Diego Ruiz Mata

slag that contains copper and little lead, while the slag from San Bartolomé has a greater content of lead and less copper. Percentages similar to those at San Bartolomé are provided by two samples of slag at Castillo de Doña Blanca from strata of the 7th century b.c. The composition of clays has been analyzed (Galván) by means of X-ray diffraction, electron microscopy, and infrared spectroscopy on handmade and red slip pottery from the towns of San Bartolomé (Almonte), Castillo de Doña Blanca, San Pedro, Valencina de la Concepción, and Cerro de la Cabeza (Santiponce, Sevilla). Similarities have been noted between the wheel-made pottery in San Bartolomé and in Castillo de Doña Blanca. The results show that the wheel-made pottery from San Bartolomé and Cerro de la Cabeza does not contain the same mineral components as the handmade pottery at these sites, but rather that their composition matches the red slip pottery from Castillo de Doña Blanca. The analyses seem to suggest imports of red slip pottery from the area of Gadir to San Bartolomé and Cerro de la Cabeza. In the strata of San Bartolomé, the earliest period corresponds to a horizon of the Chalcolithic period, with foundations of circular huts and ceramic types that are related to Huelva and the lower Guadalquivir. After a hiatus in which we do not see Middle or Late Bronze materials, the site was reoccupied during Final Bronze I, perhaps during the last part of this phase (second half of the 9th century b.c.), and it was inhabited until the end of the 7th century or beginning of the 6th century b.c. Huts XXXII/XXXIII and V provided the earliest material from the protohistoric settlement of Final Bronze I in the period preceding the Phoenician presence in the town. The ceramic types have been partially published (Ruiz Mata 1981). However, in regard to types of cooking dishes (especially the fabrication of the rims, the brown and grayish ware, and the gray and black well-burnished surfaces with a metallic look), the most direct connections may point toward the area of Huelva, the cultural nucleus for this community of metal workers. Connections with the lower Guadalquivir are not entirely absent, though, as suggested by fragments of large cups and cooking dishes with red monochrome geometric motifs generally similar to those at El Carambolo and the lower Guadalquivir. San Bartolomé has been extremely interesting because of the opportunity to extensively excavate a habitat with separate dwellings dating to this period. At times excavation is difficult in places where people have lived without interruption and where thick layers of strata have accumulated. However, the huts at San Bartolomé have a fragile structure and lack thick strata, suggesting short, temporary occupations, noted also in the uniformity of ceramic types that have provided a well-documented horizontal stratigraphy. In regard to the theme that concerns us here, we can be precise about the first Phoenician remains in the settlement, the absolute percentages (since each hut presupposes a closed grouping), and the course of development following the first conspread is 1 pica long

The Phoenician Presence in Southwestern Andalusia 1

273

2 10 11 4

3

6

5

12

7 8 9

Fig. 2. San Bartolomé (Almonte, Huelva).

tacts. We note, though, that San Bartolomé sprang up only because of silver metallurgy and probable trade with Gadir, and it did not achieve a high degree of orientalization during its existence. Rather it maintained its indigenousness and its system of living in huts, without the urban development that other towns experienced during this period in the region of Huelva and the lower Guadalquivir. The interpretation of the material data also suggests indigenous control of the metallurgic resources. Another group of huts (XXXIV, XIV.A and XIV.B) provides later handmade materials. The cooking dishes, for example, lose the carination of their exterior rim, as do the cups. The closed forms are substituted by types that show evolved or new formal aspects. However, it is basically the same cultural world, with a new look to its pottery. Materials made on the potter’s wheel are very scarce and constitute a very small percentage when compared to the indigenous pottery made by hand: six fragments from Hut XXXIV, three from Hut XIV.A, and only one from XIV.B. The forms do not show much variation: a bowl covered with a red slip (fig. 2.3), another bowl devoid of decoration (fig. 2.1), a rim of a fragmented red slip plate with an undetermined width, a small juglet with a short turned-out rim (fig. 2.5), a bottle neck with a thick rim and carination on its lower edge (fig. 2.2), a body sherd from a red slip oenochoe, a gray plate with a concave rim and burnished/polished surfaces, two fragments of urns decorated with bands, and the mouth of an amphora (fig. 2.4). Sherds of red slip plates are not very common, with a preference for bowls with a red slip on the interior or no decoration at all. During this initial period the demand for pottery was not very great. In the next phase, the gray pottery gains a much higher percentage. The presence of juglets and oenochoes as exotic materials in this indigenous environment may show that perhaps they were valued for their contents and used as gifts. This is probably the significance of the amphora in Hut XIV.A.

274

Diego Ruiz Mata

A later phase is distinguished by the appearance of a new type of handmade cooking bowl with a short and angular rim and by new styles of cups. There is greater use of impressions made with fingersprints on the shoulders of rough-surfaced pots and of decorations incised on the bell-shaped à chardon pots and vases. Wheel-made pottery is now more common, though its percentage is not high and it does not compete seriously with the usual indigenous types. Its presence is the result of more intense relationships rather than daily living side by side, and the settlement maintains its arrangement of ancestral dwellings and its own pottery. In this phase the style of wheel-made pottery is more varied. For example, the number of bowls decorated with an interior red slip increases, and the gray pottery is quite a bit more frequent (fig. 2.6, 7). We now notice a gray bowl with a rim reinforced on the inside, which was absent during the former phase. The fragments of amphorae (fig. 2.8, 10) and urns (fig. 2.12) increase considerably. In general, the demand for Phoenician products is greater, and the ceramic style is more varied. Phoenician interest in the settlement was probably irregular and exclusively commercial, related to the silver work that was done there. It also appears that San Bartolomé was a seasonal home for the indigenous inhabitants. We cannot perceive even a trace of urbanism in the whole area excavated. The settlement was abandoned while the indigenous pottery was still prevalent. As for the Phoenician presence at the site, we must emphasize the scarcity of sherds in the initial period and the absence of characteristic forms which can be dated. The first period of occupation is indigenous, perhaps an advanced phase of Final Bronze I, and lacks materials made on the potter’s wheel. During the evolution of the handmade pottery, the Phoenician imports begin with the types mentioned above. There are no red slip plates (perhaps they were not in demand), a phenomenon which also occurs at the nearby town of Tejada la Vieja, according to our information from Jesús Fernández Jurado (director of the excavations). Neither are there other forms which can be dated more precisely. However, the open forms made on the wheel are bowls (as in Tejada la Vieja), which are less common in the towns on the coast and in Huelva itself. The indigenous forms, which are typologically related to Huelva and the lower Guadalquivir, can be dated to the end of the 8th century b.c. and the beginning of the 7th, which probably was the time of the Phoenician arrival in San Bartolomé. In regard to the origin of the wheel-made pottery, we have indicated above how the composition of the clay shows the same mineral components as that of Castillo de Doña Blanca, which probably means that they originated in Gadir.

III. Castillo de Doña Blanca Castillo de Doña Blanca is the name for a watchtower constructed in the 15th century a.d. and for a tell, more or less rectangular in shape, about 300

spread is 1 pica long

The Phoenician Presence in Southwestern Andalusia

275

meters long on its east-west axis and 200 meters wide on its north-south axis. The tell has an elevation of 34 meters above the present sea level, containing from 8 to 10 meters of strictly archaeological strata. The site belongs to the city of Puerto de Santa María (province of Cádiz) and lies about three kilometers off the highway on the side road that leads to Jerez de la Frontera through El Portal. It is located at the foot of the Sierra de San Cristóbal, a range of hills with a smooth outline and low elevation, rarely exceeding 100 meters, and on the border of an ancient estuary whose waters once reached the city. At one time this estuary increased the size of the present Bay of Cádiz up to the base of these hills. However, it has filled up gradually with alluvial silt from the Guadalete River, which in ancient times flowed into the sea near the site of Castillo de Doña Blanca. This site, therefore, has meaning only as a port closely related to the Phoenician city of Gadir and the indigenous market. The first description of the site comes from F. Ventura López (1923), who was still able to see surface remains of streets, rooms, and a wall more than three meters thick. On the necropolis he saw “Phoenician tombs, and tombstones with Iberian and Tartessian characters.” F. de Ciria y Vergara (1935), a local scholar, describes the site in similar terms. César Pemán (1941) was also familiar with these ruins, suggesting that they could be some topographic point cited in the ancient sources about Tartessos, perhaps the port of Tartessos itself. A. Schulten (1943) could still see several vestiges of construction, including the wall that surrounded the town, and he identified it with the port of Menesteo, mentioned by Strabo, Ptolemy, and Marcianus. In spite of the fact that there were references to its location in the scientific bibliography, the site was never systematically excavated. The materials collected from its surface, which L. Peña Basurto had referred to at times (1952), were never published or evaluated culturally or chronologically. Only recently, due to surveys performed by G. Chic García (1979), have we realized its true significance as a pre-Roman town from the 9th–8th centuries to the 3d–2d centuries b.c. In 1979, the Subdirección General de Arqueología granted us permission to administer the excavations, which continue up to the present time. The principal motivation for this excavation was the opportunity to research a protohistoric site close to Cádiz (such as Doña Blanca) in order to provide the data that we lack for Gadir concerning its cultural characteristics and chronology. We have finished four seasons at the town in the southeast and southwest corners (1979, 1981, 1982, and 1983), and one season on the necropolis (1984). The Phoenician presence in the town is the theme that interests us in this article, but it is also appropriate to briefly present the cultural and chronological development of the site as seen through the stratigraphy excavated up to this point.

276

Diego Ruiz Mata

Chalcolithic Period The earliest ruins in the area correspond to the Chalcolithic period. They are spread around the base of the Sierra de San Cristóbal, and they occupy the whole area of Doña Blanca and the area called “La Dehesa” to the east. Circular huts have been excavated that reach 2.5–3 meters in diameter, with foundations of rubblework placed against the sides of the circular pit that had been dug to a depth of 0.40–0.50 meters. These foundations probably supported walls of mud-brick. Other smaller circular and elongated pits have been excavated that were probably used for storerooms or toilets for the dwellings. The pottery is of poor quality and consists of closed vessels, bowls of different sizes and depths, and plates with almond-shaped or thickened rims, typologically matching that of the lower Guadalquivir and Huelva. Decorated pottery is scarce, with the exception of a bowl or shallow plate that has black radial lines painted on the polished interior surface. In the southeast zone of the town of Doña Blanca, there is a stratum on top of bedrock that may correspond to a later phase of the Chalcolithic period, and from it we have gathered numerous sherds with a late Bell Beaker decoration and form. The small space did not permit us to note the features of the dwellings.

Hiatus Later superimpositions have not been found in La Dehesa. In the excavations done in the town, the strata that correspond to the Chalcolithic period are covered by a sterile layer of reddish clay 0.50/0.70 meters thick.

Final Bronze I Within the town of Doña Blanca we have not detected clean strata that correspond to this period. But to the northeast of the town, on the slopes of the Sierra de San Cristóbal and among the surface materials at La Dehesa, we have found materials from Final Bronze I that surely came from the groups of huts located in this area. We have found carinated cooking bowls, burnished decorations, and fragments of large vessels that sometimes have a thick layer of red ochre on the outside. We have also found stands (in one case, with a red line decoration on the central support), rough-surfaced pots, and a fragment of a large vessel with a burnished exterior and red geometric design, of the type that has been called “Carambolo” (from the zone known as fondo de cabaña) and that we prefer to call Guadalquivir I. These are the most characteristic materials from this period.

Final Bronze I–II Commencing with this period (the first half of the 8th century b.c., perhaps at the very beginning of the century), the founding of Doña Blanca took

The Phoenician Presence in Southwestern Andalusia

277

place, evidenced by the first remains of rectangular rooms with red clay surfaces and walls of rubblework plastered with lime. The existing data show that probably only the southern part of the site was occupied during this period, which was the terrace nearest the water. It is here that we can observe the initial contacts between the indigenous population and the Phoenicians that had settled in Gadir. The greatest percentage of pottery is fashioned by hand and corresponds to that of an advanced stage of the Tartessian Final Bronze. The Phoenician pottery is similar to the earliest pottery from Morro de Mezquitilla and Chorreras in Málaga and constitutes the earliest evidence that we have of the initiation of the Phoenican presence in Doña Blanca. We will deal further with this topic below. 7th Century b.c. This is a period of great urban activity during which the city was enlarged. The excavated remains belong to dwellings with limestone structures, walls of plastered mudbrick, and walls of well-executed rubblework, such as foundations. During this century we note two or three superimposed architectural structures, as well as greater stratigraphic depth. We have excavated a stratum that can be dated to the beginning of the 7th century b.c. From this point on and in a process of orientalization that must have occurred very rapidly in this town, the largest percentage of pottery was made on the potter’s wheel. The handmade pottery is less common and the cooking vessels lose their distinctive carinations and rims. At the same time the shallow bowls with pointed rims are abundant, with burnished motifs that are new and more poorly executed. The wheel-made pottery already shows a wide variety, and we will point out just a few examples. The plate rims, which are scarce, have widths of 4–5 centimeters. The red slip carinated bowls are of the classical type from Toscanos and Mogador. The lamps now have two wicks and lack the flat bases of those in the older strata, and they are covered by a red slip. The amphorae are of the de saco type with solid, thickened rims on the interior. We now see a large number of urns with cylindrical necks and bodies, and those with larger dimensions (or pithoi), with decorations of bands, etc. During the rest of this century, the percentage of wheel-made pottery increases, while the handmade pottery is reduced to bowls with pointed rims and some rough-surfaced pots. The rims of the red slip plates acquire a greater width (6–7 centimeters), while the other wheel-made forms show hardly any variation. Indeed, the impression is that the town of Doña Blanca, due to its proximity and contacts with Gadir acquired a high degree of orientalization in all of its aspects from the beginning of the 7th century b.c. on.

278

Diego Ruiz Mata

6th Century b.c. During this century, and more precisely at the middle, the wheel-made types of pottery continue to evolve, and at the same time (as occurs also in the colonies at Málaga and Huelva), importations from eastern Greece begin (SOS and Corinthian amphorae and Ionian and Attic cups). During this century we can point to a certain decline in construction and less stratigraphic depth. Some pottery forms appear to be of Carthaginian origin. We begin to glimpse ceramic types that are a prelude to the Turdetanic culture.

5th Century b.c. This century witnesses a resurgence of the city, which acquires a layout of streets and blocks. The ceramic types evolve toward the classic Turdetanic forms, although certain orientalizing forms are maintained. It is at this time that trading posts for the salting industry begin to function, and we note at least twenty nuclei (of families?) on the coast between Puerto de Santa María and Rota who are involved in this industry, providing materials from this period. In Doña Blanca we can document the same types of amphorae that were found in strata in the agora at Corinth before 448 b.c. (Williams II 1978: 17– 19), in a warehouse associated with the fishing industry. In addition, we see amphorae from the Greek colony of Marseilles, from the beginning or the first half of the 5th century b.c. Certain Attic products should probably be dated to the middle of this century, such as the so-called Cástulo type cups.

Later Developments At the end of the 5th century b.c. and beginning of the 4th, the Attic imports increase, while the Turdetanic ceramic type remains fully developed. During the middle and second half of the 4th century b.c., relations with northern Africa are frequent, demonstrated by the appearance of amphorae and a considerable quantity of imitation-Greek red pottery, like that found at Kouass. At the end of the 4th century b.c. and during the first half of the 3d century, we see a gradual abandonment of the town, for reasons yet unknown. It is possible that the silt deposits in the Guadalete River made navigation difficult to the port that had given the city its economic prosperity. Perhaps the population was moving toward the coast of Puerto de Santa María-Rota, where a prosperous salting industry was flourishing at a spot better suited for fishing and salt. Roman vestiges are scarce, suggesting the existence of a few homes. The medieval remains of the 11th–13th centuries a.d. can be explained in the same way. Let us return to the theme that concerns us: the first period of the Phoenician presence in Doña Blanca. Areas 4 and 5 (excavated in 1983) have pro-

spread is 6 points long

The Phoenician Presence in Southwestern Andalusia

279

vided the most data so far. Below we present the strata from Area 5 that contain the earliest elements in that part of the town. The numbers given are those of the complexes and were used during the course of the excavation.

2.035 This number corresponds to the stratum that lies on virgin soil and has provided material from the Chalcolithic period.

2.034 The earliest protohistoric materials from this area (the southeast corner of the town) lie on a stratum of sterile reddish clay. This stratum provided few fragments, mostly atypical, including body sherds of burnished cooking vessels and rough-surfaced pots. Next to them were fragments of wheel-made pottery, including a body sherd of an oenochoe, with grooves on the upper section of the sherd.

2.033/2.032 These are strata with little depth and few sherds. They have yielded a rim of a cooking vessel with a very pronounced carination on the outside and burnished surfaces—a frequent type during Final Bronze I (Ruiz Mata). Also, we found cups, rims of pots, and burnished decorations that had been reticulated and incised. The wheel-made materials correspond to bases of plates and body sherds of amphorae and red slip oenochoes.

2.031 This stratum has greater depth and more abundant materials. The rims of cooking vessels have smooth carinations on the outside (fig. 3.1–5) and belong to a type that is found at a late stage of the Final Bronze—Phase I/II—in San Bartolomé (Almonte, Huelva) (Ruiz Mata 1979, 3ff.). All of the burnished decorations are reticulated with a narrow crosshatch and divided into quadrants separated by burnished bands, as is frequent in the lower Guadalquivir. In addition, there are bowls (fig. 3.6, 7), à chardon cups (fig. 3.8, 9, 12), and rough-surfaced pots (fig. 3.11). The deep cups with an incised geometric decoration continue. Among the wheel-made pottery we can point out a body sherd of a carinated bowl (probably Samarian ware), made from a refined clay and with a red slip and grooves, with no slip on the outside surface.

2.030 The largest percentage of the pottery is still handmade. The most characteristic materials are cooking vessels with rims that are hardly distinguishable (fig. 3.13–16), reticulated decorations with a narrow crosshatch, cups with thin sides and concave rims (fig. 3.17), open vessels with bell-shaped necks

280

Diego Ruiz Mata

Fig. 3. Castillo de Doña Blanca (Puerto de Santa María, Cádiz).

The Phoenician Presence in Southwestern Andalusia

281

(fig. 3.18–20), rough-surfaced pots, etc. Among the wheel-made material, we found a saucer that was decorated on the inside with a broad black band on the rim and with other narrower black/grayish bands below, over an orange background (fig. 3.22). Fragment 21 in figure 3 belongs to a similar saucer without decorations. Other fragments are body sherds of urns and amphorae.

2.028 The ratio of handmade to wheel-made pottery appears more balanced. Additionally, we note new features on the rims of the cooking vessels (fig. 3.23– 25), which tend to lose their external carination or to get thicker on the inside (fig. 3.23). The reticulated, burnished decorations continue, and on some body sherds we note a design with a wider crosshatch. The wheel-made materials consist of a red slip plate rim (fig. 3.26), the side of a juglet (fig. 3:27), and neck and body sherds of urns with a decoration of red and black bands.

2.027 A balanced ratio is maintained between handmade and wheel-made pottery. The cooking vessels have rims that are similar to those of the previous stratum (fig. 3.28–36), with hardly-noticeable carinations. However, as a novelty, the rims of numbers 28, 31, and 32 tend to be concave. The cooking vessels with incised decorations on the burnished exterior surfaces continue (fig. 3.37), as do the rough-surfaced pots with short, concave necks and with rims or shoulders that are sometimes decorated with fingerprints. Also handmade bowls and urns with bell-shaped necks continue. The normal wheel-made pottery is decorated with a red slip, and the pottery decorated with bands has become scarce. The most frequent types are plates with a high-quality red slip (fig. 3.38–42); carinated bowls decorated in the same way, with very low sides (fig. 3.43, 44) or of a larger size with higher sides; neck and body sherds of oenochoes; amphorae with short necks (fig. 3: 46, 47) and prominent rims on the outside (fig. 3.48); juglets (fig. 3.45); flat-based lamps without a slip; a few body sherds of urns decorated with bands; and ostrich eggs, with remains of ochre inside and a line decoration on the outside.

2.026 We see characteristics similar to those of Stratum 2.027. On the cooking vessels we note types with evolved rims (fig. 4.1–8) and hardly any carination, or that become thicker on the inside, or that are concave, decorated with medium and narrow burnished crosshatching. The cups with concave rims continue (fig. 4: 10, 11), among which we should mention a cup with very thin sides, well-burnished surfaces, and an interior decoration of wide burnished bands (fig. 4.9). Other materials are pots with concave necks (fig. 4.12), á chardon vessels (fig. 4.13), and sherds of cups with an incised decoration

282

Diego Ruiz Mata

Fig. 4. Castillo de Doña Blanca (Puerto de Santa María, Cádiz).

Fig. 5 (opposite). Castillo de Doña Blanca (Puerto de Santa María, Cádiz).

The Phoenician Presence in Southwestern Andalusia

283

284

Diego Ruiz Mata

(fig. 4:14). The wheel-made pottery does not differ from the earlier forms: a red slip plate (fig. 4.15), a red slip carinated bowl with high sides (fig. 4.16), and an amphora with a short, pointed rim (fig. 4.17).

2.024 The handmade cooking vessels continue their evolution, with a tendency to lose their carinations (fig. 4.18–25). The same handmade and wheel-made materials continue (fig. 3.33–41). In Area 4 (fig. 5) we have also reached virgin rock. The lowest stratum corresponds to a horizon of the Chalcolithic period, and after a hiatus (a sterile level of reddish clay), we find the protohistoric settlement. The handmade cooking vessels have sides with slight carinations (fig. 5.1–8) and are decorated with reticulated designs (fig. 5.12, 13). The rough-surfaced pots have short, concave necks (fig. 5.9–11). The Phoenician wheel-produced material shows the same repertoire as Area 5 (fig. 5.14–29). We should point out that these strata are characterized by a low percentage of pottery that has been painted with bands, in favor of red slip pottery. In summary, these are the data for the first period of settlement and development in the southeast sector of the town of Doña Blanca. This zone, especially the strip closest to the ancient coastline, was occupied during the Chalcolithic period by a population that was culturally similar to that of the lower Guadalquivir and Huelva. The zone diminished in population toward the end of the Bell Beaker period, as shown by the degenerate ceramic forms and decorations. Probably, after a hiatus, a small community belonging to the Tartessian Final Bronze was established (perhaps in the 9th century b.c.), located on the far eastern end and on the sides of the Sierra de San Cristóbal. Scattered remains have been found from this period, but they are few in number, suggesting a small nucleus of population. It is evident that the Phoenician presence in Gadir and the initiation of organized trading activities attracted a more numerous population. The spot where the town of Doña Blanca was to be established brought together optimum conditions, since it had an excellent natural port and was close to the Guadalete River on the east, the best route toward the interior. To the north is the Sierra de San Cristóbal, low in elevation and easily crossed, and from there a route departed toward the estuaries at Mesas de Asta and Ebora. After crossing the estuary of the Guadalquivir, one could reach El Rocío and Almonte, where we have verified the existence of settlements that refined silver. The island of Cádiz required support points on the coast that were in contact with the lower Guadalquivir and the metal-producing zone, but which avoided the route that led to Tartessos. For this reason, the settlement at San Bartolomé (Almonte) was economically linked to Gadir. The for-

The Phoenician Presence in Southwestern Andalusia

285

tified town of Tejada la Vieja (Escacena, Huelva), which Jesús Fernández Jurado is now excavating, also seems to be part of this network. We should additionally mention the surveys that are being performed between Puerto de Santa María and the Guadalquivir River. These surveys are documenting a route which linked Final Bronze I–II settlements containing orientalizing materials. This route led toward the area of Ebora, where one could cross the river and reach El Rocío and Almonte. Another obvious route is the Guadalquivir River itself. Thus, the location of Doña Blanca can be justified. It was a trading center, possibly controlled by Gadir, in an enclave that was favorable for reaching the lower Guadalquivir, the Gaditanian mountains across the Guadalete River, and the metal working areas. By situating itself on the borders of Tartessos, Gadir could control agricultural and mining centers with the same economic potential as Tartessos. This is probably the meaning of the text of Posidonius when he mentions the two previous voyages (to Sexi and to Huelva) which occurred before the founding of Gadir. In other words, the purpose of these voyages was a reconnaissance of economic and colonial possibilities, information about Tartessos, and the final founding of Cádiz, with all the advantages already mentioned. As for the period in which the settlement of Doña Blanca took place, we presently rely on the data that we have described from Areas 4 and 5. The earliest remains (related to the rectangular dwellings of rubblework) appear to be found mostly on the southern side of the town, closer to the water. Perhaps they indicate a small trading post with purely commercial purposes, which later increased in size and urban organization during the 7th century b.c. The earliest strata correspond to a late period of the Final Bronze, which we have called Phase I–II at San Bartolomé and during which the first Phoenician elements appear in the indigenous environment. The handmade indigenous cooking vessels (which are excellent guides for establishing the various phases) show evolved forms, seen in the rims that have lost the prominent carinations that characterize the pre-Phoenician Final Bronze I in western Andalusia. The greatest percentage of pottery is indigenous, as can be expected. But in little time, in a rapid process of rapprochement and assimilation, the ratio becomes equal, as has been mentioned for Stratum 2.028. Since the material is still being studied, it is impossible to be more precise. However, in regard to the material produced on the potter’s wheel, it is possible to point out the almost exclusive use of red slip pottery and the scarcity of vessels decorated with bands. As chronological indexes, we want to emphasize that the red slip plates have narrow rims, between 2.1 and 3.5 centimeters. When correlated with the data obtained from the trading posts in Málaga (Schubart 1976), these plates can be dated to around the middle of the 8th century b.c. Nevertheless, these plates seem to be preceded by strata in

286

Diego Ruiz Mata

which wheel-made Phoenician materials have appeared, though this has not been documented. Moreover, in another area that we excavated (Area 4), we found plates with rims of around 2 centimeters at the bottom level, which at Morro de Mezquitilla appear to indicate an earlier period (Schubart 1983)— the first half of the 8th century b.c. The Phoenician presence in Castillo de Doña Blanca probably began during the first half of the 8th century b.c., at least according to the data provided so far by the excavations. This indicates that Gadir was probably founded around 800 b.c., since we must keep in mind that the first Phoenician settlers did not remain isolated for very long. They required coastal bases for the initiation and expansion of their activities in order to meet their economic objectives.*

IV. The Town of El Carambolo El Carambolo is a hill located on the eastern edge of the Aljarafe plateau, 91 meters above sea level and three kilometers from the city of Seville. Excavations resulted from the accidental discovery of a treasure of gold pieces on September 30, 1958 (Carriazo 1973). Excavations focused on the area where the gold was found (an area known as fondo de cabaña ‘area of huts’) and later on the hillside, which has been called poblado bajo ‘lower town’. The earliest material from pre-Phoenician Final Bronze I was excavated in the lower levels of fondo de cabaña, specifically in Strata IV and III. The upper strata (II and I) formed as fill on top of this area and therefore have little relation to it, but correspond instead to the orientalizing period. We need to make this point, since it has been frequently said that the four strata belong to the fondo when in reality the only related strata are IV and III. In these two strata are found materials characteristic of Final Bronze I, with the exception of some wheel-made elements (Ruiz Mata 1981). The later orientalizing strata are found in Strata II and I of fondo de cabaña and also in poblado bajo, where an extended area of 375 square meters was opened. In summary, four levels have been distinguished in poblado bajo (A, B, C, D, according to the excavator) that correspond to the remains of superimposed rooms. It must be pointed out that sketches of the stratigraphies and the location of materials in their corresponding strata are lacking. The difficulties are at times insurmountable for an accurate review. In addition, the sherds were identified with adhesive labels, which of course have fallen off. Paloma Cabrera Bonet has sketched the red slip material in the Archaeological Museum of Seville and has kindly given us permission to republish it.

Editor’s note: See also the author’s update on Castillo de Doña Blanca in his other article in this volume, pp. 170–198.

The Phoenician Presence in Southwestern Andalusia

287

Fig. 6. Typology of the rims of handmade cooking dishes from poblado bajo in El Carambolo.

The pottery from poblado bajo has been analyzed in a very general and succinct way, and only on a few occasions are the strata from which it comes identified. The data we have can be summarized by the following points: (1) Four levels of superimposed construction have been identified, and the earliest (Level D), on top of virgin soil, is associated with indigenous handmade material and Phoenician wheel-made pottery, probably imported to the town. (2) The painted monochrome pottery of the “Guadalquivir I” type is practically absent, except for three fragments that preserve vestiges of paint and could be from a later style. The types that characterize fondo de cabaña, such as cooking vessels, cups, stands, etc., were no longer in use. (3) Handmade pottery still persists (abundantly in Level IV), but only the cooking vessels and bowls with burnished decorations are mentioned and sketched. All together, 340 fragments of cooking vessels were collected. These are described in groups: 111 come from Level IV, 21 from Level III, 5 from II, and 9 from I. The cooking vessels have profiles and characteristics which differ from the earliest in fondo de cabaña. (4) According to the excavator, the wheel-made pottery appears to be very abundant, but he does not indicate percentages or types for the various levels. He mentions gray pottery, amphorae (that supposedly are half of all the fragments excavated), pottery decorated with bands, and red slip pottery. The great quantity of amphorae (50 percent of the wheel-made fragments) suggests the importance of the spot for commercial interests. The best-known form is the handmade cooking vessel (fig. 6), and most of these vessels come from Level IV. The rims and surface finishes are distinct from the cooking vessels in fondo de cabaña. Types I and II with concave rims are the most numerous and can be dated to 725–700 b.c., as hut XIV.B of

288

Diego Ruiz Mata

1

2

3

4

5

6

7

8

9

10 11

12

13

14

15 16

17

Fig. 7. Red slip plates from El Carambolo.

The Phoenician Presence in Southwestern Andalusia

289

San Bartolomé suggests (Ruiz Mata 1982) and to 650–625 b.c., according to the chronology that Castillo de Doña Blanca has provided. Type IV, with a short, angled rim, is not as frequent and is used more in the region of Huelva, where it is dated to the first half of the 7th century b.c. The stratigraphic position of Type V appears to be more recent than that of Types I and II. This is also true for the Type IX bowl, which in poblado bajo appears beginning at Level III. In summary, Level IV can be dated from the end of the 8th century b.c. until a little after the middle of the 7th. Rims belonging to the transition phase (Final Bronze I/II) have not been found among the cooking vessels at poblado bajo. However, they have been found at Castillo de Doña Blanca and San Bartolomé, which is the context in which the first signs of the Phoenicians appear. The oldest handmade materials at poblado bajo (Level IV) start at the end of the 8th century b.c. and beginning of the 7th, dates which are suggested also by the red slip pottery. We have already mentioned that in the publication of these materials neither the strata nor the phases to which these materials belong were specified and that the majority have lost the adhesive labels that recorded the data. However, it is possible to correlate the type and quality of their slips with materials that have better stratigraphic evidence at other sites. The plates are relatively numerous and can be compared by typology with those at the town of Doña Blanca and at the trading posts of Málaga. Figs. 7 and 8 organize the plates according to the width of the rims, and fig. 9 summarizes the data. Narrow plate rims not exceeding 3.5 centimeters (those between 1.5 and 2.9 centimeters) are present on seven occasions (fig. 7.1–7). We know that plate 5 (2.5 cm.) comes from Level D (or Stratum IV) and that plate 1 (1.5 cm.) comes from Level C (III). Both were found 3 meters down, according to the labels. It is very possible that both fragments belong to the same stratum, given the ambiguity with which the levels are described. The ware is orange (only on two occasions do we see an orange core, fragments 5 and 7) with very fine grits. The slips, of quite good quality (except on fragments 3 and 4, which have very diluted slips), are thick and form a consistent film, and they are brown or brownish-red in color. The slips cover the whole inside surface of the plates; on the outside, the slips reach somewhat lower than the width of the rim on fragments 5, 6, and 7, and they reach only to the upper area of the rim on fragments 1 and 4. The characteristics of the slips closely resemble those at the town of Doña Blanca in levels that are older or contemporary, and the pottery at El Carambolo probably comes from that region (either Doña Blanca or Cádiz). During the initial phase of Doña Blanca, the width of the rims is between 1.8/2.0 and 3.5 centimeters. However, the handmade cooking vessels at Doña Blanca are transition types (Phase I/II) from the middle of the 8th century b.c., with less pronounced carination of the rims. The cooking vessels from poblado bajo (in

290

Diego Ruiz Mata

1

2

3

4 5 6

7

8

9

10

11

12

13

14

Fig. 8. Red slip plates from El Carambolo.

The Phoenician Presence in Southwestern Andalusia

291

Level IV/D) are more evolved. When we correlate these cooking vessels with those from other sites in southwestern Andalusia, they can be dated from the end of the 8th century b.c. until a little after the middle of the 7th. The majority of the plates have rims between 3.5 and 5.0 centimeters (fig. 7.8–17 and 8.1–8). The plates with rims between 3.5 and 4.0 centimeters are covered with high-quality slips which extend over the whole interior and a large part of the exterior surface, down to a point below the edge of the rim. The plates that have rims of 5 centimeters or more are inferior in quality. The slips are dark, light, or reddish-brown in color. Another group of plates has rims between 5.5 and 8 centimeters wide (fig. 8.9–14), and they also are covered with dark, light, or reddish brown slips that are uneven and faded in most cases. Keeping in mind the lack of stratigraphic documentation for the materials, we can assume that the plates show a typology similar to the evolution observed in the trading posts and necropolises of Málaga and Granada and in the closer town of Castillo de Doña Blanca. Thus, we can date the materials from the end of the 8th century b.c. to the beginning of the 6th. Other open forms (also without stratigraphic information) are the carinated red slip bowls that show different tendencies in the formation of their rims (fig. 10.1–8), documented, for example, in Toscanos (Schubart and Lindemann 1984) and in Castillo de Doña Blanca. Fragments 3, 4, and 5 (fig. 10) have open rims with pointed edges and high carinations and are covered with high-quality slips. Another type shows an almost vertical rim (fig. 10.1), covered with a light red slip. Fragment 3, in the same figure, has a vertical rim that is shorter and pointed, with an intensely burnished light brown slip. Bowls 7 and 8 (fig. 10) have a smaller diameter; fragment 7 has a vertical rim and a low carination, and fragment 8 has an extremely concave rim. Both bowls are covered with a high-quality brown slip. Other open forms are the bowls with concave rims and high carinations (fig. 11.1–4). These bowls are characteristically covered with very faded reddish brown and dark brown slips, in contrast to the consistent, thick slips of the carinated bowls. This type is similar to the forms coming from strata of the 7th century b.c. in Doña Blanca. The urns have a cylindrical, vertical neck (fig. 11.5, 7) with broad bands of red slip on the exterior surface, or a short neck inclined slightly inward (fig. 11.6). The body tends toward a spherical shape. A geminate handle takes up almost the whole neck and sits on top of the upper area of the body. Only the edge of the rim is decorated with a covering of dark red slip. Fragment 8 (fig. 11) has a central raised band around its neck and a more prominent rim, and it is decorated with a very faded red slip. Finally we should mention that body sherds of oenochoes (fig. 11.9–11), with grooves on the upper section and with burnished reddish-brown slips of

292

Diego Ruiz Mata

1

3

2

4

5

6

Fig. 9. Widths of rims of the red slip plates from El Carambolo (black circles) and Castillo de Doña Blanca (white circles).

8 7

9

10

11

Fig. 10 (right). Red slip bowls and plates from El Carambolo.

Fig. 11 (opposite page). Phoenician bowls, urns, and oenochoes from El Carambolo.

high quality, are included in this collection of Phoenician pottery from poblado bajo. The inside surface of fragment 10 is covered with a very faded violet paint. Fragment 12 comes from the neck of a mushroom-mouth oenochoe, which is covered with a burnished cream slip applied vertically and a red slip on the upper section of the mouth. Fragment 13 belongs to an oenochoe with a trefoil mouth and is covered with an uneven light brown slip.

The Phoenician Presence in Southwestern Andalusia

293

1

2

3

4

6

5

8

7

12 9

10

13 11

294

Diego Ruiz Mata

The data provided by the excavator do not help much when we try to reach a conclusion about the stratigraphic position of the pottery. We can give only a global appraisal of the red slip ceramic type. As for the plates, we see the same typology that is already well known at the trading posts of Málaga and the current excavations at Castillo de Doña Blanca. Fig. 9 illustrates the width of the rim for this type of plate. The black circles show the widths from poblado bajo and the white circles show the widths from the earliest levels of Doña Blanca, which we have discussed earlier in this article. We see plates with narrow rims (up to 3.3–3.4 centimeters), but the majority have widths between 3.5 and 5.0 centimeters. When we correlate these data with those from the trading post at Toscanos, it appears that rims with widths of less than 3.5 centimeters (Schubart 1976) can be dated to the second half of the 8th century b.c. We can date the rim width of the imports entering El Carambolo to around the transition from the 8th to the 7th centuries b.c. or to the beginning of the 7th. We mentioned earlier that in Doña Blanca the narrow-rimmed plates are found with handmade cooking vessels that are lightly carinated on the outside, while in poblado bajo these plates are found with cooking vessels whose rims are more concave and that seem a little later. Perhaps the beginning of the imports should be placed in the second half of the 8th century b.c., and probably closer to its end. The type of ware and slip is similar to that at Castillo de Doña Blanca, which should be a faithful reflection of Cádiz due to the proximity of the two sites. Once the Phoenicians were established in Cádiz, probably not much time passed before they sailed up the Guadalquivir. One of their points of contact must have been the town of El Carambolo. The remaining forms find strict parallels in the trading posts of Málaga and Castillo de Doña Blanca. Carinated bowls like 1, 2, and 4 of figure 10 are found in strata at Doña Blanca dated to the second half of the 8th century. Number 3 appears a little later, during the first half of the 7th century b.c. The bowls with concave rims have the same dates. In summary, and keeping in mind the necessary reservations in this case, the Phoenician presence in El Carambolo probably began in the latter part of the second half of the 8th century b.c. and intensified at the beginning of the 7th. The Phoenician economic interest in this place must have been great. Thus it is very interesting that the excavator mentions that almost half the pottery that was found came from amphorae.

V. Conclusions (1) Before the arrival of the Phoenicians in western Andalusia, a complex nucleus of population (of unknown origin) had taken shape, principally in the Basin of the Guadalquivir and Huelva. The cultural horizon known as “Tar-

The Phoenician Presence in Southwestern Andalusia

295

tessian” was composed of their small settlements and economic activities. During this period we see signs of metal working, with the goal of extracting silver, in the strata of Phase I at Cabezo de San Pedro and on the floors of huts in San Bartolomé. Thus the Tartessian economic activity that provided the basic attraction for the Semitic traders is clarified. However, within this homogeneous cultural horizon, we can evaluate the differentiating characteristics of certain ceramic forms and of the burnished and painted decorations. This allows us to identify at least two large nuclei, one around Huelva and the other in the lower Guadalquivir. The stratigraphic results and analyses of slags and clays suggest signs of pre-Phoenician silver metallurgy, with considerable development under Phoenician influence. They also suggest the existence of at least two foci of metal extraction, the zone of the Río Tinto reaching down to Huelva, and the zone of Aznalcóllar, which traded directly with the Phoenician colony of Gadir, avoiding the former route. This hypothesis is also suggested by the analysis of slag and of the red slip pottery from both zones, linking Gadir with San Bartolomé. (2) When we combine these data with the information from Posidonius and passed on to us by Strabo (3.5.5), the purpose of the founding of Gadir becomes clearer. According to these accounts, the first attempts at Phoenician settlements on the Iberian Peninsula took place at Sexi (identified with Almuñécar) and later, after passing through the Gates of Hercules, at Onoba (probably Huelva). The founding of Gadir occurred on the third trip. It is clear that the purpose of the first voyage was to survey and explore the possibilities for trade and settlement. On the second try the Phoenicians reached Huelva, a center where silver-related activity probably began during the pre-Phoenician era and where it would not have been easy for the Phoenicians to initiate an advantageous trade during that initial period. However, it is obvious that relations began around this time, as suggested by the construction of the wall of San Pedro and the red slip ceramic types from the first half of the 8th century b.c. It is also probable that the Phoenicians who settled in Huelva had little to do commercially with those who founded Gadir, as seems to be indicated by the separate development of the pottery in Huelva. This pottery does not parallel the ceramic assemblage from Gadir, which we are now beginning to recognize because of the excavations at Castillo de Doña Blanca. In this regard, Schubart (1976) noted differences in the typological evolution of plates from the trading posts at Málaga and Huelva, differences that exist also in other forms. Whatever the case, the Phoenicians finally founded Gadir, and from there used a different route for transporting metal, a route that reached Aznalcóllar by way of El Rocío, Almonte, and Tejada la Vieja. This means that Gadir avoided the route of the Río Tinto to Huelva, which was under indigenous control. In fact, there are no early Phoenician settlements on this section of the Atlantic coast except for Gadir. Gadir was rather far from the centers of

296

Diego Ruiz Mata

production in Huelva, but had easy and perhaps politically favorable access to Aznalcóllar and the rich agricultural and cattle-raising regions of the Guadalquivir. All of this probably took place during the first half of the 8th century b.c., judging by the recent findings. (3) We have analyzed certain material aspects of the Phoenician presence in western Andalusia. Huelva has not yet provided clean strata of occupation from the first period of the Phoenician presence. This is not because of lack of excavation, but perhaps because of the system of nuclear settlements on hills, the current systematic destruction that has erased a large part of the early evidence, and the water table that prevents digging to greater depths. Nevertheless, the excavations at Cabezo de San Pedro provide elements that date the Phoenician presence in Huelva to the first half of the 8th century b.c. and even to the beginning of the century. This can be deduced from the retaining wall itself, its stratigraphy, and the fragment of an Attic crater from the midGeometric II, dated between 800 and 760 b.c. Castillo de Doña Blanca has provided more pottery from this first period. Because of its proximity to Cádiz (Gadir) and its location on the ancient coast, Doña Blanca may reflect the early period of Cádiz, where early strata are still unexcavated. The results of the findings at Castillo de Doña Blanca show that the Phoenician presence in the bay of Cádiz can be dated, as in Huelva, to the first half of the 8th century b.c. The town of San Bartolomé was of great interest to us because of its silver extracting activity, its clearly indigenous culture, its location in an area where there were no mines (though surely a trade route), and its probable trading relationships with Gadir. We have excavated a significant sample of about thirty huts, which contain predominantly indigenous materials. However, these huts also contain wheel-made pottery as a product of commercial contact, beginning perhaps in the second half of the 8th century b.c. at the earliest. Finally, El Carambolo has provided Phoenician materials, and its location next to the ancient mouth of the Guadalquivir makes it an interesting spot to evaluate the Phoenician presence in the rich cattle-raising and agricultural centers of the lower Guadalquivir. We have pointed out the problems that result from archaeological work that lacks proper stratigraphic records. However, the typological correlations of the red slip pottery from El Carambolo with the pottery from Doña Blanca and Málaga show that the Phoenicians probably sailed to this area beginning at the end of the 8th century, and their presence increased beginning in the first decades of the 7th century b.c.

Bibliography Amores Carredano, F. 1982 Carta arqueológica de Los Alcores (Sevilla). Series 1, no. 22. Seville: Diputación Provincial de Sevilla, Sección Historia.

The Phoenician Presence in Southwestern Andalusia

297

Aubet Semmler, M. E.; Escacena, J. L.; and Ruiz, M. M. 1983 La Mesa de Setefilla. Lora del Río (Sevilla). Campaña de 1979. EAE 122. Bikai, P. M. 1978 The Pottery of Tyre. Warminster: Aris & Phillips. Blázquez, J. M.; Ruiz Mata, D.; et al. 1979 Excavaciones en el Cabezo de San Pedro (Huelva): Campaña de 1977. EAE 102. Carriazo, J. de M. 1973 Tartesos y El Carambolo. Madrid. Chic García, G. 1979 Gades y la desembocadura del Guadalquivir. Gades 3: 7–23. Ciria y Vergara, F. 1935 Tartesos. Puerto de Santa María (typed text). Elayi, J. 1980 Remarques sur un type de mur phénicien. RSF 8.2: 165–80. Escacena, J. L. 1982 Antiguas vías de comunicación en el Bajo Guadalquivir. Gades 9: 129ff. Fernández Jurado, J. 1983 San Bartolomé de Almonte: yacimiento metalúrgico de época tartésica. Revista de Arqueología 26: 24–33. Galván, V. 1986 Análisis de pastas cerámicas. Chapter 29 in El yacimiento metalúrgico de época tartésica de San Bartolomé de Almonte (Huelva), eds. D. Ruiz Mata and J. Fernández Jurado. Huelva Arqueológica 8. Garrido, J. P., and Orta, E. 1978 Excavaciones en la necrópolis de “La Joya.” Huelva. II. (3.a, 4.a, y 5.a Campañas). EAE 96. Pemán, C. 1941 Nuevas contribuciones al estudio del problema de Tartesos. AEArq 42: 177– 87. Pellicer, M.; Escacena, J. L.; and Bendala, M. 1983 El Cerro Macareno. EAE 124. Peña Basurto, L. 1952 Tartesos, la ciudad más antigua de Occidente, descubierta. Revista Munibe. Pinedo, I. 1963 Piritas de Huelva. Madrid: Summa. Ruiz Mata, D.; Blázquez, J. M.; and Martín de la Cruz, J. C. 1981 Excavaciones en el Cabezo de San Pedro (Huelva). Campaña de 1978. Huelva Arqueológica 5: 149–316. Ruiz Mata, D. 1979 El Bronce Final – Fase Inicial – en Andalucía Occidental: Ensayo de definición de sus cerámicas.” AEArq 52: 3–19. 1981 El poblado metalúrgico de época tartésica de San Bartolomé (Almonte, Huelva). MM 22: 150–70. Ruiz Mata, D., and Fernández Jurado, J. 1985 La metalurgia de la plata en época tartésica en Huelva. Pyrenae 21: 23–44.

298

1986

Diego Ruiz Mata

El yacimiento metalúrgico de época tartésica de San Bartolomé de Almonte (Huelva). Huelva Arqueológica 8. Schubart, H. 1976 Westphönizische Teller. RSF 4: 179–96. 1983 Morro de Mezquitilla. Vorbericht über die Grabungskampagne 1981 auf dem Siedlungshügel an der Algarrobo-Múndung. MM 23: 33–45. Schubart, H., and Maass-Lindemann, G. 1984 Toscanos, el asentamiento fenicio occidental en la desembocadura del río Vélez. Excavación 1971. NAHisp 18: 39–210. Schulten, A. 1943 Forschungen in Spanien. AA 58: 19–62. Shefton, B. B. 1982 Greeks and Greek Imports in the South of the Iberian Peninsula. The archaeological evidence. Phönizier in Westen, Madrider Beiträge 8: 337ff. Ventura López, F. 1923 Articles published in the newspaper El Guadalete (Apr. 8, 1923; Apr. 10, 1923; Apr. 11, 1923; Apr. 15, 1923; Apr. 24, 1923; Apr. 25, 1923, Apr. 27, 1923; May 3, 1923; May 8,1923; May 18, 1923; May 20, 1923; Aug. 12, 1923; Dec. 2, 1923; Dec. 5, 1923; Dec. 7, 1923; Dec. 11, 1923). Williams II, C. K. 1978 Corinth, 1977: Forum Southwest. Hesperia 47.1: 1ff.

Index of Sites and Geographical Names

A Abdera 56, 80, 100 see also Adra Abul 104 Acebuchal 75, 203, 212–213, 216–217 Adra 18, 50, 80, 88, 160, 163, 202 see also Abdera Aegean Sea 159 Agde 108 Akhziv 66 Alalia 260 Alarcón 32, 192 Alcácer do Sal 104 Alcalá 205, 219, 235 Alcantarilla 213, 216 Alcaudete 213 Alcores 212–213, 216, 264 Aldovesta 108 Alemtejo 104 Algarrobo River 3–4, 17–18, 57, 71, 81, 89–90, 93 Algeciras 54, 100 Algeria 62, 64, 141, 150 Alicante xi, xiv, 101, 106, 113 Aliseda 104, 203, 205, 211, 217–219, 235 Aljarafe plateau 167, 264, 266, 286 Almanzora River 100 Almería 80, 92, 103, 160, 194 Al-Mina 159 Almonte 245, 247, 250–251, 253, 255– 260, 264–265, 270–274, 279, 284, 295 Almudaina 140, 142 Almuñécar xiv, 18, 42, 49–55, 62, 67, 74–77, 80–82, 84, 87, 100, 155, 157–158, 160–161, 188, 210, 214–215, 231, 263, 295 see also Sexi

Amarna 66 Andalusia 44, 49, 75, 80, 86–87, 94, 97–99, 101–103, 123, 151, 170, 172, 209, 222, 224–228, 231, 263–296 Angles 108 Antequera 86 Aragon 108 Arga River 225 Aroer 269 Assur 66–67 Assyria 67 Astorga 205 Asturica Augusta 205 Atalaia 212 Ategua 264 Athens 159 Attica 34 Aude River 108 Ayamonte 260 Aznalcóllar 157, 247, 249, 260–261, 271, 295–296 B Badajoz 158 Bahía-Caleta 163–166 Bahria 65–66 Baria 80, 87 Beira Alta 104 Bencarrón 75, 203, 213 Béticos (mountains) 264 Bithia 228 Botica 246–247 Bubastis 65 Byblos 42, 62 Byrsa 74 C Cabecico de Parra 101 Cabezo de la Esperanza 206 see also Esperanza

299

300

Index of Sites and Geographical Names

Cabezo de San Pedro 57, 206, 221, 229, 266–270, 295–296 see also San Pedro Cabezo del Estaño 123 Cádiz xiii–xiv, 50, 56, 75, 79, 84, 128, 155–198, 201, 230, 263–264, 266, 271, 275, 284–285, 289, 294, 296 see also Gadir, Gadeira Cádiz, Bay of 184, 190, 192 Caldero de Mojácar 215 Camarina 146 Campiña, la 264 Campo del Sur 164 Campos 215 Campos de Urnas 106, 108 Cañada de Ruiz Sánchez 75, 213, 219 Carambolo xv, 16, 102, 158, 161, 203, 205, 207, 210–211, 229–230, 232, 266, 272, 276, 286–294, 296 Carmona 42, 75, 102, 161, 194, 203, 206, 208, 212–213, 217–219, 223, 228–230, 232 Cartagena 75, 101 Carteia 54 Carthage xiii, 38, 44–45, 50, 53, 62, 64, 66–67, 70, 74, 85, 100–101, 123, 130, 156, 158–159, 192, 194, 228, 269 Castañuelo 75 Castellones de Ceal 211 Castillo de Doña Blanca xiv–xv, 161, 163, 170–198, 231, 264, 266, 272, 274–286, 289, 291, 294, 296 Castillo de Guardamar 123 Castro Marim 104 Cástulo 59, 169, 211 Catalonia 148, 186 Catujal 104 Cementerio Viejo 176, 266, 268–269 Cerca Niebla 91 Cerro de Alarcón 33, 38, 40 Cerro de la Cabeza 272 Cerro de la Mora 42

Cerro de la Tortuga 89 Cerro de los Infantes 42 Cerro de Salomón 264 Cerro de San Cristóbal 49, 59, 62, 81 Cerro del Mar 18, 67, 81 Cerro del Peñón 32, 34, 38, 40, 67 Cerro del Prado 75, 100, 161 Cerro del Trigo 169 Cerro del Villar 67, 72, 75, 84–85, 91–92, 94–95, 100, 192, 237 see also Guadalhorce Cerro Macareno 54, 75, 203, 205–206, 210, 223, 265 Ceuta 75 Chalcis 209 Chiclana 156, 169 Chios 75 Chorreras 4, 35, 80, 84, 93, 100, 123, 160, 163, 177, 201, 220, 277 Churriana 95 Coca 219 Colina de los Quemados 206, 264 see also Quemados Coll del Moro 108 Columns of Hercules 56, 155–157, 159, 295 Conímbriga 104 Córdoba 86, 157, 169, 206, 264 Coria 167, 205 Corinth 75, 209, 278 Crete 67, 159 Cruz del Negro 75, 203, 212–213, 215, 221, 232 Cueva de Siete Palacios 55, 57–58, 60–61 Cueva del Jarro 74 Cumas 74, 200 Cyprus 45, 57, 73, 159, 186, 228 D Dermech 50, 64 Doñana 167 Douïmes 50 E Ebora 203, 231, 284–285 Ebro River 106

Index of Sites and Geographical Names

Ebusus 80, 87 see also Ibiza Ecija 229–230 Egypt 66–67 El Oral 120 El Portal 275 El Rocío 157, 264, 271, 284, 295 Entremalo 203, 212–213 Erytheia 163–164 Es Codolar 130 Escacena 245, 249–250, 264, 285 Esperanza 203–204, 268 see also Cabezo de la Esperanza Espigón 197 Estremadura 104, 186, 208, 218, 226, 230, 234–235, 237, 264 Estremadura, Portuguese 104 Etruria 200, 220, 228, 232 Euboea 159, 209 F Formentera 129, 149 Frigiliana 42, 211 Fuengirola 84 G Gadeira 155–156 see also Gadir, Cádiz Gades 156 Gadir xiii–xiv, 56–57, 64, 79, 81, 86, 94, 98, 100–101, 103–108, 155– 157, 159–161, 163–167, 169– 170, 172, 174–175, 186, 192, 201–202, 204, 209–210, 219, 221, 228, 230–232, 237, 261, 263–264, 272–275, 277, 284– 286, 295–296 see also Cádiz Gaditanian islands 162–163 Gaditanian mountains 285 Gamaza mountains 175 Garonne River 108 Gebel Mlezza 62 Genil River 59 Gibalbín mountains 175 Gibraltar xi, xiv, 71, 84, 94, 101, 128, 141, 160–161, 163, 169, 199, 201, 207, 209, 217, 221–222

301

Golfe du Lion 106, 128 Granada 42, 49, 80, 86, 98–99, 101, 103, 143, 160, 188, 201–202, 230, 234, 291 Greece 64, 95, 192, 199–200, 278 Guadalete River 71, 79, 157, 163, 167, 169–171, 175, 190, 231, 265, 275, 278, 284–285 Guadalfeo River 103 Guadalhorce 18, 85, 92, 123, 160, 163, 192, 202 see also Cerro del Villar Guadalhorce River 71, 81, 84, 86, 88– 89, 91–95, 99–100, 104, 202, 237 Guadalquivir River xi, xiii–xiv, 42, 71, 79, 98, 103–104, 157–158, 161, 167, 169–170, 172, 174–175, 182, 184, 186, 196, 199, 202, 205–209, 211–212, 215, 217, 219, 223, 226, 229, 231–232, 234, 237, 250, 260, 264–266, 271–274, 276, 279, 284–285, 294–296 Guadarranque 18, 202 Guadiamar River 260 Guadiana River 104, 194, 202, 205, 208, 260 Guadiaro 100 Guadiaro River 103 Guardamar 106, 113–114, 123 see also La Fonteta H Haifa 95 Hazor 37, 178, 269 Heliopolis 66 Herault River 108 Herna 124 Herrerías 101 Hispalis 205 Huelva xi, xiv–xv, 54, 56–57, 98, 102, 155, 157, 161, 176–177, 182, 186, 194, 199, 202–203, 205– 207, 209, 211–215, 218, 220– 223, 229–230, 232, 237, 242, 244–251, 253–254, 257, 259–

302

Index of Sites and Geographical Names

260, 262–274, 276, 278–279, 284–285, 289, 294–296 see also Onoba I Iberia

xii, 56, 141, 146, 148, 150, 155, 160, 222, 225, 237 Iberian Peninsula xiii, xv, 14–15, 18, 32, 50, 52, 55, 57, 74–75, 79– 80, 84, 97, 105, 128, 150–152, 156, 159, 161, 169, 181, 186, 192, 194, 199, 201–202, 207– 209, 215, 222, 225, 232, 237, 241, 245, 260, 262–264, 266, 295 Ibiza xiv, 62, 75, 80, 106–107, 123, 127–152, 194 see also Ebusus Ibiza, Bay of xiv, 130, 142, 147 Illa Plana 142, 152 Italy 199–200, 228 J Jerez de la Frontera 264, 275 Jete River 59 Joan Tur Esquerrer, Point of 141, 146, 152 Joppa 269 Junon 50, 62, 64 K Kalpe 56, 155 Karga 65 Khaldeh 66, 178 Khios 34 Kishon River 95 Kition 159, 228 Kos 159 Kotinoussa 163 Kouass 75, 278 Kush 67 L La Dehesa 172–173, 276 La Fonteta xiv, 113–124 see also Guardamar La Joya 186, 194, 203, 208, 212–217, 219–220, 228, 232–233 La Rábita 114 Lagos 231

Las Cumbres 173, 175, 182 Latium 200, 228, 232 Laurion 159 Laurita 50, 52–55, 59, 62–68, 70–76, 82, 161, 188 Lebanon 86 Lebrija 167, 229, 231 Libya 66, 156, 159 Ligustinus, Lake 167, 231 Linares 169 Lisbon xi, 104, 161 Lixus 45, 81, 106, 228 Loma del Aeropuerto 100 Lora del Río 205 M Mailhac 108 Mainake 32, 56 Málaga xiv, 16, 32, 42, 51, 71, 80, 83– 84, 86, 89, 91–94, 98–99, 101, 103, 143, 158, 160, 163, 177, 182, 188, 201–202, 232, 237, 277–278, 285, 289, 291, 294– 296 Malaka 56, 80, 84, 86–87 see also Málaga Malta 62, 74, 228 Manzanilla 271 Mar Menor 106 Marseilles 278 Matalascañas 167 Mazagoso 75 Mazarrón 101 Medellín 104, 205, 211 Medina de las Torres 157 Megiddo 269 Mende 75 Menesteo 275 Mérida 104, 219–220 Mesa del Castillo 264 Mesas de Asta 203, 205, 210, 212, 231, 284 Mezquitilla 3, 123, 201–202 Minet-el-Beida 67 Mogador 71, 75, 106, 192, 277 Molino del Viento 268–269 Mondego River 104

Index of Sites and Geographical Names

Montemolín 229 Montilla 103 Morocco 57, 106, 141, 150, 192 Morro de Mezquitilla xiv, 3–29, 77, 80– 81, 84, 86, 93, 100, 158, 160, 163, 177, 181, 277, 286 see also Mezquitilla Motril 84 Motril River 86 Motya 37, 45, 50, 64, 74, 87, 123, 158, 269 Murcia 106 N Naukratis 137–138 Nerja 92 Niebla 219, 264 Nora 228, 269 Nubia 67 O Odiel River 204, 266 Onoba 56–57, 155, 157, 263, 266, 295 see also Huelva Osuna 203 P Palestine 40, 62 Panormo 87 Parazuelos 215 Peña Negra 106, 124, 192 Peñalosa 98 Penibetic Mountains 86, 88, 202 Phocaea 152 Phoenicia 53, 67, 159 Pitecusas 55, 74, 101, 200, 219 Pitiuses 129, 149 Portugal xi, 103, 106, 169, 186, 208, 226, 235, 242, 260 Puente de Noy 54–55, 62–64, 71–72, 74, 77 Puerto Chico 164 Puerto de Santa María 67, 169, 173, 231, 264, 275, 278, 285 Puerto Real 169 Puig de Vila 128, 130, 140–141, 151 Puig des Jondal 130 Puig des Molins 128–130, 134–140, 142, 144–147, 152

303

Punta de Vaca 166 Punta del Vapor 74 Pyrenees (mountains) 108 Q Quebrantahuesos 75, 264 Quemados 203, 210 see also Colina de los Quemados Quinta do Almaraz 104 R Rachgoun 64 Ras Shamra 42, 269 see also Ugarit Rechidiye 66 Retiro de Churriana 67 Rhodes 74, 209 Río Salado 184 Río Tinto 204, 264, 266, 268, 271, 295 Riotinto xi, 79, 157, 211, 237, 243–244, 260–261, 264, 268 Rota 278 S Sa Caleta xiv, 106, 123, 128, 130–131, 135, 137, 139, 146–152 Sa Mola 130 Sado River xi Saladares 106, 192 Salobreña 103 Samos 57 San Antioco 158, 181 San Bartolomé xv, 98, 229, 245, 247– 249, 251–260, 264, 266, 270– 274, 279, 284–285, 289, 295– 296 San Fernando 169–170 San Juan 143 San Julián 95 San Pedro 176, 203–204, 206, 266, 268–269, 272, 295 see also Cabezo de San Pedro San Vicente 169 Sancti Petri 156, 166, 175 Sanlúcar de Barrameda 157, 167 Santa Lucía 142, 152 Santa Olaia 104 Santarém 104 Santiponce 272

304

Sardinia

Index of Sites and Geographical Names

44, 62, 81, 85, 101, 134, 150, 158, 181, 194, 228, 231 Sarepta 269 s’Argentera 148 Seco River 57, 59, 81 Segura River xi, xiv, 101, 106, 118, 122–124 Setefilla 75, 102, 203, 205–206, 208, 210–213, 215–217, 220, 223, 228–230, 232, 234, 237 Setúbal 104 Sevilla 16, 54, 205–206, 212, 218, 222, 242, 247, 264, 266, 271–272 Seville 86, 157, 205, 231, 286 Sexi xiv, 49–77, 80, 82, 157–158, 160, 163, 263, 285, 295 see also Almuñécar Sicily 64, 81, 85, 87, 134, 158, 194, 228 Sidon 42, 67 Sierra de Crevillente 124 Sierra de San Cristóbal 169–170, 173– 174, 275–276, 284 Sierra Estremadura 205 Sierra Morena 205, 264 Sines 203 Siruela 205, 219–220, 235 Solunto 87 Sudan 67 Sulcis 100 Syracuse 146 Syria 62, 200 Syria–Palestine 57 T Tabarca 123 Tagus River 104, 208–209, 215 Tanis 66–67 Tapada da Ajuda 104 Tarsis 244 Tartessos xi, 79, 86, 98, 100, 103, 152, 162, 169, 188, 192, 199–262, 266, 271, 275, 284 Tas Silg 228 Tejada xv, 75, 98, 245, 249–250, 253– 254, 257, 259–260, 264–265, 270–271, 274, 285, 295 Tell Keisan 178

Tharros 45, 123, 219 Thebes (Egypt) 65 Thorikos 159 Tingentera 156 Tinto see Río Tinto Torre de Doña Blanca 71, 100, 231 Torre de Tavira 164–165, 175 Torre del Mar 32, 51 Torres Vedras 219, 235 Tortuga 90 Toscanos xiv, 17–18, 31–45, 54, 67, 71– 72, 74–75, 80–81, 86–87, 90– 91, 95, 100, 123, 160, 163, 177, 184, 186, 192, 201–202, 220, 231, 237, 277, 291, 294 Trayamar 4, 72, 81, 87, 161, 187–188, 194, 210, 215–216 Troy 156 Tunis 62, 134 see also Gebel Mlezza Tyre xiii, 42, 44–45, 67, 86, 94–95, 100–101, 155, 160, 163, 175, 178, 192, 209, 228, 230, 237, 269 U Ugarit 42 see also Ras Shamra Utica 45, 56, 62, 64, 156, 158, 228 V Valdegamas 205, 211, 219–220, 235– 236 Valdelagrana 170 Valencia 124 Valencina 203, 205, 272 Vélez River 18, 32, 42, 51, 57, 71, 81– 82, 86, 88, 90–91, 95 Vélez-Málaga 84, 91 Vélez-Zafarraya 202 Verde River 57, 103 Villalba del Alcor 264 Villanueva de la Vera 219 Villar 81, 95 Villaricos 75, 80, 82, 92, 100–101, 161 Vinarragell 106, 192 Z Zarza de Alange 235–236